BR112015030103B1 - COMPRESSION OF SOUND FIELD DECOMPOSED REPRESENTATIONS - Google Patents

Abstract

compression of decomposed representations of a sound field. These are, in general, techniques that are described to compress decomposed representations of a sound field. a device comprising one or more processors may be configured to perform the techniques. The one or more processors may be configured to obtain a bit stream comprising a compressed version of a spatial component of a sound field, the spatial component being generated by performing a vector-based synthesis with respect to a plurality of spherical harmonic coefficients.

Description

[0001] This application claims the benefit of Interim Application No. US 61/828,445 filed May 29, 2013, Interim Application No. US 61/829,791 filed May 31, 2013, Interim Application No. US 61/899,034 filed on November 1, 2013, Interim Application No. US 61/899,041 filed on November 01, 2013, Interim Application No. US 61/829,182 filed on May 30, 2013, Interim Application No. US 61/829,174 filed on 30 May 2013, Interim Application No. US 61/829,155 filed May 30, 2013, Interim Application No. US 61/933,706 filed January 30, 2014, Interim Application No. US 61/829,846 filed May 31 2013, Provisional Application No. US 61/886,605 filed on October 03, 2013, Interim Application No. US 61/886,617 filed on October 03, 2013, Provisional Application No. US 61/925,158 filed on January 8, 2014 , Provisional Application No. US 61/933,721 filed on January 30, 2014, Application Pr Interim Application No. US 61/925,074 filed on January 8, 2014, Interim Application No. US 61/925,112 filed on January 8, 2014, Interim Application No. US 61/925,126 filed on January 8, 2014 and Interim Application No. No. US 62/003,515 filed May 27, 2014 and Interim Application No. US 61/828,615 filed May 29, 2013, the entire contents of which are incorporated herein by reference.

FIELD OF TECHNIQUE

[0002] This disclosure pertains to audio data and, more specifically, the compression of audio data.

BACKGROUND

[0003] A higher order ambisonics (HOA) signal (often represented by a plurality of spherical harmonic coefficients (SHC) or other hierarchical elements) is a three-dimensional representation of a sound field. This HOA or SHC representation can represent that sound field in a way that is independent of the speaker geometry used to reproduce a multichannel audio signal rendered from that SHC signal. This SHC signal can also facilitate backwards compatibility as this SHC signal can be rendered to well-known and highly adopted multi-channel formats such as a 5.1 audio channel format or a 7.1 audio channel format. SHC representation can therefore enable a better representation of a sound field that also accommodates backwards compatibility.

SUMMARY

[0004] In general, techniques are described for compressing and decompressing higher-order ambisonic audio data.

[0005] In one aspect, a method comprises obtaining one or more first vectors that describe distinct components of a sound field and one or more second vectors that describe the background components of the sound field, both the one or more first vectors and the one or more second vectors generated by at least performing a transformation with respect to a plurality of spherical harmonic coefficients.

[0006] In another aspect, a device comprises one or more processors configured to determine one or more first vectors that describe distinct components of a sound field and one or more second vectors that describe the background components of the sound field, both the one or more first vectors than the one or more second vectors generated by at least performing a transformation with respect to a plurality of spherical harmonic coefficients.

[0007] In another aspect, a device comprises means for obtaining one or more first vectors that describe distinct components of a sound field and one or more second vectors that describe the background components of the sound field, either the one or more first vectors and the one or more second vectors generated by at least performing a transformation with respect to a plurality of spherical harmonic coefficients and means for storing the one or more first vectors.

[0008] In another aspect, a non-transient computer-readable storage medium has instructions stored therein which, when executed, cause the one or more processors to obtain one or more first vectors that describe distinct components of a sound field and one or more second vectors describing the background components of the sound field, either the one or more first vectors or the one or more second vectors generated by at least performing a transformation with respect to a plurality of spherical harmonic coefficients.

[0009] In another aspect, a method comprises selecting one of a plurality of decompression schemes based on indicating whether a compressed version of spherical harmonic coefficients representative of a sound field are generated from a synthetic audio object, and decompressing the compressed version of the spherical harmonic coefficients using that scheme selected from the plurality of decompression schemes.

[0010] In another aspect, a device comprises one or more processors configured to select one of a plurality of decompression schemes based on the indication of whether a compressed version of spherical harmonic coefficients representative of a sound field are generated from a decompression object. synthetic audio, and decompressing the compressed version of the spherical harmonic coefficients using that scheme selected from the plurality of decompression schemes.

[0011] In another aspect, a device comprises means for selecting one of a plurality of decompression schemes based on indicating whether a compressed version of spherical harmonic coefficients representative of a sound field are generated from a synthetic audio object, and means for decompressing the compressed version of the spherical harmonic coefficients using that scheme selected from the plurality of decompression schemes.

[0012] In another aspect, a non-transient computer-readable storage medium has stored instructions thereon which, when executed, cause one or more processors of an integrated decoding device to select one of a plurality of decompression schemes based on the indicating whether a compressed version of spherical harmonic coefficients representative of a sound field is generated from a synthetic audio object, and decompressing the compressed version of spherical harmonic coefficients using that scheme selected from the plurality of decompression schemes.

[0013] In another aspect, a method comprises obtaining an indication of whether representative spherical harmonic coefficients of a sound field are generated from a synthetic audio object.

[0014] In another aspect, a device comprises one or more processors configured to obtain an indication of whether representative spherical harmonic coefficients of a sound field are generated from a synthetic audio object.

[0015] In another aspect, a device comprises means for storing spherical harmonic coefficients representative of a sound field, and means for obtaining an indication of whether spherical harmonic coefficients are generated from a synthetic audio object.

[0016] In another aspect, a non-transient computer-readable storage medium has stored therein instructions which, when executed, cause one or more processors to obtain an indication of whether representative spherical harmonic coefficients of a sound field are generated. from a synthetic audio object.

[0017] In another aspect, a method comprises quantizing one or more first vectors representative of one or more components of a sound field, and compensating for the error introduced due to quantizing the one or more first vectors into one or more second vectors that are also representative of the same one or more components of the sound field.

[0018] In another aspect, a device comprises one or more processors configured to quantize one or more first vectors representative of one or more components of a sound field, and compensate for error introduced due to quantizing the one or more first vectors into one or more second vectors that are also representative of the same one or more sound field components.

[0019] In another aspect, a device comprises means for quantizing one or more first vectors representative of one or more components of a sound field and means for compensating for the error introduced due to quantizing the one or more first vectors into one or more second vectors which are also representative of the same one or more sound field components.

[0020] In another aspect, a non-transient computer-readable storage medium has instructions stored therein which, when executed, cause one or more processors to quantize one or more first vectors representing one or more components of a sound field. , and compensate for errors introduced due to the quantization of the one or more first vectors into one or more second vectors that are also representative of the same one or more components of the sound field.

[0021] In another aspect, a method comprises performing, based on a target bit rate, order reduction with respect to a plurality of spherical harmonic coefficients or decompositions thereof to generate reduced spherical harmonic coefficients or reduced decompositions thereof , wherein the plurality of spherical harmonic coefficients represents a sound field.

[0022] In another aspect, a device comprises one or more processors configured to perform, based on a target bit rate, order reduction with respect to a plurality of spherical harmonic coefficients or decompositions thereof to generate reduced spherical harmonic coefficients or reduced decompositions thereof, wherein the plurality of spherical harmonic coefficients represents a sound field.

[0023] In another aspect, a device comprises means for storing a plurality of spherical harmonic coefficients or decompositions thereof, and means for performing, based on a target bit rate, order reduction with respect to the plurality of spherical harmonic coefficients or decompositions thereof to generate reduced spherical harmonic coefficients or reduced decompositions thereof, wherein the plurality of spherical harmonic coefficients represents a sound field.

[0024] In another aspect, a non-transient computer-readable storage medium has stored therein instructions which, when executed, cause one or more processors to perform, based on a target bit rate, order reduction. with respect to a plurality of spherical harmonic coefficients or decompositions thereof to generate reduced spherical harmonic coefficients or reduced decompositions thereof, wherein the plurality of spherical harmonic coefficients represents a sound field.

[0025] In another aspect, a method comprises obtaining a first non-zero set of coefficients of a vector that represents a distinct component of the sound field, the vector being decomposed from a plurality of spherical harmonic coefficients that describe a field sound.

[0026] In another aspect, a device comprises one or more processors configured to obtain a first non-zero set of coefficients from a vector representing a distinct component of a sound field, the vector being decomposed from a plurality of spherical harmonic coefficients that describe the sound field.

[0027] In another aspect, a device comprises means for obtaining a first non-zero set of coefficients from a vector representing a distinct component of a sound field, the vector being decomposed from a plurality of spherical harmonic coefficients that describe the sound field and means for storing the first non-zero set of coefficients.

[0028] In another aspect, a non-transient computer-readable storage medium has stored therein instructions which, when executed, cause one or more processors to determine a non-zero first set of coefficients from a vector representing a distinct component of a sound field, the vector being decomposed from a plurality of spherical harmonic coefficients that describe the sound field.

[0029] In another aspect, a method comprises obtaining, from a bit stream, at least one of one or more vectors decomposed from the spherical harmonic coefficients that have been recombined with spherical background harmonic coefficients, in which the coefficients spherical harmonics describe a sound field, and where the spherical background harmonic coefficients describe one or more background components of the same sound field.

[0030] In another aspect, a device comprises one or more processors configured to determine, from a stream of bits, at least one of one or more vectors decomposed from spherical harmonic coefficients that have been recombined with spherical harmonic coefficients of second plane, where spherical harmonic coefficients describe a sound field, and where background spherical harmonic coefficients describe one or more background components of the same sound field.

[0031] In another aspect, a device comprises means for obtaining, from a bit stream, at least one of one or more vectors decomposed from spherical harmonic coefficients that have been recombined with spherical background harmonic coefficients, wherein spherical harmonic coefficients describe a sound field, and where background spherical harmonic coefficients describe one or more background components of the same sound field.

[0032] In another aspect, a non-transient computer-readable storage medium has instructions stored therein which, when executed, cause one or more processors to obtain, from a stream of bits, at least one of one or more more vectors decomposed from the spherical harmonic coefficients that were recombined with spherical background harmonic coefficients, where the spherical harmonic coefficients describe a sound field, and where the spherical harmonic coefficients of the background describe one or more of the background components of the sound field. same sound field.

[0033] In another aspect, a method comprises identifying one or more distinct audio objects from one or more spherical harmonic coefficients (SHC) associated with the audio objects based on a given directionality for one or more of the audio objects.

[0034] In another aspect, a device comprises one or more processors configured to identify one or more distinct audio objects from one or more spherical harmonic coefficients (SHC) associated with the audio objects based on a directionality determined to one or more more of the audio objects.

[0035] In another aspect, a device comprises means for storing one or more spherical harmonic coefficients (SHC) and means for identifying one or more distinct audio objects from the one or more spherical harmonic coefficients (SHC) associated with the audio objects based on a given directionality to one or more of the audio objects.

[0036] In another aspect, a non-transient computer-readable storage medium has instructions stored therein which, when executed, cause one or more processors to identify one or more distinct audio objects from one or more harmonic coefficients. spheres (SHC) associated with the audio objects based on a given directionality to one or more of the audio objects.

[0037] In another aspect, a method comprises performing vector-based synthesis against a plurality of spherical harmonic coefficients to generate decomposed representations of the plurality of spherical harmonic coefficients representative of one or more audio objects and corresponding to directional information, where spherical harmonic coefficients are associated with an order and describe a sound field, determine distinct and background directional information from the directional information, reduce an order of directional information associated with background audio objects to generate directional information from background transforms, apply compensation to increase values of the transformed directional information to preserve an overall sound field energy.

[0038] In another aspect, a device comprises one or more processors configured to perform vector-based synthesis against a plurality of spherical harmonic coefficients to generate decomposed representations of the plurality of spherical harmonic coefficients representative of one or more audio objects and that correspond to directional information, where the spherical harmonic coefficients are associated with an order and describe a sound field, determine distinct and background directional information from the directional information, reduce an order of the directional information associated with the second audio objects plane to generate transformed background directional information, apply compensation to increase values of the transformed directional information to preserve an overall sound field energy.

[0039] In another aspect, a device comprises means for performing vector-based synthesis against a plurality of spherical harmonic coefficients to generate decomposed representations of the plurality of spherical harmonic coefficients representative of one or more audio objects and corresponding information directional, where spherical harmonic coefficients are associated with an order and describe a sound field, means for determining distinct directional information and background from the directional information, means for reducing an order of directional information associated with second audio objects plane to generate transformed background directional information and means to apply compensation to increase values of the transformed directional information to preserve an overall sound field energy.

[0040] In another aspect, a non-transient computer-readable storage medium has stored therein instructions which, when executed, cause one or more processors to perform vector-based synthesis with respect to a plurality of spherical harmonic coefficients to generate decomposed representations of the plurality of spherical harmonic coefficients representative of one or more audio objects and corresponding directional information, wherein the spherical harmonic coefficients are associated with an order and describe a sound field, determine distinct and background directional information from of directional information, reduce an order of the directional information associated with the background audio objects to generate transformed background directional information, and apply compensation to increase the values of the transformed directional information to preserve an overall sound field energy.

[0041] In another aspect, a method comprises obtaining interpolated spherical harmonic coefficients decomposed by a time segment, at least in part, by performing an interpolation with respect to a first decomposition of a first plurality of spherical harmonic coefficients and a second decomposition of a second plurality of spherical harmonic coefficients.

[0042] In another aspect, a device comprises one or more processors configured to obtain interpolated spherical harmonic coefficients decomposed by a time segment, at least in part, performing an interpolation with respect to a first decomposition of a first plurality of coefficients spherical harmonics and a second decomposition of a second plurality of spherical harmonics.

[0043] In another aspect, a device comprises means for storing a first of spherical harmonic coefficients and a second plurality of spherical harmonic coefficients and means for obtaining interpolated spherical harmonic coefficients decomposed by a time segment, at least in part, performing an interpolation with respect to a first decomposition of the first plurality of spherical harmonic coefficients and the second decomposition of a second plurality of spherical harmonic coefficients.

[0044] In another aspect, a non-transient computer-readable storage medium has stored therein instructions which, when executed, cause comprising one or more processors to obtain interpolated spherical harmonic coefficients decomposed by a time segment, at least in part, an interpolation being performed with respect to a first decomposition of a first plurality of spherical harmonic coefficients and a second decomposition of a second plurality of spherical harmonic coefficients.

[0045] In another aspect, a method comprises obtaining a bit stream comprising a compressed version of a spatial component of a sound field, the spatial component generated by performing vector-based synthesis against a plurality of harmonic coefficients spherical.

[0046] In another aspect, a device comprises one or more processors configured to obtain a bit stream comprising a compressed version of a spatial component of a sound field, the spatial component generated by performing vector-based synthesis against to a plurality of spherical harmonic coefficients.

[0047] In another aspect, a device comprises means for obtaining a bit stream comprising a compressed version of a spatial component of a sound field, the spatial component generated by performing vector-based synthesis against a plurality of spherical harmonic coefficients and means for storing the bit stream.

[0048] In another aspect, a non-transient computer-readable storage medium has stored therein instructions which, when executed, cause one or more processors to obtain a stream of bits comprising a compressed version of a spatial component of a sound field, the spatial component generated by performing vector-based synthesis against a plurality of spherical harmonic coefficients.

[0049] In another aspect, a method comprises generating a bit stream comprising a compressed version of a spatial component of a sound field, the spatial component generated by performing vector-based synthesis against a plurality of harmonic coefficients spherical.

[0050] In another aspect, a device comprises one or more processors configured to generate a bit stream comprising a compressed version of a spatial component of a sound field, the spatial component generated by performing vector-based synthesis against to a plurality of spherical harmonic coefficients.

[0051] In another aspect, a device comprises means for generating a bit stream comprising a compressed version of a spatial component of a sound field, the spatial component generated by performing vector-based synthesis against a plurality of spherical harmonic coefficients and means for storing the bit stream.

[0052] In another aspect, a non-transient computer-readable storage medium has instructions that, when executed, cause one or more processors to generate a bit stream comprising a compressed version of a spatial component of a sound field, the spatial component generated by performing vector-based synthesis against a plurality of spherical harmonic coefficients.

[0053] In another aspect, a method comprises identifying a Huffman codebook to use when decompressing a compressed version of a spatial component from a plurality of compressed spatial components based on an order of the compressed version of the spatial component relative to those remaining of the plurality of compressed spatial components, the spatial component generated by performing vector-based synthesis with respect to a plurality of spherical harmonic coefficients.

[0054] In another aspect, a device comprises one or more processors configured to identify a Huffman codebook for use when decompressing a compressed version of a spatial component from a plurality of compressed spatial components based on an order of the compressed version of the spatial component with respect to those remaining of the plurality of compressed spatial components, the spatial component generated by performing vector-based synthesis with respect to a plurality of spherical harmonic coefficients.

[0055] In another aspect, a device comprises means for identifying a Huffman codebook for use when decompressing a compressed version of a spatial component from a plurality of compressed spatial components based on an order of the compressed version of the spatial component with respect to to those remaining of the plurality of compressed spatial components, the spatial component generated by performing vector-based synthesis with respect to a plurality of spherical harmonic coefficients, and means for chaining the plurality of compressed spatial components.

[0056] In another aspect, a non-transient computer-readable storage medium has stored therein instructions which, when executed, cause one or more processors to identify a Huffman codebook to use when decompressing a spatial component of a plurality of spatial components based on an order of the spatial component with respect to those remaining from the plurality of spatial components, the spatial component generated by performing vector-based synthesis with respect to a plurality of spherical harmonic coefficients.

[0057] In another aspect, a method comprises identifying a Huffman codebook to use when compressing a spatial component of a plurality of spatial components based on an order of the spatial component relative to those remaining in the plurality of spatial components, the spatial components generated by performing vector-based synthesis against a plurality of spherical harmonic coefficients.

[0058] In another aspect, a device comprises one or more processors configured to identify a Huffman codebook for use when compressing a spatial component of a plurality of spatial components based on an order of the spatial component relative to those remaining in the plurality of spatial components, the spatial components generated by performing vector-based synthesis with respect to a plurality of spherical harmonic coefficients.

[0059] In another aspect, a device comprises means for storing a Huffman codebook, and means for identifying the Huffman codebook for use when compressing a spatial component of a plurality of spatial components based on an order of the spatial component with respect to those remaining of the plurality of spatial components, the spatial components generated by performing vector-based synthesis with respect to a plurality of spherical harmonic coefficients.

[0060] In another aspect, a non-transient computer-readable storage medium has stored therein instructions which, when executed, cause one or more processors to identify a Huffman codebook to use when compressing a spatial component of a plurality of spatial components based on an order of the spatial component with respect to those remaining from the plurality of spatial components, the spatial component generated by performing vector-based synthesis with respect to a plurality of spherical harmonic coefficients.

[0061] In another aspect, a method comprises determining a quantization step size to be used when compressing a spatial component of a sound field, the spatial component generated by performing vector-based synthesis against a plurality of spherical harmonic coefficients.

[0062] In another aspect, a device comprises one or more processors configured to determine a quantization step size to be used when compressing a spatial component of a sound field, the spatial component generated by performing vector-based synthesis. with respect to a plurality of spherical harmonic coefficients.

[0063] In another aspect, a device comprises means for determining a quantization step size to be used when compressing a spatial component of a sound field, the spatial component generated by performing vector-based synthesis against a plurality of spherical harmonic coefficients, and means for storing the quantization step size.

[0064] In another aspect, a non-transient computer-readable storage medium has stored therein instructions which, when executed, cause one or more processors to determine a quantization step size to be used when compressing a spatial component. of a sound field, the spatial component generated by performing vector-based synthesis against a plurality of spherical harmonic coefficients.

[0065] Details of one or more aspects of the techniques are set out in the accompanying drawings and description below. Other features, objects and advantages of these techniques will be apparent from the description and drawings, and from the claims.

BRIEF DESCRIPTION OF THE DRAWINGS

[0066] Figures 1 and 2 are diagrams that illustrate spherical harmonic basis functions of various orders and suborders.

[0067] Figure 3 is a diagram illustrating a system that can perform various aspects of the techniques described in this disclosure.

[0068] Figure 4 is a block diagram illustrating, in more detail, an example of the audio encoding device shown in the example of Figure 3 that can perform various aspects of the techniques described in this disclosure.

[0069] Figure 5 is a block diagram illustrating the audio decoding device of Figure 3 in more detail.

[0070] Figure 6 is a flowchart illustrating the exemplary operation of a content analysis unit of an audio encoding device in performing various aspects of the techniques described in this disclosure.

[0071] Figure 7 is a flowchart illustrating the exemplary operation of an audio encoding device in performing various aspects of the vector-based synthesis techniques described in this disclosure.

[0072] Figure 8 is a flowchart illustrating the exemplary operation of an audio decoding device in performing various aspects of the techniques described in this disclosure.

[0073] Figures 9A to 9L are block diagrams illustrating various aspects of the audio encoding device of the Figure 4 example in more detail.

[0074] Figures 10A to 10O(ii) are diagrams illustrating a portion of the bit stream or auxiliary channel information that may specify the compressed spatial components in more detail.

[0075] Figures 11A to 11G are block diagrams illustrating, in more detail, various units of the audio decoding device shown in the example of Figure 5.

[0076] Figure 12 is a diagram illustrating an exemplary audio ecosystem that can perform various aspects of the techniques described in this disclosure.

[0077] Figure 13 is a diagram illustrating an example of the audio ecosystem of Figure 12 in more detail.

[0078] Figure 14 is a diagram illustrating an example of the audio ecosystem of Figure 12 in more detail.

[0079] Figures 15A and 15B are diagrams that illustrate other examples of the audio ecosystem of Figure 12 in more detail.

[0080] Figure 16 is a diagram illustrating an exemplary audio encoding device that can perform various aspects of the techniques described in this disclosure.

[0081] Figure 17 is a diagram illustrating an example of the audio encoding device of Figure 16 in more detail.

[0082] Figure 18 is a diagram illustrating an exemplary audio decoding device that can perform various aspects of the techniques described in this disclosure.

[0083] Figure 19 is a diagram illustrating an example of the audio decoding device of Figure 18 in more detail.

[0084] Figures 20A through 20G are diagrams illustrating exemplary audio acquisition devices that can perform various aspects of the techniques described in this disclosure.

[0085] Figures 21A to 21G are diagrams illustrating exemplary audio reproduction devices that may perform various aspects of the techniques described in this disclosure.

[0086] Figures 22A to 22H are diagrams illustrating exemplary audio playback environments in accordance with one or more techniques described in this disclosure.

[0087] Figure 23 is a diagram illustrating an exemplary use case in which a user can experience a 3D sound field of a sports game while wearing headphones in accordance with one or more of the techniques described in this disclosure.

[0088] Figure 24 is a diagram illustrating a sports stadium in which a 3D sound field can be recorded according to one or more techniques described in this disclosure.

[0089] Figure 25 is a flow diagram illustrating a technique for rendering a 3D sound field based on a local audio landscape in accordance with one or more techniques described in this disclosure.

[0090] Figure 26 is a diagram illustrating an exemplary game studio in accordance with one or more techniques described in this disclosure.

[0091] Figure 27 is a diagram illustrating a plurality of game systems that include rendering engines in accordance with one or more techniques described in this disclosure.

[0092] Figure 28 is a diagram illustrating a speaker configuration that can be simulated through headphones in accordance with one or more of the techniques described in this disclosure.

[0093] Figure 29 is a diagram illustrating a plurality of mobile devices that can be used to acquire and/or edit a 3D sound field in accordance with one or more techniques described in this disclosure.

[0094] Figure 30 is a diagram illustrating a video frame associated with a 3D sound field that can be processed according to one or more techniques described in this disclosure.

[0095] Figures 31A to 31M are diagrams illustrating graphs showing various simulation results of performing synthetic or recorded sound field categorization in accordance with various aspects of the techniques described in this disclosure.

[0096] Figure 32 is a diagram illustrating a graph of single values from a matrix S decomposed from higher order ambisonic coefficients in accordance with the techniques described in this disclosure.

[0097] Figures 33A and 33B are diagrams illustrating respective graphs showing a potential impact that reordering has when encoding vectors describing foreground components of the sound field in accordance with the techniques described in this disclosure.

[0098] Figures 34 and 35 are conceptual diagrams illustrating differences between energy-only and directionality-based identification of distinct audio objects, in accordance with this disclosure.

[0099] Figures 36A to 36G are diagrams illustrating projections of at least a portion of the decomposed version of spherical harmonic coefficients into the spatial domain in order to perform interpolation in accordance with various aspects of the techniques described in this disclosure.

[0100] Figure 37 illustrates a representation of techniques to obtain a spatiotemporal interpolation as described in this document.

[0101] Figure 38 is a block diagram illustrating artificial U.S. matrices, USi and US2, for sequential SVD blocks for a multidimensional signal in accordance with techniques described herein.

[0102] Figure 39 is a block diagram illustrating the decomposition of subsequent frames of a higher order ambisonics (HOA) signal using Single Value Decomposition and smoothing of spatiotemporal components according to the techniques described in this revelation.

[0103] Figures 40A to 40J are each a block diagram illustrating audio encoding devices that can perform various aspects of the techniques described in this disclosure to compress spherical harmonic coefficients that describe two- or three-dimensional sound fields.

[0104] Figures 41A to 41D are block diagrams each illustrating an exemplary audio decoding device that can perform various aspects of the techniques described in this disclosure to decode spherical harmonic coefficients that describe two- or three-dimensional sound fields.

[0105] Figures 42A to 42C are each block diagrams illustrating the order reduction unit shown in the examples of Figures 40B to 40J in more detail.

[0106] Figure 43 is a diagram illustrating the compression unit V shown in Figure 401 in more detail.

[0107] Figure 44 is a diagram illustrating exemplary operations performed by the audio encoding device to compensate for quantization error in accordance with various aspects of the techniques described in this disclosure.

[0108] Figures 45A and 45B are diagrams illustrating the interpolation of subframes from portions of two frames in accordance with various aspects of the techniques described in this disclosure.

[0109] Figures 46A to 46E are diagrams illustrating a cross-section of a projection of one or more vectors of a decomposed version of a plurality of spherical harmonic coefficients that have been interpolated in accordance with the techniques described in this disclosure.

[0110] Figure 47 is a block diagram illustrating, in more detail, the extraction unit of the audio decoding devices shown in the examples of Figures 41A to 41D.

[0111] Figure 48 is a block diagram illustrating the audio rendering unit of the audio decoding device shown in the examples of Figures 41A to 41D in more detail.

[0112] Figures 49A to 49E(ii) are diagrams illustrating respective audio coding systems that can implement various aspects of the techniques described in this disclosure.

[0113] Figures 50A and 50B are block diagrams each illustrating one of two different approaches to potentially reducing the order of background content in accordance with the techniques described in this disclosure.

[0114] Figure 51 is a block diagram  illustrating examples of a distinct component compression path of an audio encoding device that can implement various aspects of the techniques described in this disclosure to compress spherical harmonic coefficients.

[0115] Figures 52 is a block diagram illustrating another example of an audio decoding device that can implement various aspects of the techniques described in this disclosure to reconstruct or substantially reconstruct spherical harmonic coefficients (SHC).

[0116] Figure 53 is a block diagram illustrating another example of an audio encoding device that can perform various aspects of the techniques described in this disclosure.

[0117] Figure 54 is a block diagram illustrating, in more detail, an exemplary implementation of the audio encoding device shown in the example of Figure 53.

[0118] Figures 55A and 55B are diagrams illustrating an example of performing various aspects of the techniques described in this disclosure to rotate a sound field.

[0119] Figure 56 is a diagram illustrating an exemplary sound field captured in accordance with a first frame of reference which is then rotated in accordance with the techniques described in this disclosure to express the sound field in terms of a second frame of reference. reference.

[0120] Figures 57A to 57E are each a diagram illustrating bit streams formed in accordance with the techniques described in this disclosure.

[0121] Figure 58 is a flowchart illustrating the exemplary operation of the audio encoding device shown in the example of Figure 53 in implementing the rotation aspects of the techniques described in this disclosure.

[0122] Figure 59 is a flowchart illustrating the exemplary operation of the audio encoding device shown in the example of Figure 53 in performing the transforming aspects of the techniques described in this disclosure.

DETAILED DESCRIPTION

[0123] The evolution of surround sound has made many output formats available for entertainment today. Examples of such consumer surround sound formats are primarily "channel" based on the fact that they implicitly specify feeds to the speakers in certain geometric coordinates. These include the popular 5.1 format (which includes the following six channels: front left (FL), front right (FR), front center or center, back left or surround left, right back or surround right, and low frequency effects ( LFE)), the growing 7.1 format, various formats that include loudspeakers such as the 7.1.4 format and the 22.2 format (e.g. for use with the Ultra High Definition Television standard). Non-consumer formats can span any number of speakers (in symmetrical and non-symmetrical geometries) often referred to as "surround arrays". An example of such an arrangement includes 32 speakers positioned in coordinates of the corners of a truncated icosahedron.

[0124] Input to a future MPEG encoder is optionally one of three possible formats: (i) traditional channel-based audio (as discussed above), which is intended to be played through speakers at pre-specified positions ; (ii) object-based audio, which involves discrete pulse code modulation (PCM) data for single audio objects with associated metadata that contain their location coordinates (among other information); and (iii) scene-based audio, which involves representing the sound field using coefficients of spherical harmonic basis functions (also called “spherical harmonic coefficients” or SHC, “higher order ambisonics” or HOA, and of HOA”). This future MPEG encoder can be described in more detail in a document entitled “Call for Proposals for 3D Audio” by the International Organization for Standardization/International Electrotechnical Commission (ISO)/(IEC) JTC1/SC29/WG11/N13411, released in January 2013 in Geneva, Switzerland, and available at http://mpeg.chiariglione.org/sites/default/files/files/stan dards/parts/docs/w13411.zip

[0125] There are several channel-based formats of "surround sound" on the market. They range, for example, from the 5.1 home theater system (which was the most successful in terms of making advances in living rooms beyond stereo) to the 22.2 system developed by NHK (Nippon Hoso Kyokai or Japan Broadcasting Corporation). Content creators (eg Hollywood studios) would like to produce the soundtrack for a movie once and without spending the effort remixing it for each speaker configuration. Recently, Standards Development Organizations have considered ways in which to provide an encoding for a standardized bitstream and subsequent decoding that is adaptive and agnostic to speaker geometry (and number) and acoustic conditions at the location of the speaker. playback (which involves a renderer).

[0126] To provide such flexibility to content creators, a hierarchical set of elements can be used to represent a sound field. The hierarchical set of elements can refer to a set of elements in which the elements are ordered in such a way that a basic set of lower-ordered elements provides a complete representation of the modeled sound field. As the set is extended to include higher-order elements, the representation becomes more detailed, increasing resolution.

[0127] An example of a hierarchical set of elements is a set of spherical harmonic coefficients (SHC). The following expression demonstrates a description or representation of a sound field using SHC:

[0128] This expression shows that the pressure pi at any point {rr, r, r} of the sound field, at time t, can be represented solely by the SHC, Amn(k). At present, k = , c the speed of sound (-343 m/s), {rr, r, r} is a reference point (or observation point), jn(•) is the spherical Bessel function of order n , and are the spherical harmonic basis functions of order n and suborder m. It can be recognized that the term in square brackets is a frequency domain representation of the signal (i.e. coefficients of multi-resolution based functions

[0129] Figure 1 is a diagram that illustrates spherical harmonic basis functions from zero order (n = 0) to fourth order (n = 4). As can be seen, for each order, there is an expression of suborders m which are shown, but not explicitly noted in the example of Figure 1 for ease of illustration purposes.

[0130] Figure 2 is another diagram that illustrates spherical harmonic basis functions from zero order (n = 0) to fourth order (n = 4). In Figure 2, the spherical harmonic basis functions are shown in three-dimensional coordinate space with both order and suborder shown.

pode ou ser fisicamente adquirido (por exemplo, gravado) através de várias configurações de arranjo de microfone ou, alternativamente, pode ser derivado de descrições com base em canal ou com base em objeto do campo sonoro. O SHC representa o áudio com base em cena, em que o SHC pode ser inserido em um codificador de áudio para obter o SHC codificado que pode promover a transmissão ou armazenamento mais eficaz. Por exemplo, uma representação de quarta ordem que envolve (1+4)2 (25, e, por isso, quarta ordem) os coeficientes podem ser usados.[0131] The SHC

it can either be physically acquired (eg, recorded) through various microphone array configurations, or alternatively it can be derived from channel-based or object-based descriptions of the sound field. SHC stands for scene-based audio, where SHC can be inserted into an audio encoder to get encoded SHC that can promote more efficient transmission or storage. For example, a fourth-order representation involving (1+4)2(25, and therefore fourth-order) coefficients can be used.

[0132] As noted above, SHC can be derived from microphone recording using a microphone. Several examples of how SHC can be derived from microphone arrays are described in Poletti, M., “Three-Dimensional Surround Sound Systems Based on Spherical Harmonics,” J. Audio Eng. Soc, Volume 53, No. 11, Nov. 2005, pages 1004 to 1025.

para o campo sonoro que corresponde a um objeto de áudio individual podem ser expressos como:

[0133] To illustrate how these SHCs can be derived from an object-based description, consider the following equation. The coefficients

for the sound field corresponding to an individual audio object can be expressed as:

é a função de Hankel esférica (do segundo tipo) da ordem n, e {rs, θs, tps} é o local do objeto. Conhecendo a energia da fonte do objeto g(ío) como uma função da frequência (por exemplo, com o uso das técnicas de análise de tempo e frequência, como realizar uma transformação rápida de Fourier no fluxo de PCM) permite converter cada objeto de PCM e seu local para

o SHC 1 . Ademais, pode ser mostrado (uma vez que o exposto acima é uma decomposição linear e ortogonal) que os

coeficientes de para cada objeto são aditivos. Dessa maneira, uma multiplicidade dos objetos de PCM podem ser

, representados pelos coeficientes de (por exemplo, como uma soma dos vetores de coeficiente para os objetos individuais). Essencialmente, esses coeficientes contêm informações sobre o campo sonoro (a pressão como uma fu não de coordenadas em 3D), e o exposto acima representa a transformação a partir de objetos individuais em uma representação do campo sonoro geral, nas redondezas do ponto de observação {rr, θr, <Pr}- As Figuras restantes são descritas abaixo no contexto da codificação de áudio com base em objeto e com base em SHC.[0134] where i is

is the spherical Hankel function (of the second type) of order n, and {rs, θs, tps} is the object location. Knowing the source energy of the object g(ío) as a function of frequency (e.g. using time and frequency analysis techniques such as performing a fast Fourier transform on the PCM stream) allows converting each PCM object and your location for

the SHC 1 . Furthermore, it can be shown (since the above is a linear and orthogonal decomposition) that the

coefficients for each object are additive. In this way, a multitude of PCM objects can be

, represented by the coefficients of (for example, as a sum of the coefficient vectors for the individual objects). Essentially, these coefficients contain information about the sound field (pressure as a fu n of 3D coordinates), and the above represents the transformation from individual objects into a representation of the overall sound field, in the vicinity of the observation point { rr, θr, <Pr}- The remaining Figures are described below in the context of object-based and SHC-based audio encoding.

[0135] Figure 3 is a diagram illustrating a system 10 that can perform various aspects of the techniques described in this disclosure. As shown in the example in Figure 3, the system 10 includes a content creator 12 and a content consumer 14. Although described in the context of the content creator 12 and the content consumer 14, the techniques can be deployed in any context where SHCs (which may be referred to as HOA coefficients) or any other hierarchical representation of a sound field are encoded to form a bit stream representative of the audio data. Furthermore, the content creator 12 may represent any form of computing device capable of implementing the techniques described in this disclosure, which includes a handset (or cell phone), a tablet computer, a smart phone or a computer of the desktop type to provide some examples. Likewise, the content consumer 14 may represent any form of computing device capable of implementing the techniques described in this disclosure, which includes a handset (or cell phone), a tablet computer, a smart phone, a set-top box or a computer. desktop type to provide some examples.

[0136] Content creator 12 may represent a film studio or other entity that can generate multi-channel audio content for consumption by content consumers, such as content consumer 14. In some instances, content creator 12 may represent an individual user who would like to compress HOA coefficients 11. This content creator will often generate audio content in conjunction with video content. Content consumer 14 represents an individual who owns or has access to an audio reproduction system, which may refer to any form of audio reproduction system capable of rendering the SHC to play as multi-channel audio content. In the example of Figure 3, the content consumer 14 includes an audio playback system 16.

[0137] Content creator 12 includes an audio editing system 18. Content creator 12 obtains live recordings 7 in various formats (including directly as HOA coefficients) and audio objects 9, which content creator 12 can edit using audio editing system 18. The content creator may, during the editing process, render HOA coefficients 11 from audio objects 9, listen to the rendered speaker feeds in an attempt to identify various aspects of the sound field that require further editing. The content creator 12 can then edit the HOA coefficients 11 (potentially indirectly by manipulating different objects from the audio objects 9 from which the source HOA coefficients can be derived in the manner described above). Content creator 12 may employ audio editing system 18 for HOA coefficients 11. Audio editing system 18 represents any system capable of editing audio data and outputting that audio data as one or more harmonic coefficients source spheres.

[0138] When the editing process is complete, content creator 12 can generate a bit stream 21 based on HOA coefficients 11. That is, content creator 12 includes an audio encoding device 20 that represents a device configured to encode or otherwise compress HOA coefficients 11 in accordance with various aspects of the techniques described in this disclosure for generating bit stream 21. Audio encoding device 20 may generate bit stream 21 for transmission, such as an example, through a transmission channel, which may be a wired or wireless channel, a data storage device, or the like. Bitstream 21 may represent an encoded version of HOA coefficients 11 and may include a primary bitstream and another auxiliary bitstream, which may be referred to as auxiliary channel information.

[0139] Although described in more detail below, the audio encoding device 20 can be configured to encode the HOA coefficients 11 based on either a vector based synthesis or a directional based synthesis. To determine whether to perform a vector-based synthesis methodology or a directional-based synthesis methodology, the audio encoding device 20 can determine, based at least in part on the HOA coefficients 11, regardless of whether the HOA coefficients 11 being generated through a natural recording of a sound field (e.g. live recording 7) or artificially produced (i.e. synthetically) from, as an example, audio objects 9, as a PCM object. When the HOA coefficients 11 have been generated from the audio objects 9, the audio encoding device 20 can encode the HOA coefficients 11 using directional based synthesis methodology. When the HOA coefficients 11 have been captured live using, for example, an eigenmike, the audio encoding device 20 can encode the HOA coefficients 11 based on vector-based synthesis methodology. The above distinction represents an example of where vector-based or directional-based synthesis methodology can be installed. There may be other cases where either or both can be useful for natural recordings, artificially generated content, or a mixture of the two (hybrid content). Furthermore, it is also possible to use both methodologies simultaneously to encode a single time frame of HOA coefficients.

[0140] Assuming that, for purposes of illustration, the audio encoding device 20 determines that the HOA coefficients 11 were captured live or otherwise represent live recordings, such as the live recording 7, the audio encoding 20 can be configured to encode the HOA coefficients 11 using a vector-based synthesis methodology that involves applying a linear invertible transform (LIT). An example of the linear invertible transform is referred to as a “single-value decomposition” (or “SVD”). In this example, the audio coding device 20 can apply the SVD to the HOA coefficients 11 to determine a decomposed version of the HOA coefficients 11. The audio coding device 20 can then analyze the decomposed version of the HOA coefficients 11 to identify various parameters, which can facilitate reordering of the decomposed version of the HOA coefficients 11. The audio encoding device 20 can then reorder the decomposed version of the HOA coefficients 11 based on the identified parameters, wherein such reordering, as described in more detail below, can improve coding efficiency as the transformation can reorder the HOA coefficients across the HOA coefficients frames (where a frame commonly includes M samples of the HOA coefficients 11 and M is, in some examples, defined in 1024). After reordering the decomposed version of the HOA coefficients 11, the audio encoding device 20 can select those of the decomposed version of the HOA coefficients 11 representative of foreground (or, in other words, distinct, predominant, or protruding). Audio encoding device 20 may specify the decomposed version of HOA coefficients 11 representative of foreground components as an audio object and associated directional information.

[0141] The audio coding device 20 may also perform a sound field analysis against the HOA 11 coefficients in order to, at least in part, identify those of the HOA 11 coefficients representative of one or more components of the sound field background (or, in other words, environment). The audio encoding device 20 can perform power compensation against the background components as, in some instances, the background components may only include a subset of any given sample of the HOA coefficients 11 (e.g., as those that correspond to zero and first-order spherical-based functions and not those that correspond to second-order or higher-order spherical-based functions). When order reduction is performed, in other words, the audio encoding device 20 can increase (e.g., add/subtract energy to/from) the background HOA coefficients remaining from the HOA coefficients 11 to compensate for the change in the general energy resulting from the realization of order reduction.

[0142] The audio encoding device 20 may then perform a form of psychoacoustic encoding (such as MPEG surround forms, MPEG-AAC, MPEG-USAC or other known forms of psychoacoustic encoding) with respect to each of the coefficients of HOA 11 representative of the background components and each of the foreground audio objects. The audio encoding device 20 can perform a form of interpolation with respect to the foreground directional information and then can perform an order reduction with respect to the interpolated foreground directional information to generate reduced order foreground directional information. . The audio encoding device 20 may additionally perform, in some examples, a quantization with respect to low-order foreground directional information, which outputs encoded foreground directional information. On some occasions, this quantization may comprise a scalar/entropy quantization. Audio encoding device 20 can then form bit stream 21 to include encoded background components, encoded foreground audio objects, and quantized directional information. Audio encoding device 20 can then transmit or otherwise send bitstream 21 to content consumer 14.

[0143] Although shown in Figure 3 as being directly transmitted to the content consumer 14, the content creator 12 can output the bit stream 21 to an intermediate device positioned between the content creator 12 and the content consumer 14. This intermediary device can store the bit stream 21 for later delivery to the content consumer 14, which can request that bit stream. The intermediary device may comprise a file server, a network server, a desktop-type computer, a laptop-type computer, a tablet-type computer, a mobile phone, a smart phone, or any other device capable of storing the stream. 21 bits for later retrieval via an audio decoder. This intermediary device may reside on a content delivery network capable of continuously transmitting the 21 bit stream (and possibly in conjunction with the transmission of a corresponding video data bit stream) to subscribers such as the consumer of content 14, which requests bitstream 21.

[0144] Alternatively, the content creator 12 may store the bit stream 21 on a storage medium such as a compact disc, digital video disc, high definition video disc or other storage media, most of which have ability to be read by a computer and therefore may be referred to as computer-readable storage media or non-transient computer-readable storage media. In this context, broadcast channel may refer to those channels through which content stored on these media is transmitted (and may include retail stores and other store-based delivery mechanisms). In any event, the techniques of this disclosure should therefore not be limited in this regard to the example in Figure 3.

[0145] As further shown in the example of Figure 3, the content consumer 14 includes the audio reproduction system 16. The audio reproduction system 16 can represent any audio reproduction system capable of playing multi-channel audio data. . The audio reproduction system 16 may include a number of different Competitors 22. The Competitors 22 may each provide a different form of rendering, where the different forms of rendering may include one or more of several modes of performing stepwise variation of vector-based amplitude (VBAP) and/or one or more of several ways to perform sound field synthesis. As used herein, "A and/or B" means "A or B" or both "A and B".

[0146] Audio reproduction system 16 may additionally include an audio decoding device 24. Audio decoding device 24 may represent a device configured to decode HOA coefficients 11' from bit stream 21 into that HOA 11' coefficients may be similar to HOA 11 coefficients, but differ due to loss operations (eg, quantization) and/or transmission through the transmission channel. That is, the audio decoding device 24 can dequantize the foreground directional information specified in the bitstream 21, while also performing psychoacoustic decoding with respect to the foreground audio objects specified in the bitstream 21 and the frequency coefficients. HOA encoded representative of background components. Audio decoding device 24 may additionally perform interpolation against the decoded foreground directional information and then determine representative HOA coefficients of foreground components based on the decoded foreground audio objects and directional information. interpolated foreground. The audio decoding device 24 can then determine the HOA coefficients 11' based on the determined HOA coefficients representative of the foreground components and the decoded HOA coefficients representative of the background components.

[0147] Audio reproduction system 16 may, after decoding bit stream 21, obtain HOA coefficients 11' and render HOA coefficients 11' to output speaker feeds 25. Speaker 25 can drive one or more speakers (which are not shown in the example of Figure 3 for ease of illustration purposes).

[0148] In order to select the proper renderer or, on some occasions, generate a suitable renderer, the audio playback system 16 may obtain speaker information 13 indicative of a number of speakers and/or a spatial geometry of the loudspeakers. -speakers. On some occasions, the audio reproduction system 16 may obtain information from speaker 13 using a reference microphone and which drives the speakers in such a way as to dynamically determine information from speaker 13. In At other times or in conjunction with dynamically determining speaker information 13, audio reproduction system 16 may prompt a user to interface with audio reproduction system 16 and input speaker information 16.

[0149] Audio playback system 16 may then select one of Audio Competitors 22 based on speaker information 13. On some occasions, audio playback system 16 may, when none of the Audio Competitors 22 is within some limiting similarity measure (in the sense of speaker geometry) to that specified in speaker information 13, generate the one among audio competitors 22 based on speaker information 13. The system 16 may, on some occasions, generate the one of Audio Competitors 22 based on speaker information 13 without first trying to select an existing Competitor from among the Audio Competitors 22.

[0150] Figure 4 is a block diagram illustrating, in more detail, an example of the audio encoding device 20 shown in the example of Figure 3 that can perform various aspects of the techniques described in this disclosure. The audio encoding device 20 includes a content analysis unit 26, a vector-based synthesis methodology unit 27 and a directional-based synthesis methodology unit 28.

[0151] Content analysis unit 26 represents a unit configured to analyze the content of HOA coefficients 11 to identify whether HOA coefficients 11 represents content generated from a live recording or an audio object. The content analysis unit 26 can determine whether the HOA coefficients 11 were generated from a recording of an actual sound field or an artificial audio object. The content analysis unit 26 can make this determination in several ways. For example, the content analysis unit 26 can encode (N+1)2-1 the channels and predict the last remaining channel (which can be represented as a vector). The content analysis unit 26 can apply scalars to at least some of the (N+1)2-1 channels and sum the resulting values to determine the last remaining channel. Furthermore, in this example, the content analysis unit 26 can determine an accuracy of the predicted channel. In this example, if the precision of the predicted channel is relatively high (for example, the precision exceeds a specific threshold), the HOA 11 coefficients are likely generated from a synthetic audio object. In contrast, if the predicted channel accuracy is relatively low (eg, the accuracy is below the specific threshold), the HOA 11 coefficients are more likely to represent a recorded sound field. For example, in this example, if a signal-to-noise ratio (SNR) of the predicted channel is above 100 decibels (dbs), the HOA 11 coefficients are more likely to represent a sound field generated from a synthetic audio object. . On the other hand, the SNR of a sound field recorded using its own microphone can be from 5 to 20 dBs. Thus, there may be a clear demarcation in SNR ratios between the sound field represented by the HOA 11 coefficients generated from a real direct recording and a synthetic audio object.

[0152] More specifically, the content analysis unit 26 can, when determining whether the HOA coefficients 11 representative of a sound field are generated from a synthetic audio object, obtain coefficients with a frame of HOA coefficients, which can be 25 by 1024 in size for a fourth-order representation (ie N = 4). After obtaining the framed HOA coefficients (which may be denoted herein as a framed SHC matrix 11 and subsequent framed SHC matrices may be denoted as framed SHC matrices 27B, 27C, etc.). The content analysis unit 26 can then exclude the first vector of framed HOA coefficients 11 to generate reduced frame HOA coefficients. In some examples, this first vector excluded from the HOA coefficients with frame 11 may correspond to those of the HOA coefficients 11 associated with the zero-order, zero-order spherical harmonic basis function.

[0153] The content analysis unit 26 can then predict the first non-zero vector of the framed reduced HOA coefficients from the remaining vectors of the framed reduced HOA coefficients. The first non-zero vector can refer to a first vector that goes from the first order (and that considers each of the order-dependent suborders) to the fourth order (and that considers each of the order-dependent suborders) that has values in addition to zero. In some examples, the first non-zero vector of the frame-reduced HOA coefficients refers to those of the 11 HOA coefficients associated with the zero-order, first-order spherical harmonic basis function. Although described with respect to the first non-zero vector, the techniques can predict other vectors of the frame-reduced HOA coefficients from the remaining vectors of the frame-reduced HOA coefficients. For example, the content analysis unit 26 can predict those of frame-reduced HOA coefficients associated with a first-order, first-order, spherical harmonic basis function or a first-order, first-order, spherical harmonic basis function. negative. As yet other examples, the content analysis unit 26 can predict those of the frame-reduced HOA coefficients associated with a zero-order second-order spherical harmonic basis function.

[0154] To predict the first non-zero vector, the content analysis unit 26 can operate according to the following equation:

, que é 23 para uma representação de quarta ordem, cxi denota alguma constante para o z’-ésimo vetor, e V1 se refere ao z’-ésimo vetor. Após predizer o primeiro vetor diferente de zero, a unidade de análise de conteúdo 26 pode obter um erro com base no primeiro vetor diferente de zero predito e no vetor diferente de zero real. Em alguns exemplos, a unidade de análise de conteúdo 26 subtrai o primeiro vetor diferente de zero predito do primeiro vetor diferente de zero real para derivar o erro. A unidade de análise de conteúdo 26 pode computar o erro como uma soma do valor absoluto das diferenças entre cada entrada no primeiro vetor diferente de zero predito e o primeiro vetor diferente de zero real.[0155] where z is from 1 to

, which is 23 for a fourth-order representation, cxi denotes some constant for the z'-th vector, and V1 refers to the z'-th vector. After predicting the first nonzero vector, the content analysis unit 26 may obtain an error based on the predicted first nonzero vector and the actual nonzero vector. In some examples, the content analysis unit 26 subtracts the first predicted nonzero vector from the first actual nonzero vector to derive the error. The content analysis unit 26 may compute the error as a sum of the absolute value of the differences between each entry in the first predicted nonzero vector and the first actual nonzero vector.

[0156] Once the error is obtained, the content analysis unit 26 can compute a ratio based on an actual non-zero first vector energy and the error. The content analysis unit 26 can determine this energy by squaring each input of the first non-zero vector by adding the squared inputs together. The content analysis unit 26 can then compare this ratio to a threshold. When the ratio does not exceed the threshold, the content analysis unit 26 can determine that the HOA coefficients with frame 11 are generated from a recording and indicate in the bit stream that the corresponding encoded representation of the HOA coefficients 11 has been generated. from a recording. When the ratio exceeds the threshold, the content analysis unit 26 can determine that the HOA coefficients with frame 11 are generated from a synthetic audio object and indicate in the bit stream that the corresponding encoded representation of the HOA coefficients 11 with frame was generated from a synthetic audio object.

[0157] The indication of whether the HOA coefficients with frame 11 were generated from a recording or a synthetic audio object may comprise a single bit for each frame. The single bit can indicate that different encodings were used for each frame effectively switching between different ways in which to encode the corresponding frames. On some occasions, when HOA coefficients with frame 11 were generated from a recording, the content analysis unit 26 passes the HOA coefficients 11 to the vector-based synthesis unit 27. On some occasions, when the HOA coefficients with frame 11 were generated from a synthetic audio object, the content analysis unit 26 passes the HOA coefficients 11 to the directional-based synthesis unit 28. The directional-based synthesis unit 28 can represent a unit configured to perform a directional based synthesis of HOA coefficients 11 to generate a directional based 21 bit stream.

[0158] In other words, the techniques are based on encoding the HOA coefficients with or using a front-end classifier. The classifier can work as follows:

[0159] Start with a framed SH matrix (say 4th order, 1024 frame size, which can also be referred to as framed HOA coefficients or HOA coefficients) - where a 25 x 1024 size matrix is obtained.

[0160] Delete the 1st vector (zero order SH) - so there is a matrix of size 24 x 1024.

[0161] Predict the first non-zero vector in the matrix (a vector of size 1 x 1024) - from the rest of the vectors in the matrix (23 vectors of size 1 x 1024).

[0162] The prediction is as follows: predicted vector = sum-over-i [alpha-i x vector-I] (where the sum over I is done in 23 indices, i=l...23)

[0163] Then check the error: actual vector - predicted vector = error.

[0164] If the vector energy/error ratio is large (ie Error is smaller), then the underlying sound field (in that frame) is sparse/synthetic. Or, the underlying sound field is a recorded sound field (using, say, a microphone array).

[0165] Depending on the recording versus synthetic decision, perform encoding/decoding (which may refer to bandwidth compression) in different ways. The decision is a 1-bit decision, which is sent through the bit stream for each frame.

[0166] As shown in the example of Figure 4, the vector-based synthesis unit 27 may include a linear invertible transform (LIT) unit 30, a parameter calculation unit 32, a reorder unit 34, a foreground selection 36, a power compensation unit 38, a psychoacoustic audio encoder unit 40, a bit stream generation unit 42, a sound field analysis unit 44, a coefficient reduction unit 46, a background selection unit (BG) 48, a spatiotemporal interpolation unit 50 and a quantization unit 52.

[0167] The linear invertible transform unit (LIT) 30 receives the HOA coefficients 11 in the form of HOA channels, each channel is representative of a block or frame of a coefficient associated with a given order, suborder of spherical base functions ( which can be denoted as HOA[k], where k can denote the current frame or sample block). The HOA coefficient matrix 11 can have dimensions D: Mx (N+1)2.

[0168] That is, the unit of LIT 30 may represent a unit configured to perform a form of analysis referred to as single-value decomposition. While described with respect to SVD, the techniques described in this disclosure can be performed with respect to any similar transformation or decomposition that provides linearly uncorrected energy compressed output sets. Also, the reference to "sets" in this disclosure is generally intended to refer to non-zero sets unless specifically stated otherwise and is not intended to refer to the classical mathematical definition of sets that include the so-called "empty set". .

[0169] An alternative transformation may comprise a principal component analysis, which is often referred to as “PCA.” PCA refers to a mathematical procedure that employs an orthogonal transformation to convert a set of observations of possibly correlated variables into a set of linearly uncorrected variables referred to as principal components. Linearly uncorrected variables represent variables that have no statistical relationship (or dependence) on each other. These main components can be described as having a small degree of statistical correctness with each other. In any event, the number of named principal components is less than or equal to the number of original variables. In some examples, the transformation is defined such that the first principal component has as much variance as possible (or, in other words, accounts for as much variability in the data as possible), and each succeeding component, in turn, has the highest possible variance under the constraint that this successive component is orthogonal to the preceding components (which can be reset as uncorrected with them). PCA can perform a form of order reduction, which in terms of HOA 11 coefficients can result in compression of HOA 11 coefficients. Depending on the context, PCA can be referred to by a number of different names, such as discrete Karhunen-Loeve transform , the Hotellmg transform, appropriate orthogonal decomposition (POD) and eigenvalue decomposition (EVD) to name a few examples. Properties of such operations that are conducive to the underlying purpose of audio data compression are "power compression" and "decorrelation" of multichannel audio data.

[0170] In any event, the LIT 30 unit performs a single-value decomposition (which, again, may be referred to as “SVD”) to transform the HOA 11 coefficients into two or more sets of transformed HOA coefficients. These “sets” of transformed HOA coefficients may include vectors of transformed HOA coefficients. In the example of Figure 4, the unit of LIT 30 can perform the SVD against the coefficients of HOA 11 to generate a so-called matrix V, a matrix S and a matrix U. SVD, in linear algebra, can represent a factorization of a matrix y-by-z real or complex X (where X can represent multichannel audio data, such as HOA coefficients 11) in the following form:

[0171] U can represent a real or complex y-by-y unitary matrix, where the y columns of U are commonly known as the unique left vectors of multichannel audio data. S can represent a rectangular diagonal y-by-z matrix with non-negative real numbers on the diagonal, where the diagonal values of S are commonly known as the unique values of multichannel audio data. V* (which can denote a conjugate transpose of V) can represent a real or complex z-by-z unitary matrix, where the z columns of V* are commonly known as the unique right vectors of multichannel audio data.

[0172] Although described in this disclosure as being applied to multichannel audio data comprising HOA coefficients 11, the techniques can be applied to any form of multichannel audio data. Thereby, the audio coding device 20 can perform a single-value decomposition against the multi-channel audio data representative of at least a portion of the sound field to generate a matrix U representative of left-unique vectors of the multi-channel audio data. , a matrix S representative of unique values of the multichannel audio data and a matrix V representative of unique right vectors of the multichannel audio data, and representing the multichannel audio data as a function of at least a portion of one or more among the U matrix, the S matrix and the V matrix.

[0173] In some examples, the matrix V* in the mathematical expression referred to above is denoted as the conjugate transpose of the matrix V to reflect that SVD can be applied to matrices comprising complex numbers. When applied to matrices that comprise only real numbers, the complex conjugate of the matrix V (or, in other words, the matrix V*) can be considered to be the transpose of the matrix V. Below, it is assumed that, for ease of purposes for illustration, the HOA coefficients 11 comprise real numbers with the result that matrix V is output via SVD instead of matrix V*. Furthermore, although denoted as the matrix V in this disclosure, the reference to the matrix V should be understood to refer to the transpose of the matrix V where appropriate. Although it is assumed to be the V matrix, the techniques can be applied in a similar way to HOA 11 coefficients that have complex coefficients, where the output of the SVD is the V* matrix. In this way, the techniques should not be limited in this sense to just provide the application of SVD to generate a matrix V, but may include the application of SVD on the HOA 11 coefficients that have complex components to generate a matrix V*.

[0174] At any time the LIT unit 30 may perform a SVD block shape with respect to each block (which may refer to a frame) of higher order ambisonics (HOA) audio data (wherein this data ambisonics audio include blocks or samples of HOA 11 coefficients or any other form of multichannel audio data). As noted above, a variable M can be used to denote the length of an audio frame in samples. For example, when an audio frame includes 1024 audio samples, M equals 1024. Although described in relation to this typical value for M, the techniques of this disclosure should not be limited to this typical value for M. The LIT 30 unit can therefore perform a block SVD against a block of HOA 11 coefficients. which have HOA coefficients of M-per-(N+1)2, where N, again, denotes the order of the HOA audio data. The LIT unit 30 can generate, by performing such SVD, a matrix V, a matrix S and a matrix U, wherein each of the matrices can represent the respective matrices V, S and U described above. Thereby, the linear invertible transform unit 30 can perform the SVD against the HOA coefficients 11 to output XJS[k] vectors 33 (which can represent a combined version of the S vectors and the U vectors) that have dimensions D:M x (N+1)2, and vectors V[k] 35 that have dimensions D: (N+1)2 x (N+1)2. Individual vector elements in the matrix U.S.[k] may also be designated XPS(k) while the individual vectors in the matrix V[k] may also be designated v(k).

[0175] An analysis of the U, S and V matrices can reveal that these matrices carry or represent spatial and temporal characteristics of the underlying sound field represented above by X. Each of the vectors N in U (of sample length M) can represent signals separate audio streams normalized as a function of time (for the time period represented by M samples), which are orthogonal to each other and which have been decoupled from any spatial features (which may also be referred to as directional information). The spatial features, which represent the spatial shape and position width (r, theta, phi) can instead be represented by individual vectors ith, v(i) (k), in the matrix V (each of the length (N+ 1)2). Both the vectors in the matrix U and in the matrix V are normalized in such a way that their energies with the square of the root mean are equal to unity. The energy of the audio signals in U are then represented by the diagonal elements in S. Multiplying U and S to form XPS[k] (with individual vector elements XPS(k)), then represents the signal of audio with true energies. The ability of the SVD decomposition to decouple signals by audio time (U), their energies (S), and their spatial characteristics (V) can support many aspects of the techniques described in this disclosure. Furthermore, this model of synthesizing the underlying HOA[k] coefficients, X, by a vector multiplication of XJS[k] and V[k] gives rise to the term “vector-based synthesis methodology”, which is used throughout this document.

[0176] Although described as being performed directly with respect to the HOA 11 coefficients, the LIT 30 unit can apply the linear invertible transform to derivatives of the HOA 11 coefficients. For example, the LIT 30 unit can apply the SVD with respect to to a power spectral density matrix derived from the HOA coefficients 11. The power spectral density matrix may be denoted as PSD and obtained by matrix multiplication of the transpose from the hoaFrame to the hoaFrame, as outlined in the pseudocode below. The hoaFrame notation refers to a frame of HOA 11 coefficients.

[0177] The LIT 30 unit can, after applying the SVD (svd) to the PSD, obtain a matrix S[k]2 (S square) and a matrix V[k]. The matrix S[k]2 can denote a square matrix S[k], where the unit of LIT 30 can apply a square root operation to the matrix S[k]2 to obtain the matrix S[k]. The LIT unit 30 can, on some occasions, quantize against the matrix V[k] to obtain a quantized matrix V[k] (which can be denoted as matrix V[k]'). The LIT unit 30 can obtain the matrix U[k] by first multiplying the matrix S[k] by the quantized matrix V[k]' to obtain a matrix SV[k]'. The unit of LIT 30 can then take the pseudoinverse (pinv) of the matrix SV[k]' and then multiply the coefficients of HOA 11 by the pseudoinverse of the matrix SV[k]' to obtain the matrix U[k] . What was exposed earlier can be represented by the following pseudocode:

[0178] By performing SVD against the power spectral density (PSD) of the HOA coefficients rather than the coefficients themselves, the LIT 30 unit can potentially reduce the computational complexity of performing SVD in terms of one or more of processor cycles and storage space, while achieving the same source audio encoding efficiency as if SVD were applied directly to the HOA coefficients. That is, the PSD type SVD described above can potentially be less computationally demanding because the SVD is done in an F*F matrix (with F being the number of HOA coefficients). Compared to an M * F matrix with M being the frame length, i.e. 1024 or more samples. The complexity of an SVD can now, by applying to PSD instead of HOA 11 coefficients, be about O(LA3) compared to O(M*LA2) when applied to HOA 11 coefficients (where O(* ) denotes the notation of big-O computational complexity notation common to computer science techniques).

[0179] Parameter calculation unit 32 represents the unit configured to calculate various parameters such as a correlation parameter (R), directional property parameters (θ, cp, r) and an energy property (e). Each of these parameters for the current frame can be denoted as R[k], θ[k], cp [k], e[k] and e[k]. The parameter calculation unit 32 can perform an energy analysis and/or correlation (or so-called cross-correlation) against the U.S.[k] vectors 33 to identify these parameters. The parameter calculation unit 32 can also determine these parameters for the previous frame, where the previous frame parameters can be denoted R[k-1], θ[kl], <p[kl], r[kl] and ee [k-l], based on the previous table of the US[k-1] vector and V[k-1] vectors. The parameter calculation unit 32 can output the current parameters 37 and the previous parameters 39 to the reordering unit 34.

[0180] That is, the parameter calculation unit 32 can perform an energy analysis against each of the first vectors US[k] 33 that correspond to a first time and each of the second vectors US[k] 33 which correspond to a second time, which computes a root-mean squared energy for at least a portion of (but often all) the first audio frame and a portion of (but often all) the second audio frame, and hence mode, it can generate 2L energies, one for each of the first L vectors US[k] 33 of the first audio frame and one for each of the second vectors US[k -1] 33 of the second audio frame.

[0181] In other examples, the parameter calculation unit 32 may perform a cross-correlation between some portion of the sample set (if not all the same) for each of the first vectors XJS[k] 33 and each of the second vectors US[k-1] 33. Cross-correlation can refer to cross-correlation as understood in signal processing techniques. In other words, cross-correlation can refer to a measure of similarity between two waveforms (which in this case is defined as a discrete set of M samples) as a function of a delay applied to one of them. In some examples, to perform cross-correlation, the calculation unit of parameter 32 compares the last sample L of each of the first vectors XJS[k] 27, in terms of time, to the first samples L of each of the remaining of the second US[k-1] 33 vectors to determine a correlation parameter. As used herein, an operation "in terms of turn" refers to an element through element operation performed on a first set of elements and a second set of elements, where the operation extracts one element from each between the first and second sets of elements “at a time” according to an ordering of the sets.

[0182] Parameter calculation unit 32 can also analyze vectors V[k] and/or V[k-1] 35 to determine directional property parameters. These directional property parameters can provide an indication of motion and location of the audio object represented by the corresponding XJS[k] and/or U.S.[k-1] 33 vectors. The parameter calculation unit 32 can provide any combination of current foreground parameters 37 (determined with respect to XJS[k] 33 vectors and/or V[k] 35 vectors) and any combination of the above parameters 39 (determined in relative to US[k-1] 33 vectors and/or V[k-1] 35 vectors) for the reordering unit 34.

o vetor - mesmo sinal de áudio/objeto (avançado no tempo) representado pelo p-ésimo vetor nos vetores XJS[k] 33, que também podem ser denotados como vetores U.S.[k][p] 33 (ou, XΦJffr'j’j alternativamente como

Os parâmetros calculados pela unidade de cálculo de parâmetro 32 podem ser usados pela unidade de reordenação 34 para reordenar os objetos de áudio para representar sua avaliação natural ou continuidade com o tempo.[0183] SVD decomposition does not guarantee that the audio signal/object represented by the p-th vector into US[k-1] 33 vectors, which can be denoted as

the vector - same audio signal/object (time-advanced) represented by the pth vector in the vectors XJS[k] 33, which can also be denoted as vectors US[k][p] 33 (or, XΦJffr'j' j alternatively as

The parameters calculated by the parameter calculation unit 32 can be used by the reordering unit 34 to reorder the audio objects to represent their natural evaluation or continuity with time.

[0184] That is, the reordering unit 34 can then compare each of the parameters 37 from the first vectors US[k] 33 in terms of turn against each of the parameters 39 to the second vectors US[k-1] 33 The reordering unit 34 can reorder (using, as an example, a Hungarian algorithm) the various vectors within the matrix US[k] 33 and the matrix V[k] 35 based on the current parameters 37 and the parameters 39 to output a reordered matrix US[k] 33' (which can be mathematically denoted as US[k]) and a reordered matrix V[k] 35' (which can be mathematically denoted as V[k]) for a unit foreground sound selection (or predominant sound - PS) 36 (“foreground selection unit 36”) and a power compensation unit 38.

(em que, novamente, cada vetor dos vetores U.S.[k] 33, que novamente podem ser X r/ri alternativamente denotados como , pode representar um ou mais objetos distintos (ou, em outras palavras, predominantes) de monoáudio presentes no campo sonoro) pode variar a partir de porções dos dados de áudio. Ou seja, dado que o dispositivo de codificação de áudio 12, em alguns exemplos, opera nessas porções dos dados de áudio geralmente referidos como quadros de áudio, a posição de vetores que correspondem a esses objetos de monoáudio distintos conforme representados na matriz U.S.[k] 33 como derivados, pode variar de quadro de áudio para quadro de áudio devido à aplicação de SVD aos quadros e a saliência variante de cada objeto de áudio de quadro para quadro.[0185] In other words, the reordering unit 34 can represent a unit configured to reorder the vectors within the matrix US[k] 33 to generate the reordered matrix US[k] 33'. The reordering unit 34 can sort the matrix XJS[k] 33 because of the order of the vectors

(where, again, each vector of US[k] 33 vectors, which again can be X r/ri alternatively denoted as , can represent one or more distinct (or, in other words, predominant) monoaudio objects present in the sound field ) may vary from portions of the audio data. That is, given that the audio encoding device 12, in some examples, operates on those portions of the audio data generally referred to as audio frames, the position of vectors that correspond to these distinct monoaudio objects as represented in the matrix US[k ] 33 as derived, may vary from audio frame to audio frame due to the application of SVD to the frames and the varying salience of each audio object from frame to frame.

[0186] Passing the vectors within the XJ matrix S[k] 33 directly to the psychoacoustic audio encoding unit 40 without reordering the vectors within the matrix US[k] 33 from audio frame to audio frame can reduce the extent of compression achievable by some compression schemes, such as inheritance compression schemes that perform best when monoaudio objects are continuous (in terms of channel, which is defined in this example by the position order of vectors within the related US[k] 33 matrix each other) through audio frames. Furthermore, when not reordered, encoding the vectors within the U.S.[k] 33 matrix can reduce the quality of the audio data when decoded. For example, the AAC encoders, which can be represented in the example of Figure 3 by the psychoacoustic audio encoder unit 40, can more efficiently compress one or more reordered vectors within the US[k] matrix 33' from frame to frame compared to compression achieved when directly encoding the vectors into the US[k] 33 matrix frame by frame. While described above in relation to AAC encoders, the techniques can be performed in relation to any encoder that provides better compression when monoaudio objects are specified through frames in a specific order or position (in terms of channel).

[0187] Various aspects of the techniques may thereby enable the audio encoding device 12 to reorder one or more vectors (e.g. the vectors within the US[k] matrix 33 to generate one or more reordered vectors within the reordered matrix US[k] 33' and thereby facilitate compression of the vectors within the matrix US[k] 33 by a legacy audio encoder, such as the psychoacoustic audio encoder unit 40).

[0188] For example, the reordering unit 34 may reorder one or more vectors within the matrix US[k] 33 of a first audio frame subsequent in time to the second frame at which one or more second vectors within the matrix US[k] -1] 33 corresponds based on current parameters 37 and previous parameters 39. Although described in the context of a first audio frame that is subsequent in time to the second audio frame, the first audio frame may precede the second in time audio frame. Thus, the techniques should not be limited to the example described in this disclosure.

[0189] For the sake of illustration, consider Table 1 below in which each of the vectors p within the matrix US[k] 33 is denoted as US [k][p], where k denotes whether the corresponding vector is from the -th or previous frame (k-1) and p denotes the vector row with respect to the vectors of the same audio frame (where the US[k] matrix has (N+1)2 as vectors). As noted above, assuming that N is determined to be one, p can denote vectors one (1) to (4). TABLE 1

[0190] In Table 1 above, the reordering unit 34 compares the computed energy for US[£-1][1] to the computed energy for each of US[k][1], US[k][2], US[k][3], US[k][4], the energy computed for US[k-1][2] the energy computed for each of US[k][1], US[k][2 ], US[k][3], US[k][4], etc. The reordering unit 34 may then discard one or more of the second vectors US[k-1] 33 of the preceding second audio frame (in terms of time). For the sake of illustration, consider Table 2 below which shows the remaining second vectors US[k-1] 33: TABLE 2

[0191] In Table 2 above, the reordering unit 34 can determine, based on the energy comparison, that the energy computed for US[k-1][1] is similar to the energy computed for each of US[k][ 1] and US[k][2], the energy computed for US[k-1][2] is similar to the energy computed for each of US[k][1] and US[k][2], the energy computed for XJS[k-1][3] is similar to the energy computed for each of US[k][3] and US[k][4], and the energy computed for US[k-1][4 ] is similar to the computed energy for each of US[k][3] and US[k][4]. In some examples, the reordering unit 34 may perform additional energy analysis to identify a similarity between each of the first vectors of the matrix U.S.[k] 33 and each of the second vectors of the matrix U.S.[k-1] 33.

[0192] In other examples, the reordering unit 32 can reorder vectors based on current parameters 37 and previous parameters 39 that refer to cross correlation. In these examples, with reference again to Table 2 above, the reordering unit 34 can determine the following exemplary correlation expressed in Table 3 based on these cross-correlation parameters: TABLE 3

[0193] From Table 3 above, the reordering unit 34 determines, as an example, that the US[k-1][1] vector correlates to the differently positioned US[k][2] vector, the XJS vector [k-1][2] correlates to the differently positioned US[k][1] vector, the US[k-1][3] vector correlates to the similarly positioned US[k][3] vector, and the US vector [k-1][4] correlates to the similarly positioned US[k][4] vector. In other words, the reordering unit 34 determines what can be referred to as reordered information that describes how to reorder the first vectors of the matrix XJ S[k] 33 such that the vector US[k][2] is repositioned in the first row of the first vectors of the matrix US[k] 33 and the vector US[k][1] is repositioned in the second row of the first vectors XJS[k] 33. The reordering unit 34 can then reorder the first vectors of the matrix US[k] 33 based on this reorder information to generate the reordered US[k] 33' array.

[0194] Additionally, the reordering unit 34 may, although not shown in the example of Figure 4, provide this reordering information to the bitstream generating device 42, which can generate the bitstream 21 to include this reordering information. reordering so that the audio decoding device, such as the audio decoding device 24 shown in the example of Figures 3 and 5, can determine how to reorder the reordered vectors of the US[k] matrix 33' in order to recover the vectors of the matrix US[k] 33.

[0195] Although described above as performing a two-step process that involves an analysis based first on energy-specific parameters and then on cross-correlation parameters, the reordering unit 32 can only perform this analysis only in relation to to the energy parameters to determine the reorder information, perform this analysis only against the cross-correlation parameters to determine the reorder information, or perform the analysis against both the energy parameters and the cross-correlation parameters in the manner described above . Additionally, the techniques may employ other types of processes to determine correlation that do not involve performing one or both of an energy comparison and/or a cross-correlation. Thus, the techniques should not be limited in this sense to the examples set out above. Furthermore, other parameters obtained from the calculation unit of parameter 32 (such as spatial position parameters derived from vectors V or correlation of vectors in V[k] and V[k-1]) can also be used (or concurrently/jointly or sequentially) with the energy and cross-correlation parameters obtained from US[k] and US[kl] to determine the correct ordering of the vectors in US.

[0196] As an example of using the correlation of vectors in matrix V, the calculation unit of parameter 34 can determine that vectors of matrix Y[k] 35 are correlated as specified in Table 4 below: TABLE 4

[0197] From Table 4 above, the reordering unit 34 determines, as an example, that the vector V[k-1][1] correlates to the differently positioned vector V[k][2], the vector V [k-1][2] correlates to the differently positioned vector V[k][1], the vector V[k-1][3] correlates to the similarly positioned vector V[k][3] and the vector V [&-1][4] correlates to similarly positioned vector V[&][4]. The reordering unit 34 can output the reordered version of the vectors of the matrix Y[k] 35 as a reordered matrix Y[k] 35'.

[0198] In some examples, the same reordering that is applied to the vectors in the US matrix is also applied to the vectors in the V matrix. In other words, any analysis used in the reordering of the V vectors can be used in conjunction with any analysis used to reorder the US vectors. To illustrate an example where the reordering information is not only determined with respect to energy parameters and/or cross-correlation parameters with respect to US[&] vectors 35, the reordering unit 34 can also perform this analysis against Y[k] 35 vectors based on cross-correlation parameters and energy parameters in a similar manner as described above with respect to Y[k] 35 vectors. Furthermore, although XJS[k] ] 33 do not have any directional properties, the Y[k] 35 vectors can provide information related to the directionality of the corresponding XJS[k] 33 vectors. In this sense, the reordering unit 34 can identify correlations between vectors V[k] 35 and vectors V[k-1] 35 based on an analysis of corresponding directional property parameters. That is, in some examples, the audio object moves within a sound field in a continuous manner when in motion or when it is in a relatively stable location. As such, the reordering unit 34 can identify those vectors from the matrix V[k] 35 and the matrix V[k-1] 35 that exhibit some known physically realistic motion or that are stationary within the sound field as related by reordering the vectors. XJS[k] 33 and vectors V[k] 35 based on the correlation of directional properties. In any event, the reordering unit 34 may output the reordered vectors U.S.[k] 33' and the reordered vectors V[k] 35' to the foreground selection unit 36.

[0199] Additionally, the techniques may employ other types of processes to determine the correct order that does not involve performing one or both of an energy comparison and/or a cross-correlation. Thus, the techniques should not be limited in this sense to the examples set out above.

[0200] Although described above as the reordering of matrix V vectors to mirror the reordering of US matrix vectors, on certain occasions V vectors may be reordered differently from US vectors where separate syntax elements can be generated to indicate the reordering of the US vectors and the reordering of the V vectors. On some occasions, the V vectors may not be reordered and only the US vectors may be reordered since the V vectors may not be psychoacoustically encoded.

[0201] A modality in which the reordering of the V matrix vectors and the US matrix vectors is different when the intention is to swap the audio objects in space - that is, move them away from the original recorded position (when the field the underlying sound field is a natural recording) or the artistically intended position (when the underlying sound field is an artificial mixture of objects). As an example, suppose there are two audio sources A and B, A might be the sound of a cat “meow” emanating from the “left” part of the sound field, and B might be the sound of a dog “au- au" emanating from the “right” part of the sound field. When the reordering of the V and US are different, the position of the two sound sources is switched. After the switch, A (the “meow”) emanates from the right part of the field sound and B (the "whoop") emanates from the left part of the sound field.

[0202] Sound field analysis unit 44 may represent a unit configured to perform sound field analysis against HOA coefficients 11 in order to potentially achieve a target bit rate 41. Analysis unit sound field 44 may, based on such analysis and/or a received target bit rate 41, determine the total amount of psychoacoustic encoder instance formations (which may be a function of the total amount of ambient channels or second plane (BGTOT) and the number of foreground channels or, in other words, predominant channels. The total number of formations of psychoacoustic encoder instances can be denoted as numHOATransportChannels. The sound field analysis unit 44 can also determine, again to potentially achieve target bit rate 41, total amount of foreground channels (nFG) 45, minimum sound field order (NBG or alternatively MmAmbHoaOrder) of second of the plane (or, in other words, ambient), the corresponding amount of actual channels representative of the minimum background sound field order (NBGa = (MinAmbHoaOrder+1)2) and indices (i) of additional BG HOA channels to send (which can be collectively denoted as background channel information 43 in the example in Figure 4). Background channel information 42 can also be called ambient channel information 43. Each of the channels that remains from numHOATransportChannels - NBGa, can either be an "additional background/environment channel", a "predominant channel with vector base active”, a “predominant signal with active directional base” or “completely inactive”. In one embodiment, such channel types may be indicated (as a "ChannelType") syntax element via two bits (e.g., 00: additional background channel; 01: predominant signal on a vector basis; 10: signal inactive ; 11: directional based signal). The total amount of background or ambient signals, NBGa, can be determined from (MmAmbHoaOrder +1)2 + the amount of times index 00 (in the example above) appears as a channel type in the bitstream for such a picture.

[0203] In any case, the sound field analysis unit 44 can select the amount of background (or, in other words, ambient) channels and the amount of foreground (or, in other words, predominant) channels based on target bit rate 41, so as to select more background and/or foreground channels when target bit rate 41 is relatively higher (e.g. when target bit rate 41 is equal or greater than 512 Kbps). In one embodiment, the numHOATransportChannels can be set to 8 while the MmAmbHoaOrder can be set to 1 in the bitstream header section (which is described in more detail with reference to Figures 10 to 100(ii)). In this scenario, in every frame, four channels can be dedicated to representing the background or ambient portion of the sound field while the other 4 channels can, on a frame-by-frame basis, vary in channel type - for example, both used as an additional background/environment channel or as a foreground/predominant channel. Foreground/predominant signals can be either a vector-based signal or a directional-based signal as described above.

[0204] In some instances, the total amount of vector-based predominant signals for a frame can be determined by the number of times the ChannelType index is 01 in the bitstream of such a frame in the example above. In the above modality, for all additional background/environment channels (eg corresponding to a ChannelType of 00), corresponding information on which of the possible HOA coefficients (in addition to the first four) can be represented on the channel. This information, for a fourth order HOA content, can be an index to indicate between 5 to 25 (the first four 1 to 4 can be sent all the time when minAmbHoaOrder is set to 1, so it needs to indicate only one of 5 to 25). Such information can therefore be sent using a 5-bit syntax element (for 4th order content), which can be denoted as “CodedAmbCoeffldx”.

[0205] In a second embodiment, all foreground/mainstream signals are vector-based signals. In such a second embodiment, the total amount of foreground/predominant signals can be determined through nFG = numHOATransportChannels - [(MinAmbHoaOrder +1)2 + the amount of times the index 00].

[0206] Sound field analysis unit 44 outputs background channel information 43 and HOA coefficients 11 to background selection unit (BG) 46, background channel information 43 to unit reduction coefficient 46 and the bit stream generation unit 42 and the nFG 45 to a foreground selection unit 36.

[0207] In some examples, the sound field analysis unit 44 may select, based on an analysis of the matrix vectors 33 US[k] and the target bit rate 41, a variable amount of such components that has the highest value. In other words, the sound field analysis unit 44 can determine a value for a variable A (which may be similar or substantially similar to NBG), which separates two subspaces, by analyzing the slope of the curve created through the descending diagonal values. of the vectors of the matrix 33 S[k], where the large singular values represent foreground or distinct sounds and the low singular values represent background components of the sound field. That is, variable A can segment the overall sound field to form a foreground subspace and a background subspace.

[0208] In some examples, the sound field analysis unit 44 may use a first and a second derivative of the singular value curve. The sound field analysis unit 44 may also limit the value so that variable A is between one and five. As another example, the sound field analysis unit 44 may limit the value of variable A to be between one and (N+1)2. Alternatively, the sound field analysis unit 44 may preset the value for variable A, such as to a value of four. Anyway, based on the value of A, the sound field analysis unit 44 determines the total amount of foreground channels (nFG) 45, the order of the background sound field (NBG) and the amount (NBGa) and the indices (i) of additional BG HOA channels to be sent.

[0209] Furthermore, the sound field analysis unit 44 can determine the energy of the vectors in the matrix 35 V[k] on a per vector basis. The sound field analysis unit 44 can determine the energy for each of the vectors in the matrix 35 V[k] and identify those having a high energy as foreground components.

[0210] In addition, the sound field analysis unit 44 can perform various other analyzes in relation to the HOA coefficients 11, including a spatial energy analysis, a spatial masking analysis, a diffusion analysis or other forms of auditory analysis. The sound field analysis unit 44 can perform spatial energy analysis by transforming the HOA coefficients 11 into high energy spatial identification and domain areas representative of directional components of the sound field that must be preserved. Sound field analysis unit 44 can perform perceptual spatial masking analysis in a similar manner to spatial energy analysis, with the exception that sound field analysis unit 44 can identify spatial areas that are masked by sounds. of spatially close superior energy. The sound field analysis unit 44 may then, based on the perceptually masked areas, identify fewer foreground components in some cases. The sound field analysis unit 44 may additionally perform a scatter analysis against the HOA coefficients 11 to identify areas of diffuse energy that may represent background components of the sound field.

[0211] Sound field analysis unit 44 may also represent a unit configured to determine salience, distinctiveness, or dominance of audio data representing a sound field, using directionality-based information associated with the audio data. While energy-based determinations can enhance the rendering of a sound field decomposed through SVD to identify audio components distinct from the sound field, energy-based determinations can also cause a device to misidentify background audio components. as distinct audio components, in cases where Background audio components exhibit a high level of power. That is, a power-only separation of distinct and background audio components may not be robust, while energetic (eg, louder) background audio components may be incorrectly identified as distinct audio components. To more robustly distinguish between distinct and background audio components of the sound field, various aspects of the techniques described in the present disclosure enable the sound field analysis unit 44 to perform a directionality-based analysis of the HOA coefficients 11 to separate the foreground and ambient components from the decomposed versions of the HOA 11 coefficients.

[0212] In this regard, the sound field analysis unit 44 may represent a unit configured or otherwise operable to identify distinct (or foreground) elements from background elements included in one or more of the vectors in the matrix 33 US[k] and the vectors in the matrix 35 V[k]. According to some SVD-based techniques, the most energetic components (e.g. the first few vectors of one or more of the 33 US[k] matrix and the 35 V[k] matrix or vectors derived from them) can be treated as separate components. However, the most energetic components (which are represented by vectors) of one or more of the vectors in the 33 US[k] matrix and the vectors in the 35 V[k] matrix may, in all scenarios, not represent the components/signals that are the most directional.

[0213] The sound field analysis unit 44 may implement one or more aspects of the techniques described herein to identify foreground/direct/predominant elements based on the directionality of the vectors of one or more of the vectors in the matrix 33 US[ k] and the vectors in the matrix 35 V[k] or vectors derived therefrom. In some examples, the sound field analysis unit 44 may identify or select as distinct audio components (where the components may also be referred to as "objects"), one or more vectors based on both the energy and directionality of the vectors. . For example, the sound field analysis unit 44 can identify the vectors of one or more of the vectors in the 33 US[k] matrix and the vectors in the 35 V[k] matrix (or vectors derived therefrom) that exhibit both high energy and high directionality (eg represented as a directionality quotient) as distinct audio components. As a result, if the sound field analysis unit 44 determines that a particular vector is relatively less directional when compared to other vectors of one or more of the vectors in the matrix 33 US[k] and the vectors in the matrix 35 V[k] (or vectors derived therefrom), then, regardless of the energy level associated with the particular vector, the sound field analysis unit 44 can determine that the particular vector represents background (or ambient) audio components of the sound field. represented through the coefficients of HOA 11.

[0214] In some examples, the sound field analysis unit 44 can identify distinct audio objects (which, as noted above, can also be referred to as “components”) based on directionality by performing the following operations. The sound field analysis unit 44 can multiply (for example, using one or more matrix multiplication processes) vectors in the matrix S[k] (which can be derived from the vectors 33 US[k] or, although not is shown in the example of Figure 4, separately emitted by the LIT unit 30) through the vectors in the matrix 35 V[k]. By multiplying the matrix 35 V[k] and the vectors S[k], the sound field analysis unit 44 can obtain matrix VS[K]. Additionally, the sound field analysis unit 44 can square (i.e., exponentiate to a power of two) at least a portion of the entries of each of the vectors in the matrix VS[K]. In some instances, the sound field analysis unit 44 may sum the squared inputs of each vector that are associated with an order greater than 1.

[0215] As an example, if each vector of the matrix VS[K], which includes 25 entries, the sound field analysis unit 44 can, with respect to each vector, square the entries of each vector, starting at fifth entry and ending with the fifteenth entry, so as to sum the squared entries to determine a directionality quotient (or an indicator of directionality). Each summation operation can result in a directionality quotient for a corresponding vector. In this example, the sound field analysis unit 44 can determine that the inputs of each row that are associated with an order less than or equal to 1, namely, the first through the fourth inputs, are more generally directed to the amount of energy and less to the directionality of such inputs. That is, the lowest ambisonic order associated with an order of zero or one corresponds to the spherical basis functions which, as illustrated in Figure 1 and Figure 2, do not give many terms of the pressure wave direction, but instead give volume (which is representative of energy).

[0216] The operations described in the example above can also be expressed according to the following pseudocode. The pseudocode below includes annotations, in the form of comment statements that are enclosed in consecutive cases of the strings “/*” and “*/” (without quotes).

[0217] /* The next line is directed to the analysis of each row independently and the sum of the values in the first row (as an example) from the fifth input to the fifteenth input to determine a directionality quotient or directionality metric for a corresponding vector. Square the entries before the sum. Entries in each row that are associated with an order greater than 1 are associated with a higher ambisonic order and are therefore more likely to be directional.

[0218] /* The next line is directed to sorting the sum of squares for the generated VS matrix and selecting a set of the highest values (e.g. three or four of the highest values)

[0219] In other words, according to the above pseudocode, the sound field analysis unit 44 can select entries from each vector of the matrix VS[K] decomposed from the HOA coefficients 11 that correspond to a spherical basis function which has an order greater than one. The sound field analysis unit 44 can then square these inputs for each vector of the S[k] matrix, so as to sum the squared inputs to identify, compute, or otherwise determine a directionality metric or quotient for each vector of the matrix VS[K]. Then, the sound field analysis unit 44 can sort the vectors from the matrix VS[K] based on the respective directionality metrics of each of the vectors. The sound field analysis unit 44 can sort such vectors in a descending order of directionality metrics so that the vectors with the highest corresponding directionality are first and the vectors with the lowest corresponding directionality are last. The sound field analysis unit 44 can then select the non-zero subset of the vectors that have the greatest relative directionality metric.

[0220] The sound field analysis unit 44 can perform any combination of the foreground analyzes to determine the total amount of psychoacoustic encoder instance formations (which may be a function of the total amount of ambient or background channels (BGTOT) ) and the number of foreground channels. The sound field analysis unit 44 can, based on any combination of the foreground analyses, determine the total number of foreground channels (nFG) 45, the order of the sound field (NBG) and the amount (NBGa) and indices (i) of additional BG HOA channels to send (which may collectively be denoted as background channel information 43 in the example of Figure 4).

[0221] In some examples, the sound field analysis unit 44 may perform such analysis every M-samples, which may be reprocessed as on a frame-by-frame basis. In this respect, the value for A may vary from frame to frame. A case of a bit stream where the decision is made every M-samples is shown in Figures 10 to 100(ii). In other examples, the sound field analysis unit 44 may perform such analysis more than once per frame so as to analyze two or more portions of the frame. Accordingly, the techniques should not be limited in this regard to the examples described in the present disclosure.

[0222] Background selection unit 48 may represent a unit configured to determine HOA coefficients of background or environments 47 based on background channel information (e.g., background sound field (NBG) and the quantity (NBGa) and indices (i) of additional BG HOA channels to be sent). For example, when NBG is equal to one, the background selection unit 48 can select HOA coefficients 11 for each sample of the audio frame that has an order equal to or less than one. The background selection unit 48 can, in this example, then select the HOA coefficients 11 that have an index identified through one of the indices (i) as additional BG HOA coefficients, and that NBGa is provided to the unit generation of bitstream 42 to be specified in bitstream 21 so as to enable the audio decoding device, such as the audio decoding device 24 shown in the example of Figure 3, to analyze the HOA coefficients of BG 47 from bitstream 21. Background select unit 48 may then output ambient HOA coefficients 47 to power offset unit 38. Ambient HOA coefficients 47 may have dimensions D:Mx [(NBG+1)2 + NBGa].

49) para a unidade de codificador de áudio psicoacústico 40, em que os sinais de nFG 49 podem ter as dimensões D: M x nFG e cada um representa objetos de monoáudio. A unidade de seleção de primeiro plano 36 também pode emitir a matriz 35 V[k]’ reordenada (ou v''-""'') (k) 35’) que corresponde aos componentes de primeiro plano do campo sonoro para a unidade de interpolação espaço-temporal 50, em que aquelas das matrizes 35 V[k]’ reordenadas correspondentes aos componentes de primeiro plano podem ser denotadas como matriz 511 V[k] de primeiro plano (que podem ser matematicamente denotadas como

) que têm dimensões D:

tadas como matriz 5"" V[k] de primeiro plano (que podem ser matematicamente denotadas como ) que têm dimensões D: (N+")2 x nF. A unidade de compensação de energia 38 pode representar uma unidade configurada para realizar compensação de energia em relação aos coeficientes de HOA ambientes 47 para compensar pela perda de energia devido à remoção de vários dentre os canais de HOA através da unidade de seleção de segundo plano 48. A unidade de compensação de energia 38 pode realizar uma análise de energia em relação a uma ou mais dentre a matriz 33 U.S.[k] ‘ reordenada, a matriz 35 V[k]’ reordenada, os sinais de nFG 49, os vetores 5"k V[k] reordenados e os coeficientes de HOA ambientes 47 e, em seguida, realizar a compensação de energia com base em tal análise de energia para gerar os coeficientes de HOA ambientes 47’ de energia compensada. A unidade de compensação de energia 38 pode emitir os coeficientes de HOA ambientes 47’ de energia compensada para a unidade de codificador de áudio psicoacústico 40.[0223] The foreground selection unit 36 may represent a unit configured to select those among the reordered matrix 33 US[k]' and the reordered matrix 35 V[k]' representing the foreground or distinct components of the field sound based on nFG 45 (which may represent one or more indices that identify such foreground vectors). The foreground selection unit 36 may output signals from nFG 49 (which may be denoted as a rearranged US[k]1, NFG 49, FG1, ) nfG[k] 49, or

49) to the psychoacoustic audio encoder unit 40, wherein the nFG signals 49 may have dimensions D: M x nFG and each represents monoaudio objects. The foreground selection unit 36 can also output the reordered matrix 35 V[k]' (or v''-""'') (k) 35') which corresponds to the foreground components of the sound field for the unit of spatiotemporal interpolation 50, wherein those of the reordered 35 V[k]' matrices corresponding to the foreground components can be denoted as the foreground 511 V[k] matrix (which can be mathematically denoted as

) that have dimensions D:

The foreground 5"" V[k] matrix (which can be mathematically denoted as ) having dimensions D: (N+")2 x nF. Power compensation unit 38 may represent a unit configured to perform power compensation. power against ambient HOA coefficients 47 to compensate for power loss due to the removal of several of the HOA channels through the background selection unit 48. The power compensation unit 38 can perform power analysis against to one or more of the reordered 33 US[k]' matrix, the reordered 35 V[k]' matrix, the nFG signs 49, the reordered 5"k vectors V[k] and the ambient HOA coefficients 47 and, then perform power compensation based on such power analysis to generate the 47' ambient HOA coefficients of compensated power. Power compensation unit 38 can output power compensated ambient HOA coefficients 47' to psychoacoustic audio encoder unit 40.

[0224] Effectively, the power compensation unit 38 can be used to compensate for possible reductions in the overall power of the background sound components of the sound field generated by reducing the order of the ambient sound field components described through HOA coefficients. "" to generate the reduced-order ambient HOA 47 coefficients (which, in some examples, have an order less than N in terms of just included coefficients corresponding to spherical basis functions that have the following orders/suborders: [(NBG +1)2 + NBGa]). In some examples, the power compensation unit 38 compensates for this power loss by determining a compensation gain in the form of amplification values to apply to each of the columns [(NBG+1)2 + NBGa] of the coefficients 47 ambient HOA coefficients in order to increase the root mean square (RMS) energy of the ambient 47 HOA coefficients to be equal to or at least closer to the RMS of the HOA 11 coefficients (as determined through aggregate energy analysis of one or more of the reordered 33 US[k]' matrix, the reordered 35 V[k]' matrix, the 49 nFG signals, the foreground 511 V[k] vectors, and the order 47 ambient HOA coefficients reduced), before outputting ambient HOA coefficients 47 to the psychoacoustic audio encoder unit 40.

[0225] In some instances, the power compensation unit 38 may identify the RMS for each row and/or column of one or more of the reordered 33 U.S.[k]' matrix and the reordered 35 V[k]' matrix. The power compensation unit 38 can also identify the RMS for each row and/or column of one or more of the selected foreground channels, which may include the nFG signals 49 and the foreground vectors 511 V[k] and reduced-order ambient HOA coefficients 47. The RMS for each row and/or column of one or more of the reordered 33 US[k]' matrix and the reordered 35 V[k]' matrix can be stored to a vector denoted RMSFULL, while the RMS for each row and/or or column of one or more of the nFG signals 49, the foreground vector 511 V[k] and the reduced-order ambient HOA coefficients 47 may be stored to a vector denoted RMSREDUCED- The power compensation unit 38 may , then compute a vector of amplification value Z, according to the following equation: Z = RMSFULL/RMSREDUCED. The power compensation unit 38 may then apply that amplification value vector Z or various portions thereof to one or more of the nFG signals 49, the foreground 511 V[k] vectors, and the HOA coefficients. reduced-order environments. In some instances, the amplification value vector Z is applied only to reduced-order ambient 47 HOA coefficients by the following equation HOABG-RED = HOABG_REDZT, where HOABG-RED denotes reduced-order ambient 47 HOA coefficients, HOABG -RED' denotes the offset energy, the reduced ambient HOA coefficients 47', and ZT denotes the transpose of the Z vector.

[0226] [0214]. In some examples, to determine each RMS of respective rows and/or columns of one or more of the reordered matrix 33 US[k]', the reordered matrix 35 V[k]', the signs of nFG 49, the vectors 51k V [k] and reduced-order ambient HOA coefficients 47, the power compensation unit 38 can first apply a reference spherical harmonic coefficients (SHC) renderer to the columns. The application of the reference SHC renderer through the energy compensation unit 38 allows the determination of RMS in the SHC domain to determine the overall sound field energy described through each row and/or column of the frame represented through rows and/or or columns of one or more of the reordered 33' XJS[k] matrix, the reordered 35 V[k]' matrix, the nFG signals 49, the foreground 511 V[k] vectors, and the ambient HOA coefficients 47 of reduced order, as described in more detail below.

[0227] The spatiotemporal interpolation unit 50 may represent a unit configured to receive the foreground 5 V[k] vectors for the k'-th frame and the foreground 51k V[k-1] vectors for the previous frame (hence the k-1 notation) and perform a spatiotemporal interpolation to generate interpolated foreground vectors V[k]. The spatiotemporal interpolation unit 50 can recombine the nFG signals 49 with the foreground 51k V[k] vectors to recover the reordered foreground HOA coefficients. The spatiotemporal interpolation unit 50 can then divide the reordered foreground HOA coefficients across the interpolated vectors V[k] to generate interpolated nFG signals 49'. The spatiotemporal interpolation unit 50 can also output those of the 511 foreground V[k] vectors that were used to generate the interpolated foreground V[k] vectors, so that an audio decoding device such as as the audio decoding device 24, can generate the interpolated foreground vectors V[k] and thus recover the 51k foreground vectors V[k]. Those of the foreground 51k V[k] vectors used to generate the interpolated foreground V[k] vectors are denoted as the remainder of the foreground 53 V[k] vectors. In order to ensure that the same V[k] and V[k-1] are used in the encoder and decoder, (to create the interpolated V[k] vectors) quantized/dequantized versions of them can be used in the encoder and decoder.

[0228] In that regard, the spatiotemporal interpolation unit 50 may represent a unit that interpolates a first portion of a first audio frame from some other portion of the first audio frame and a temporally subsequent or previous second audio frame . In some examples, the portions may be denoted as subframes, where interpolation as performed with respect to the subframes is described in greater detail below with reference to Figures 45 to 46E. In other examples, the spatiotemporal interpolation unit 50 may operate against a certain last number of samples from the previous frame and a certain first number of samples from the subsequent frame, as described in more detail with reference to Figures 37 to 39. The spatiotemporal interpolation unit 50 can, while performing such interpolation, reduce the amount of samples of the foreground 5 V[k] vectors that need to be specified in the bit stream 21, due to the fact that only those among the foreground 51k V[k] vectors that are used to generate the interpolated V[k] vectors represent a subset of the foreground 51k V[k] vectors. That is, in order to make the compression of HOA coefficients 11 potentially more efficient (by reducing the amount of foreground 51k V[k] vectors that are specified in bitstream 21), various aspects of the techniques described in the present disclosure may provide interpolation of one or more portions of the first audio frame, each portion may represent decomposed versions of the HOA 11 coefficients.

[0229] Spatio-temporal interpolation can result in several benefits. First, the nFG 49 signals may not be continuous from frame to frame due to the block nature in which the SVD or other LIT is performed. In other words, since the LIT unit 30 applies the SVD on a frame-by-frame basis, certain discontinuities may exist in the resulting transformed HOA coefficients as evidence, for example, by the unordered nature of the matrix 33 US[k] and of the matrix 35 V[k]. By performing this interpolation, discontinuity can be reduced, as the interpolation can have a smoothing effect that potentially reduces any artifacts introduced due to frame boundaries (or, in other words, segmentation of the HOA 11 coefficients so to form frames). Using the foreground 51k V[k] vectors to perform this interpolation and then generating the interpolated nFG 49' signals based on the foreground V[k] 51k vectors interpolated from the retrieved reordered HOA coefficients can smooth at least some of the effects due to frame-by-frame operation, as well as due to the reordering of nFG 49 signals.

[0230] During operation, the spatiotemporal interpolation unit 50 may interpolate one or more subframes of a first audio frame from a first decomposition, e.g. the foreground 5 V[k] vectors, from a portion of a first plurality of HOA coefficients 11 included in the first frame and a second decomposition, e.g., foreground vectors 51k-1 V[k], of a portion of a second plurality of HOA coefficients 11 included in a second frame to generate the decomposed interpolated spherical harmonic coefficients for the one or more subframes.

[0231] In some examples, the first decomposition comprises the first 51k V[k] foreground vectors representative of singular vectors to the right of the HOA coefficients portion 11. Similarly, the second decomposition comprises the second 51k V[ vectors k] of foreground representative of singular vectors to the right of the portion of the coefficients of  HOA 11.

[0232] In other words, spherical harmonic based 3D audio can be a parametric representation of the 3D pressure field in terms of orthogonal basis functions on a sphere. The higher the N order of the representation, the potentially higher the spatial resolution and often the greater the amount of spherical harmonic (SH) coefficients (for a total of (N+1)2 coefficients). For many applications, bandwidth compression of the coefficients may be required in order to efficiently transmit and store the coefficients. These techniques addressed in the present disclosure can provide a frame-based dimensionality reduction process using Singular Value Decomposition (SVD). SVD analysis can decompose each frame of coefficients into three matrices U, S, and V. In some examples, the techniques can manipulate some of the vectors in the U.S.[k] matrix, as the foreground components of the underlying sound field. However, when manipulated in such a way, these vectors (in the matrix S[k] U) are discontinuous from frame to frame - even though they represent the same distinct audio component. Such discontinuities can lead to significant artifacts when components are fed through transform audio encoders.

[0233] The techniques described in the present disclosure can address this discontinuity. That is, the techniques can be based on the observation that the matrix V can be interpreted as orthogonal spatial axes in the Harmonic Spatial domain. The matrix U[k] can represent a projection of the Spherical Harmonics (HOA) data in terms of such basis functions, where the discontinuity can be attributed to orthogonal spatial geometric axes (V[k]) that are changed in every frame - and are therefore discontinuous in themselves. This is different from a similar decomposition, such as the Fourier Transform, where the basis functions are, in some instances, constants from frame to frame. In these terms, SVD can be considered as a compatibility search algorithm. The techniques described in the present disclosure enable the spatiotemporal interpolation unit 50 to maintain continuity between the base functions (V[k]) from frame to frame - by interpolation therebetween.

[0234] As noted above, interpolation can be performed against samples. This case is generalized in the description above when the subframes comprise a single set of samples. Both in the case of interpolation over the samples and over the subframes, the interpolation operation can take the form of the following equation:

são necessários e também indica que a saída de tal processo produz / de tais vetores). Alternativamente, / pode indicar subquadros que consistem em múltiplas amostras. Quando, por exemplo, um quadro é dividido em quatro subquadros, / pode compreender valores de 1, 2, 3 e 4, para cada um dos subquadros. O valor de / pode ser sinalizado como um campo denominado “CodedSpatiallnterpolationTime” através de um fluxo de bits - de modo que a operação de interpolação possa ser replicada no decodificador. A w(Z) pode compreender valores das ponderações de interpolação. Quando a interpolação é linear, w(7) pode variar linearmente e monotonicamente entre 0 e 1, como uma função de /. Em outros casos, w(Z) pode variar entre 0 e 1 de uma maneira não linear, porém, monotônica (tal como um quarto de ciclo de um cosseno elevado) como uma função de /. A função, w(Z), pode ser indexada entre algumas diferentes possibilidades de funções e sinalizada no fluxo de bits como um campo denominado “SpatiallnterpolationMethod” de modo que a operação de interpolação idêntica possa ser replicada pelo decodificador. Quando w(Z) é um valor próximo de 0, a saída,

) pode ser altamente ponderada ou influenciada por v(k— 1). Enquanto quando w(Z) é um valor próximo de 1, isso assegura que a saída,

, seja altamente ponderada ou influenciada por v(k— 1).[0235] In this equation above, the interpolation can be performed with respect to the single V-vector v(k) from the single V-vector v(k— 1) which, in one embodiment, could represent V-vectors of adjacent frames ke k- 1. In the above equation, /, represents the resolution through which the interpolation is performed, where / can indicate an integer sample, and / = 1, T (where T is the length of samples through which the interpolation is performed and through which the emitted interpolated vectors,

are needed and also indicates that the output of such a process produces / of such vectors). Alternatively, / can indicate subframes consisting of multiple samples. When, for example, a frame is divided into four subframes, / may comprise values of 1, 2, 3 and 4 for each of the subframes. The value of / can be signaled as a field named “CodedSpatiallnterpolationTime” via a bitstream - so that the interpolation operation can be replicated at the decoder. The w(Z) can comprise values of the interpolation weights. When the interpolation is linear, w(7) can vary linearly and monotonically between 0 and 1, as a function of /. In other cases, w(Z) can vary between 0 and 1 in a nonlinear but monotonic manner (such as a quarter cycle of a high cosine) as a function of /. The function, w(Z), can be indexed among a few different possible functions and flagged in the bitstream as a field called “SpatiallnterpolationMethod” so that the identical interpolation operation can be replicated by the decoder. When w(Z) is close to 0, the output,

) can be highly weighted or influenced by v(k— 1). While when w(Z) is a value close to 1, this ensures that the output,

, either highly weighted or influenced by v(k— 1).

[0236] Coefficient reduction unit 46 may represent a unit configured to perform coefficient reduction against the remaining foreground 53 V[k] vectors based on background channel information 43 to output 55 V[k] vectors ] reduced foreground to the quantization unit 52. The reduced foreground 55 V[k] vectors can have dimensions D:

. A unidade de análise de campo sonoro 44 pode analisar os coeficientes de HOA 11 para determinar BGTOT, o que pode identificar, não apenas a (NBG+1)2, mas o TotalOfAddAmbHOAChan, que pode ser chamado, coletivamente, de as informações de canal de segundo plano 43. A unidade de redução de coeficiente 46 pode, em seguida, remover os coeficientes correspondentes a (NBG+1)2 e o TotalOfAddAmbHOAChan dos vetores 53 V[k] de primeiro plano restantes para gerar uma matriz 55 dimensional V[k] menor de tamanho

, que também pode ser denominada como os vetores 55 V[k] de primeiro plano.[0237] The coefficient reduction unit 46 may, in this regard, represent a unit configured to reduce the amount of coefficients from the remaining 53 V[k] foreground vectors. In other words, the coefficient reduction unit 46 can represent a unit configured to eliminate the coefficients of the foreground V[k] vectors (which form the remaining foreground 53 V[k] vectors) that have little or no information. directional. As described above, in some examples, the coefficients of distinct vectors V[k] or, in other words, foreground vectors V[k] corresponding to first-order and zero-order basis functions (which can be denoted as NBG) provide little directional information and therefore can be removed from the foreground V vectors (through a process that might be called “coefficient reduction”). In this example, greater flexibility can be provided, not only to identify such coefficients that correspond to NBG, but to identify additional HOA channels (which can be called the variable TotalOfAddAmbHOAChan) from the set of

. The sound field analysis unit 44 can analyze the HOA coefficients 11 to determine BGTOT, which can identify not only the (NBG+1)2, but the TotalOfAddAmbHOAChan, which may collectively be called the channel information. 43. The coefficient reduction unit 46 can then remove the coefficients corresponding to (NBG+1)2 and the TotalOfAddAmbHOAChan from the remaining 53 V[k] foreground vectors to generate a 55 dimensional matrix V[ k] smaller in size

, which can also be referred to as the foreground 55 V[k] vectors.

[0238] The quantization unit 52 may represent a unit configured to perform any form of quantization to compress the reduced foreground 55 V[k] vectors to generate coded foreground 57 V[k] vectors so as to output such encoded foreground vectors 57 V[k] to the bitstream generating unit 42. During operation, the quantizing unit 52 may represent a unit configured to compress a spatial component of the sound field, i.e., one or more of the reduced foreground 55 V[k] vectors in the present example. For example purposes, the reduced foreground 55 V[k] vectors are assumed to include two rows of vectors that have, as a result of the coefficient reduction, less than 25 elements each (which implies a representation fourth-order HOA of the sound field). Although described in relation to two-row vectors, any number of vectors can be included in the foreground 55 V[k] vectors reduced to (N+1)2, where n denotes the order of the HOA representation of the sound field . Furthermore, although it is described below to perform a scale and/or entropy quantization, the quantization unit 52 can perform any form of quantization that results in compression of the reduced foreground vectors 55 V[k].

[0239] The quantization unit 52 can receive the reduced foreground 55 V[k] vectors and perform a compression scheme to generate coded foreground 57 V[k] vectors. This compression scheme can involve any conceivable compression scheme for compressing elements of an array or data generically and should not be limited to the example described below in greater detail. The quantization unit 52 may perform, as an example, a compression scheme that includes one or more of: transforming floating point representations of each element of the reduced foreground vectors 55 V[k] into integer representations of each element of reduced foreground 55 V[k] vectors, uniformly quantize the integer representations of the reduced foreground 55 V[k] vectors, and categorizing and encoding the quantized integer representations of the foreground 55 V[k] vectors remaining.

[0240] In some examples, several of the one or more processes of the present compression scheme may be dynamically controlled via parameters to achieve or approximately achieve, as an example, a target bit rate for the resulting bit stream 21. Since each of the reduced foreground 55 V[k] vectors is orthonormal to the others, each of the reduced foreground 55 V[k] vectors can be coded independently. In some examples, as described in more detail below, each element of each of the reduced foreground 55 V[k] vectors can be encoded using the same encoding mode (defined via various sub-modes).

[0241] Anyway, as noted above, this coding scheme may first involve transforming the floating point representations of each element (which is, in some examples, a 32-bit floating point number) of each of the foreground 55 V[k] vectors reduced to a 16-bit integer representation. The quantization unit 52 can perform this floating-point-to-integer transformation by multiplying each element of a specific one of the reduced foreground 55 V[k] vectors by 215, which is, in some examples, performed by shifting right by 15.

[0242] The quantization unit 52 can then perform uniform quantization with respect to all elements of the specific among the reduced foreground 55 V[k] vectors. The quantization unit 52 can identify a quantization step size based on a value, which can be denoted with a parameter of nbits. The quantization unit 52 can dynamically determine that nbit parameter based on the target bit rate 41. The quantization unit 52 can determine the quantization step size as a function of such nbit parameter. As an example, the quantization unit 52 can determine the quantization step size (denoted as "delta" or "" in the present disclosure) to be equal to 216-nbits. In this example, if nbits equals six, delta equals 210 and there are 26 levels of quantization. In this respect, for a vector element v, the quantized vector element q is equal to

[0243] The quantization unit 52 can then perform the categorization and residual encoding of the quantized vector elements. As an example, the quantization unit 52 can, for a given quantized vector element q, identify a category (by determining a cid category identifier) to which that element corresponds using the following equation:

[0244] The quantization unit 52 can then use the Huffman code in this category index cid, while also identifying a sign bit that indicates whether q is a positive value or a negative value. The quantization unit 52 can then identify a residue in that category. As an example, the quantization unit 52 can determine this residue according to the following equation:

0245] The quantization unit 52 can then block encode this residue with cid-1 bits.

[0246] The following example illustrates a simplified example of this categorization and residual encoding process. First, nbits is assumed to be six, so q [-31,31]. Then the following is assumed:

[0247] In addition, the following is assumed:

[0248] Thus, for a q = [6, -17, 0, 0, 3], the following can be determined:

[0249] » cid = 3,5,0,0,2

[0250] » sign = 1,0,x,x,l

[0251] » residual = 2,1,x,x, 1

[0252] » Bits for 6 = '0010' + 1 ' + ' 10'

[0253] » Bits for -17 = '00111 ' + ' ' + '0001'

[0254] » Bits for 0 = '0'

[0255] » Bits for 0 = '0'

[0256] » Bits for 3 = '000' + 'l' + 'l'

[0257] » Total bits = 7 + 10+ 1 + 1 +5 = 24

[0258] » Average bits = 24/5 = 4.8 Although not shown in the above simplified example, the quantization unit 52 can select different Huffman codebooks for different nbit values during cid encoding. In some examples, the quantization unit 52 may provide a different Huffman codebook for nbit values 6,..., 15. In addition, the quantization unit 52 may include five different Huffman codebooks for each of the different values of nbits in the range of 6,...15 for a total of 50 Huffman codebooks. In this regard, the quantization unit 52 may include a plurality of different Huffman codebooks to accommodate cid encoding in a number of different statistical contexts.

[0259] To illustrate, the quantization unit 52 may, for each of the nbit values, include a first Huffman codebook for encoding vector elements one through four, a second Huffman codebook for encoding the vector elements five through nine, a third Huffman codebook to encode vector elements nine onwards. These first three Huffman codebooks can be used when one of the 55 V[k] reduced foreground vectors to be compressed is not predicted from one of the corresponding temporally subsequent 55 V[k] reduced foreground vectors and is not representative of spatial information from a synthetic audio object (one defined, for example, originally by a pulse code modulated (PCM) audio object). The quantization unit 52 may additionally include, for each of the nbit values, a fourth Huffman codebook to encode one of the reduced foreground 55 V[k] vectors when the same among the 55 V[k] vectors reduced foreground vectors is predicted from one of the corresponding temporally subsequent reduced foreground vectors 55 V[k]. The quantization unit 52 may also include, for each of the nbit values, a fifth Huffman codebook to encode the one of the 55 V[k] reduced foreground vectors when the same among the 55 V[k] vectors ] reduced foreground is representative of a synthetic audio object. Huffman's various codebooks can be developed for each of these different statistical contexts, that is, the non-predicted and non-synthetic context, the predicted context, and the synthetic context in this example.

[0260] The following table illustrates the Huffman table selection and the bits to be specified in the bitstream to allow the decompression unit to select the appropriate Huffman table:

[0261] In the above table, the prediction mode (“Prev Mode”) indicates whether the prediction was performed for the current vector, while the Huffman Table (“HT info”) indicates the codebook information (or Additional Huffman tables) used to select one of Huffman tables one through five.

[0262] The following table further illustrates this Huffman table selection process considering various statistical contexts or scenarios.

[0263] In the above table, the column “Record” indicates the encoding context when the vector is representative of an audio object that has been recorded while the column “Synthetic” indicates an encoding context for when the vector is representative of an object of synthetic audio. The row “Without Prediction” indicates encoding context when prediction is not performed against vector elements, while row “With prediction” indicates encoding context when prediction is performed against vector elements. As shown in this table, the quantization unit 52 selects HT{1, 2, 3} when the vector is representative of a recorded audio object and prediction is not performed with respect to vector elements. The quantization unit 52 selects HT5 when the audio object is representative of a synthetic audio object and prediction is not performed with respect to vector elements. The quantization unit 52 selects HT4 when the vector is representative of a recorded audio object and prediction is performed with respect to vector elements. The quantization unit 52 selects HT5 when the audio object is representative of a synthetic audio object and prediction is performed with respect to vector elements.

[0264] In this regard, the quantization unit 52 can perform the scale quantization and/or Huffman coding noted above to compress the reduced foreground 55 V[k] vectors so as to output the 57 V[k] vectors encoded foreground information, which may be called auxiliary channel information 57. This auxiliary channel information 57 may include syntax elements used to encode the remaining foreground vectors 55 V[k]. The quantizing unit 52 can output the auxiliary channel information 57 in a similar manner as shown in the example of one of Figures 10B and 10c.

[0265] As noted above, quantization unit 52 may generate syntax elements for auxiliary channel information 57. For example, quantization unit 52 may specify a syntax element in a header of an access unit (which may include one or more frames) that denotes which of the plurality of configuration modes has been selected. Although described as specified on a per-unit basis, the quantization unit 52 may specify that syntax element on a per-frame basis or any other periodic basis or non-periodic basis (such as once for the entire bit stream) . In any case, this syntax element may comprise two bits that indicate which of the frame configuration modes have been selected to specify the set of non-zero coefficients of the reduced foreground 55 V[k] vectors to represent the directional aspects of this component. distinct. The syntax element can be denoted as “codedVVecLength”. In this way, the quantization unit 52 can signal or otherwise specify in the bitstream which of the frame configuration modes were used to specify the foreground vectors 57 V[k] encoded in the bitstream. Although it is described in relation to frame configuration modes, the techniques should not be limited to frame configuration modes, but any other number of configuration modes, including a single configuration mode or a plurality of configuration modes. The scale/entropy quantization unit 53 may also specify the flag 63 as another syntax element in the auxiliary channel information 57.

[0266] The psychoacoustic audio encoder unit 40 included in the audio encoding device 20 can represent multiple instances of a psychoacoustic audio encoder, each of which is used to encode an audio object or HOA channel different from each other of the compensated ambient HOA 47' coefficients and the interpolated nFG 49' signals to generate encoded ambient HOA 59 coefficients and encoded nFG 61 signals. The psychoacoustic audio encoder unit 40 can output the encoded ambient HOA coefficients 59 and the encoded nFG signals 61 to the bit stream generation unit 42.

[0267] In some instances, this psychoacoustic audio encoder unit 40 may represent one or more cases of an advanced audio coding encryption (AAC) unit. The psychoacoustic audio encoder unit 40 can encode each column or row of the energy compensated ambient HOA coefficients 47' and the interpolated nFG signals 49'. Often, the psychoacoustic audio encoder unit 40 may invoke an instance of an AAC encryption unit for each of the order/suborder combinations remaining in the energy compensated ambient HOA coefficients 47' and the interpolated nFG signals 49' . More information regarding how spherical background 31 harmonic coefficients can be encoded using an AAC encryption unit can be found in a convention work by Eric Hellerud, et al, entitled “Encoding Higher Order Ambisonics with AAC ”, presented at the 124th Convention, May 17-20, 2008 and available at: http://ro.uow.edu.au/cgi/viewcontent.cgi?article=8025&conte xt=engpapers. In some instances, the audio encryption unit 14 may encode the audio of the energy compensated ambient 47' HOA coefficients using a lower target bit rate than that used to encode the interpolated nFG 49' signals, potentially compressing more so, the ambient HOA coefficients 47' of compensated energy compared to the interpolated nFG signals 49'. [0240] The bitstream generation unit 42 included in the audio encoding device 20 represents a unit that formats data to conform to a known format (which may refer to a format known by a decoding device), so as to thus generate the vector-based bit stream 21. The bit stream generation unit 42 can represent a multiplexer in some examples, which can receive the coded foreground vectors 57 V[k], the coded ambient HOA coefficients 59, the coded nFG signals 61 and the background channel 43. The bit stream generation unit 42 may then generate a bit stream 21 based on the coded foreground vectors 57 V[k], the coded ambient HOA coefficients 59, the signals of encoded nFG 61 and in the background channel information 43. The bit stream 21 may include a primary or main bit stream and one or more auxiliary channel bit streams.

[0268] Although not shown in the example of Figure 4, the audio encoding device 20 may also include a bitstream output unit that switches the bitstream output of the audio encoding device 20 (e.g., between directional-based bitstream 21 and vector-based bitstream 21) based on whether a current frame needs to be encoded using directional-based synthesis or vector-based synthesis. This bitstream output unit can perform such switching based on the syntax element output via the content analysis unit 26 which indicates whether a directional based synthesis has been performed (as a result of detecting that the HOA coefficients 11 were generated from a synthetic audio object) or a vector-based synthesis was performed (as a result of detecting that HOA coefficients were recorded). The bitstream output unit can specify the correct header syntax to indicate that current switching or encoding used for the current frame along with the respective one of the 21 bitstreams.

[0269] In some instances, various aspects of the techniques also enable the audio encoding device 20 to determine whether HOA coefficients 11 are generated from a synthetic audio object. Such aspects of the techniques enable the audio encoding device 20 to be configured to obtain an indication of whether representative spherical harmonic coefficients of a sound field are generated from a synthetic audio object.

[0270] In these and other cases, the audio encoding device 20 is further configured to determine whether spherical harmonic coefficients are generated from the synthetic audio object.

[0271] In these and other cases, the audio encoding device 20 is configured to exclude a first vector from a framed spherical harmonic coefficient matrix that stores at least a portion of the spherical harmonic coefficients representative of the sound field to obtain a matrix of spherical harmonic coefficient with reduced frame.

[0272] In these and other cases, the audio encoding device 20 is configured to exclude a first vector from a framed spherical harmonic coefficient matrix that stores at least a portion of the spherical harmonic coefficients representative of the sound field to obtain a matrix spherical harmonic coefficient matrix with reduced frame and predict a vector of the spherical harmonic coefficient matrix with reduced frame based on the remaining vectors of the spherical harmonic coefficient matrix with reduced frame.

[0273] In these and other cases, the audio encoding device 20 is configured to exclude a first vector from a framed spherical harmonic coefficient matrix that stores at least a portion of the spherical harmonic coefficients representative of the sound field to obtain a matrix reduced-frame spherical harmonic coefficient matrix and predicting a reduced-frame spherical harmonic coefficient matrix vector based, at least in part, on a sum of the remaining vectors of the reduced-frame spherical harmonic coefficient matrix.

[0274] In these and other cases, the audio encoding device 20 is configured to predict a vector of a framed spherical harmonic coefficient matrix that stores at least a portion of the spherical harmonic coefficients based, at least in part, on a sum of the remaining vectors of the framed spherical harmonic coefficient matrix.

[0275] In these and other cases, the audio encoding device 20 is further configured to predict a vector of a framed spherical harmonic coefficient matrix that stores at least a portion of the spherical harmonic coefficients based, at least in part, on into a sum of the remaining vectors of the framed spherical harmonic coefficient matrix and compute an error based on the predicted vector.

[0276] In these and other cases, the audio encoding device 20 is configured to predict a vector of a framed spherical harmonic coefficient matrix that stores at least a portion of the spherical harmonic coefficients based, at least in part, on a sum of the remaining vectors of the framed spherical harmonic coefficient matrix and compute an error based on the predicted vector and the corresponding vector of the framed spherical harmonic coefficient matrix.

[0277] In these and other cases, the audio encoding device 20 is configured to predict a vector of a framed spherical harmonic coefficient matrix that stores at least a portion of the spherical harmonic coefficients based, at least in part, on a sum of the remaining vectors of the framed spherical harmonic coefficient matrix and compute an error as a sum of the absolute value of the predicted vector difference and the corresponding vector of the framed spherical harmonic coefficient matrix.

[0278] In these and other cases, the audio encoding device 20 is configured to predict a vector of a framed spherical harmonic coefficient matrix that stores at least a portion of the spherical harmonic coefficients based, at least in part, on a sum of the remaining vectors of the framed spherical harmonic coefficient matrix, compute an error based on the predicted vector and the corresponding vector of the framed spherical harmonic coefficient matrix, compute a ratio based on an energy of the corresponding vector of the coefficient matrix spherical harmonic with frame and error and comparing the ratio to a threshold to determine if spherical harmonic coefficients representative of the sound field are generated from the synthetic audio object.

[0279] In these and other cases, the audio encoding device 20 is configured to specify the indication in a bit stream 21 that stores a compressed version of the spherical harmonic coefficients.

[0280] In some instances, the various techniques allow the audio encoding device 20 to perform a transformation with respect to the HOA coefficients 11. In these and other cases, the audio encoding device 20 can be configured to obtain one or more plus first vectors describing distinct components of the sound field and one or more second vectors describing background components of the sound field, either the one or more first vectors or the one or more second vectors generated by at least performing a transformation into in relation to the plurality of spherical harmonic coefficients 11.

[0281] In these and other cases, the audio encoding device 20, wherein the transformation comprises a singular value decomposition that generates a matrix U representative of the left singular vectors of the plurality of spherical harmonic coefficients, a matrix S representative of singular values of the plurality of spherical harmonic coefficients and a matrix V representative of singular vectors to the right of the plurality of spherical harmonic coefficients 11.

[0282] In these and other cases, audio encoding device 20, wherein the first one or more vectors comprise one or more audio-encoded UDIST * SDIST vectors which, prior to audio encoding, were generated by multiplying one or more audio-encoded SDIST vectors of a matrix U through one or more SDIST vectors of a matrix S and wherein the matrix U and matrix S are generated by at least performing singular value decomposition with respect to the plurality of coefficients spherical harmonics.

[0283] In these and other cases, audio encoding device 20, wherein the first one or more vectors comprise one or more audio-encoded UDIST * SDIST vectors which, prior to audio encoding, were generated by multiplying one or more audio-encoded UDIST vectors of a matrix U through one or more audio-encoded SDIST vectors of a matrix S and one or more VTDIST vectors of a transform of a matrix V and where the matrix U and the matrix S and the matrix V are generated at least by performing the singular value decomposition with respect to the plurality of spherical harmonic coefficients 11.

[0284] In these and other cases, audio encoding device 20, wherein the first one or more vectors comprise one or more UDIST * SDIST vectors which, prior to audio encoding, were generated by multiplying one or more vectors Audio encoded UDIST of a matrix U through one or more SDIST vectors of a matrix S and one or more vectors VTDIST of a transform of a matrix V, where the matrix U, matrix S and matrix V have been generated at least performing singular value decomposition with respect to the plurality of spherical harmonic coefficients and wherein the audio coding device 20 is further configured to obtain a D-value indicating the amount of vectors to be extracted from a bit stream to form o one or more UDIST * SDIST vectors and o one or more DIST vectors.

[0285] In these and other cases, audio encoding device 20, wherein the first one or more vectors comprise one or more UDIST * SDIST vectors which, prior to audio encoding, were generated by multiplying one or more vectors Audio encoded UDIST of a matrix U through one or more SDIST vectors of a matrix S and one or more vectors VTDIST of a transform of a matrix V, where the matrix U, matrix S and matrix V have been generated at least performing singular value decomposition with respect to the plurality of spherical harmonic coefficients and wherein the audio encoding device 20 is further configured to obtain a D value that indicates, on an audio frame per audio frame basis, the number of vectors to extract from a bit stream to form one or more UDIST * SDIST vectors and one or more VtDIST vectors.

[0286] In these and other cases, the audio encoding device 20, wherein the transformation comprises a principal component analysis to identify the distinct components of the sound field and the background components of the sound field.

[0287] Various aspects of the techniques described in the present disclosure may provide the audio encoding device 20 configured to compensate for quantization error.

[0288] In some instances, the audio encoding device 20 may be configured to quantize one or more first vectors representative of one or more components of a sound field and compensate for the error introduced due to quantizing the one or more first vectors into a or more second vectors which are also representative of the same one or more sound field components.

[0289] In these and other cases, the audio coding device is configured to quantize one or more vectors from a transform of a generated V matrix, at least in part, by performing a singular value decomposition with respect to a plurality of spherical harmonic coefficients that describe the sound field.

[0290] In these and other cases, the audio encoding device is further configured to perform a singular value decomposition with respect to a plurality of spherical harmonic coefficients representative of a sound field to generate a matrix U representative of left singular vectors of the plurality of spherical harmonic coefficients, a matrix S representative of singular values of the plurality of spherical harmonic coefficients and a matrix V representative of singular vectors to the right of the plurality of spherical harmonic coefficients and configured to quantize one or more vectors from a transform of the matrix v.

[0291] In these and other cases, the audio encoding device is further configured to perform a singular value decomposition with respect to a plurality of spherical harmonic coefficients representative of a sound field to generate a matrix U representative of left singular vectors of the plurality of spherical harmonic coefficients, a matrix S representative of singular values of the plurality of spherical harmonic coefficients and a matrix V representative of singular vectors to the right of the plurality of spherical harmonic coefficients, configured to quantize one or more vectors from the matrix transform V is set to compensate for the error introduced due to quantization on one or more U*S vectors computed by multiplying one or more U vectors from the U matrix by one or more S vectors from the S matrix.

[0292] In these and other cases, the audio encoding device is further configured to perform a singular value decomposition with respect to the plurality of spherical harmonic coefficients representative of a sound field to generate a matrix U representative of left singular vectors of the sound field. plurality of spherical harmonic coefficients, a matrix S representative of singular values of the plurality of spherical harmonic coefficients and a matrix V representative of singular vectors to the right of the plurality of spherical harmonic coefficients, determine one or more UDIST vectors of the matrix U, each of which of them corresponds to a distinct component of the sound field, determine one or more SDIST vectors of the matrix S, each of which corresponds to the same distinct component of the sound field and determine one or more VTDIST vectors of a transform of the matrix V, in that each of them corresponds to the same distinct component of the sound field oro, set to quantize the one or more VTDIST vectors to generate one or more VTQ_DIST vectors and set to compensate for the error introduced due to quantization in one or more UDIST * SDIST vectors computed by multiplying the one or more UDIST vectors from the matrix U through of one or more SDIST vectors of the matrix S so as to generate one or more error compensated UDIST * SDIST vectors.

[0293] In these and other cases, the audio encoding device is configured to determine distinct spherical harmonic coefficients based on the one or more UDIST vectors, where the one or more SDIST vectors and the one or more VTDIST vectors and perform pseudoinversion against the VTQ_DIST vectors to divide the distinct spherical harmonic coefficients by the one or more VTQ_DIST vectors and thus generate one or more error compensated UC_DIST * SC_DIST vectors that compensate, at least in part, for the error introduced through the quantization of the VTDIST vectors .

[0294] In these and other cases, the audio encoding device is further configured to perform a singular value decomposition with respect to a plurality of spherical harmonic coefficients representative of a sound field to generate a matrix U representative of left singular vectors of the plurality of spherical harmonic coefficients, a matrix S representative of singular values of the plurality of spherical harmonic coefficients and a matrix V representative of singular vectors to the right of the plurality of spherical harmonic coefficients, determine one or more vectors UBG of the matrix U that describe one or more background components of the sound field and one or more UDIST vectors of the matrix U that describe one or more distinct components of the sound field, determine one or more SBG vectors of the matrix S that describe the one or more background components of the sound field and one or more SDIST vectors of the matrix S that describe the one or more components distinct elements of the sound field and determine one or more VTDIST vectors and one or more VTBG vectors from a transform of the matrix V, where the VTDIST vectors describe the one or more distinct components of the sound field and the VTBGs describe the one or more components of background of the sound field, set to quantize the one or more VTDIST vectors to generate one or more VTQ_DIST vectors and set to compensate for the error introduced due to quantization into spherical background harmonic coefficients formed by multiplying the one or more UBG vectors by the one or more SBG vectors and then by the one or more VTBG vectors so as to generate error-compensated background spherical harmonic coefficients.

[0295] In these and other cases, the audio encoding device is configured to determine the error based on the VTDIST vectors and one or more UDIST*SDIST vectors formed by multiplying the UDIST vectors by the SDIST vectors and adding the determined error to the spherical background harmonic coefficients to generate the error compensated spherical background harmonic coefficients.

[0296] In these and other cases, the audio encoding device is configured to compensate for the error introduced due to the quantization of one or more first vectors into one or more second vectors that are also representative of the same one or more sound field components. to generate one or more second compensated error vectors and further configured to generate a bit stream to include the one or more second compensated error vectors and the one or more first quantized vectors.

[0297] In these and other cases, the audio encoding device is configured to compensate for the error introduced due to the quantization of one or more first vectors into one or more second vectors that are also representative of the same one or more sound field components. to generate one or more second compensated error vectors and additionally configured to audio encode the one or more second compensated error vectors and generate a bit stream to include the one or more second audio encoded compensated error vectors and the one or plus first quantized vectors.

[0298] The various aspects of the techniques may additionally enable the audio encoding device 20 to generate reduced spherical harmonic coefficients or decompositions thereof. In some instances, the audio coding device 20 may be configured to perform, based on a target bit rate, order reduction with respect to a plurality of spherical harmonic coefficients or decompositions thereof to generate reduced or reduced spherical harmonic coefficients. the reduced decompositions thereof, wherein the plurality of spherical harmonic coefficients represents a sound field.

[0299] In these and other cases, the audio encoding device 20 is further configured to, prior to performing order reduction, perform a singular value decomposition with respect to a plurality of spherical harmonic coefficients to identify one or more first vectors that describe distinct components of the sound field and one or more second vectors that identify background components of the sound field and configured to perform order reduction with respect to the one or more first vectors, the one or more second vectors or the one or more more first vectors and the one or more second vectors.

[0300] In these and other cases, the audio encoding device 20 is additionally configured to perform a content analysis with respect to the plurality of spherical harmonic coefficients or decompositions thereof and configured to perform, based on the target rate of bits and in content analysis, order reduction with respect to the plurality of spherical harmonic coefficients or decompositions thereof to generate reduced spherical harmonic coefficients or reduced decompositions thereof.

[0301] In these and other cases, the audio encoding device 20 is configured to perform a spatial analysis with respect to the plurality of spherical harmonic coefficients or decompositions thereof.

[0302] In these and other cases, the audio encoding device 20 is configured to perform a scatter analysis with respect to the plurality of spherical harmonic coefficients or decompositions thereof.

[0303] In these and other cases, the audio encoding device 20 is the one or more processors that are configured to perform a spatial analysis and a scatter analysis with respect to the plurality of spherical harmonic coefficients or decompositions thereof.

[0304] In these and other cases, the audio encoding device 20 is further configured to specify one or more orders and/or one or more suborders of spherical basis functions to which those of the reduced spherical harmonic coefficients or the reduced decompositions of the same correspond in a stream of bits that includes the reduced spherical harmonic coefficients or the reduced decompositions thereof.

[0305] In these and other cases, the reduced spherical harmonic coefficients or the reduced decompositions thereof have less values than the plurality of spherical harmonic coefficients or the decompositions thereof.

[0306] In these and other cases, the audio encoding device 20 is configured to remove those from among the plurality of spherical harmonic coefficients or vectors of the decompositions thereof that have a specified order and/or suborder to generate the reduced spherical harmonic coefficients or their reduced decompositions.

[0307] In these and other cases, the audio encoding device 20 is configured to zero out those among the plurality of spherical harmonic coefficients or those vectors of the decomposition thereof that have a specified order and/or suborder to generate the spherical harmonic coefficients reduced or reduced decomposition thereof.

[0308] Various aspects of the techniques may also allow the audio encoding device 20 to be configured to represent distinct components of the sound field. In these and other cases, the audio encoding device 20 is configured to obtain a first set of non-zero coefficients of a vector to be used to represent a distinct component of a sound field, where the vector is decomposed from a plurality of spherical harmonic coefficients that describe the sound field.

[0309] In these and other cases, the audio encoding device 20 is configured to determine the first non-zero set of vector coefficients to include all coefficients.

[0310] In these and other cases, the audio encoding device 20 is configured to determine the first non-zero set of coefficients as those among the coefficients that correspond to an order greater than an order of a base function, at the which one or more of the plurality of spherical harmonic coefficients corresponds.

[0311] In these and other cases, the audio encoding device 20 is configured to determine the first non-zero set of coefficients to include those among the coefficients that correspond to an order greater than an order of a base function to the which one or more of the plurality of spherical harmonic coefficients corresponds and exclude at least one of the coefficients which correspond to an order greater than the order of the basis function to which the one or more of the plurality of spherical harmonic coefficients corresponds.

[0312] In these and other cases, the audio encoding device 20 is configured to determine the first non-zero set of coefficients to include all coefficients with the exception of at least one of the coefficients that correspond to an order greater than the that an order of a basis function to which one or more of the plurality of spherical harmonic coefficients corresponds.

[0313] In these and other cases, the audio encoding device 20 is further configured to specify the first non-zero set of vector coefficients in auxiliary channel information.

[0314] In these and other cases, the audio encoding device 20 is further configured to specify the first non-zero set of vector coefficients in non-audio-encoded auxiliary channel information of the first non-zero set of vector coefficients .

[0315] In these and other cases, the vector comprises a vector decomposed from the plurality of spherical harmonic coefficients using vector-based synthesis.

[0316] In these and other cases, vector-based synthesis comprises a singular value decomposition.

[0317] In these and other cases, the vector comprises a vector V decomposed from the plurality of spherical harmonic coefficients using singular value decomposition.

[0318] In these and other cases, the audio encoding device 20 is further configured to select one of a plurality of configuration modes through which to specify the non-zero set of vector coefficients and to specify the non-zero set of the vectors. vector coefficients based on selected from the plurality of configuration modes.

[0319] In these and other cases, the one of the plurality of configuration modes indicates that the non-zero set of coefficients includes all coefficients.

[0320] In these and other cases, the one among the plurality of configuration modes indicates that the non-zero set of coefficients includes those among the coefficients that correspond to an order greater than an order of a basis function to which a or more of the plurality of spherical harmonic coefficients corresponds.

[0321] In these and other cases, the one of the plurality of configuration modes indicates that the non-zero set of coefficients includes those among the coefficients that correspond to an order greater than an order of a basis function to which a or more of the plurality of spherical harmonic coefficients corresponds to and excludes at least one of the coefficients that correspond to an order greater than the order of the basis function to which the one or more of the plurality of spherical harmonic coefficients corresponds.

[0322] In these and other cases, the one of the plurality of configuration modes indicates that the non-zero set of coefficients includes all coefficients with the exception of at least one of the coefficients.

[0323] In these and other cases, the audio encoding device 20 is further configured to specify selected from among a plurality of configuration modes in a bit stream.

[0324] Various aspects of the techniques described in the present disclosure also allow the audio encoding device 20 to be configured to represent such a distinct component of the sound field in various ways. In these and other cases, the audio encoding device 20 is configured to obtain a first set of non-zero coefficients of a vector representing a distinct component of a sound field, wherein the vector has been decomposed from a plurality of spherical harmonic coefficients that describe the sound field.

[0325] In these and other cases, the first non-zero set of coefficients includes all coefficients in the vector.

[0326] In these and other cases, the first non-zero set of coefficients includes those among the coefficients that correspond to an order greater than an order of a basis function, to which one or more of the plurality of spherical harmonic coefficients matches.

[0327] In these and other cases, the first non-zero set of coefficients includes those among the coefficients that correspond to an order greater than an order of a basis function to which one or more of the plurality of spherical harmonic coefficients corresponds and excluding at least one of the coefficients that correspond to an order greater than the order of the basis function to which the one or more of the plurality of spherical harmonic coefficients correspond.

[0328] In these and other cases, the first non-zero set of coefficients includes all coefficients with the exception of at least one of the coefficients identified for not having sufficient directional information.

[0329] In these and other cases, the audio encoding device 20 is further configured to extract the first non-zero set of coefficients as a first portion of the vector.

[0330] In these and other cases, the audio encoding device 20 is further configured to extract the first set of the non-zero vector from the auxiliary channel information and obtain a recomposed version of the plurality of spherical harmonic coefficients based on the first nonzero set of vector coefficients.

[0331] In these and other cases, the vector comprises a vector decomposed from the plurality of spherical harmonic coefficients using vector-based synthesis.

[0332] In these and other cases, vector-based synthesis comprises a singular value decomposition.

[0333] In these and other cases, the audio encoding device 20 is further configured to determine one of a plurality of configuration modes through which to extract the non-zero set of vector coefficients according to that of the plurality of configuration modes and extract the non-zero set of vector coefficients based on that obtained from among the plurality of configuration modes.

[0334] In these and other cases, the one of the plurality of configuration modes indicates that the non-zero set of coefficients includes all coefficients.

[0335] In these and other cases, the one among the plurality of configuration modes indicates that the non-zero set of coefficients includes those among the coefficients that correspond to an order greater than an order of a basis function to which a or more of the plurality of spherical harmonic coefficients corresponds.

[0336] In these and other cases, the one of the plurality of configuration modes indicates that the non-zero set of coefficients includes those among the coefficients that correspond to an order greater than an order of a basis function to which a or more of the plurality of spherical harmonic coefficients corresponds to and excludes at least one of the coefficients that correspond to an order greater than the order of the basis function to which the one or more of the plurality of spherical harmonic coefficients corresponds.

[0337] In these and other cases, the one of the plurality of configuration modes indicates that the non-zero set of coefficients includes all coefficients with the exception of at least one of the coefficients.

[0338] In these and other cases, the audio encoding device 20 is configured to determine one of a plurality of configuration modes based on a signaled value in a bit stream.

[0339] Various aspects of the techniques may also, in some cases, enable the audio encoding device 20 to identify one or more distinct audio objects (or, in other words, predominant audio objects). In some instances, the audio encoding device 20 may be configured to identify one or more distinct audio objects from one or more spherical harmonic coefficients (SHC) associated with the audio objects based on a given directionality to one or more of the audio objects.

[0340] In these and other cases, the audio encoding device 20 is further configured to determine the directionality of the one or more audio objects based on the spherical harmonic coefficients associated with the audio objects.

[0341] In these and other cases, the audio encoding device 20 is further configured to perform a singular value decomposition with respect to spherical harmonic coefficients to generate a matrix U representative of left singular vectors of the plurality of spherical harmonic coefficients, a matrix S representative of singular values of the plurality of spherical harmonic coefficients and a matrix V representative of singular vectors to the right of the plurality of spherical harmonic coefficients and representing the plurality of spherical harmonic coefficients as a function of at least a portion of one or more of the U matrix, the S matrix and the V matrix, wherein the audio encoding device 20 is configured to determine the respective directionality of the one or more audio objects is based, at least in part, on the V matrix.

[0342] In these and other cases, the audio encoding device 20 is additionally configured to reorder one or more vectors of the matrix V, so that vectors that have a higher directionality quotient are positioned above vectors that have a lower directionality quotient in the reordered V matrix.

[0343] In these and other cases, the audio encoding device 20 is further configured to determine that vectors that have the highest directionality quotient include more directional information than vectors that have less directionality quotient.

[0344] In these and other cases, the audio encoding device 20 is further configured to multiply the matrix V by the matrix S to generate a matrix VS, where the matrix VS includes one or more vectors.

[0345] In these and other cases, audio encoding device 20 is additionally configured to select inputs from each row of the VS matrix that are associated with an order greater than 14, square each of the selected inputs to form inputs corresponding squared and for each row of the VS matrix, sum all squared entries to determine a directionality quotient for a corresponding vector.

[0346] In these and other cases, the audio encoding device 20 is configured to select inputs from each row of the VS matrix associated with the order greater than 14, where this comprises selecting all inputs starting at the 18th input of each row of the VS matrix and ending in the 38th entry of each row of the VS matrix.

[0347] In these and other cases, the audio encoding device 20 is further configured to select a subset of the VS matrix vectors to represent the distinct audio objects. In these and other cases, the audio encoding device 20 is configured to select four vectors from the VS matrix and wherein the four selected vectors have the four largest directionality quotients among all vectors from the VS matrix.

[0348] In these and other cases, the audio encoding device 20 is configured to determine that the selected subset of vectors representing the distinct audio objects is based on both the directionality and an energy of each vector.

[0349] In these and other cases, the audio encoding device 20 is further configured to perform an energy comparison between one or more first vectors and one or more second vectors representative of distinct audio objects to determine the one or more first vectors. reordered vectors, where the one or more first vectors describe the discrete audio objects in a first portion of the audio data and the one or more second vectors describe the discrete audio objects in a second portion of the audio data.

[0350] In these and other cases, the audio encoding device 20 is further configured to perform a cross-relationship between one or more first vectors and one or more second vectors representative of distinct audio objects to determine the one or more first vectors reordered, wherein the one or more first vectors describe the discrete audio objects in a first portion of the audio data and the one or more second vectors describe the discrete audio objects in a second portion of the audio data.

[0351] Various aspects of the techniques may also, in some cases, enable the audio coding device 20 to be configured to perform power compensation with respect to decompositions of HOA coefficients 11. In these and other cases, the audio coding device 11 audio coding 20 may be configured to perform vector-based synthesis against a plurality of spherical harmonic coefficients to generate decomposed representations of the plurality of spherical harmonic coefficients representative of one or more audio objects and corresponding directional information, wherein the spherical harmonic coefficients are associated with an order and describe a sound field, determine distinct directional and background information from the directional information, reduce an order of directional information associated with background audio objects to generate transformed background directional information , apply compensation They are used to increase the transformed directional information values to preserve an overall sound field energy.

[0352] In these and other cases, the audio encoding device 20 can be configured to perform a singular value decomposition with respect to a plurality of spherical harmonic coefficients to generate a matrix U and a matrix S representative of the audio objects and a matrix V representative of the directional information, determine distinct column vectors from matrix V and background column vectors from matrix V, reduce an order of background column vectors from matrix V to generate transformed background column vectors from matrix V and apply compensation to increase the values of the transformed background column vectors of the matrix V to preserve an overall sound field energy.

[0353] In these and other cases, the audio coding device 20 is additionally configured to determine the amount of salient singular values of the matrix S, in which the amount of vectors of distinct columns of the matrix V is the amount of salient singular values of the matrix V. matrix S.

[0354] In these and other cases, the audio encoding device 20 is configured to determine a reduced order for the spherical harmonic coefficients and zero values for the rows of the matrix V background column vectors associated with an order that is larger than the reduced order.

[0355] In these and other cases, the audio encoding device 20 is additionally configured to combine U matrix background columns, S matrix background columns, and a transpose of the transformed matrix V background columns to generate modified spherical harmonic coefficients.

[0356] In these and other cases, the modified spherical harmonic coefficients describe one or more background components of the sound field.

[0357] In these and other cases, the audio encoding device 20 is configured to determine a first energy of a vector among the background column vectors of the matrix V and a second energy of a vector among the column vectors of background transforms of the matrix V and applying an amplification value to each element of the vector among the transformed background column vectors of the matrix V, wherein the amplification value comprises a ratio of the first energy to the second energy.

[0358] In these and other cases, the audio encoding device 20 is configured to determine a first squared average root energy of a vector among the background column vectors of matrix V and a second square average of energy root of a vector among the transformed background column vectors of the matrix V and applying an amplification value to each element of the vector among the transformed background column vectors of the matrix V, wherein the amplification value comprises a ratio of the first energy to the second energy.

[0359] Various aspects of the techniques described in the present disclosure also enable the audio encoding device 20 to perform an interpolation against decomposed versions of the HOA coefficients 11. In some instances, the audio encoding device 20 may be configured to obtaining interpolated spherical harmonic coefficients decomposed by a time segment, at least in part, by performing an interpolation with respect to a first decomposition of a first plurality of spherical harmonic coefficients and a second decomposition of a second plurality of spherical harmonic coefficients.

[0360] In these and other cases, the first decomposition comprises a first matrix V representative of singular vectors to the right of the first plurality of spherical harmonic coefficients.

[0361] In these and other examples, the second decomposition comprises a second matrix V representative of singular vectors to the right of the second plurality of spherical harmonic coefficients.

[0362] In these and other cases, the first decomposition comprises a first matrix V representative of singular vectors to the right of the first plurality of spherical harmonic coefficients and the second decomposition comprises a second matrix V representative of singular vectors to the right of the second plurality of coefficients spherical harmonics.

[0363] In these and other cases, the time segment comprises a subframe of an audio frame.

[0364] In these and other cases, the time segment comprises a time sample of an audio frame.

[0365] In these and other cases, the audio encoding device 20 is configured to obtain an interpolated decomposition of the first decomposition and the second decomposition to a spherical harmonic coefficient of the first plurality of spherical harmonic coefficients.

[0366] In these and other cases, the audio encoding device 20 is configured to obtain interpolated decompositions from the first decomposition to a first portion of the first plurality of spherical harmonic coefficients included in the first frame and the second decomposition to a second portion of the second plurality of spherical harmonic coefficients included in the second frame, and the audio encoding device 20 is further configured to apply the interpolated decays to a first time component of the first portion of the first plurality of spherical harmonic coefficients included in the first frame to generating a first artificial time component of the first plurality of spherical harmonic coefficients, and applying respective interpolated decompositions to a second time component of the second portion from among the second plurality of spherical harmonic coefficients included in the second frame to generate a second artificial time component of the second plurality of included spherical harmonic coefficients.

[0367] In these and other cases, the first time component is generated by performing a vector-based synthesis with respect to the first plurality of spherical harmonic coefficients.

[0368] In these and other cases, the second time component is generated by performing a vector-based synthesis with respect to the second plurality of spherical harmonic coefficients.

[0369] In these and other cases, the audio encoding device 20 is further configured to receive the first artificial time component and the second artificial time component, compute the interpolated decompositions from the first decomposition to the first portion of the first plurality of spherical harmonic coefficients and the second decomposition to the second portion of the second plurality of spherical harmonic coefficients, and applying inverses of the interpolated decompositions to the first artificial time component to recover the first time component and to the second artificial time component to recover the second time component.

[0370] In these and other cases, the audio encoding device 20 is configured to interpolate a first spatial component of the first plurality of spherical harmonic coefficients and the second spatial component of the second plurality of spherical harmonic coefficients.

[0371] In these and other cases, the first spatial component comprises a first matrix U representative of singular vectors to the left of the first plurality of spherical harmonic coefficients.

[0372] In these and other cases, the second spatial component comprises a second matrix U representative of singular vectors to the left of the second plurality of spherical harmonic coefficients.

[0373] In these and other cases, the first spatial component is representative of M time segments of spherical harmonic coefficients for the first plurality of spherical harmonic coefficients and the second spatial component is representative of M time segments of spherical harmonic coefficients spherical to the second plurality of spherical harmonic coefficients.

[0374] In these and other cases, the first spatial component is representative of M time segments of spherical harmonic coefficients for the first plurality of spherical harmonic coefficients and the second spatial component is representative of M time segments of spherical harmonic coefficients spherical to the second plurality of spherical harmonic coefficients, and the audio encoding device 20 is configured to interpolate the last N elements of the first spatial component and the first N elements of the second spatial component.

[0375] In these and other cases, the second plurality of spherical harmonic coefficients is subsequent to the first plurality of spherical harmonic coefficients in the time domain.

[0376] In these and other cases, the audio encoding device 20 is further configured to decompose the first plurality of spherical harmonic coefficients to generate the first decomposition of the first plurality of spherical harmonic coefficients.

[0377] In these and other cases, the audio encoding device 20 is further configured to decompose the second plurality of spherical harmonic coefficients to generate the second decomposition of the second plurality of spherical harmonic coefficients.

[0378] In these and other cases, the audio encoding device 20 is further configured to perform a singular value decomposition with respect to the first plurality of spherical harmonic coefficients to generate a matrix U representative of left singular vectors of the first plurality of spherical harmonic coefficients, a matrix S representative of singular values of the first plurality of spherical harmonic coefficients and a matrix V representative of singular vectors to the right of the first plurality of spherical harmonic coefficients.

[0379] In these and other cases, the audio encoding device 20 is further configured to perform a singular value decomposition with respect to the second plurality of spherical harmonic coefficients to generate a matrix U representative of left singular vectors of the second plurality of spherical harmonic coefficients, a matrix S representative of singular values of the second plurality of spherical harmonic coefficients and a matrix V representative of singular vectors to the right of the second plurality of spherical harmonic coefficients.

[0380] In these and other cases, each of the first and second pluralities of spherical harmonic coefficients represents a plane wave representation of the sound field.

[0381] In these and other cases, each of the first and second pluralities of spherical harmonic coefficients represents one or more mono audio objects mixed together.

[0382] In these and other cases, each of the first and second pluralities of spherical harmonic coefficients comprises respective first and second spherical harmonic coefficients representing a three-dimensional sound field.

[0383] In these and other cases, the first and second pluralities of spherical harmonic coefficients are each associated with at least one spherical basis function that has an order greater than one(1).

[0384] In these and other cases, the first and second pluralities of spherical harmonic coefficients are each associated with at least one spherical basis function that has an order equal to four.

[0385] In these and other cases, the interpolation is a weighted interpolation of the first decomposition and the second decomposition, where the weighted interpolation weights applied to the first decomposition are inversely proportional to a time represented by vectors of the first and second decompositions and in that the weighted interpolation weights applied to the second decomposition are proportional to a time represented by vectors of the first and second decompositions.

[0386] In these and other cases, the decomposed interpolated spherical harmonic coefficients smooth at least one of the spatial components and the time components of the first plurality of spherical harmonics coefficients and the second plurality of spherical harmonics coefficients.

[0387] In these and other cases, audio encoding device 20 is configured to compute Us[n] = HOA(n) * (V_vec[n])-1 to obtain a scalar.

[0388] In these and other cases, the interpolation comprises a linear interpolation. In these and other cases, the interpolation comprises a non-linear interpolation. In these and other cases, the interpolation comprises a cosine interpolation. In these and other cases, the interpolation comprises a weighted cosine interpolation. In these and other cases, the interpolation comprises a cubic interpolation. In these and other cases, the interpolation comprises an Adaptive Spline Interpolation. In these and other cases, the interpolation comprises a minimum curvature interpolation.

[0389] In these and other cases, the audio encoding device 20 is further configured to generate a bit stream that includes a representation of the decomposed interpolated spherical harmonic coefficients for the time segment, and an indication of a type of the interpolation. .

[0390] In these and other cases, the indication comprises one or more bits that map to the interpolation type.

[0391] In this way, various aspects of the sets of procedures described in this disclosure can enable the audio encoding device 20 to be configured to obtain a bit stream that includes a representation of the interpolated spherical harmonic coefficients decomposed for the time segment, and an indication of an interpolation type.

[0392] In these and other cases, the indication comprises one or more bits that map to the interpolation type.

[0393] In this regard, the audio encoding device 20 may represent an embodiment of the sets of procedures in which the audio encoding device 20 may, in some cases, be configured to generate a bit stream comprising a version compression of a spatial component of a sound field, the spatial component being generated by performing vector-based synthesis with respect to a plurality of spherical harmonic coefficients.

[0394] In these and other cases, the audio encoding device 20 is further configured to generate the bit stream to include a field that specifies a prediction mode used when compressing the spatial component.

[0395] In these and other cases, the audio encoding device 20 is configured to generate the bit stream to include Huffman table information that specifies a Huffman table used when compressing the spatial component.

[0396] In these and other cases, the audio encoding device 20 is configured to generate the bit stream to include a field that indicates a value that expresses a quantization step size or a variable thereof used when compressing the spatial component.

[0397] In these and other cases, the value comprises an nbits value.

[0398] In these and other cases, the audio encoding device 20 is configured to generate the bit stream to include a version of a plurality of spatial components of the sound field whose compressed version of the spatial component is included, where the value expresses the quantization step size or a variable thereof used when compressing the plurality of spatial components.

[0399] In these and other cases, the audio encoding device 20 is further configured to generate the bit stream to include a Huffman code to represent a category identifier that identifies a compression category to which the spatial component corresponds.

[0400] In these and other cases, the audio encoding device 20 is configured to generate the bit stream to include a sign bit that identifies whether the spatial component is a positive value or a negative value.

[0401] In these and other cases, the audio encoding device 20 is configured to generate the bit stream to include a Huffman code to represent a residual value of the spatial component.

[0402] In these and other cases, vector-based synthesis comprises a singular value decomposition.

[0403] In this regard, the audio encoding device 20 may additionally implement various aspects of sets of procedures where the audio encoding device 20 may, in some cases, be configured to identify a Huffman codebook for use when compressing a spatial component of a plurality of spatial components based on an order of the spatial component related to the remaining components among the plurality of spatial components, where the spatial component is generated by performing a vector-based synthesis with respect to to a plurality of spherical harmonic coefficients.

[0404] In these and other cases, the audio encoding device 20 is configured to identify the Huffman codebook based on a prediction mode used when compressing the spatial component.

[0405] In these and other cases, a compressed version of the spatial component is represented in a bitstream using, at least in part, Huffman table information that identifies the Huffman codebook.

[0406] In these and other cases, a compressed version of the spatial component is represented in a bitstream using, at least in part, a field that indicates a value that expresses a quantization step size or a variable of the same used when compressing the spatial component.

[0407] In these and other cases, the value comprises an nbits value.

[0408] In these and other cases, the bitstream comprises a compressed version of a plurality of spatial components of the sound field whose compressed version of the spatial component is included, and the value expresses the quantization step size or a variable thereof used when compressing the plurality of spatial components.

[0409] In these and other cases, a compressed version of the spatial component is represented in a bitstream using, at least in part, a Huffman code selected from the identified Huffman codebook to represent an identifier category that identifies a compression category to which the spatial component corresponds.

[0410] In these and other cases, a compressed version of the spatial component is represented in a bit stream using, at least in part, a sign bit that identifies whether the spatial component is a positive value or a negative value. .

[0411] In these and other cases, a compressed version of the spatial component is represented in a bitstream using, at least in part, a Huffman code selected from the identified Huffman codebook to represent a value residual of the spatial component.

[0412] In these and other cases, the audio encoding device 20 is further configured to compress the spatial component based on the identified Huffman codebook to generate a compressed version of the spatial component, and generate the bit stream to include the compressed version of the spatial component.

[0413] Furthermore, the audio encoding device 20 may, in some cases, implement various aspects of the sets of procedures where the audio encoding device 20 may be configured to determine a quantization step size to be used when compresses a spatial component of a sound field, the spatial component being generated by performing vector-based synthesis with respect to a plurality of spherical harmonic coefficients.

[0414] In these and other cases, the audio encoding device 20 is further configured to determine the quantization step size based on a target bitrate.

[0415] In these and other cases, the audio encoding device 20 is configured to determine an estimate of a number used to represent the spatial component, and to determine the quantization step size based on a difference between the estimate and a target bitrate.

[0416] In these and other cases, the audio encoding device 20 is configured to determine an estimate of a number of bits used to represent the spatial component, determine a difference between the estimate and a target bit rate, and determine the quantization step size by adding the difference to the target bitrate.

[0417] In these and other cases, the audio encoding device 20 is configured to calculate the estimate of the number of bits that should be generated for the spatial component given a codebook corresponding to the target bit rate.

[0418] In these and other cases, the audio encoding device 20 is configured to calculate the estimate of the number of bits that should be generated for the spatial component given an encoding mode used when compressing the spatial component.

[0419] In these and other cases, the audio encoding device 20 is configured to calculate a first estimate of the number of bits that should be generated for the spatial component given a first encoding mode to be used when compressing the spatial component , calculate a second estimate of the number of bits that should be generated for the spatial component given a second encoding mode to be used when compressing the spatial component, select one of the first estimate and the second estimate that has the fewest bits to be used as the given estimate of the number of bits.

[0420] In these and other cases, the audio encoding device 20 is configured to identify a category identifier that identifies a category to which the spatial component corresponds, identifies a bit length of a residual value for the spatial component that would result when compressing the spatial component corresponding to the category, and determining the estimate of the number of bits, at least in part, by adding a number of bits used to represent the category identifier to the bit length of the residual value.

[0421] In these and other cases, the audio encoding device 20 is further configured to select one of a plurality of codebooks to be used when compressing the spatial component.

[0422] In these and other cases, the audio encoding device 20 is further configured to determine an estimate of a number of bits used to represent the spatial component using each of the plurality of codebooks, and select the book among the plurality of codebooks that resulted in the given estimate having the fewest bits.

[0423] In these and other cases, the audio encoding device 20 is further configured to determine an estimate of a number of bits used to represent the spatial component using one or more of the plurality of codebooks, in that the one or more of the plurality of codebooks selected based on an order of the spatial component elements to be compressed relative to other spatial component elements.

[0424] In these and other cases, the audio encoding device 20 is further configured to determine an estimate of a number of bits used to represent the spatial component using one of the plurality of codebooks designed to be used when the spatial component is not predicted from a subsequent spatial component.

[0425] In these and other cases, the audio encoding device 20 is further configured to determine an estimate of a number of bits used to represent the spatial component using one of the plurality of codebooks designed to be used when the spatial component is predicted from a subsequent spatial component.

[0426] In these and other cases, the audio encoding device 20 is further configured to determine an estimate of a number of bits used to represent the spatial component using the plurality of codebooks designed to be used when the component space is representative of a synthetic audio object in the sound field.

[0427] In these and other cases, the synthetic audio object comprises a pulse code modulated (PCM) audio object.

[0428] In these and other cases, the audio encoding device 20 is further configured to determine an estimate of a number of bits used to represent the spatial component using the plurality of codebooks designed to be used when the component space is representative of an audio object recorded in the sound field.

[0429] In each of the various cases described above, it should be understood that the audio encoding device 20 may perform a method or otherwise comprises means for performing each step of the method for which the audio encoding device 20 is configured to perform in some cases, these means may comprise one or more processors. In some cases, the one or more processors may represent a special-purpose processor configured through instructions stored on a non-transient computer-readable storage medium. In other words, various aspects of the sets of procedures in each of the sets of coding examples can provide a non-transient computer-readable storage medium that has stored instructions in it that, when executed, cause the one or more processors to perform the method for which the audio encoding device 20 has been configured to perform.

[0430] Figure 5 is a block diagram illustrating the audio decoding device 24 of Figure 3 in more detail. As shown in the example of Figure 5, the audio decoding device 24 may include an extraction unit 72, a directionality-based reconstruction unit 90 and a vector-based reconstruction unit 92.

[0431] Extractor unit 72 may represent a unit configured to receive bitstream 21 and extract the various encoded versions (e.g. a directionality-encoded version or a vector-based encoded version) of the HOA coefficients 11. The extraction unit 72 can determine from the syntax element noted above (e.g., the ChannelType syntax element shown in the examples of Figures 10E and 10H(i)-100(ii)) whether the HOA coefficients 11 were coded through the various versions. When directionality-based encoding has been performed, the extraction unit 72 can extract the directionality-based version of the HOA coefficients 11 and the syntax elements associated with that encoded version (which is denoted as directionality-based information 91 in example in Figure 5), which passes this directionality-based information 91 to the directionality-based reconstruction unit 90. This directionality-based reconstruction unit 90 can represent the unit configured to reconstruct the HOA coefficients in the form of  coefficients. HOA 11' based on directionality-based information 91. The bit stream and arrangement of syntax elements within the bit stream are described below in greater detail with respect to the example of Figures 10-100(ii) and 11 .

[0432] When the syntax element indicates that the HOA coefficients 11 were encoded using vector-based synthesis, the extraction unit 72 can extract the encoded foreground V[k] vectors 57, the HOA coefficients of coded environment 59 and the coded nFG signals 59. The extraction unit 72 may pass the coded foreground vectors V[k] 57 to the quantization unit 74 and the coded environment HOA coefficients 59 together with the coded environment signals 59. nFG coded 61 to psychoacoustic decoding unit 80.

[0433] To extract the coded foreground vectors V[k] 57, the coded ambient HOA coefficients 59 and the coded nFG signals 59, the extraction unit 72 can obtain the auxiliary channel information 57, which includes the syntax element denoted codedVVecLength. The extraction unit 72 can parse the codedVVecLength from the auxiliary channel information 57. The extraction unit 72 can be configured to operate in one of the configuration modes described above based on the codedVVecLength syntax element.

[0434] Extractor unit 72 then operates in any of the configuration modes to analyze a compressed form of reduced foreground vectors V[k] 55k from auxiliary channel information 57. Extractor unit 72 can operate according to the switch statement presented in the following pseudo-code with the syntax presented in the following syntax table for VectorData:

[0435] In the syntax table above, the first switch statement with the four cases (case 0-3) provides a way by which to determine the length of the VTDIST array with respect to number (WecLength) and coefficient indices (VVecCoeffld ). The first case, case 0, indicates that all coefficients for the VTDIST vectors (NumOfHoaCoeffs) are specified. The second case, case 1, indicates that only those coefficients of the VTDIST vector corresponding to the number greater than a MmNumOfCoeffsForAmbHOA are specified, which can denote what is called (NDIST + 1)2 - (NBG + 1)2 above. Additionally the coefficients of NumOfContAddAmbHoaChan identified in ContAddAmbHoaChan are subtracted. The ContAddAmbHoaChan list specifies additional channels (where "channels" refers to a particular coefficient corresponding to a given order, suborder combination) corresponding to an order that exceeds the MinAmbHoaOrder order. The third case, case 2, indicates that those coefficients of the VTDIST vector corresponding to the number greater than a MmNumOfCoeffsForAmbHOA are specified, which can denote what is called (NDIST + 1)2 - (NBG + 1)2 above. The fourth case, case 3, indicates that those coefficients of the VTDIST vector remaining after removing the coefficients identified by NumOfContAddAmbHoaChan are specified. Both VVecLength and the VVecCoeffld list are valid for all VVectors within HOAFrame.

[0436] After this switch instruction, the decision whether to perform uniform dequantization can be controlled by NbitsQ (or, as denoted above, nbits), which if equal to 5, a uniform 8-bit scalar dequantization is performed. In contrast, an NbitsQ value greater than or equal to 6 may result in Huffman decoding being applied. The value of cid referred to above can be equal to the two least significant bits of the value of NbitsQ. The prediction mode discussed above is denoted as PFlag in the syntax table above, while the HT info bit is denoted as CbFlag in the syntax table above. The remaining syntax specifies how decoding takes place substantially similar to that described above. Various examples of bitstream 21 conforming to each of the above various cases are described in greater detail below with reference to Figures 10H(i)-100(ii).

[0437] The vector-based reconstruction unit 92 represents a unit configured to perform reciprocal operations to those described above with respect to the vector-based synthesis unit 27 for the purpose of reconstructing the HOA coefficients 11'. The vector-based reconstruction unit 92 may include a quantization unit 74, a spatio-temporal interpolation unit 76, a foreground formulation unit 78, a psychoacoustic decoding unit 80, an HOA coefficient formulation unit. 82 and a reordering unit 84.

[0438] The quantization unit 74 can represent a unit configured to operate reciprocally to the quantization unit 52 shown in the example of Figure 4 in order to dequantify the coded foreground vectors V[k] 57 and thus generate vectors of reduced foreground V[k] 55k. The dequantization unit 74 may, in some instances, perform a form of entropy decoding and scalar dequantization reciprocally to that described above with respect to the quantization unit 52. The dequantization unit 74 may route the reduced foreground vectors V[ k] 55k for reorder unit 84.

[0439] The psychoacoustic decoding unit 80 may reciprocally operate with the psychoacoustic audio encoding unit 40 shown in the example of Figure 4 for the purpose of decoding the coded ambient HOA coefficients 59 and the coded nFG signals 61 and generate thus energy compensated ambient HOA coefficients 47' and interpolated nFG signals 49' (which may also be called interpolated nFG audio objects 49'). Psychoacoustic decoding unit 80 may pass energy compensated ambient HOA coefficients 47' to HOA coefficient formulation unit 82 and nFG signals 49' to reordering 84.

[0440] Reordering unit 84 may represent a unit configured to operate in a reciprocal manner similar to that described above with respect to reordering unit 34. Reordering unit 84 may receive syntax elements indicative of the original order of foreground components of the HOA coefficients 11. The reordering unit 84 can, based on these reordering syntax elements, rearrange the interpolated nFG signals 49' and the reduced foreground vectors 55A V[k] to generate 49" reordered nFG signals and reordered foreground 55k'V[k] vectors. The reordering unit 84 may output the reordered nFG signals 49" to the foreground formulation unit 78 and the reordered foreground vectors 55k' V[k] to the spatiotemporal interpolation unit 76.

[0441] The spatiotemporal interpolation unit 76 may operate in a similar manner to that described above with respect to the spatiotemporal interpolation unit 50. The spatiotemporal interpolation unit 76 may receive the vectors 55k' V[k] of reordered foreground vectors and perform spatiotemporal interpolation against the reordered foreground 55k' V[k] and reordered foreground 55k-i' V[k-1] vectors to generate 55k'” V[k] interpolated foreground. The spatiotemporal interpolation unit 76 may transmit the interpolated foreground vectors 55k” V[k] to the foreground formulation unit 78.

[0442] The foreground formulation unit 78 may represent a unit configured to perform matrix multiplication with respect to interpolated foreground 55k"V[k] vectors and 49" nFG signals reordered to generate the first HOA coefficients plane 65. The foreground formulation unit 78 may perform a matrix multiplication of the reordered nFG signals 49" by the interpolated foreground vectors V[k] 55k".

[0443] The HOA coefficient formulation unit 82 may represent a unit configured to add the foreground HOA coefficients 65 to the ambient HOA channels 47' for the purpose of obtaining the HOA coefficients 11', where The application reflects that these HOA 11' coefficients may be similar to, but not equal to, the HOA 11 coefficients. Differences between the 11 and 11' HOA coefficients may result from loss due to transmission over a lossy transmission medium, quantization or other lossy operations.

[0444] In this way, procedure sets can enable an audio decoding device, such as audio decoding device 24, to determine, from a bit stream, quantized directional information, a foreground audio object encoded, and encoded ambient higher-order ambisonic (HOA) coefficients, where the quantized directional information and the encoded foreground audio object represent foreground HOA coefficients that describe a foreground component of a sound field, and where encoded ambient HOA coefficients describe an ambient component of the sound field, dequantify quantified directional information to generate directional information, perform spatiotemporal interpolation against directional information to generate interpolated directional information, audio decode the object foreground audio encoded to generate an object foreground audio and ambient HOA coefficients encoded to generate ambient HOA coefficients, determine foreground HOA coefficients as a function of interpolated directional information and the foreground audio object, and determine HOA coefficients as a function of the foreground HOA coefficients and the ambient HOA coefficients.

[0445] In this way, various aspects of the procedure sets can enable a unified audio decoding device 24 to switch between two different decompression schemes. In some cases, the audio decoding device 24 may be configured to select one of a plurality of decompression schemes based on an indication of whether a compressed version of spherical harmonic coefficients representative of a sound field are generated from a sound object. synthetic audio, and decompressing the compressed version of the spherical harmonic coefficients using that scheme selected from the plurality of decompression schemes. In these and other cases, the audio decoding device 24 comprises an integrated decoder.

[0446] In some cases, the audio decoding device 24 may be configured to obtain an indication of whether representative spherical harmonic coefficients of a sound field are generated from a synthetic audio object.

[0447] In these and other cases, the audio decoding device 24 is configured to obtain the indication from a bit stream that stores a compressed version of the spherical harmonic coefficients.

[0448] In this way, various aspects of the procedure sets can enable the audio decoding device 24 to obtain vectors that describe background and distinct components of the sound field. In some cases, the audio decoding device 24 may be configured to determine one or more first vectors that describe distinct components of a sound field and one or more second vectors that describe the background components of the sound field, either one or more more first vectors than the one or more second vectors generated by at least performing a transformation with respect to a plurality of spherical harmonic coefficients.

[0449] In these and other cases, the audio decoding device 24, wherein the transformation comprises a singular value decomposition that generates a matrix U representative of singular vectors to the left of the plurality of spherical harmonic coefficients, a matrix S representative of singular values of the plurality of spherical harmonic coefficients and a matrix V representative of singular vectors to the right of the plurality of spherical harmonic coefficients.

[0450] In these and other cases, the audio decoding device 24, where the first one or more vectors comprise one or more audio-encoded UDIST * SDIST vectors that, prior to audio encoding, were generated by multiplying a or more audio-encoded UDIST vectors of a matrix U by one or more SDIST vectors of a matrix S, and wherein the matrix U and matrix S are generated by at least performing singular value decomposition with respect to the plurality of coefficients of spherical harmonics.

[0451] In these and other cases, the audio decoding device 24 is additionally configured to audio-decode the one or more audio-encoded UDIST*SDIST arrays to generate an audio-decoded version of the one or more encoded UDIST*SDIST arrays by audio.

[0452] In these and other cases, the audio decoding device 24, where the first one or more vectors comprise one or more audio-encoded UDIST * SDIST vectors that, prior to audio encoding, were generated by multiplying a or more audio-encoded UDIST vectors of a matrix U by one or more SDIST vectors of a matrix S, and one or more VTDIST vectors of a transpose of a matrix V, and where the matrix U and the matrix S and the matrix V are generated at least by performing the singular value decomposition with respect to the plurality of spherical harmonic coefficients.

[0453] In these and other cases, the audio decoding device 24 is additionally configured to audio-decode the one or more audio-encoded UDIST*SDIST arrays to generate an audio-decoded version of the one or more encoded UDIST*SDIST arrays by audio.

[0454] In these and other cases, the audio decoding device 24 is further configured to multiply the UDIST * SDIST vectors by the VTDIST vectors to recover those from among the plurality of spherical harmonics representative of the distinct components of the sound field.

[0455] In these and other cases, the audio decoding device 24, where the one or more second vectors comprise one or more audio-encoded UBG * SBG * VTBG vectors which, prior to audio encoding, were generating by multiplying them. if vectors UBG enclosed within a matrix U by vectors SBG enclosed within a matrix S and then by vectors VTBG enclosed within a transpose of a matrix V, and where matrix S, matrix U and matrix V were each one, generated at least by performing the singular value decomposition with respect to the plurality of spherical harmonic coefficients.

[0456] In these and other cases, the audio decoding device 24, where the one or more second vectors comprise one or more audio-encoded UBG * SBG * VTBG vectors which, prior to audio encoding, were generating by multiplying them. if vectors UBG included within a matrix U by vectors SBG included within a matrix S and then by vectors VTBG included within a transpose of a matrix V, where matrix S, matrix U and matrix V have been generated at least performing singular value decomposition with respect to the plurality of spherical harmonic coefficients, and wherein the audio decoding device 24 is further configured to audio decode the one or more audio-encoded UBG * SBG * VTBG vectors to generate one or more audio-decoded UBG * SBG * VTBG vectors.

[0457] In these and other cases, the audio decoding device 24, where the first one or more vectors comprise one or more audio-encoded UDIST * SDIST vectors which, prior to audio encoding, were generated by multiplying a or more audio-encoded UDIST vectors of a matrix U by one or more SDIST vectors of a matrix S, and one or more VTDIST vectors of a transpose of a matrix V, where the matrix U, matrix S, and matrix V have been generated by at least performing singular value decomposition with respect to the plurality of spherical harmonic coefficients, and wherein the audio decoding device 24 is further configured to audio decode the one or more audio-encoded UDIST * SDIST vectors to generate the one or more UDIST * SDIST vectors, and multiply the UDIST * SDIST vectors by the VTDIST vectors to retrieve those of the plurality of spherical harmonic coefficients that describe the distinct components of the sound field, wherein the one or more second vectors comprise one or more audio-encoded UBG * SBG * VTBG vectors that, prior to audio encoding, were generated by multiplying UBG vectors included in the matrix U by SBG vectors included in the matrix S and then by VTBG vectors included in the transpose of matrix V, and wherein the audio decoding device 24 is further configured to audio decode the one or more audio-encoded UBG * SBG * VTBG vectors to recover at least a portion of the plurality of spherical harmonic coefficients that describe background components of the sound field, and adding the plurality of spherical harmonic coefficients that describe the distinct components of the sound field to the at least portion of the plurality of spherical harmonic coefficients that describe the components of background of the sound field to generate a reconstructed version of the plurality of spherical harmonic coefficients you.

[0458] In these and other cases, audio decoding device 24, where the first one or more vectors comprise one or more UDIST*SDIST vectors which, prior to audio encoding, were generated by multiplying one or more vectors audio-encoded UDIST of a matrix U by one or more SDIST vectors of a matrix S, and one or more vectors VTDIST of a transpose of a matrix V, where the matrix U, matrix S, and matrix V have been generated at least performing singular value decomposition with respect to the plurality of spherical harmonic coefficients, and wherein the audio decoding device 20 is further configured to obtain a D value indicating the number of vectors to be extracted from a bit stream to form the one or more UDIST * SDIST vectors and the one or more DiST vectors.

[0459] In these and other cases, audio decoding device 24, where the first one or more vectors comprise one or more UDIST*SDIS vectors which, prior to audio encoding, were generated by multiplying one or more vectors audio-encoded UDIST of a matrix U by one or more SDIST vectors of a matrix S, and one or more vectors VTDIST of a transpose of a matrix V, where the matrix U, matrix S, and matrix V have been generated at least performing singular value decomposition with respect to the plurality of spherical harmonic coefficients, and wherein the audio decoding device 24 is further configured to obtain a D value on an audio frame per audio frame basis that indicates the number of vectors to extract from a bit stream to form the one or more UDIST * SDIST vectors and the one or more VTDIST vectors.

[0460] In these and other cases, the audio decoding device 24, wherein the transformation comprises a principal component analysis to identify the distinct components of the sound field and the background components of the sound field.

[0461] Various aspects of the sets of procedures described in this disclosure may also enable the audio encoding device 24 to perform interpolation against decomposed versions of the HOA coefficients. In some cases, the audio decoding device 24 may be configured to obtain interpolated spherical harmonic coefficients decomposed by a time segment, at least in part, by performing an interpolation with respect to a first decomposition of a first plurality of coefficients. of spherical harmonics and a second decomposition of a second plurality of spherical harmonics coefficients.

[0462] In these and other cases, the first decomposition comprises a first matrix V representative of singular vectors to the right of the first plurality of spherical harmonic coefficients.

[0463] In these and other examples, the second decomposition comprises a second matrix V representative of singular vectors to the right of the second plurality of spherical harmonic coefficients.

[0464] In these and other cases, the first decomposition comprises a first matrix V representative of singular vectors to the right of the first plurality of spherical harmonic coefficients, and the second decomposition comprises a second matrix V representative of singular vectors to the right of the second plurality of spherical harmonic coefficients.

[0465] In these and other cases, the time segment comprises a subframe of an audio frame.

[0466] In these and other cases, the time segment comprises a time sample of an audio frame.

[0467] In these and other cases, the audio decoding device 24 is configured to obtain an interpolated decomposition of the first decomposition and the second decomposition to a spherical harmonic coefficient of the first plurality of spherical harmonic coefficients.

[0468] In these and other cases, the audio decoding device 24 is configured to obtain interpolated decompositions from the first decomposition to a first portion of the first plurality of spherical harmonic coefficients included in the first frame and the second decomposition to a second portion of the second plurality of spherical harmonic coefficients included in the second frame, and the audio decoding device 24 is further configured to apply the interpolated decays to a first time component of the first portion of the first plurality of spherical harmonic coefficients included in the first frame to generating a first artificial time component of the first plurality of spherical harmonic coefficients, and applying the respective interpolated decompositions to a second time component of the second portion from among the second plurality of spherical harmonic coefficients included in the second frame to generate a second artificial time component of the second plurality of included spherical harmonic coefficients.

[0469] In these and other cases, the first time component is generated by performing a vector-based synthesis with respect to the first plurality of spherical harmonic coefficients.

[0470] In these and other cases, the second time component is generated by performing a vector-based synthesis with respect to the second plurality of spherical harmonic coefficients.

[0471] In these and other cases, the audio decoding device 24 is further configured to receive the first artificial time component and the second artificial time component, compute the interpolated decompositions from the first decomposition to the first portion of the first plurality of spherical harmonic coefficients and the second decomposition to the second portion of the second plurality of spherical harmonic coefficients, and applying inverses of the interpolated decompositions to the first artificial time component to recover the first time component and to the second artificial time component to recover the second time component.

[0472] In these and other cases, the audio decoding device 24 is configured to interpolate a first spatial component of the first plurality of spherical harmonic coefficients and the second spatial component of the second plurality of spherical harmonic coefficients.

[0473] In these and other cases, the first spatial component comprises a first matrix U representative of singular vectors to the left of the first plurality of spherical harmonic coefficients.

[0474] In these and other cases, the second spatial component comprises a second matrix U representative of singular vectors to the left of the second plurality of spherical harmonic coefficients.

[0475] In these and other cases, the first spatial component is representative of M time segments of spherical harmonic coefficients for the first plurality of spherical harmonic coefficients and the second spatial component is representative of M time segments of spherical harmonic coefficients spherical to the second plurality of spherical harmonic coefficients.

[0476] In these and other cases, the first spatial component is representative of M time segments of spherical harmonic coefficients for the first plurality of spherical harmonic coefficients and the second spatial component is representative of M time segments of spherical harmonic coefficients to the second plurality of spherical harmonic coefficients, and the audio decoding device 24 is configured to interpolate the last N elements of the first spatial component and the first N elements of the second spatial component.

[0477] In these and other cases, the second plurality of spherical harmonic coefficients is subsequent to the first plurality of spherical harmonic coefficients in the time domain.

[0478] In these and other cases, the audio decoding device 24 is further configured to decompose the first plurality of spherical harmonic coefficients to generate the first decomposition of the first plurality of spherical harmonic coefficients.

[0479] In these and other cases, the audio decoding device 24 is further configured to decompose the second plurality of spherical harmonic coefficients to generate the second decomposition of the second plurality of spherical harmonic coefficients.

[0480] In these and other cases, the audio decoding device 24 is further configured to perform a singular value decomposition with respect to the first plurality of spherical harmonic coefficients to generate a matrix U representative of left singular vectors of the first plurality of spherical harmonic coefficients, a matrix S representative of singular values of the first plurality of spherical harmonic coefficients and a matrix V representative of singular vectors to the right of the first plurality of spherical harmonic coefficients.

[0481] In these and other cases, the audio decoding device 24 is further configured to perform a singular value decomposition with respect to the second plurality of spherical harmonic coefficients to generate a matrix U representative of left singular vectors of the second plurality of spherical harmonic coefficients, a matrix S representative of singular values of the second plurality of spherical harmonic coefficients and a matrix V representative of singular vectors to the right of the second plurality of spherical harmonic coefficients.

[0482] In these and other cases, each of the first and second pluralities of spherical harmonic coefficients represents a plane wave representation of the sound field.

[0483] In these and other cases, each of the first and second pluralities of spherical harmonic coefficients represents one or more mono audio objects mixed together.

[0484] In these and other cases, each of the first and second pluralities of spherical harmonic coefficients comprises respective first and second spherical harmonic coefficients representing a three-dimensional sound field.

[0485] In these and other cases, the first and second pluralities of spherical harmonic coefficients are each associated with at least one spherical basis function that has an order greater than one(1).

[0486] In these and other cases, the first and second pluralities of spherical harmonic coefficients are each associated with at least one spherical basis function that has an order equal to four.

[0487] In these and other cases, the interpolation is a weighted interpolation of the first decomposition and second decomposition, where the weights of the weighted interpolation applied to the first decomposition are inversely proportional to a time represented by vectors of the first and second decompositions and in which the weights of the weighted interpolation applied to the second decomposition are proportional to a time represented by the vectors of the first and second decompositions.

[0488] In these and other cases, the decomposed interpolated spherical harmonic coefficients smooth at least one of the spatial components and the time components of the first plurality of spherical harmonic coefficients and the second plurality of spherical harmonic coefficients.

[0489] In these and other cases, the audio decoding device 24 is configured to compute Us[n] = HOA(n) * (V_vec[n])-1 to obtain a scalar.

[0490] In these and other cases, the interpolation comprises a linear interpolation. In these and other cases, the interpolation comprises a non-linear interpolation. In these and other cases, the interpolation comprises a cosine interpolation. In these and other cases, the interpolation comprises a weighted cosine interpolation. In these and other cases, the interpolation comprises a cubic interpolation. In these and other cases, the interpolation comprises an Adaptive Spline Interpolation. In these and other cases, the interpolation comprises a minimum curvature interpolation.

[0491] In these and other cases, the audio decoding device 24 is further configured to generate a bit stream that includes a representation of the decomposed interpolated spherical harmonic coefficients for the time segment, and an indication of a type of the interpolation. .

[0492] In these and other cases, the indication comprises one or more bits that map to the interpolation type.

[0493] In these and other cases, the audio encoding device 24 is further configured to obtain a bit stream that includes a representation of the decomposed interpolated spherical harmonic coefficients for the time segment, and an indication of a type of the interpolation. .

[0494] In these and other cases, the indication comprises one or more bits that map to the interpolation type.

[0495] Various aspects of the procedure sets may, in some cases, additionally enable the audio decoding device 24 to be configured to obtain a bit stream comprising a compressed version of a spatial component of a sound field, the spatial component is generated by performing vector-based synthesis with respect to a plurality of spherical harmonic coefficients.

[0496] In these and other cases, the compressed version of the spatial component is represented in the bitstream using, at least in part, a field that specifies a prediction mode used when compressing the spatial component.

[0497] In these and other cases, the compressed version of the spatial component is represented in the bitstream using, at least in part, Huffman table information that specifies a Huffman table used when compressing the spatial component.

[0498] In these and other cases, the compressed version of the spatial component is represented in the bit stream using, at least in part, a field that indicates a value that expresses a quantization step size or a variable thereof used when compressing the spatial component.

[0499] In these and other cases, the value comprises an nbits value.

[0500] In these and other cases, the bit stream comprises a compressed version of a plurality of spatial components of the sound field whose compressed version of the spatial component is included, and the value expresses the quantization step size or a variable thereof used when compressing the plurality of spatial components.

[0501] In these and other cases, the compressed version of the spatial component is represented in the bitstream using, at least in part, a Huffman code to represent a category identifier that identifies a compression category to which the spatial component corresponds.

[0502] In these and other cases, the compressed version of the spatial component is represented in the bit stream using, at least in part, a sign bit that identifies whether the spatial component is a positive value or a negative value.

[0503] In these and other cases, the compressed version of the spatial component is represented in the bitstream using, at least in part, a Huffman code to represent a residual value of the spatial component.

[0504] In these and other cases, the device comprises an audio decoding device.

[0505] Various aspects of procedure sets can also enable the audio decoding device 24 to identify a Huffman codebook to use when decompressing a compressed version of a spatial component from a plurality of compressed spatial components based on an order of the compressed version of the spatial component with respect to those remaining of the plurality of compressed spatial components, the spatial component generated by performing vector-based synthesis with respect to a plurality of spherical harmonic coefficients.

[0506] In these and other cases, the audio decoding device 24 is configured to obtain a bit stream comprising the compressed version of a spatial component of a sound field, and decompresses the compressed version of the spatial component with use, at least in part, from the Huffman codebook identified to obtain the spatial component.

[0507] In these and other cases, the compressed version of the spatial component is represented in the bitstream using, at least in part, a field that specifies a prediction mode used when compressing the spatial component, and the device decoding system 24 is configured to decompress the compressed version of the spatial component based at least in part on the prediction mode to obtain the spatial component.

[0508] In these and other cases, the compressed version of the spatial component is represented in the bitstream using, at least in part, Huffman table information that specifies a Huffman table used when compressing the spatial component, and the audio decoding device 24 is configured to decompress the compressed version of the spatial component based at least in part on the Huffman table information.

[0509] In these and other cases, the compressed version of the spatial component is represented in the bit stream using, at least in part, a field that indicates a value that expresses a quantization step size or a variable thereof. used when compressing the spatial component, and the audio decoding device 24 is configured to decompress the compressed version of the spatial component based at least in part on the value.

[0510] In these and other cases, the value comprises an nbits value.

[0511] In these and other cases, the bit stream comprises a compressed version of a plurality of spatial components of the sound field whose compressed version of the spatial component is included, the value expresses the quantization step size or a variable thereof used when the plurality of spatial components is compressed and the audio decoding device 24 is configured to decompress the plurality of compressed version of the spatial component based at least in part on the value.

[0512] In these and other cases, the compressed version of the spatial component is represented in the bitstream using, at least in part, a Huffman code to represent a category identifier that identifies a compression category to which the spatial component corresponds and the audio decoding device 24 is configured to decompress the compressed version of the spatial component based, at least in part, on the Huffman code.

[0513] In these and other cases, the compressed version of the spatial component is represented in the bit stream using, at least in part, a sign bit that identifies whether the spatial component is a positive value or a negative m value, and the audio decoding device 24 is configured to decompress the compressed version of the spatial component based at least in part on the signal bit.

[0514] In these and other cases, the compressed version of the spatial component is represented in the bitstream using, at least in part, a Huffman code to represent a residual value of the spatial component and the audio decoding device. 24 is configured to decompress the compressed version of the spatial component based, at least in part, on the Huffman code included in the identified Huffman codebook.

[0515] In each of the various cases described above, it should be understood that the audio decoding device 24 may perform a method or otherwise comprises means for performing each step of the method for which the audio decoding device 24 is configured to perform in some cases, these means may comprise one or more processors. In some cases, the one or more processors may represent a special-purpose processor configured through instructions stored on a non-transient computer-readable storage medium. In other words, various aspects of the sets of procedures in each of the sets of coding examples can provide a non-transient computer-readable storage medium that has stored instructions that, when executed, cause the one or more processors to perform the method for which the audio decoding device 24 was configured to perform.

[0516] Figure 6 is a flowchart illustrating the exemplary operation of a content analysis unit of an audio encoding device, such as the content analysis unit 26 shown in the example of Figure 4, in carrying out various aspects of the sets of procedures described in this disclosure.

[0517] The content analysis unit 26 can, when determining whether representative HOA coefficients 11 of a sound field are generated from a synthetic audio object, obtain a table of HOA coefficients (93), which can be of size 25 by 1024 for a fourth-order representation (that is, N = 4). After obtaining the framed HOA coefficients (which may also be denoted herein as a framed SHC matrix 11 and subsequent framed SHC matrices may be denoted as framed SHC matrices 27B, 27C, etc.), the content analysis unit 26 can then exclude the first vector of framed HOA coefficients 11 to generate reduced frame HOA coefficients(94).

[0518] The content analysis unit 26 can then predict the first non-zero vector of reduced frame HOA coefficients from the remaining vectors of reduced frame HOA coefficients(95). After predicting the first non-zero vector, the content analysis unit 26 may obtain an error based on the first predicted non-zero vector and the actual non-zero vector (96). Once the error is obtained, the content analysis unit 26 can compute a ratio based on an energy of the real non-zero first vector and the error (97). The content analysis unit 26 can then compare this ratio to a threshold (98). When the ratio does not exceed the threshold ("NO" 98), the content analysis unit 26 can determine that the framed SHC matrix 11 is generated from a recording and indicates in the bitstream that the corresponding encoded representation of the SHC 11 array was generated from a recording (100, 101). When the ratio exceeds the threshold ("YES" 98), the content analysis unit 26 can determine that the SHC matrix with frame 11 is generated from a synthetic audio object and indicates in the bitstream that the encoded representation corresponding matrix of SHC 11 was generated from a synthetic audio object (102, 103). In some cases, when the frame 11 SHC matrix has been generated from a recording, the content analysis unit 26 passes the frame 11 SHC matrix to the unit vector based synthesis 27 (101). In some cases, when the frame 11 SHC matrix was generated from a synthetic audio object, the content analysis unit 26 passes the frame 11 SHC matrix to the directional-based synthesis unit 28 (104 ).

[0519] Figure 7 is a flowchart illustrating the exemplary operation of an audio encoding device, such as the audio encoding device 20 shown in the example of Figure 4, in carrying out various aspects of the sets of based synthesis procedures. vectors described in this disclosure. Initially, the audio encoding device 20 receives the HOA coefficients 11 (106). The audio coding device 20 may invoke the LIT unit 30, which may apply an LIT against the HOA coefficients to output transformed HOA coefficients (e.g. in the case of SVD, where the transformed HOA coefficients may comprise the US[k] 33 vectors and the V[k] 35 vectors) (107).

[0520] Audio encoding device 20 may then invoke parameter calculation unit 32 to perform the analysis described above with respect to any combination of US[k]33 vectors, US[k-1]33 vectors , the vectors V[k] and/or V[k-1] 35 to identify various parameters as described above. That is, the parameter calculation unit 32 can determine at least one parameter based on an analysis of the transformed HOA coefficients 33/35 (108).

[0521] The audio encoding device 20 can then invoke the reordering unit 34, which can reorder the transformed HOA coefficients (which, again in the context of SVD, can refer to the XJS[k] vectors 33 and the vectors V[k] 35) based on the parameter to generate reordered transformed HOA coefficients 33735' (or, in other words, the vectors XJS[k] 33' and the vectors V[k] 35'), as described above ( 109). The audio encoding device 20 may, during any of the foreground operations or subsequent operations, also invoke the sound field analysis unit 44. The sound field analysis unit 44 may, as described above, perform a sound field analysis unit 44. sound field in relation to HOA coefficients 11 and/or transformed HOA coefficients 33/35 to determine the total number of foreground channels (nFG) 45, the order of the background sound field (NBG) and the number ( NBGa) and indices (i) of BG HOA channels to send (which may collectively be denoted as background channel information 43 in the example of Figure 4) (109).

[0522] Audio encoding device 20 may also invoke background selection unit 48. Background selection unit 48 may determine ambient or background HOA coefficients 47 based on background channel information 43 (110). The audio encoding device 20 can additionally invoke the foreground selection unit 36, which can select those among the reordered XJS[k] vectors 33' and the reordered V[k] vectors 35' representing distinct components or foreground of the sound field based on nFG 45 (which may represent one or more indices that identify these foreground vectors) (112).

[0523] Audio encoding device 20 may invoke power compensation unit 38. Power compensation unit 38 may perform power compensation against ambient HOA coefficients 47 to compensate for power loss due to removal of several channels among the HOA channels by the background selection unit 48 (114) and thereby generate the energy compensated ambient HOA coefficients 47'.

[0524] The audio encoding device 20 may then invoke the spatiotemporal interpolation unit 50. The spatiotemporal interpolation unit 50 may perform spatiotemporal interpolation against the reordered transformed HOA coefficients 33735' to obtain the interpolated foreground signals 49' (which may also be called the "interpolated nFG signals 49"') and the remaining foreground directional information 53 (which may also be called the "Y[k] vectors 53") ( 116). The audio encoding device 20 can then invoke a coefficient reduction unit 46. The coefficient reduction unit 46 can perform coefficient reduction with respect to the remaining foreground vectors V[k] 53 based on the information 43 to obtain reduced foreground directional information 55 (which may also be called reduced foreground vectors V[k] 55) (118).

[0525] The audio encoding device 20 can then invoke the quantization unit 52 to compress, in the manner described above, the reduced foreground vectors V[k] 55 and generate coded foreground vectors V[k] 57 (120).

[0526] Audio encoding device 20 may also invoke psychoacoustic audio encoding unit 40. Psychoacoustic audio encoding unit 40 may psychoacoustically encode each vector of energy compensated ambient HOA coefficients 47' and nFG signals interpolated 49' to generate encoded ambient HOA coefficients 59 and encoded nFG signals 61. The audio encoding device may then invoke the bitstream generation unit 42. The bitstream generation unit 42 may generating bitstream 21 based on encoded foreground directional information 57, encoded ambient HOA coefficients 59, encoded nFG signals 61, and background channel information 43.

[0527] Figure 8 is a flowchart illustrating the exemplary operation of an audio decoding device, such as the audio decoding device 24 shown in Figure 5, in performing various aspects of the sets of procedures described in this disclosure. Initially, the audio decoding device 24 can receive the bit stream 21 (130). Upon receiving the bitstream, the audio decoding device 24 may invoke the extraction unit 72. Assuming for purposes of discussion that the bitstream 21 indicates that vector-based reconstruction is to be performed, the extraction device 72 can parse this bit stream to retrieve the information noted above, passing that information to the vector-based reconstruction unit 92.

[0528] In other words, the extraction unit 72 can extract the coded foreground directional information 57 (which, again, can also be called the coded foreground vectors V[k] 57), the ambient HOA coefficients 59 and the encoded foreground signals (which may also be called the encoded foreground nFG signals 59 or the encoded foreground audio objects 59) from the bit stream 21 as described above (132).

[0529] The audio decoding device 24 may additionally invoke the quantization unit 74. The quantization unit 74 may entropy dequantize and decode the encoded foreground directional information 57 to obtain reduced foreground directional information 55k ( 136). The audio decoding device 24 can also invoke the psychoacoustic decoding unit 80. The psychoacoustic audio coding unit 80 can decode the encoded ambient HOA coefficients 59 and the encoded foreground signals 61 to obtain ambient HOA coefficients energy compensated 47' and interpolated foreground signals 49' (138). Psychoacoustic decoding unit 80 may pass energy compensated ambient HOA coefficients 47' to HOA coefficient formulation unit 82 and nFG signals 49' to reordering unit 84.

[0530] The reordering unit 84 may receive syntax elements indicative of the original order of the foreground components of the HOA coefficients 11. The reordering unit 84 may, based on these reordering syntax elements, rearrange the interpolated nFG signals 49' and reduced foreground V[k] 55k vectors to generate reordered 49” nFG signals and reordered foreground V[k] 55k' vectors (140). The reordering unit 84 may output the reordered nFG signals 49" to the foreground formulation unit 78 and the reordered foreground vectors 55k' V[k] to the spatiotemporal interpolation unit 76.

[0531] The audio decoding device 24 may then invoke the spatiotemporal interpolation unit 76. The spatiotemporal interpolation unit 76 may receive the reordered foreground directional information 55k' and perform the spatiotemporal interpolation. time relative to the reduced foreground directional information to generate the 55k interpolated foreground directional information” (142). The spatiotemporal interpolation unit 76 may transmit the interpolated foreground vectors 55k” V[k] to the foreground formulation unit 718.

[0532] Audio decoding device 24 may invoke foreground formulation unit 78. Foreground formulation unit 78 may perform matrix multiplication of interpolated foreground signals 49" by interpolated foreground directional information 55k” to get the foreground HOA coefficients 65 (144). The audio decoding device 24 may also invoke the HOA coefficient formulation unit 82. The HOA coefficient formulation unit 82 may add the foreground HOA coefficients 65 to the ambient HOA channels 47' for the purpose of to obtain the HOA coefficients 11' (146).

[0533] Figures 9A to 9L are block diagrams illustrating various aspects of the audio encoding device 20 of the example of Figure 4 in more detail. Figure 9A is a block diagram illustrating the LIT unit 30 of the audio encoding device 20 in greater detail. As shown in the example of Figure 9A, the LIT unit 30 can include multiple different linear invertible transforms 200-200N. The LIT unit 30 may include, to provide some examples, a singular value decomposition transform (SVD) 200A ("SVD 200A"), a principal component analysis transform (PCA) 200B ("PCA 200B"), a Karhunen-Loeve transform (KLT) 200C ("KLT 200C"), a fast Fourier transform (FFT) 200D ("FFT 200D") and a discrete cosine transform (DCT) 200N ("DCT 200N"). The LIT unit 30 can invoke any of the linear invertible transforms 200 to apply the respective transform with respect to the HOA coefficients 11 and generate the respective transformed HOA coefficients 33/35.

[0534] Although described as being performed directly with respect to HOA 11 coefficients, the LIT unit 30 can apply the linear invertible transforms 200 to derivatives of the HOA coefficients 11. For example, the LIT unit 30 can apply the SVD 200 against a power spectral density matrix derived from the HOA coefficients 11. The power spectral density matrix can be denoted as PSD and obtained by matrix multiplication of the transpose of the hoaFrame to the hoaFrame, as highlighted in the pseudocode that follow below. The hoaFrame notation refers to a frame of HOA 11 coefficients.

[0535] The unit of LIT 30 can, after applying the SVD 200 (svd) to the PSD, can obtain a matrix S[k]2 (Squadratic) and a matrix V[k]. The matrix S[k]2 can denote a quadratic matrix S[k] whereby the unit of LIT 30 (or alternatively the unit of SVD 200 as an example) can apply a square root operation to the matrix S[ k]2 to obtain the matrix S[k]. The SVD unit 200 can, in some cases, quantize against the matrix V[k] to obtain a quantized matrix V[k] (which can be denoted as the matrix V[&]'). The LIT unit 30 can obtain the matrix U[k] by first multiplying the matrix S[k] by the quantized matrix V[&]' to obtain a matrix SV[k]'. The unit of LIT 30 can then obtain the pseudoinverse (pinv) of the matrix SV[k]' and then multiply the coefficients of HOA 11 by the pseudoinverse of the matrix SV[k]' to obtain the matrix U[k]. The foreground can be represented by the following pseudo-code:

[0536] By performing SVD against the power spectral density (PSD) of the HOA coefficients rather than the coefficients themselves, the LIT 30 unit can potentially reduce the computational complexity of performing the SVD in terms of one or more of the cycles of processor and storage space, while achieving the same source audio coding efficiency as if SVD were applied directly to the HOA coefficients. That is, the PSD-type SVD described above can potentially be less computationally demanding because the SVD is done on an F*F matrix (with F the number of HOA coefficients). Compared to an M * F matrix with M as the frame length, i.e. 1024 or more samples. The complexity of an SVD can now, through application to PSD rather than HOA 11 coefficients, be around 0(LA3) compared to 0(M*LA2) when applied to HOA 11 coefficients (where 0( *) denotes the large 0 computing complexity notation common to the computer science arts).

[0537] Figure 9B is a block diagram illustrating a LIT unit 32 of audio encoding device 20 in greater detail. the parameter calculation unit 32 may include an energy analysis unit 202 and a cross correlation unit 204. The energy analysis unit 202 may perform the energy analysis described above with respect to one or more of the vectors US[k ] 33 and the vectors V[k] 35 to generate one or more of the correlation parameter (R), the directional property parameters (θ, cp, r), and the energy property (e) for one or more of the current frame (k) or the previous frame (-1). Likewise, the cross-correlation unit 204 can perform the cross-correlation described above against one or more of the vectors XJS[k] 33 and the vectors V[k] 35 to generate one or more of the correlation parameter (R ), the parameters of directional properties (θ, cp, r), and the energy property (e) for one or more of the current frame (k) or the previous frame (-1). The parameter calculation unit 32 can output current frame parameters 37 and previous frame parameters 39.

[0538] Figure 9A is a block diagram illustrating the LIT unit 34 of the audio encoding device 20 in greater detail. The reordering unit 34 includes a parameter evaluating unit 206 and a vector reordering unit 208. The parameter evaluating unit 206 represents a unit configured to evaluate the previous frame parameters 39 and the current frame parameters 37 in the form described above for generating reordering indices 205. Reordering indices 205 include indices that identify how vectors among vectors US[A:] 33 and vectors among vectors V[k] 35 are to be reordered (e.g., pairwise). index with the first index of the pair identifying the index of the current vector location and the second index of the pair identifying the reordered location of the vector). Vector reordering unit 208 represents a unit configured to reorder vectors XJS[k] 33 and vectors V[k] 35 according to reorder indices 205. Reorder unit 34 can output vectors XJS[k] reordered 33' and the reordered vectors V[k] 35', while also passing the reordering indices 205 as one or more syntax elements to the bitstream generating unit 42.

[0539] Figure 9D is a block diagram illustrating the sound field analysis unit 44 of the audio encoding device 20 in greater detail. As shown in the example of Figure 9D, the sound field analysis unit 44 may include a single value analysis unit 210A, a power analysis unit 210B, a spatial analysis unit 210c, a spatial masking analysis unit 210B. 210D, a diffusion analysis unit 210E and a directional analysis unit 21 OF. The singular value analysis unit 210A can represent a unit configured to analyze the slope of the curve created by the descending diagonal values of S vectors (forming part of the XJS[k] 33 vectors), where the large singular values represent foreground sounds. or distinct and the low singular values represent background components of the sound field as described above. The energy analysis unit 21 OB may represent a unit configured to determine the energy of vectors V[k] 35 on a per vector basis.

[0540] Spatial analysis unit 2 10c may represent a unit configured to perform the spatial energy analysis described above by transforming the HOA coefficients 11 into the spatial domain and identifying high energy areas representative of directional components of the sound field that should be preserved. The 210D spatial masking analysis unit can represent a unit configured to perform spatial masking analysis in a manner similar to that of the spatial energy analysis, except that the 210D spatial masking analysis unit can identify spatial areas that are masked by sounds of superior energy spatially close. The scatter analysis unit 210E may represent a unit configured to perform the scatter analysis described above in relation to HOA coefficients 11 to identify areas of diffuse energy that may represent background components of the sound field. The directional analysis unit 21 OF may represent a unit configured to perform the directional analysis noted above that involves computing the vectors \S[k], and performing the frame and sum of each input of each of these vectors \S [k] to identify a directionality quotient. The directional analysis unit 21 OF can provide this directionality quotient for each of the vectors \S[k] to the foreground/background identification (ID) unit (BG/FG) 212.

[0541] Sound field analysis unit 44 may also include BG/FG ID unit 212, which may represent a unit configured to determine the total number of foreground channels (nFG) 45, sound field order (NBG) and the number (NBGa) and indices (i) of additional BG HOA channels to send (which can be collectively denoted as background channel information 43 in the example in Figure 4) based on any combination of analysis output by any combination of analysis units 210-21 OF. BG/FG ID unit 212 can determine nFG 45 and background channel information 43 in order to achieve target bit rate 41.

[0542] Figure 9E is a block diagram illustrating the foreground selection unit 36 of the audio encoding device 20 in greater detail. The foreground selection unit 36 includes a vector analysis unit 214 that can analyze or otherwise extract the foreground US[k] vectors 49 and the foreground V[k] vectors 51k identified by the foreground element. nFG 45 syntax from the 33' reordered XJS[k] vectors and 35' reordered V[k] vectors. The vector analysis unit 214 can analyze the various vectors representative of the foreground components of the sound field identified by the sound field analysis unit 44 and specified by the nFG syntax element 45 (which may also be referred to as audio channel information). foreground 45). As shown in the example of Figure 9E, the vector analysis unit 214 may select, in some cases, non-consecutive vectors within the foreground XJS[k] vectors 49 and the foreground V[k] vectors 51k to represent the foreground components of the sound field. Furthermore, the vector analysis unit 214 may select, in some cases, the same vectors (by position) from the foreground XJS[k] vectors 49 and the foreground V[k] vectors 51k to represent the foreground components. sound field plane.

[0543] Figure 9A is a block diagram illustrating a LIT unit 48 of audio encoding device 20 in greater detail. Background selection unit 48 can determine ambient or background HOA coefficients 47 based on background channel information (e.g., background sound field (NBG) and number (NBGa) and indices (i) additional BG HOA channels to send). For example, when NBG is equal to one(1), the background selection unit 48 can select HOA coefficients 11 for each audio frame sample that has an order equal to or less than one(1). The background selection unit 48 can, in this example, then select the HOA coefficients 11 that have an index identified by one of the indices (i) as additional BG HOA coefficients, where the NBGa is provided to the generating unit bitstream 42 to be specified in bitstream 21 for the purpose of enabling the audio decoding device, such as the audio decoding device 24 shown in the example of Figure 5, to analyze the HOA coefficients from BG 47 to from the bit stream 21. The background selection unit 48 may then output the ambient HOA coefficients 47 to the power compensation unit 38. The ambient HOA coefficients 47 may have dimensions DM x [( NBG+1)2 + NBGa].

[0544] As further shown in the example of Figure 10C(i), portion 258B' may include bit substreams for vl to vn, each of which includes a Huffman Table information field 263 and a corresponding one from the directional components compressed vl to vn without including predict mode field 262. In all other respects, portion 258B' may be similar to portion 258B.

[0545] The energy determination unit 218 may represent a unit configured to identify the RMS for each row and/or column of one or more of the 33' reordered US[k] matrix and 35' reordered matrix V[k] . The energy determination unit 38 may also identify the RMS for each row and/or column of one or more of the selected foreground channels, which may include the nFG signals 49 and the foreground V[k] vectors 511, and the reduced-order ambient HOA coefficients 47. The RMS for each row and/or column among the one or more 33' reordered US[k] matrix and 35' reordered matrix V[k] can be stored in a vector denoted RMSCOMPLETE, while the RMS for each row and/or column of one or more of the nFG signals 49, the foreground V[k] vectors 511, and the reduced ambient HOA coefficients in order 47 can be stored in a vector denoted RMSREDUCED- [0519] In some examples, to determine each RMS of the respective rows and/or columns of one or more of the 33' reordered US[k] matrix, 35' reordered V[k] matrix, the signals of nFG 49, the foreground V[k] vectors 511, and the reduced ambient HOA coefficients in order 47, the unit and power determination 218 may first apply a reference spherical harmonic coefficient (SHC) renderer to the columns. Application of the reference SHC renderer by the energy determination unit 218 allows the determination of RMS in the SHC domain to determine the energy of the overall sound field described by each row and/or column of the frame represented by rows and/or columns of one or more of the reordered XJS[k] matrix 33', the reordered matrix V[k] 35', the nFG signals 49, the foreground vectors V[k] 51k, and the reduced ambient HOA coefficients in order 47. The energy determination unit 38 can pass these FULL RMS and LOW RMS vectors to the energy analysis unit 220.

[0546] The energy analysis unit 220 may represent a unit configured to compute a Z amplification value vector, according to the following equation: Z = FULL RMS / LOW RMS - The energy analysis unit 220 may then , passing this Z amplification value vector to the power amplification unit 222. The power amplification unit 222 may represent a unit configured to apply this Z amplification value vector or various portions thereof to one or more of the signals of nFG signals 49, the foreground V[k] vectors 511, and the reduced ambient HOA coefficients in order 47. In some cases, the amplification value vector Z is applied only to the reduced ambient HOA coefficients in order 47 according to the following equation HOABG_RED' = HOABG_REDZT, where HOABG-RED denotes reduced ambient HOA coefficients in order 47, HOABG-RED denotes compensated energy, reduced ambient HOA coefficients 47’ and Zr denotes the transpose of the vector Z.

[0547] Figure 9H is a block diagram illustrating, in greater detail, the spatiotemporal interpolation unit 50 of the audio encoding device 20 shown in the example of Figure 4. The spatiotemporal interpolation unit 50 may represent a unit configured to receive the foreground V[k] vectors 51k for the k'th frame and the foreground V[k-1] vectors 51k-1 for the previous frame (hence the k-1 notation) and perform spatiotemporal interpolation to generate interpolated foreground V[k] vectors. The spatiotemporal interpolation unit 50 may include a V interpolation unit 224 and a foreground adaptation unit 226.

[0548] The V interpolation unit 224 can select a portion of the current foreground vectors V[k] 51k to interpolate based on the remaining portions of the current foreground vectors V[k] 51k and the vectors V[k-1 ] foreground earlier 51k. The V interpolation unit 224 may select the portion to be one or more of the subframes noted above or just a single undefined portion which may vary on a frame-by-frame basis. The V interpolation unit 224 may, in some cases, select a single sample portion 128 from among the 1024 samples of the current foreground V[k] vectors 51k to interpolate. The V interpolation unit 224 can then convert each of the vectors into the current foreground V[k] vectors 511 and the previous foreground vectors V[k-1] 51k to separate the separate spatial maps by projecting the vectors on a sphere (using a projection matrix as a design matrix T). The V interpolation unit 224 can then interpret the vectors in V as shapes on a sphere. To interpolate the V matrices for the sample portion 256, the V interpolation unit 224 can then interpolate these spatial formats - and then transform them back into spherical harmonic domain vectors via the inverse of the projection matrix. The sets of procedures from this disclosure can thus provide a smooth transition between the V matrices. The V interpolation unit 224 can then generate the remaining V[k] vectors 53, which represent the V[k] vectors of first 51k plane after being modified to remove the interpolated portion of the 51k foreground V[k] vectors. The V interpolation unit 224 can then pass the interpolated foreground V[k] vectors 51k' to the nFG adaptation unit 226.

[0549] When a single portion is selected for interpolation, the interpolation unit V 224 can generate a syntax element denoted CodedSpatiallnterpolationTime 254, which identifies the duration or, in other words, the time of the interpolation (for example, in what refers to the number of samples). When a single portion is selected to perform subframe interpolation, the V interpolation unit 224 may also generate another syntax element denoted SpatiallnterpolationMethod 255, which may identify a type of interpolation performed (or, in some cases, whether interpolation was performed or no). The spatiotemporal interpolation unit 50 may output these syntax elements 254 and 255 to the bitstream generating unit 42.

[0550] The nFG adaptation unit 226 may represent a unit configured to generate the adapted nFG signals 49'. The nFG adaptor unit 226 can generate the adapted nFG signals 49' by first obtaining the foreground HOA coefficients by multiplying the nFG signals 49 by the foreground vectors V[k] 51k. After obtaining the foreground HOA coefficients, the nFG adaptation unit 226 can divide the foreground HOA coefficients by the interpolated foreground V[k] vectors 53 to obtain the adapted nFG signals 49' (which can be called interpolated nFG signals 49' since these signals are derived from interpolated foreground V[k] vectors 51k').

[0551] Figure 91 is a block diagram illustrating in more detail the coefficient reduction unit 46 of the audio encoding device 20 shown in the example of Figure 4. The coefficient reduction unit 46 may represent a configured unit. to perform coefficient reduction with respect to the remaining foreground vectors V[k] 53 based on the background channel information 43 to output reduced foreground vectors V[k] 55 to the quantization unit 52. The vectors foregrounds V[k] 55 can have dimensions D: [(N+1)2 - (NBG+1)2-NBGa] x nFG.

[0552] The coefficient reduction unit 46 may include a coefficient minimization unit 228, which may represent a unit configured to reduce or otherwise minimize the size of each of the remaining foreground vectors V[k] 53 by removing - if any coefficients that are taken into account in background HOA coefficients 47 (as identified by background channel information 43). The coefficient minimization unit 228 can remove those coefficients identified by the background channel information 43 to obtain the reduced foreground vectors V[k] 55.

[0553] Figure 9J is a block diagram illustrating, in greater detail, the psychoacoustic audio encoder unit 40 of the audio encoder device 20 shown in the example of Figure 4. The psychoacoustic audio encoder unit 40 may represent a configured unit. to perform psychoacoustic coding against energy compensated background HOA coefficients 47' and interpolated nFG signals 49'. As shown in the example of Figure 9H, the psychoacoustic audio encoder unit 40 can invoke multiple instances of psychoacoustic audio encoders 40A-40N to audio encode each of the channels of the energy compensated background HOA coefficients 47' (wherein a channel in this context refers to coefficients for all samples in the frame corresponding to a particular suborder/order spherical basis function) and each signal among the interpolated nFG signals 49'. In some examples, the psychoacoustic audio encoder unit 40 instantiates or otherwise includes (when implemented in hardware) sufficient numbers of audio encoders 40A-40N to separately encode each channel of the energy compensated background HOA coefficients 47' ( or NBGa plus the total number of indices (i)) and each signal among the 49' interpolated nFG signals (or nFG) for a total of NBGa plus the total number of indices (i) of additional room HOA channels plus nFG . Audio encoders 40A-40N can output 59 encoded background HOA coefficients and 61 encoded nFG signals.

[0554] Figure 9K is a block diagram illustrating in greater detail the quantization unit 52 of the audio encoding device 20 shown in the example of Figure 4 In the example of Figure 9K, the quantization unit 52 includes a unit quantization unit 230, an nbit unit 232, a prediction unit 234, a prediction mode unit 236 ("Pred Mode Unit 236"), a category or residual encoding unit 238, and a selection unit 240. Uniform quantization unit 230 represents a unit configured to perform the uniform quantization described above with respect to one of the spatial components (which may represent any of the reduced foreground vectors V[k] 55). Unit of nbits 232 represents a unit configured to determine the value or parameter of nbits.

[0555] Prediction unit 234 represents a unit configured to perform prediction with respect to the quantized spatial component. The prediction unit 234 can perform the prediction by performing a per-element subtraction of the current vectors among the reduced foreground vectors V[k] 55 by a corresponding temporally subsequent vector among the reduced foreground vectors V[k] 55 ( which can be denoted as reduced foreground vectors Y[k-1] 55). The result of this prediction can be called the predicted spatial component.

[0556] Prediction mode unit 236 may represent a unit configured to select prediction mode. The Huffman table selection unit 240 may represent a unit configured to select a suitable Huffman table for cid encoding. The prediction mode unit 236 and the Huffman table selection unit 240 may operate, as an example, according to the following pseudo-code:

[0557] For a given nbits, retrieve all Huffman Tables that have nbits

[0558] BOO = 0; B01 = initialize to compute expected bits per encoding mode

[0559] for m = 1:(Number of elements in the vector)

[0560] // calculate expected number of bits for an array element

[0561] // without prediction and using Huffman's Table 5

[0562] B00 = B00 + calculate_bits(v(m), HT5);

[0563] // without prediction and using Huffman's Table {1,2,3}

[0564] B01 = B01 + calculate_bits(v(m), HTq); q in {1,2,3}

[0565] // calculate expected number of bits for residual prediction e(m)

[0566] e(m) = v(m) - vp(m); // vp(m): previous frame vector element

[0567] // with prediction and using Huffman's Table 4

[0568] B10 = B10 + calculate_bits(e(m), HT4);

[0569] // with prediction and using Huffman's Table 5

[0570] B11 = B11 + calculate_bits(e(m), HT5);

[0571] end

[0572] // find a best prediction mode and Huffman table that yield minimum bits // best prediction mode and Huffman table are flagged by pflag and Htflag, respectively

[0573] [Be, id] = min([B00 B01 B10 B11]);

[0574] Change id

[0575] case 1: pflag = 0; HTflag = 0;

[0576] case 2: pflag = 0; HTflag = 1;

[0577] case 3: pflag = 1; HTflag = 0;

[0578] case 4: pflag = 1; HTflag = 1;

[0579] end

[0580] Category and residual coding unit 238 may represent a unit configured to perform categorization and residual coding of a predicted spatial component or the quantized spatial component (when prediction is disabled) in the manner described in greater detail above.

[0581] As shown in the example of Figure 9K, the quantization unit 52 may output various parameters or values for inclusion in the bit stream 21 or sub-information (which may itself be a separate bit stream from the bit stream 21). Assuming the information is specified in the auxiliary channel information, the scalar/entropy quantization unit 50 can output the nbits value as nbits value 233, the prediction mode as prediction mode 237, and the Huffman table information as Huffman table information 241 for the bitstream generation unit 42 together with the compressed version of the spatial component (shown as coded foreground vectors V[k] 57 in the example of Figure 4), which in this example can be refer to the selected Huffman code to encode the cid, the sign bit, and the block-encoded residual. The value of nbits can be specified once in the auxiliary channel information for all V[k] 57 encoded foreground vectors, while the prediction mode and Huffman table information can be specified for each of the foreground vectors coded plane V[k] 57. The portion of the bit stream that specifies the compressed version of the spatial component is shown in greater detail in the example of Figures 10B and/or 10c.

[0582] Figure 9L is a block diagram illustrating, in greater detail, the bitstream generating unit 42 of the audio encoding device 20 shown in the example of Figure 4. The bitstream generating unit 42 may include a main channel information generating unit 242 and the auxiliary channel information generating unit 244. The main channel information generating unit 242 may generate a main bit stream 21 that includes one or more, if not all of reordering indices 205, CodedSpatiallnterpolationTime syntax element 254, SpatiallnterpolationMethod syntax element 255, coded background HOA coefficients 59, and coded nFG signals 61. Auxiliary channel information generation units 244 may represent a unit configured to generate a sidechannel 2 IB bit stream that may include one or more, if not all, of the nbit value 233, the prediction mode 237, the information Huffman table s 241 and the coded foreground vectors V[k] 57. Bitstreams 21 and 2 IB may be collectively referred to as bitstream 21. In some contexts, bitstream 21 may only be refer to main channel bit stream 21, while bit stream 2 IB may be called auxiliary channel information 2 IB.

[0583] Figures 10A to 100(ii) are diagrams illustrating portions of the bit stream or auxiliary channel information that may specify the compressed spatial components in more detail. In the example in Figure 10A, a portion 250 includes a renderer identifier field (“render ID”) 251 and a HOADecoderConfig field 252. The renderer ID field 251 can represent a field that stores a renderer ID that was used for mixing HOA content. The HOADecoderConfig 252 field can represent a field configured to store information to initialize the HOA spatial decoder.

[0584] The HOADecoderConfig field 252 additionally includes a directional information field (“direction information”) 253, a CodedSpatiallnterpolationTime field 254, a SpatiallnterpolationMethod field 255, a CodedVVecLength field 256 and a gain information field 257. The directional information field 253 can represent a field that stores information for configuring the directionality-based synthesis decoder. The CodedSpatiallnterpolationTime field 254 can represent a field that stores a time of spatiotemporal interpolation of vector-based signals. The SpatiallnterpolationMethod field 255 can represent a field that stores an indication of the type of interpolation applied during spatiotemporal interpolation of vector-based signals. The CodedVVecLength field 256 can represent a field that stores a vector length of transmitted data used to synthesize vector-based signals. Gain information field 257 represents a field that stores information indicative of a gain correction applied to signals.

[0585] In the example of Figure 10B, portion 258A represents a portion of the auxiliary information channel, wherein portion 258A includes a frame header 259 that includes several byte field 260 and an nbit field 261. The number of byte field 260 may represent a field to express the number of bytes included in the frame to specify spatial components vl to vn which include the zeros for byte alignment field 264. The nbit field 261 represents a field that may specify the value of nbits identified for use when decompressing spatial components vl to vn.

[0586] As further shown in the example of Figure 10B, portion 258A may include bit substreams for vl to vn, each of which includes a prediction mode field 262, a Huffman Table information field 263, and a corresponding of the special compressed components vl to vn. Prediction mode field 262 may represent a field for storing an indication of whether the prediction has been performed with respect to a corresponding one of the compressed special components vl through vn. The Huffman table information field 263 represents a field to indicate, at least in part, that the Huffman table should be used to decode various aspects of the corresponding one of the special compressed components vl through vn.

[0587] In this regard, techniques can enable audio encoding device 20 to obtain a bit stream comprising a compressed version of a spatial component of a sound field, the spatial component generated by performing a vector based synthesis with respect to a plurality of spherical harmonic coefficients.

[0588] Figure 10C is a diagram illustrating an alternative example of a portion 258B of the auxiliary channel information that may specify the compressed special components in greater detail. In the example of Figure 10C , portion 258B includes a frame header 259 that includes an Nbit field 261. Nbit field 261 represents a field that can specify an identified nbit value for use in decompressing spatial components vl through vn.

[0589] As further shown in the example of Figure 10C, portion 258B may include bit substreams for vl to vn, each of which includes a predictive mode field 262, a Huffman Table information field 263, and a corresponding one of the special components compressed vl to vn. Prediction mode field 262 may represent a field for storing an indication of whether the prediction has been performed with respect to a corresponding one of the compressed special components vl through vn. The Huffman table information field 263 represents a field to indicate, at least in part, that the Huffman table should be used to decode various aspects of the corresponding one of the special compressed components vl through vn.

[0590] The Nbit field 261 in the illustrated example includes subfields A 265, B 266, and C 267. In this example, A 265 and B 266 are each 1-bit subfields, while C 267 is a 2-bit subfield. Other examples may include differently sized subfields 265, 266, and 267. Field A 265 and field B 266 may represent fields that store first and second most significant bits of Nbit field 261, while field C 267 may represent a field which stores the least significant bits of the 261 Nbits field.

[0591] Portion 258B may also include an AddAmbHoalnfoChannel field 268. The AddAmbHoalnfoChannel field 268 may represent a field that stores information for additional environment HOA coefficients. As shown in the example in Figure 10C, the AddAmbHoalnfoChannel 268 includes a CodedAmbCoeffldx field 246, an AmbCoeffldxTransition field 247. The CodedAmbCoeffldx field 246 can represent a field that stores an index of an additional ambient HOA coefficient. The field configured to store data indicative if, in this frame, an additional ambient HOA coefficient is either gradually increasing or decreasing.

[0592] Figure 10C(i) is a diagram illustrating an alternative example of a portion 258B' of auxiliary channel information that may specify the compressed special components in greater detail. In the example of Figure 10C(i), portion 258B' includes a frame header 259 that includes an Nbits field 261. Nbits field 261 represents a field that can specify an identified nbits value for use in decompressing the spatial components. vl to vn.

[0593] Figure 10D is a diagram illustrating a portion 258C of bit stream 21 in greater detail. Portion 258C is similar to portion 258, except that frame header 259 and byte-to-zero alignment 264 have been removed, while Nbit field 261 has been added before each of the bit streams for vl through vn , as shown in the example in Figure 10D.

[0594] Figure 10D(i) is a diagram illustrating a portion 258C of bit stream 21 in greater detail. Portion 258C is similar to portion 258C except that portion 258C does not include prediction mode field 262 for each of vectors V vl through vn.

[0595] Figure 10E is a diagram illustrating a portion 258D of bit stream 21 in greater detail. Portion 258D is similar to portion 258B, except that frame header 259 and byte-to-zero alignment 264 have been removed, while Nbit field 261 has been added before each of the bit streams for vl through vn , as shown in the example in Figure 10E.

[0596] Figure 10E(i) is a diagram illustrating a portion 258D' of bit stream 21 in greater detail. Portion 258D' is similar to portion 258D except that portion 258D' does not include prediction mode field 262 for each of vectors V vl through vn. In this regard, the audio coding device 20 can generate a bit stream 21 that does not include the prediction mode field 262 for each compressed vector V, as demonstrated with reference to the examples of Figures 10C(i), 10D(i). ) and 10E(i).

[0597] Figure 1 OF is a diagram illustrating, in a different way, the portion 250 of the bit stream 21 shown in the example of Figure 10A. The portion 250 shown in the example in Figure 10D, includes a field HOAOrder (which was not shown in the example in Figure 10F for ease of illustration purposes), a field MinAmbHoaOrder (which, again, was not shown in the example in Figure 10 for ease of for illustration purposes), the direction information field 253, the CodedSpatiallnterpolationTime field 254, the SpatiallnterpolationMethod field 255, the CodedVVecLength field 256, and the gain information field 257. As shown in the example in Figure 10F, the CodedSpatiallnterpolationTime field 254 can comprises a three-bit field, the SpatiallnterpolationMethod field 255 may comprise a one-bit field, and the CodedVVecLength field 256 may comprise a two-bit field.

[0598] Figure 10G is a diagram illustrating a portion 248 of bit stream 21 in greater detail. Portion 248 represents a unified speech/audio encoder (USAC) three-dimensional (3D) payload that includes a HOAframe field 249 (which may also be denoted as the auxiliary track information, auxiliary channel information, or channel bitstream information). assistant). As shown in the example of Figure 10E, the exploded view of the HOAFrame field 249 may be similar to the portion 258B of the bit stream 21 shown in the example of Figure 10C. The “ChannelSidelnfoData” includes a ChannelType field 269, which was not shown in the example of Figure 10C for ease of illustration purposes, field A 265 indicated as “ba” in the example of Figure 10E, field B 266 indicated as “bb” in the example of Figure 10E and the C field 267 indicated as “unitC” in the example of Figure 10E. The ChannelType field indicates whether the channel is a direction-based signal, a vector-based signal, or an additional ambient HOA coefficient. Between different ChannelSidelnfoData there are AddAmbHoalnfoChannel fields 268 with the different V vector bit streams indicated in gray (eg “bit stream for vl” and “bit stream for V2”).

[0599] Figures 10H-100(ii) are diagrams illustrating several other exemplary portions 248H to 2480 of bitstream 21 along with attached HOAconfig portions 250H to 2500 in greater detail. Figures 10H(i) and 10H(ii) illustrate a first exemplary bit stream 248H and attached HOA configuration portion 250H that were generated to correspond to case 0 in the above pseudocode. In the example of Figure 10H(i), the HOAconfig 25 OH portion includes a CodedVVecLength 256 syntax element set to indicate that all elements of a vector V are encoded, for example, all 16 elements of vector V. The HOAconfig portion 25 The H also includes a SpatiallnterpolationMethod 255 syntax element established to indicate that the spatiotemporal interpolation interpolation function is a raised cosine. The HOAconfig 25 OH portion further includes a CodedSpatiallnterpolationTime 254 set to indicate an interpolated sample duration of 256. The HOAconfig 25 OH portion additionally includes a MinAmbHoaOrder 150 syntax element set to indicate that the MinimumHOA order of the ambient HOA content is one, where the audio decoding device 24 can derive a MinNumofCoeffsForAmbHOA syntax element that is equal to (1+1)2 or four. The HOAconfig 25 OH portion includes a HoaOrder syntax element 152 set to indicate the HOA order of the content is equal to three (or, in other words, N = 3), where the audio decoding device 24 can derive from a NumOfHoaCoeffs to be equal to (N+ 1)2 or 16.

[0600] As further shown in the example of Figure 10H(i), portion 248H includes a unified speech and audio coding (USAC) three-dimensional (USAC-3D) audio frame in which two HOA frames 249A and 249B are stored in a USAC extension payload due to the fact that two audio frames are stored within one USAC-3D frame when spectral band replication (SBR) is enabled. The audio decoding device 24 may derive from several flexible transport channels as a function of a numHOATransportChannels syntax element and a MinNumOfCoeffsForAmbHOA syntax element. In the following examples, it is assumed that the numHOATransportChannels syntax element is equal to 7 and the MinNumOfCoeffsForAmbHOA syntax element is equal to four, where the number of flexible transport channels is equal to the numHOATransportChannels syntax element minus the MinNumOfCoeffsForAmbHOA syntax element (or three).

[0601] Figure 10H(ii) illustrates frames 249A and 249B in greater detail. As shown in the example of Figure 10H(ii), frame 249A includes ChannelSidelnfoData (CSID) fields 154 to 154C, a HOAGainCorrectionData (HOAGCD) field, VVectorData fields 156 and 156B, and HOAPredictionlnfo fields. The CSID field 154 includes the C unit 267, bb 266 and ba265 along with the ChannelType 269, each of which are set to the corresponding values 01, 1, 0 and 01 shown in the example of Figure 10H(i). The CSID field 154B includes the C unit 267, bb 266 and ba265 along with the ChannelType 269, each of which are set to the corresponding values 01, 1, 0 and 01 shown in the example of Figure 10H(ii). The CSID field 154C includes the ChannelType field 269 which has a value of 3. Each of the CSID fields 154 to 154C correspond to those of the respective transport channels 1, 2 and 3. Effectively, each CSID field 154 to 154C indicates whether the payload corresponding 156 and 156B are direction-based signals (when the corresponding ChannelType is equal to zero), vector-based signals (when the corresponding ChannelType is equal to one), an additional Environment HOA coefficient (when the corresponding ChannelType is equals two) or empty (when the ChannelType equals three).

[0602] In the example of Figure 10H(ii), frame 249A includes two vector-based signals (due to ChannelType 269 equal to 1 in CSID fields 154 and 154B) and an empty one (due to ChannelType 269 equal to 3 in fields CSID 154C). Due to the preceding HOAconfig portion 25 OH, the audio decoding device 24 can determine that all 16 V vector elements are encoded. Accordingly, VVectorData 156 and 156B each include all 16 vector elements, each of which is uniformly quantized to 8 bits. As noted by footnote 1, the number and indices of encoded VVectorData elements are specified by the CodedVVecLength=0 parameter. Also, as verified by the single asterisk (*), the encoding scheme is signaled by NbitsQ = 5 in the CSID field for the corresponding transport channel.

[0603] In frame 249B, the CSID field 154 and 154B are the same as those in frame 249, while the CSID field 154C of frame 249B is switched to a ChannelType of one. The CSID field 154C of frame 249B therefore includes Cbflag 267, Pflag 267 (which indicates Huffman encoding), and Nbits 261 (equals twelve). As a result, frame 249B includes a third VVectorData field 156C that includes 16 V vector elements, each of which is uniformly quantized to 12 bits and Huffman encoded. As seen above, the number and indices of encoded VVectorData elements are specified by the parameter CodedVVecLength = 0, while the Huffman encoding scheme is signaled by NbitsQ = 12, CbFlag = 0, and Pflag = 0 in the CSID field 154C for that channel. private transport (eg transport channel #3) .

[0604] The example of Figures 10I(i) and 10I(ii) illustrates a second exemplary bit stream 2481 and attached HOA configuration portion 2501 that were generated to correspond to case 0 above in the above pseudocode. In the example of Figure 10I(i), the HOAconfig portion 2501 includes a CodedVVecLength syntax element 256 set to indicate that all elements of a V vector are encoded, for example, all 16 elements of a V vector. The HOAconfig portion 2501 also includes a SpatiallnterpolationMethod 255 syntax element established to indicate that the spatiotemporal interpolation interpolation function is a raised cosine. The HOAconfig portion 2501, in addition, includes a CodedSpatiallnterpolationTime 254 set to indicate an interpolated sample duration of 256.

[0605] HOAconfig portion 2501 additionally includes a MinAmbHoaOrder syntax element 150 established to indicate that the MinimumHOA order of the ambient HOA content is one, where the audio decoding device 24 may derive from a MinNumofCoeffsForAmbHOA syntax element that is equals (1+1)2 or four. The audio decoding device 24 can also derive from a MaxNoofAddActiveAmbCoeffs syntax element as established for a difference between the NumOfHoaCoeff syntax element and the MmNumOfCoeffsForAmbHOA, which is assumed in this example to be equal to 16-4 or 12. audio 24 can also derive from an AmbAsignmBits syntax element as established for ceil (log2(MaxNoOfAddActiveAmbCoeffs)) = ceil(log2(12)) = 4. The HOAconfig 25 portion includes a HoaOrder syntax element 152 set to indicate the HOA order of content is equal to three (or, in other words, N = 3), where the audio decoding device 24 can derive a NumOfHoaCoeffs to be equal to (N+ 1)2 or 16.

[0606] As further shown in the example of Figure 10I(i), portion 248H includes a USAC-3D audio frame in which two HOA frames 249C and 249D are stored in a USAC extension payload due to those two frames audio files to be stored within a USAC-3D frame when spectral band replication (SBR) is enabled. The audio decoding device 24 may derive from several flexible transport channels as a function of a numHOATransportChannels syntax element and a MinNumOfCoeffsForAmbHOA syntax element. In the following examples, it is assumed that the numHOATransportChannels syntax element is equal to 7 and the MinNumOfCoeffsForAmbHOA syntax element is equal to four, where the number of flexible transport channels is equal to the numHOATransportChannels syntax element minus the MinNumOfCoeffsForAmbHOA syntax element (or three).

[0607] Figure 10I(ii) illustrates frames 249C and 249D in greater detail. As shown in the example of Figure 10I(ii), frame 249C includes CSID fields 154-154C and VVectorData fields 156. CSID field 154 includes CodedAmbCoeffldx 246, AmbCoeffldxTransition 247 (where the double asterisk (**) indicates that , for flexible transport channel #1, the internal state of the decoder is assumed in this document to be AmbCoeffldxTransitionState = 2, which results in the bitfield CodedAmbCoeffldx being signaled or otherwise specified in the bitstream) and the ChannelType 269 (which is equal to two, signaling that the corresponding payload is an additional ambient HOA coefficient). Audio decoding device 24 can derive from AmbCoeffldx as equal to CodedAmbCoeffldx+l+MinNumOfCoeffsForAmbHOA or 5 in this example. CSID field 154B includes C unit 267, bb 266 and ba265 along with ChannelType 269, each of which are set to the corresponding values 01, 1, 0 and 01 shown in the example of Figure 10I(ii). The CSID field 154C includes the ChannelType field 269 which has a value of 3.

[0608] In the example of Figure 10I(ii), frame 249C includes a single vector-based signal (due to ChannelType 269 equals 1 in CSID fields 154B) and an empty one (due to ChannelType 269 equals 3 in CSID fields 154C). Due to the preceding HOAconfig portion 2501, the audio decoding device 24 can determine that all 16 V-vector elements are encoded. Accordingly, the VVectorData 156 includes all 16 vector elements, each of which is uniformly quantized to 8 bits. As noted by footnote 1, the number and indices of encoded VVectorData elements are specified by the CodedVVecLength=0 parameter. Also, as verified by footnote 2, the encoding scheme is signaled by NbitsQ = 5 in the CSID field for the corresponding transport channel.

[0609] In frame 249D, field CSID 154 includes an AmbCoeffldxTransition 247 which indicates that no transition has taken place and therefore CodedAmbCoeffldx 246 can be implied from the previous frame and need not be signaled or otherwise specified again. The CSID fields 154B and 154C of frame 249D are the same as those for frame 249C, and thus, similar to frame 249C, frame 249D includes a single VVectorData field 156, which includes all 16 vector elements, each of which are uniformly quantized with 8 bits.

[0610] Figures 10J(i) and 10J(ii) illustrate a first exemplary bit stream 248 J and HO configuration portion A append 250 J that was generated to correspond to case 1 in the above pseudocode. In the example of Figure s10J(i), the HOAconfig 250J portion includes a CodedVVecLength 256 syntax element set to indicate that all elements of a V array are encoded, except for elements 1 to an MmNumOfCoeffsForAmbHOA syntax element and those elements specified in a ContAddAmbHoaChan syntax element (assumed to be zero in this example). The HOAconfig portion 250J also includes a SpatiallnterpolationMethod 255 syntax element established to indicate that the interpolation function of the spatiotemporal interpolation is a raised cosine. The HOAconfig portion 250J further includes a CodedSpatiallnterpolationTime 254 set to indicate an interpolated sample duration of 256. The HOAconfig portion 250J additionally includes a MinAmbHoaOrder syntax element 150 set to indicate that the MinimumHOA order of the ambient HOA content is one, where the audio decoding device 24 can derive a MinNumofCoeffsForAmbHOA syntax element that is equal to (1+1)2 or four. The HOAconfig portion 250J includes a HoaOrder syntax element 152 set to indicate the HOA order of the content to be equal to three (or, in other words, N = 3), where the audio decoding device 24 can derive from a NumOfHoaCoeffs to be equal to (N + 1)2 or 16.

[0611] As further shown in the example of Figure 10J(i), portion 248J includes a USAC-3D audio frame in which two HOA frames 249E and 249F are stored in a USAC extension payload due to those two audio frames that are stored within a USAC-3D frame when spectral band replication (SBR) is enabled. The audio decoding device 24 may derive from several flexible transport channels as a function of a numHOATransportChannels syntax element and an MmNumOfCoeffsForAmbHOA syntax element. In the following examples, it is assumed that the numHOATransportChannels syntax element is equal to 7 and the MinNumOfCoeffsForAmbHOA syntax element is equal to four, where the number of flexible transport channels is equal to the numHOATransportChannels syntax element minus the MinNumOfCoeffsForAmbHOA syntax element (or three).

[0612] Figure 10J(ii) illustrates frames 249E and 249F in greater detail. As shown in the example of Figure 10J(ii), frame 249E includes CSID fields 154-154C and VVectorData fields 156 and 156B. The CSID field 154 includes the C unit 267, bb 266 and ba265 along with the ChannelType 269, each of which are set to the corresponding values 01, 1, 0 and 01 shown in the example of Figure 10J). The CSID field 154B includes the C unit 267, bb 266 and ba265 along with the ChannelType 269, each of which are set to the corresponding values 01, 1, 0 and 01 shown in the example of Figure 10J). The CSID field 154C includes the ChannelType field 269 which has a value of 3. Each of the CSID fields 154 to 154C correspond to those of the respective transport channels 1, 2 and 3.

[0613] In the example of Figure 10J(ii), frame 249E includes two vector-based signals (due to ChannelType 269 equal to 1 in CSID fields 154 and 154B) and an empty one (due to ChannelType 269 equal to 3 in fields CSID 154C). Due to the preceding HOAconfig 25 OH portion, the audio decoding device 24 can determine that all 12 V vector elements are encoded (where 12 is derived as (HoaOrder + 1)2 - (MinNumOfCoeffsForAmbHOA) - (ContAddAmbHoaChan) = 16 -4-0 = 12). Accordingly, VVectorData 156 and 156B each include all 12 vector elements, each of which is uniformly quantized to 8 bits. As noted by footnote 1, the number and indices of encoded VVectorData elements are specified by the CodedVVecLength=0 parameter. Also, as verified by the single asterisk (*), the encoding scheme is signaled by NbitsQ = 5 in the CSID field for the corresponding transport channel.

[0614] In frame 249F, the CSID field 154 and 154B are the same as those in frame 249E, while the CSID field 154C of frame 249F is switched to a ChannelType of one. The CSID field 154C of frame 249B therefore includes Cbflag 267, Pflag 267 (which indicates Huffman encoding), and Nbits 261 (equals twelve). As a result, frame 249F includes a third VVectorData field 156C that includes 12 V vector elements, each of which is uniformly quantized to 12 bits and Huffman encoded. As seen above, the number and indices of encoded VVectorData elements are specified by the parameter CodedVVecLength = 0, while the Huffman encoding scheme is signaled by NbitsQ = 12, CbFlag = 0, and Pflag = 0 in the CSID field 154C for that channel. private transport (eg transport channel #3) .

[0615] The example of Figures 10K(i) and 10K(ii) illustrates a second example bit stream 248K and HOA configuration portion appended 250K that were generated to correspond to case 1 in the above pseudocode. In the example of Figure s10J(i), the HOAconfig 25 OK portions include a CodedVVecLength 256 syntax element set to indicate that all elements of a V array are encoded, except for elements 1 to an MmNumOfCoeffsForAmbHOA syntax element and those elements specified in a ContAddAmbHoaChan syntax element (assumed to be one in this example). The HOAconfig 250K portion also includes a SpatiallnterpolationMethod 255 syntax element established to indicate that the spatiotemporal interpolation interpolation function is a high cosine. The HOAconfig 250K portion furthermore includes a CodedSpatiallnterpolationTime of 254 set to indicate an interpolated sample duration of 256.

[0616] The HOAconfig 25 OK portion additionally includes a MinAmbHoaOrder syntax element 150 set to indicate that the MinimumHOA order of the ambient HOA content is one, where the audio decoding device 24 may derive from a MinNumofCoeffsForAmbHOA syntax element that is equal to (1+1)2 or four. The audio decoding device 24 can also derive from a MaxNoOfAddActiveAmbCoeffs syntax element as established for a difference between the NumOfHoaCoeff syntax element and the MmNumOfCoeffsForAmbHOA, which is assumed in this example to be equal to 16-4 or 12. audio 24 can also be derived from an AmbAsignmBits syntax element as established for ceil (log2(MaxNoOfAddActiveAmbCoeffs)) = ceil(log2(12)) = 4. The HOAconfig 250K portion includes a HoaOrder 152 syntax element set to indicate the HOA order of the content to be equal to three (or, in other words, N = 3), where the audio decoding device 24 can derive a NumOfHoaCoeffs to be equal to (N + 1)2 or 16.

[0617] As further shown in the example of Figure 10K(i), the 248K portion includes a USAC-3D audio frame in which two HOA frames 249G and 249H are stored in a USAC extension payload due to those two frames audio that are stored within a USAC-3D frame when spectral band replication (SBR) is enabled. The audio decoding device 24 may derive from several flexible transport channels as a function of a numHOATransportChannels syntax element and an MmNumOfCoeffsForAmbHOA syntax element. In the following examples, it is assumed that the numHOATransportChannels syntax element is equal to 7 and the MmNumOfCoeffsForAmbHOA syntax element is equal to four, where the number of flexible transport channels is equal to the numHOATransportChannels syntax element minus the MinNumOfCoeffsForAmbHOA syntax element (or three).

[0618] Figure 10K(ii) illustrates frames 249G and 249H in greater detail. As shown in the example in Figure 10K(ii), frame 249G includes CSID fields 154-154C and VVectorData fields 156. CSID field 154 includes CodedAmbCoeffldx 246, AmbCoeffldxTransition 247 (where the double asterisk (**) indicates that , for flexible transport channel #1, the internal state of the decoder is assumed in this document to be AmbCoeffldxTransitionState = 2, which results in the bitfield CodedAmbCoeffldx being signaled or otherwise specified in the bitstream) and the ChannelType 269 (which is equal to two, signaling that the corresponding payload is an additional ambient HOA coefficient). Audio decoding device 24 can derive from AmbCoeffldx as equal to CodedAmbCoeffldx+l+MinNumOfCoeffsForAmbHOA or 5 in this example. CSID field 154B includes C unit 267, bb 266 and ba265 along with ChannelType 269, each of which are set to the corresponding values 01, 1, 0 and 01 shown in the example of Figure 10K). The CSID field 154C includes the ChannelType field 269 which has a value of 3.

[0619] In the example of Figure 10K(ii), the 249G frame includes a single vector-based signal (due to ChannelType 269 equals 1 in CSID 154B fields) and an empty one (due to ChannelType 269 equals 3 in CSID fields 154C). Due to the preceding portion HOAconfig 25 OK, the audio decoding device 24 can determine that the 11 V vector elements are encoded (where 11 is derived as (HoaOrder + 1)2 -(MinNumOfCoeffsForAmbHOA) - (ContAddAmbHoaChan) = 16- 4-1 = 12). Accordingly, the VVectorData 156 includes all 11 vector elements, each of which is uniformly quantized to 8 bits. As noted by footnote 1, the number and indices of encoded VVectorData elements are specified by the CodedVVecLength=0 parameter. Also, as verified by footnote 2, the encoding scheme is signaled by NbitsQ = 5 in the CSID field for the corresponding transport channel.

[0620] In frame 249H, field CSID 154 includes an AmbCoeffldxTransition 247 which indicates that no transition has taken place and therefore CodedAmbCoeffldx 246 can be implied from the previous frame and need not be signaled or otherwise specified again. The CSID fields 154B and 154C of frame 249H are the same as those for frame 249G, and thus, similar to frame 249G, frame 249H includes a single VVectorData field 156, which includes 11 vector elements, each of which is quantized. uniformly with 8 bits.

[0621] Figures 10L(i) and 10L(ii) illustrate a first exemplary bit stream 248L and attached HOA configuration portion 25 OL that were generated to correspond to case 2 in the above pseudocode. In the example in Figure 10L(i), the HOAconfig 25 OL portion includes a CodedVVecLength 256 syntax element set to indicate that all elements of an array V are encoded, except those elements from zero order through the order specified by the MinAmbHoaOrder syntax element. 150 (which is equal to (HoaOrder + 1)2 - (MinAmbHoaOrder + 1)2 = 16 - 4 = 12, in this example). The HOAconfig 250L portion also includes a SpatiallnterpolationMethod 255 syntax element established to indicate that the spatiotemporal interpolation interpolation function is a raised cosine. The HOAconfig portion 250L further includes a CodedSpatiallnterpolationTime 254 set to indicate an interpolated sample duration of 256. The HOAconfig portion 250L additionally includes a MinAmbHoaOrder syntax element 150 set to indicate that the MinimumHOA order of the ambient HOA content is a , wherein the audio decoding device 24 can derive a MinNumofCoeffsForAmbHOA syntax element that is equal to (1+1)2 or four. The HOAconfig portion 250L includes a HoaOrder syntax element 152 set to indicate the HOA order of the content to be equal to three (or, in other words, N = 3), where the audio decoding device 24 may derive from a NumOfHoaCoeffs to be equal to (N + 1)2 or 16.

[0622] As further shown in the example of Figure 10L(i), portion 248L includes a USAC-3D audio frame in which two HOA frames 2491 and 249J are stored in a USAC extension payload due to those two frames audio files to be stored within a USAC-3D frame when spectral band replication (SBR) is enabled. The audio decoding device 24 may derive from several flexible transport channels as a function of a numHOATransportChannels syntax element and a MinNumOfCoeffsForAmbHOA syntax element. In the following examples, it is assumed that the numHOATransportChannels syntax element is equal to 7 and the MinNumOfCoeffsForAmbHOA syntax element is equal to four, where the number of flexible transport channels is equal to the numHOATransportChannels syntax element minus the MinNumOfCoeffsForAmbHOA syntax element (or three).

[0623] Figure 10L(ii) illustrates frames 2491 and 249J in greater detail. As shown in the example of Figure 10L(ii), frame 2491 includes CSID fields 154-154C and VVectorData fields 156 and 156B. The CSID field 154 includes the C unit 267, bb 266 and ba265 along with the ChannelType 269, each of which are set to the corresponding values 01, 1, 0 and 01 shown in the example of Figure 10J). The CSID field 154B includes the C unit 267, bb 266 and ba265 along with the ChannelType 269, each of which are set to the corresponding values 01, 1, 0 and 01 shown in the example of Figure 10L(ii). The CSID field 154C includes the ChannelType field 269 which has a value of 3. Each of the CSID fields 154 to 154C correspond to those of the respective transport channels 1, 2 and 3.

[0624] In the example of Figure 10L(ii), frame 2491 includes two vector-based signals (due to ChannelType 269 equal to 1 in CSID fields 154 and 154B) and an empty one (due to ChannelType 269 equal to 3 in CSID fields 154C). Due to the preceding HOAconfig portion 25 OH, the audio decoding device 24 can determine that 12 V vector elements are encoded. Accordingly, VVectorData 156 and 156B each include 12 vector elements, each of which is uniformly quantized to 8 bits. As noted by footnote 1, the number and indices of encoded VVectorData elements are specified by the CodedVVecLength=0 parameter. Also, as verified by the single asterisk (*), the encoding scheme is signaled by NbitsQ = 5 in the CSID field for the corresponding transport channel.

[0625] In frame 249J, the CSID field 154 and 154B are the same as those in frame 2491, while the CSID field 154C of frame 249F is switched to a ChannelType of one. The CSID field 154C of frame 249B therefore includes Cbflag 267, Pflag 267 (which indicates Huffman encoding), and Nbits 261 (equals twelve). As a result, frame 249F includes a third VVectorData field 156C that includes 12 V vector elements, each of which is uniformly quantized to 12 bits and Huffman encoded. As seen above, the number and indices of encoded VVectorData elements are specified by the parameter CodedVVecLength = 0, while the Huffman encoding scheme is signaled by NbitsQ = 12, CbFlag = 0, and Pflag = 0 in the CSID field 154C for that channel. private transport (eg transport channel #3) .

[0626] The example of Figures 10M(i) and 10M(ii) illustrates a second exemplary bit stream 248M and appended HOA configuration portion 250M that were generated to correspond to case 2 in the above pseudocode. In the example of Figure 10M(i), the HOAconfig 25OM portion includes a CodedVVecLength 256 syntax element set to indicate that all elements of an array V are encoded, except those elements from zero order through the order specified by the MinAmbHoaOrder syntax element 150 (which is equal to (HoaOrder + 1)2 - (MinAmbHoaOrder + 1)2 = 16 - 4 = 12, in this example). The HOAconfig 250M portion also includes a SpatiallnterpolationMethod 255 syntax element established to indicate that the interpolation function of the spatiotemporal interpolation is high cosine. The HOAconfig 250M portion furthermore includes a CodedSpatiallnterpolationTime of 254 set to indicate an interpolated sample duration of 256.

[0627] Portion HOAconfig 250M additionally includes a MinAmbHoaOrder syntax element 150 set to indicate that the MinimumHOA order of the ambient HOA content is one, where the audio decoding device 24 may derive from a MinNumofCoeffsForAmbHOA syntax element that is equals (1+1)2 or four. The audio decoding device 24 can also derive from a MaxNoOfAddActiveAmbCoeffs syntax element as established for a difference between the NumOfHoaCoeff syntax element and the MinNumofCoeffsForAmbHOA, which is assumed in this example to be equal to 16-4 or 12. audio 24 can also be derived from an AmbAsignmBits syntax element as established for ceil (log2(MaxNoOfAddActiveAmbCoeffs)) = ceil(log2(12)) = 4. The HOAconfig 250M portion includes a HoaOrder 152 syntax element set to indicate the HOA order of the content to be equal to three (or, in other words, N = 3), where the audio decoding device 24 can derive a NumOfHoaCoeffs to be equal to (N + 1)2 or 16.

[0628] As further shown in the example of Figure 10M(i), portion 248M includes a USAC-3D audio frame in which two HOA frames 249K and 249L are stored in a USAC extension payload due to those two frames audio to be stored within a USAC-3D frame when spectral band replication (SBR) is enabled. The audio decoding device 24 may derive from several flexible transport channels as a function of a numHOATransportChannels syntax element and a MinNumofCoeffsForAmbHOA syntax element. In the following examples, it is assumed that the numHOATransportChannels syntax element is equal to 7 and the MinNumofCoeffsForAmbHOA syntax element is equal to four, where the number of flexible transport channels is equal to the numHOATransportChannels syntax element minus the MinNumofCoeffsForAmbHOA syntax element (or three).

[0629] Figure 10M(ii) illustrates frames 249K and 249L in greater detail. As shown in the example in Figure 10M(ii), frame 249K includes CSID fields 154-154C and a VVectorData field 156. CSID field 154 includes CodedAmbCoeffldx 246, AmbCoeffldxTransition 247 (where the double asterisk (**) indicates that , for flexible transport channel #1, the internal state of the decoder is assumed in this document to be AmbCoeffldxTransitionState = 2, which results in the bitfield CodedAmbCoeffldx being signaled or otherwise specified in the bitstream) and the ChannelType 269 (which is equal to two, signaling that the corresponding payload is an additional ambient HOA coefficient). Audio decoding device 24 can derive from AmbCoeffldx as equal to CodedAmbCoeffldx+l+MinNumOfCoeffsForAmbHOA or 5 in this example. The CSID field 154B includes C unit 267, bb 266 and ba265 along with the ChannelType 269, each of which are set to the corresponding values 01, 1, 0 and 01 shown in the example of Figure 10M(ii). The CSID field 154C includes the ChannelType field 269 which has a value of 3.

[0630] In the example in Figure 10M(ii), the 249K frame includes a single vector-based signal (due to ChannelType 269 equals 1 in CSID 154B fields) and an empty one (due to ChannelType 269 equals 3 in CSID fields 154C). Due to the preceding HOAconfig portion 250M, the audio decoding device 24 can determine that 12 V vector elements are encoded. Accordingly, the VVectorData 156 includes 12 vector elements, each of which is uniformly quantized to 8 bits. As noted by footnote 1, the number and indices of encoded VVectorData elements are specified by the CodedVVecLength=0 parameter. Also, as verified by footnote 2, the encoding scheme is signaled by NbitsQ = 5 in the CSID field for the corresponding transport channel.

[0631] In frame 249L, field CSID 154 includes an AmbCoeffldxTransition 247 which indicates that no transition has taken place and therefore CodedAmbCoeffldx 246 can be implied from the previous frame and need not be signaled or otherwise specified again. The CSID fields 154B and 154C of frame 249l are the same as those for frame 249k, and thus, similar to frame 249k, frame 249l includes a single VVectorData field 156, which includes 12 vector elements, each of which is quantized. uniformly with 8 bits.

[0632] Figures 10N(i) and 10N(ii) illustrate a first exemplary bit stream 248N and HO configuration portion A append 250N that was generated to correspond to case 3 in the above pseudocode. In the example of Figure 10N(i), the HOAconfig 250N portion includes a CodedVVecLength 256 syntax element set to indicate that all elements of a V array are encoded, except those elements specified in a ContAddAmbHoaChan syntax element (which is assumed to be zero in this example). The HOAconfig 250N portion also includes a SpatiallnterpolationMethod 255 syntax element established to indicate that the spatiotemporal interpolation interpolation function is a raised cosine. The HOAconfig portion 250N further includes a CodedSpatiallnterpolationTime 254 set to indicate an interpolated sample duration of 256. The HOAconfig portion 250N additionally includes a MinAmbHoaOrder 150 syntax element set to indicate that the MinimumHOA order of the HO environment A content is one, wherein the audio decoding device 24 can derive a MinNumOfCoeffsForAmbHOA syntax element to be equal to (1+1)2 or four. The HOAconfig portion 250N includes a HoaOrder syntax element 152 set to indicate the HOA order of the content is equal to three (or, in other words, N = 3), where the audio decoding device 24 can derive from a NumOfHoaCoeffs to be equal to (N+ if or 16).

[0633] As further shown in the example of Figure 10N(i), portion 248N includes a USAC-3D audio frame in which two HOA frames 249M and 249N are stored in a USAC extension payload due to those two frames audio that are stored within a USAC-3D frame when spectral band replication (SBR) is enabled. The audio decoding device 24 may derive from several flexible transport channels as a function of a numHOATransportChannels syntax element and a MinNumOfCoeffsForAmbHOA syntax element. In the following examples, it is assumed that the numHOATransportChannels syntax element is equal to 7 and the MinNumOfCoeffsForAmbHOA syntax element is equal to four, where the number of flexible transport channels is equal to the numHOATransportChannels syntax element minus the MinNumOfCoeffsForAmbHOA syntax element (or three).

[0634] Figure 10N(ii) illustrates frames 249M and 249N in greater detail. As shown in the example of Figure 10N(ii), frame 249M includes CSID fields 154-154C and VVectorData fields 156 and 156B. The CSID field 154 includes the C unit 267, bb 266 and ba265 along with the ChannelType 269, each of which are set to the corresponding values 01, 1, 0 and 01 shown in the example of Figure 10J). The CSID field 154B includes the C unit 267, bb 266 and ba265 along with the ChannelType 269, each of which are set to the corresponding values 01, 1, 0 and 01 shown in the example of Figure 10N). The CSID field 154C includes the ChannelType field 269 which has a value of 3. Each of the CSID fields 154 to 154C correspond to those of the respective transport channels 1, 2 and 3.

[0635] In the example of Figure 10N(ii), frame 249M includes two vector-based signals (due to ChannelType 269 equal to 1 in CSID fields 154 and 154B) and an empty one (due to ChannelType 269 equal to 3 in fields CSID 154C). Due to the preceding HOAconfig portion 250M, the audio decoding device 24 can determine that 16 V vector elements are encoded. Accordingly, VVectorData 156 and 156B each include 16 vector elements, each of which is uniformly quantized to 8 bits. As noted by footnote 1, the number and indices of encoded VVectorData elements are specified by the CodedVVecLength=0 parameter. Also, as verified by the single asterisk (*), the encoding scheme is signaled by NbitsQ = 5 in the CSID field for the corresponding transport channel.

[0636] In frame 249N, the CSID field 154 and 154B are the same as those in frame 249M, while the CSID field 154C of frame 249F is switched to a ChannelType of one. The CSID field 154C of frame 249B therefore includes Cbflag 267, Pflag 267 (which indicates Huffman encoding), and Nbits 261 (equals twelve). As a result, frame 249F includes a third VVectorData field 156C that includes 16 V vector elements, each of which is uniformly quantized to 12 bits and Huffman encoded. As seen above, the number and indices of encoded VVectorData elements are specified by the parameter CodedVVecLength = 0, while the Huffman encoding scheme is signaled by NbitsQ = 12, CbFlag = 0, and Pflag = 0 in the CSID field 154C for that channel. private transport (eg transport channel #3) .

[0637] The example of Figures 100(i) and 100(ii) illustrates a second exemplary bit stream 2480 and appended HOA configuration portion 2500 that were generated to correspond to case 3 in the above pseudocode. In the example in Figure 100(i), the HOAconfig portion 250O includes a CodedVVecLength 256 syntax element set to indicate that all elements of an array V are encoded, except those elements specified in a ContAddAmbHoaChan syntax element (which is assumed to be one in this example). The HOAconfig portion 250O also includes a SpatiallnterpolationMethod 255 syntax element established to indicate that the spatiotemporal interpolation interpolation function is a raised cosine. The HOAconfig 250O portion furthermore includes a CodedSpatiallnterpolationTime 254 set to indicate an interpolated sample duration of 256.

[0638] The HOAconfig portion 250O additionally includes a MinAmbHoaOrder syntax element 150 set to indicate that the MinimumHOA order of the ambient HOA content is one, where the audio decoding device 24 may derive from a MinNumofCoeffsForAmbHOA syntax element that is equals (1+1)2 or four. The audio decoding device 24 can also derive from a MaxNoOfAddActiveAmbCoeffs syntax element as established for a difference between the NumOfHoaCoeff syntax element and the MinNumofCoeffsForAmbHOA, which is assumed in this example to be equal to 16-4 or 12. audio 24 can also be derived from an AmbAsignmBits syntax element as established for ceil (log2(MaxNoOfAddActiveAmbCoeffs)) = ceil(log2(12)) = 4. The HOAconfig 250O portion includes a HoaOrder 152 syntax element set to indicate the HOA order of the content to be equal to three (or, in other words, N = 3), where the audio decoding device 24 can derive a NumOfHoaCoeffs to be equal to (N + 1)2 or 16.

[0639] As further shown in the example of Figure 100(i), portion 248O includes a USAC-3D audio frame in which two HOA frames 2491 and 249P are stored in a USAC extension payload due to those two frames audio files to be stored within a USAC-3D frame when spectral band replication (SBR) is enabled. The audio decoding device 24 may derive from several flexible transport channels as a function of a numHOATransportChannels syntax element and a MinNumOfCoeffsForAmbHOA syntax element. In the following examples, it is assumed that the numHOATransportChannels syntax element is equal to 7 and the MinNumOfCoeffsForAmbHOA syntax element is equal to four, where the number of flexible transport channels is equal to the numHOATransportChannels syntax element minus the MinNumOfCoeffsForAmbHOA syntax element (or three).

[0640] Figure 100(ii) illustrates frames 2490 and 249P in greater detail. As shown in the example in Figure 100(ii), frame 2490 includes CSID fields 154-154C and a VVectorData field 156. CSID field 154 includes CodedAmbCoeffldx 246, AmbCoeffldxTransition 247 (where the double asterisk (**) indicates which, for flexible transport channel #1, the internal state of the decoder is assumed in this document to be AmbCoeffldxTransitionState = 2, which results in the bitfield CodedAmbCoeffldx being signaled or otherwise specified in the bitstream) and the ChannelType 269 (which equals two, signaling that the corresponding payload is an additional ambient HOA coefficient). Audio decoding device 24 can derive from AmbCoeffldx as equal to CodedAmbCoeffldx+l+MinNumOfCoeffsForAmbHOA or 5 in this example. CSID field 154B includes C unit 267, bb 266 and ba265 along with ChannelType 269, each of which are set to the corresponding values 01, 1, 0 and 01 shown in the example of Figure 100(ii). The CSID field 154C includes the ChannelType field 269 which has a value of 3.

[0641] In the example in Figure 100(ii), frame 2490 includes a single vector-based signal (due to ChannelType 269 equals 1 in CSID fields 154B) and an empty one (due to ChannelType 269 equals 3 in CSID fields 154C). Due to the preceding portion HOAconfig 250O, the audio decoding device 24 may determine that 16 minus the one specified by the ContAddAmbHoaChan syntax elements (for example, where the vector element associated with an index of 6 is specified as the ContAddAmbHoaChan syntax element ) or 15 V vector elements are encoded. Accordingly, the VVectorData 156 includes 15 vector elements, each of which is uniformly quantized to 8 bits. As noted by footnote 1, the number and indices of encoded VVectorData elements are specified by the CodedVVecLength=0 parameter. Also, as verified by footnote 2, the encoding scheme is signaled by NbitsQ = 5 in the CSID field for the corresponding transport channel.

[0642] In frame 249P, field CSID 154 includes an AmbCoeffldxTransition 247 which indicates that no transition has taken place and therefore CodedAmbCoeffldx 246 can be implied from the previous frame and need not be signaled or otherwise specified again. The CSID fields 154B and 154C of frame 249P are the same as those for frame 2490 and thus similar to frame 2490, frame 249P includes a single VVectorData field 156, which includes 15 vector elements, each of which is uniformly quantized with 8 bits.

[0643] Figures 11A to 11G are block diagrams illustrating, in greater detail, various units of the audio decoding device 24 shown in the example of Figure 5. Figure 11A is a block diagram illustrating, in greater detail, the extraction unit 72 of the audio decoding device 24. As shown in the example of Figure 11A, the extraction unit 72 may include a 270-mode analysis unit, a 272-mode configuration unit ("default configuration unit"). mode 272”) and a configurable extraction unit 274.

[0644] Mode parsing unit 270 may represent a unit configured to parse the syntax element checked above indicative of an encoding mode (e.g., the ChannelType syntax element shown in the example in Figure 10E) used to encode the coefficients A of HO 11 to form bit stream 21. Mode parsing unit 270 may pass the determined syntax element to mode setting unit 272. Mode setting unit 272 may represent a unit configured to configure the configurable extraction unit 274 based on parsed syntax element. Mode setting unit 272 may configure configurable extraction unit 274 to extract a direction-based encoded representation of HOA coefficients 11 from bitstream 21 or extract a vector-based encoded representation of HOA coefficients 11 from bitstream 21 based on the parsed syntax elements.

[0645] When a direction-based encoding has been performed, the configurable extraction unit 274 can extract the direction-based version of the HOA coefficients 11 and the syntax elements associated with that encoded version (which is indicated as information based on direction 91 in the example of Figure 11 A). This direction-based information 91 may include the direction-based information 253 shown in the example of Figure 10D and the direction-based SideChannellnfoData shown in the example of Figure 10E as defined by a ChannelType of zero.

[0646] When the syntax element indicates that the HOA 11 coefficients were encoded using vector-based synthesis (e.g. when the ChannelType syntax element is equal to one), the configurable extraction unit 274 can extract the coded foreground V[k] vectors 57, the coded environment HOA coefficients 59 and the coded nFG signals 59. The configurable extraction unit 274 can also, upon determining that the syntax element indicates that the HOA coefficients 11 were encoded using a vector-based synthesis, extract the CodedSpatiallnterpolationTime syntax element 254 and the SpatiallnterpolationMethod 255 syntax element from bitstream 21, parsing these syntax elements 254 and 255 for the spatiotemporal interpolation unit 76.

[0647] Figure 11B is a block diagram illustrating, in greater detail, the quantization unit 74 of the audio decoding device 24 shown in the example of Figure 5. The quantization unit 74 may represent a unit configured to operate in a different manner. a reciprocal way to the quantization unit 52 shown in the example of Figure 4 so as to decode and dequantize the entropy of the coded foreground vectors V[k] 57 and thereby generate reduced foreground 55k V[k] vectors. The scalar/entropy dequantization unit 984 may include a category/residual decoding unit 276, a prediction unit 278, and a uniform dequantization unit 280.

[0648] Category/residual decoding unit 276 may represent a unit configured to perform Huffman decoding against encoded foreground V[k] vectors 57 using the Huffman table identified by the Huffman table information 241 ( which is, as noted above, expressed as a syntax element in bitstream 21). The category/residual decoding unit 276 may output quantized foreground vectors V[k] to the prediction unit 278. The prediction unit 278 may represent a unit configured to perform prediction with respect to the foreground vectors V[k] plane quantized based on prediction mode 237, output quantized foreground vectors V[k] augmented to uniform dequantization unit 280. Uniform dequantization unit 280 may represent a unit configured to perform dequantization with respect to vectors V[k quantized foreground ] augmented based on the value of nbits 233, which outputs the reduced foreground V[k] 55k vectors.

[0649] Figure 11C is a block diagram illustrating, in greater detail, the psychoacoustic decoding unit 80 of the audio decoding device 24 shown in the example of Figure 5. As seen above, the psychoacoustic decoding unit 80 can operate in a reciprocal manner to the psychoacoustic audio coding unit 40 shown in the example of Figure 4 so as to decode the encoded ambient HOA coefficients 59 and the encoded nFG signals 61 and thereby generate energy compensated ambient HOA coefficients 47' and interpolated nFG signals 49' (which may also be called interpolated nFG audio objects 49'). Psychoacoustic decoding unit 80 may pass energy compensated ambient HOA coefficients 47' to HOA coefficient formulation unit 82 and nFG signals 49' to reordering 84. Psychoacoustic decoding unit 80 may include a plurality of 80-80N audio decoders similar to the psychoacoustic audio coding unit 40. The 80-80N audio decoders may be instantiated by or otherwise included within the psychoacoustic audio coding unit 40 in sufficient quantity to support, as noted above , simultaneous decoding of each channel of the background HOA coefficients 47' and each signal of the nFG signals 49'.

[0650] Figure 11D is a block diagram illustrating, in greater detail, the reordering unit 84 of the audio decoding device 24 shown in the example of Figure 5. The reordering unit 84 may represent a unit configured to operate in reverse order. a reciprocal manner similar to that described above with respect to the reordering unit 34. The reordering unit 84 may include a vector reordering unit 282, which may represent a unit configured to receive syntax elements 205 indicative of the original order of the first components. plane of HOA coefficients 11. The extractor unit 72 can parse these syntax elements 205 of the bit stream 21 and pass the syntax element 205 to the reorder unit 84. The vector reorder unit 282 can, based on these elements reorder syntax 205, rearrange the interpolated nFG signals 49' and the reduced foreground 55k V[k] vectors to generate 49” re nFG signals ordinates and reordered foreground 55k'V[k] vectors. The reordering unit 84 may output the reordered nFG signals 49" to the foreground formulation unit 78 and the reordered foreground vectors 55k' V[k] to the spatiotemporal interpolation unit 76.

[0651] Figure HE is a block diagram illustrating in greater detail the spatiotemporal interpolation unit 76 of the audio decoding device 24 shown in the example of Figure 5. The spatiotemporal interpolation unit 76 may operate in a manner similar to that described above with respect to the spatiotemporal interpolation unit 50. The spatiotemporal interpolation unit 76 may include a V interpolation unit 284, which may represent a unit configured to receive vectors 55k' V[k ] reordered foreground vectors and perform spatiotemporal interpolation against the reordered foreground 55k' V[k] and reordered foreground 55k-i' V[k-1] vectors to generate 55k” V[k] ] interpolated foreground. The V interpolation unit 284 can perform interpolation based on the CodedSpatiallnterpolationTime syntax element 254 and the SpatiallnterpolationMethod syntax element 255. In some examples, the V interpolation unit 285 can interpolate the V vectors for the duration specified by the CodedSpatiallnterpolationTime syntax element. 254 using the interpolation type identified by the SpatiallnterpolationMethod syntax element 255. The spatiotemporal interpolation unit 76 can transmit the interpolated foreground 55k” V[k] vectors to the foreground formulation unit 78.

[0652] Figure 1 IF is a block diagram illustrating in greater detail the foreground formulation unit 78 of the audio decoding device 24 shown in the example of Figure 5. The foreground formulation unit 78 may include a multiplication unit 286, which may represent a unit configured to perform matrix multiplication with respect to interpolated foreground vectors 55A"V[k] and reordered nFG signals 49" to generate foreground HOA coefficients 65.

[0653] Figure 11G is a block diagram illustrating in greater detail the HOA coefficient formulation unit 82 of the audio decoding device 24 shown in the example of Figure 5. The HOA coefficient formulation unit 82 may include an addition unit 288, which may represent a unit configured to add foreground HOA coefficients 65 to room HOA channels 47' in order to obtain HOA coefficients 11'.

[0654] Figure 12 is a diagram illustrating an exemplary audio ecosystem that can perform various aspects of the techniques described in this disclosure. As illustrated in Figure 12, the audio ecosystem 300 may include acquisition 301, editing 302, encoding 303, transmission 304, and playback 305.

[0655] Acquisition 301 may represent audio ecosystem techniques 300 where audio content is acquired. Examples of acquisition 301 include, but are not limited to, recording sound (e.g., live sound), audio generation (e.g., audio objects, foley production, sound synthesis, simulations), and the like. In some examples, the sound may be recorded at shows, sporting events and when conducting surveillance. In some examples, audio may be generated when performing simulations and creation/mixing (eg movies, games). Audio objects can be used in Hollywood (eg IMAX studios). In some examples, acquisition 301 may be performed by a content creator, such as content creator 12 in Figure 3.

[0656] Edit 302 may represent audio ecosystem techniques 300, where audio content is edited and/or modified. As an example, audio content can be edited by combining multiple units of audio content into a single unit of audio content. As another example, audio content can be edited by adjusting the actual audio content (for example, adjusting the levels of one or more frequency components of the audio content). In some examples, 302 editing can be performed by an audio editing system, such as audio editing system 18 of Figure 3. In some examples, 302 editing can be performed on a mobile device, such as one or more of the devices furniture illustrated in Figure 29.

[0657] Encoding 303 may represent audio ecosystem techniques 300, where audio content is encoded into a representation of audio content. In some examples, the representation of the audio content may be a bitstream, such as bitstream 21 in Figure 3. In some examples, encoding 302 may be performed by an audio encoding device, such as an audio encoding device. 20 of Figure 3.

[0658] The stream 304 may represent the audio ecosystem elements 300, where audio content is transported from a content creator to a content consumer. In some examples, the audio content may be transported in real-time or near-real time. For example, audio content can be streamed continuously to the content consumer. In some examples, audio content can be transported by encoding the audio content onto a medium, such as a computer-readable storage medium. For example, audio content can be stored on a disc, drive and the like (e.g. Blu-ray disc, memory card, hard drive, etc.)

[0659] Playback 305 may represent audio ecosystem techniques 300 where audio content is processed and played back to the content consumer. In some examples, playback 305 may include processing a 3D sound field based on one or more aspects of a playback environment. In other words, playback 305 can be based on a local acoustic environment.

[0660] Figure 13 is a diagram illustrating an example of the audio ecosystem of Figure 12 in greater detail. As illustrated in Figure 13, the audio ecosystem 300 may include audio content 308, movie studios 310, music studios 311, game audio studios 312, channel-based audio content 313, encoding engines 314, game audio 315, game audio encoding/rendering mechanisms 316, and delivery systems 317. An exemplary game audio studio 312 is illustrated in Figure 26. Some exemplary game audio encoding/rendering mechanisms 316 are illustrated in Fig. Figure 27.

[0661] As illustrated by Figure 13, movie studios 310, music studios 311, and game audio studios 312 may receive audio content 308. In some examples, audio content 308 may represent acquisition broadcast 301 of Figure 12 Film studios 310 can output audio content based on channel 313 (eg, 2.0, 5.1, and 7.1) as using a digital audio workstation (DAW). Music studios 310 can output audio content based on channel 313 (eg, 2.0 and 5.1) as using a DAW. However, encoding engines 314 can receive and encode channel-based audio content 313 based on one or more codecs (e.g., AAC, AC3, Dolby True HD, Dolby Digital Plus, and DTS Master Audio) for output by delivery systems 317. Thus, encoding mechanisms 314 may be an example of encoding 303 of Figure 12. Game audio studios 312 may output one or more game audio trunks 315, as using a DAW . Game audio encoding/rendering engines 316 can encode and/or render audio trunks 315 into channel-based audio content for broadcast by delivery systems 317. In some examples, broadcast movie studios 310, studios 311 and game audio studios 312 may represent broadcast edition 302 of Figure 12. In some examples, broadcast 314 encoding mechanisms and/or game audio encoding/rendering mechanisms 316 may be ported to systems delivery 317 via the transmission techniques 304 of Figure 12.

[0662] Figure 14 is a diagram illustrating another example of the audio ecosystem of Figure 12 in greater detail. As illustrated in Figure 14, the audio ecosystem 300B may include broadcast recording audio objects 319, professional audio systems 320, consumer device capture 322, HOA audio format 323, on-device rendering 324, consumer audio , TV and accessories 325 and car audio systems 326.

[0663] As illustrated in Figure 14, broadcast recording audio objects 319, professional audio systems 320, and consumer device capture 322 can all encode their broadcast using HOA 323 audio format. In this way, audio content can be encoded using HOA 323 audio format into a single replayable representation using on-device 324 rendering, consumer audio, 325 TV and accessories, and 325 audio systems. car 326. In other words, the single representation of the audio content can be played back on a generic audio playback system (that is, as opposed to requiring a particular setting like 5.1, 7.1, etc.) .

[0664] Figures 15A and 15B are diagrams illustrating other examples of the audio ecosystem of Figure 12 in greater detail. As illustrated in Figure 15A, the audio ecosystem 300C may include acquisition elements 331 and playback elements 336. Acquisition elements 331 may include wired and/or wireless acquisition device 332 (e.g., Eigen microphones), capture surround sound on device 334 and mobile devices 335 (e.g. smart phones and tablet computers). In some examples, wired and/or wireless acquisition device 332 may be coupled to mobile device 335 via wired and/or wireless communication channel(s) 333.

[0665] According to one or more techniques of this disclosure, mobile device 335 can be used to acquire a sound field. For example, mobile device 335 may acquire a sound field via wired and/or wireless acquisition device 332 and/or capture surround sound on device 334 (e.g., a plurality of microphones integrated into mobile device 335) . Mobile device 335 may then encode the acquired sound field at HO As 337 for playback by one or more of playback elements 336. For example, a user of mobile device 335 may record (acquire a sound field of) an event at a time. (eg, a meeting, conference, game, show, etc.) and encode the recording into HOAs.

[0666] Mobile device 335 may also utilize one or more of playback elements 336 to reproduce the HOA encoded sound field. For example, mobile device 335 may decode the HOA encoded sound field and output a signal to one or more of the playback elements 336 which causes the one or more of the playback elements 336 to recreate the sound field. As an example, mobile device 335 may utilize wireless and/or wired communication channels 338 to output the signal to one or more speakers (e.g., speaker arrays, soundbars, etc.). As another example, mobile device 335 may utilize docking solutions 339 to output the signal to one or more docking stations and/or one or more docking speakers (e.g., sound systems in smart cars and/or homes) . As another example, mobile device 335 may use headphone rendering 340 to output the signal to a set of headphones, for example, to create realistic binaural sound.

[0667] In some examples, a particular mobile device 335 can both acquire a 3D sound field and reproduce the same 3D sound field at a later time. In some examples, mobile device 335 may acquire a 3D sound field, encode the 3D sound field in HOA, and transmit the encoded 3D sound field to one or more other devices (e.g., other mobile devices and/or other devices). non-mobile) for playback.

[0668] As illustrated in Figure 15B, the 300D audio ecosystem may include 343 audio content, 344 game studios, 345 encoded audio content, 345 rendering engines, and 347 delivery systems. In some examples, game studios 344 modules may include one or more DAWs that can support HOA signal editing. For example, one or more DAWs may include HOA plug-ins and/or tools that can be configured to operate with (eg work with) one or more game audio systems. In some examples, 344 game studios may issue new trunk formats that support HOA. In either case, game studios 344 can output encoded audio content 345 to render engines 346 that can render a sound field for playback by delivery systems 347.

[0669] Figure 16 is a diagram illustrating an exemplary audio encoding device that can perform various aspects of the techniques described in that disclosure. As illustrated in Figure 16, the audio ecosystem 300E may include original 3D audio content 351, encoder 352, bitstream 353, decoder 354, renderer 355, and playback elements 356. As further illustrated by Figure 16, encoder 352 may include sound field analysis and decomposition 357, background extraction 358, background saliency determination 359, audio encoding 360, background/distinctive audio extraction 361, and audio encoding 362. In some examples, the encoder 352 may be configured to perform similar operations to the audio encoding device 20 3of Figures 3 and 4. In some examples, sound field analysis and decomposition 357 may be configured to perform similar operations to the sound field analysis unit 44 of Figure 4. In some examples, background extraction 358 and background bump determination 359 can be configured to perform s similar to the BG selection unit 48 of Figure 4. In some examples, the audio encoding 360 and the audio encoding 362 can be configured to perform similar operations to the psychoacoustic audio encoder unit 40 of Figure 4. In some examples, background/distinctive audio extraction 361 can be configured to perform similar operations to background selection unit 36 in Figure 4.

[0670] In some examples, background/distinctive 361 audio extraction may parse audio content that corresponds to video frame 390 in Figure 33. For example, background/distinctive 361 audio extraction may determine that audio content that corresponds to regions 391 A-391 C is foreground audio.

[0671] As illustrated in Figure 16, encoder 352 can be configured to encode original content 351, which can have a bit rate of 25 to 75 Mbps, into bitstream 353, which can have a bit rate of 256kbps at 1.2 Mbps. Figure 17 is a diagram illustrating an example of the audio encoding device of Figure 16 in greater detail.

[0672] Figure 18 is a diagram illustrating an exemplary audio encoding device that can perform various aspects of the techniques described in that disclosure. As illustrated in Figure 18, the audio ecosystem 300E may include original 3D audio content 351, encoder 352, bitstream 353, decoder 354, renderer 355, and playback elements 356. As further illustrated by Figure 16, encoder 354 may include audio decoder 363, audio decoder 364, foreground reconstruction 365, and mixing 366. In some instances, decoder 354 may be configured to perform similar operations to the audio decoding device 24 of Figures 3 and 5. In some For examples, audio decoder 363, audio decoder 364 can be configured to perform similar operations to psychoacoustic decoding unit 80 of Figure 5. In some examples, foreground reconstruction 365 can be configured to perform similar operations to the psychoacoustic decoding unit 80 of Figure 5. foreground formulation 78 of Figure 5.

[0673] As illustrated in Figure 16, the decoder 354 can be configured to receive and decode bitstream 353 and output the resulting reconstructed 3D sound field to render 355, which can then cause one or more of the elements 356 output 3D original representation content 351. Figure 19 is a diagram illustrating an example of the audio decoding device of Figure 18 in greater detail.

[0674] Figures 20A through 20G are diagrams illustrating exemplary audio acquisition devices that can perform various aspects of the techniques described in this disclosure. Figure 20A illustrates the Eigen 370 microphone which may include a plurality of microphones that are collectively configured to record a 3D sound field. In some examples, the plurality of microphones of Eigen 370 microphones may be located on the surface of a substantially spherical ball with a radius of approximately 4 cm. In some examples, the audio encoding device 20 can be integrated into the Eigen microphone so as to output a stream of bits 17 directly from the microphone 370.

[0675] Figure 20B illustrates a production truck 372 that can be configured to receive a signal from one or more microphones, such as one or more Eigen 370 microphones. The production truck 372 may also include an audio encoder, such as an audio encoder. audio 20 of Figure 3.

[0676] Figures 20C to 20E illustrate mobile device 374 that may include a plurality of microphones that are collectively configured to record a 3D sound field. In other words, the plurality of microphones may have X, Y, Z diversity. In some examples, mobile device 374 may include microphone 376 which may be rotated to provide X, Y, Z diversity with respect to one or more other microphones of mobile device 374. Mobile device 374 may also include an audio encoder, such as audio encoder 20 of Figure 3.

[0677] Figure 20F illustrates a ruggedized video capture device 378 that can be configured to record a 3D sound field. In some examples, the ruggedized video capture device 378 may be attached to a helmet of a user performing an activity. For example, the ruggedized video capture device 378 can be attached to a white water rafting user's helmet. In this way, the 378 rugged video capture device can capture a 3D sound field that represents the action all around the user (e.g. water falling behind the user, another rafter speaking in front of the user, etc... ).

[0678] Figure 20G illustrates enhanced mobile auxiliary device 380 that can be configured to record a 3D sound field. In some examples, mobile device 380 may be similar to mobile device 335 of Figure 15, with the addition of one or more accessories. For example, an Eigen microphone can be attached to the mobile device 335 of Figure 15 to form the Mobile Enhanced Auxiliary 380. In this way, the Mobile Enhanced Auxiliary 380 can capture a higher quality version of the 3D sound field than with just the use of sound capture components built into the mobile device Enhanced Auxiliary 380.

[0679] Figures 21A through 21E are diagrams illustrating exemplary audio reproduction devices that may perform various aspects of the techniques described in that disclosure. Figures 21A and 2IB illustrate a plurality of speakers 382 and soundbars 384. In accordance with one or more techniques of this disclosure, speakers 382 and/or soundbars 384 may be arranged in any arbitrary configuration. while still reproducing a 3D sound field. Figures 21C to 21E illustrate a plurality of headphone playback devices 386 to 386C. Headphone playback devices 386 to 386C can be attached to a set top box either through a wired connection or through a wireless connection. According to one or more techniques of this disclosure, a single generic representation of a sound field can be used to render the sound field in any combination of 382 speakers, 384 soundbars, and 386 to 386C headphone playback devices. .

[0680] Figures 22A to 22H are diagrams illustrating exemplary audio playback environments in accordance with one or more techniques described in that disclosure. For example, Figure 22A illustrates a 5.1 speaker playback environment, Figure 22B illustrates a 2.0 (e.g. stereo) speaker playback environment, Figure 22C illustrates a 2.0 speaker playback environment. 9.1 speaker with full-height anterior acoustic speakers, Figures 22D and 22E each illustrate a 22.2 speaker playback environment, Figure 22F illustrates a speaker playback environment 16.0, Figure 22G illustrates an automotive speaker playback environment, and Figure 22H illustrates a mobile device with earplug playback environment.

[0681] According to one or more techniques of this disclosure, a single generic representation of a sound field can be used to render the sound field in any of the playback environments illustrated in Figures 22A to 22H. Additionally, the techniques of this disclosure enable a rendering to render a sound field of a generic representation for playback in playback environments other than those illustrated in Figures 22A through 22H. For example, if design considerations prohibit proper placement of speakers according to the 7.1 speaker playback environment (for example, if it is not possible to place a right surround speaker), the techniques of this disclosure enable a rendering to compensate with the other 6 speakers so that playback can be achieved in a 6.1 speaker playback environment.

[0682] As illustrated in Figure 23, a user can watch a sports game while wearing 386 headphones. According to one or more techniques of this revelation, the 3D sound field of the sports game can be acquired (for example, a or more Eigen microphones can be placed in and/or around the baseball stadium illustrated in Figure 24), the HOA coefficients corresponding to the 3D sound field can be obtained and transmitted to a decoder, where the decoder can determine the reconstruction of the 3D sound field based on the HOA coefficients and output the reconstructed to the 3D sound field to a renderer, where the renderer can obtain an indication as to the type of playback environment (e.g. headphones) and render the reconstructed to the 3D sound field in signals that cause the headphones to emit a 3D sound field representation of the sports game. In some examples, the renderer can get an indication as to the type of playback environment according to the techniques in Figure 25. In this way, the renderer can “adapt” for various speaker locations, type of numbers, size and also optimally equalize to the local environment.

[0683] Figure 28 is a diagram illustrating a speaker configuration that can be simulated through headphones in accordance with one or more of the techniques described in this disclosure. As illustrated by Figure 28, the techniques of such disclosure can enable a user who wears headphones 389 to experience a sound field as if the sound field were reproduced again by speakers 388. In this way, a user can hear a sound field in 3D without sound being output to a large area.

[0684] Figure 30 is a diagram illustrating a video frame associated with a 3D sound field that can be processed according to one or more techniques described in this disclosure.

[0685] Figures 31A to 31M are diagrams illustrating graphs 400A to 400M showing various simulation results of performing synthetic or recorded sound field categorization in accordance with various aspects of the techniques described in this disclosure. In the examples in Figure 31A through 31M, each of the graphs 400A through 400M includes a boundary 402 that is indicated by a dashed line and a respective audio object 404A through 404M (collectively, “the audio objects 404”) indicated by a line dashed.

[0686] When the audio objects 404 through the analyzes described above in relation to the content analysis unit 26 are determined to be below the threshold 402, the content analysis unit 26 determines that the corresponding one of the audio objects 404 represents a audio object that was recorded. As shown in the examples of Figures 3IB, 31D-31H and 31J-31L, the content analysis unit 26 determines that the audio objects 404B, 404D-404H, 404J-404L are below threshold 402 (at least +90% of time and usually 100% of the time) and therefore represents recorded audio objects. As shown in the examples of Figures 31A, 31C and 311, the content analysis unit 26 determines that the audio objects 404A, 404C and 4041 exceed the threshold 402 and therefore represent synthetic audio objects.

[0687] In the example in Figure 31M, the audio object 404M represents a mixed synthetic/recorded audio object, which has some synthetic portions (e.g. above threshold 402) and some synthetic portions (e.g. below threshold 402 ). The content analysis unit 26, in this example, identifies the synthetic and recorded portions of the audio object 404M with the result that the audio encoding device 20 generates the bit stream 21 to include both audio data encoded based on directionality and vector-based encoded audio data.

[0688] Figure 32 is a diagram illustrating a graph 406 of single values from a matrix S decomposed from higher order ambisonic coefficients in accordance with the techniques described in this disclosure. As shown in Figure 32, nonzero singular values that have large values are few. The sound field analysis unit 44 of Figure 4 can analyze these singular values to determine the background (or, in other words, predominant) components of nFG (generally represented by vectors) of the reordered vectors 33' XJS[k] and the reordered 35' V[k] vectors.

[0689] Figures 33A and 33B are diagrams illustrating respective graphs 410A and 410B showing a potential impact that reordering has when encoding vectors describing foreground components of the sound field in accordance with the techniques described in this disclosure. Graph 410A shows the result of encoding at least some of the vectors (or, in other words, the original) unordered 33 XJS[k] vectors, while graph 410B shows the result of encoding those corresponding to the 33' XJS[k] vectors ] ordered. The top plot on each of the 410A and 410B graphs shows the error in encoding, where there is probably only noticeable error in the 410B graph at frame boundaries. Accordingly, the reordering techniques described in that disclosure may facilitate or otherwise promote the encoding of mono audio objects using a legacy audio encoder.

[0690] Figures 34 and 35 are conceptual diagrams illustrating differences between energy-only and directionality-based identification of distinct audio objects, in accordance with this disclosure. In the example in Figure 34, vectors that exhibit higher energy are identified as being distinct audio objects, regardless of directionality. As shown in Figure 34, audio objects that are positioned according to higher energy values (plotted on a y-axis) are determined to be “in the foreground”, regardless of directionality (e.g. represented by directionality quotients). plotted on an x-axis).

[0691] Figure 35 illustrates the identification of distinct audio objects based on both directionality and energy, as well as according to the techniques implemented by the sound field analysis unit 44 of Figure 4. As shown in Figure 35, quotients of greater directionality are plotted heading to the left of the x-axis, and higher energy levels are plotted heading toward the top of the y-axis. In this example, the sound field analysis unit 44 can determine that distinct audio objects (e.g., that are "in the foreground") are associated with plotted vector data relatively to the top left of the graph. As an example, the sound field analysis unit 44 can determine that those vectors that are plotted in the top left quadrant of the graph are associated with distinct audio objects.

[0692] Figures 36A to 36F are diagrams illustrating projections of at least a portion of the decomposed version of spherical harmonic coefficients into the spatial domain in order to perform interpolation in accordance with various aspects of the techniques described in this disclosure. Figure 36A is a diagram illustrating projection of one or more of the 35 V[k] vectors onto a sphere 412. In the example of Figure 36A, each number identifies a different spherical harmonic coefficient projected onto the sphere (possibly associated with a row e/ or column of matrix V 19'). Different colors suggest a direction of a distinct audio component, where lighter (and progressively darker) colors indicate the primary direction of the distinct component. The spatiotemporal interpolation unit 50 of the audio encoding device 20 shown in the example of Figure 4 can perform spatiotemporal interpolation between each of the red dots to generate the sphere shown in the example of Figure 36A.

[0693] Figure 36B is a diagram illustrating projection of one or more of the vectors V[k] 35 onto a beam. The spatiotemporal interpolation unit 50 can project a row and/or column of vectors V[k] 35 or multiple rows and/or columns of vectors V[k] 35 to generate the beam 414 shown in the example of Figure 36B.

[0694] Figure 36C is a diagram illustrating a cross-section of a projection of one or more vectors of one or more of the vectors V[k] 35 onto a sphere, such as sphere 412 shown in the example of Figure 36.

[0695] Shown in Figures 36D to 36G are examples of snapshots of time (through 1 frame of about 20 milliseconds) when different sound sources (bee, helicopter, electronic music and people in a stadium) can be illustrated in a three-dimensional space.

[0696] The techniques described in this disclosure allow the representation of these different sound sources to be identified and represented using a single U.S.[k] vector and a single V[k] vector. The temporal variability of sound sources is represented by the vector U.S.[k] while the spatial distribution of each sound source is represented by the unique vector V[k]. A vector V[k] can represent the widths, location and size of the sound source. Furthermore, the single vector V[k] can be represented as a linear combination of spherical harmonic basis functions. In the plots of Figures 36D to 36G, the representation of sound sources is based on transforming the single vector V into a spatial coordinate system. Similar methods of illustrating sound sources are used in Figures 36 to 36C.

[0697] Figure 37 illustrates a representation of techniques to obtain a spatiotemporal interpolation as described in this document. The spatiotemporal interpolation unit 50 of the audio encoding device 20 shown in the example of Figure 4 can perform the spatiotemporal interpolation described below in greater detail. Spatio-temporal interpolation can include obtaining higher resolution spatial components in both the spatial and temporal dimensions. Spatial components can be based on an orthogonal decomposition of a multidimensional signal comprised of higher-order ambisonic (HOA) coefficients (or, as HOA coefficients can also be called, “spherical harmonic coefficients”).

[0698] In the illustrated graph, vectors V1 and V2 represent corresponding vectors of two different spatial components of a multidimensional signal. The spatial components can be obtained by a block decomposition of the multidimensional signal. In some examples, the spatial components result from making a SVD block shape with respect to each block (which may refer to a frame) of higher order ambisonics audio data (HOA) (where that ambisonics audio data include blocks, samples, or any other form of multichannel audio data). A variable M can be used to indicate the length of an audio frame in samples.

[0699] Accordingly, V1 and V2 can represent corresponding vectors of the foreground 51k-1 V[k] vectors and the foreground 51k V[k-1] vectors for sequential blocks of the HOA 11 coefficients. V1 can, for example, example, represents a first vector of foreground 51k V[k-1] vectors to a first frame (k-1), while V2 can represent a first vector of foreground 51k-1 V[k] vectors to a second frame and subsequent frame (k). V1 and V2 can represent a spatial component for a single audio object included in the multidimensional signal.

, em que isso representa um vetor de fileira multiplicado por um vetor de coluna produzindo, desse modo, um elemento escalar para U.S. Vx[n]-1 pode ser obtido como um pseudoinverso de Vx[n]. Em relação à ponderação de V1 e V2, V1 é ponderado proporcionalmente inferior ao longo da dimensão de tempo devido ao V2 que ocorre subsequente no tempo para Vi. Ou seja, embora os vetores 51k-1 V[k-1] de primeiro plano são componentes espaciais da decomposição, vetores 51k V[k] de primeiro plano temporariamente sequenciais representam valores diferentes do componente espacial ao longo do tempo. Em conformidade, o peso de V1 diminui enquanto o peso de V2 cresce conforme x aumenta ao longo de t. No presente documento, d1 e d2 representam pesos.[0700] The interpolated vectors Vx for each x are obtained by weighting V1 and V2 according to several time segments or “time samples”, x, for a time component of the multidimensional signal which the interpolated vectors Vx can be applied to smooth the temporal (and hence, in some cases, the spatial) component. Assuming a composite of SVD as described above, smoothing of the nFG 49 signals can be obtained by performing a vector division of each time sample vector (e.g. a sample of the HOA 11 coefficients) with the Vx corresponding interpolated. That is,

, where this represents a row vector multiplied by a column vector thus yielding a scalar element for US Vx[n]-1 can be obtained as a pseudoinverse of Vx[n]. Relative to the weighting of V1 and V2, V1 is weighted proportionately lower along the time dimension due to V2 occurring subsequent in time to Vi. That is, although the foreground 51k-1 V[k-1] vectors are spatial components of the decomposition, temporarily sequential 51k-1 V[k] vectors represent different values of the spatial component over time. Accordingly, the weight of V1 decreases while the weight of V2 increases as x increases over t. In the present document, d1 and d2 represent weights.

[0701] Figure 38 is a block diagram illustrating artificial U.S. matrices, USi and US2, for sequential SVD blocks for a multidimensional signal in accordance with techniques described herein. The interpolated V vectors can be applied to the row vectors of the artificial U.S. matrices to recover the original multidimensional signal. More specifically, the spatiotemporal interpolation unit 50 can multiply the pseudo-inverse of the interpolated foreground vectors V[k] 53 to the result of multiplication nFG signals 49 by the foreground vectors 51k V[k] (which can be indicated as foreground HOA coefficients) to obtain interpolated K 2 samples, which can be used in place of the K/2 samples of the nFG signals as the first K/2 samples, as shown in the example of Figure 38 of the U2 matrix .

[0702] Figure 39 is a block diagram illustrating the decomposition of subsequent frames of a higher order ambisonics (HOA) signal using Single Value Decomposition and smoothing of spatiotemporal components according to the techniques described in this revelation. Frame n-1 and frame n (which may also be denoted as frame n and frame n+1) represent subsequent frames in time, where each frame comprises 1024 time segments and has an HOA order of 4, due to the coefficients ( 4+1)2 = 25. US matrices which are artificially smoothed U matrices in frame n-1 and frame n can be obtained by applying interpolated V vectors as illustrated. Each gray row or column vectors represents an audio object.

HOA'S REPRESENTATION OF ACTIVE VECTOR-BASED SIGNALS

[0703] The instantaneous CVECk is created by taking each of the vector-based signals represented in XVECk and multiplying them with their corresponding (dequantized) spatial vector, WECk. Each WECk is represented in MWECk. Thus, for an L-th order HOA signal and M-vector-based signals, there will be M-vector-based signals, each of which will have dimension given by the frame length. These signals can thus be represented as: CVECkmn, n=0,..P-1; m=0,..M-1. Correspondingly, there will be spatial vectors, V ECk of dimension (1+1)2. This can be represented as asMVECkml, 1=0,.., (1+1)2-1 ;m=0,..,M-1. The representation of HOA for each sign-based vector, CVECkm, is a matrix vector multiplication given by:

[0704] which produces a matrix of (1+1)2 by P. The complete HOA representation is due to the sum of each vector's contribution based on sign as follows:

SPACE-TEMPORAL INTERPOLATION OF VECTORS V

[0705] However, in order to maintain smooth spatiotemporal continuity, the above computation is only performed for part of the frame length, -5. The first samples of an HOA matrix are instead performed using an interpolated set of MVECkml, m=0,..,M-1, =0,.., (1+1)2, derived from the Current MVECkm and MWECk-lm from previous values. This results in a higher time density space vector as a vector is derived for each time sample p , as follows:

[0706] For each time sample p, a new HOA vector of dimension (L+1)2 is computed as:

[0707] These first samples are augmented with the P-B samples from the previous section to result in the complete HOA representation, CVECkm, of the nth vector based on signal.

[0708] In the decoder (e.g. the audio decoding device 24 shown in the example of Figure 5), for certain distinct, foreground or predominant vector-based sound, the vector V of the previous frame and the vector V of the current frame can be interpolated using linear (or non-linear) interpolation to produce a higher resolution (in time) interpolated vector V over a particular time segment. The spatiotemporal interpolation unit 76 may perform such interpolation, wherein the spatiotemporal interpolation unit 76 may then multiply the vector US in the current frame with the interpolated vector V of higher resolution to produce the HOA matrix through that particular time segment.

[0709] Alternatively, the spatiotemporal interpolation unit 76 can multiply the U.S. vector with the current frame V vector to create a first HOA matrix. The decoder may additionally multiply the U.S. vector with the V vector from the previous frame to create a second HOA matrix. The spatiotemporal interpolation unit 76 can then apply linear (or non-linear) interpolation to the first and second HOA matrices over a particular time segment. The output of this interpolation may correspond to that of multiplying the U.S. vector with an interpolated V vector, due to common insertion matrices/vectors.

[0710] In this regard, the techniques may enable the audio encoding device 20 and/or the audio decoding device 24 to be configured to operate in accordance with the following clauses.

[0711] Clause 135054-1C. A device, such as audio encoding device 20 or audio decoding device 24, comprising: one or more processors configured to obtain a plurality of spatial components of higher resolution in both space and time, wherein the spatial components are based on an orthogonal decomposition of a multidimensional signal comprising spherical harmonic coefficients.

[0712] Clause 135054-1D. A device, such as audio encoding device 20 or audio decoding device 24, comprising: one or more processors configured to smooth at least one of the spatial components and time components of the first plurality of spherical harmonic coefficients and the second plurality of spherical harmonic coefficients.

[0713] Clause 135054-1E. A device, such as audio encoding device 20 or audio decoding device 24, comprising: one or more processors configured to obtain a plurality of spatial components of higher resolution in both space and time, wherein the spatial components are based on an orthogonal decomposition of a multidimensional signal comprising spherical harmonic coefficients.

[0714] Clause 135054-1G. A device, such as audio encoding device 20 or audio decoding device 24, comprising: one or more processors configured to obtain increased resolution spherical harmonic coefficients decomposed for a time segment increasing, at least in part , a resolution with respect to a first decomposition of a first plurality of spherical harmonic coefficients and a second decomposition of a second plurality of spherical harmonic coefficients.

[0715] Clause 135054-2G. The device of clause 135054-1G, wherein the first decomposition comprises a first matrix V representative of right singular vectors of the first plurality of spherical harmonic coefficients.

[0716] Clause 135054-3G. The device of clause 135054-1G, wherein the second decomposition comprises a second matrix V representative of right singular vectors of the second plurality of spherical harmonic coefficients.

[0717] Clause 135054-4G. The device of clause 135054-1G, wherein the first decomposition comprises a first matrix V representative of right singular vectors of the first plurality of spherical harmonic coefficients and, wherein the second decomposition comprises a second matrix V representative of right singular vectors of the second plurality of spherical harmonic coefficients.

[0718] Clause 135054-5G. The device of clause 135054-1G, wherein the time segment comprises a subframe of an audio frame.

[0719] Clause 135054-6G. The device of clause 135054-1G, wherein the time segment comprises a time sample of an audio frame.

[0720] Clause 135054-7G. The device of clause 135054-1G, wherein the one or more processors are configured to obtain an interpolated decomposition of the first decomposition and the second decomposition to a spherical harmonic coefficient of the first plurality of spherical harmonic coefficients.

[0721] Clause 135054-8G. The device of clause 135054-1G, wherein the one or more processors are configured to obtain interpolated decompositions from the first decomposition to a first portion of the first plurality of spherical harmonic coefficients included in the first frame and the second decomposition to a second portion of the second plurality of spherical harmonic coefficients included in the second frame, wherein the one or more processors are further configured to apply the interpolated decompositions to a first time component of the first portion of the first plurality of spherical harmonic coefficients included in the first frame to generate a first time component. artificial time of the first plurality of spherical harmonic coefficients and applying the respective interpolated decompositions to a second time component of the second portion of the second plurality of spherical harmonic coefficients included in the second frame to generate a second time component artificial po of the second plurality of spherical harmonic coefficients included.

[0722] Clause 135054-9G. The device of clause 135054-8G, wherein the first time component is generated by performing vector-based synthesis with respect to the first plurality of spherical harmonic coefficients.

[0723] Clause 135054-10G. The device of clause 135054-8G, wherein the second time component is generated by performing vector-based synthesis with respect to the second plurality of spherical harmonic coefficients.

[0724] Clause 135054-11G. The device of clause 135054-8G, wherein the one or more processors are further configured to receive the first artificial time component and the second artificial time component, compute interpolated decompositions from the first decomposition to the first portion of the first plurality of harmonic coefficients and the second decomposition to the second portion of the second plurality of spherical harmonic coefficients and applying inverses of the interpolated decompositions to the first artificial time component to recover the first time component and to the second artificial time component to recover the second time component.

[0725] Clause 135054-12G. The device of clause 135054-1G, wherein the one or more processors are configured to interpolate a first spatial component of the first plurality of spherical harmonic coefficients and the second spatial component of the second plurality of spherical harmonic coefficients.

[0726] Clause 135054-13G. The device of clause 135054-12G, wherein the first spatial component comprises a first matrix U representative of left singular vectors of the first plurality of spherical harmonic coefficients.

[0727] Clause 135054-14G. The device of clause 135054-12G, wherein the second spatial component comprises a second matrix U representative of singular vectors to the left of the second plurality of spherical harmonic coefficients.

[0728] Clause 135054-15G. The device of clause 135054-12G, wherein the first spatial component is representative of time segments M of spherical harmonic coefficients for the first plurality of spherical harmonic coefficients and the second spatial component is representative of time segments M of spherical harmonic coefficients for the second plurality of spherical harmonic coefficients.

[0729] Clause 135054-16G. The device of clause 135054-12G, wherein the first spatial component is representative of time segments M of spherical harmonic coefficients for the first plurality of spherical harmonic coefficients and the second spatial component is representative of time segments M of spherical harmonic coefficients for the second plurality of spherical harmonic coefficients and wherein the one or more processors which are configured to obtain the decomposed interpolated spherical harmonic coefficients for the time segment comprise interpolating the last N elements of the first spatial component and the first N elements of the second spatial component .

[0730] Clause 135054-17G. The device of clause 135054-1G, wherein the second plurality of spherical harmonic coefficients is subsequent to the first plurality of spherical harmonic coefficients in the time domain.

[0731] Clause 135054-18G. The device of clause 135054-1G, wherein the one or more processors are further configured to decompose the first plurality of spherical harmonic coefficients to generate the first decomposition of the first plurality of spherical harmonic coefficients.

[0732] Clause 135054-19G. The device of clause 135054-1G, wherein the one or more processors are further configured to decompose the second plurality of spherical harmonic coefficients to generate the second decomposition of the second plurality of spherical harmonic coefficients.

[0733] Clause 135054-20G. The device of clause 135054-1G, wherein the one or more processors are further configured to perform a singular value decomposition with respect to the first plurality of spherical harmonic coefficients to generate a matrix U representative of left singular vectors of the first plurality of coefficients spherical harmonics, a matrix S representative of singular values of the first plurality of spherical harmonic coefficients and a matrix V representative of singular vectors to the right of the first plurality of spherical harmonic coefficients.

[0734] Clause 135054-21G. The device of clause 135054-1G, wherein the one or more processors are further configured to perform a singular value decomposition with respect to the second plurality of spherical harmonic coefficients to generate a matrix U representative of left singular vectors of the second plurality of coefficients spherical harmonics, a matrix S representative of singular values of the second plurality of spherical harmonic coefficients and a matrix V representative of singular vectors to the right of the second plurality of spherical harmonic coefficients.

[0735] Clause 135054-22G. The device of clause 135054-1G, wherein each of the first and second plurality of spherical harmonic coefficients represents a plane wave representation of the sound field.

[0736] Clause 135054-23G. The device of clause 135054-1G, wherein each of the first and second plurality of spherical harmonic coefficients represents one or more mixed mono audio objects.

[0737] Clause 135054-24G. The device of clause 135054-1G, wherein each of the first and second plurality of spherical harmonic coefficients comprises respective first and second spherical harmonic coefficients representing a three-dimensional sound field. Clause 135054-25G. The device of clause 135054-1G, wherein each of the first and second plurality of spherical harmonic coefficients is associated with at least one spherical basis function that has an order greater than one (1).

[0738] Clause 135054-26G. The device of clause 135054-1G, wherein each of the first and second plurality of spherical harmonic coefficients is associated with at least one spherical basis function having an order of four.

[0739] Clause 135054-27G. The provision of clause 135054-1G, where the interpolation is a weighted interpolation of the first decomposition and second decomposition, where weighted interpolation weights applied to the first decomposition are inversely proportional to a time represented by vectors of the first and second decomposition, and where weighted interpolation weights applied to the second decomposition are proportional to a time represented by vectors of the first and second decomposition.

[0740] Clause 135054-28G. The device of clause 135054-1G, wherein the decomposed interpolated spherical harmonic coefficients smooth at least one of the spatial components and temporal components of the first plurality of spherical harmonic coefficients and the second plurality of spherical harmonic coefficients.

[0741] Figures 40A to 40J are each a block diagram illustrating exemplary audio encoding devices 510A to 510J that can perform various aspects of the techniques described in this disclosure to compress spherical harmonic coefficients that describe two- or three-dimensional sound fields. In each of the examples of Figures 40A to 40J, each of the audio encoding devices 510A and 510B, in some examples, represents any device capable of encoding audio data, such as a desktop computer, a portable computer, a station desktop, a tablet or slate computer, a dedicated audio recording device, a cell phone (including so-called “smart phones”), a personal media player, a personal gaming device, or any other type device capable of encoding audio data.

[0742] Although shown as a single device, i.e., devices 510A to 510J in the examples of Figures 40A to 40J, the various components or units referenced below as included within devices 510A to 510J may actually form separate devices that are external of devices 510A to 510J. In other words, although described in this disclosure as performed by a single device, i.e., devices 510A to 510J in the examples of Figures 40A to 40J, the techniques may be implemented or otherwise performed by a system comprising multiple devices, where each one of these devices may include one or more of the various components or units described in more detail below. Accordingly, the techniques should not be limited to the examples in Figure 40A through 40J.

[0743] In some examples, audio encoding devices 510A to 510J represent alternative audio encoding devices to those described above with respect to the examples in Figures 3 and 4. For the entire discussion below of audio encoding devices 510A to 510J several similarities in terms of operation are observed with respect to the various units 30 to 52 of the audio coding device 20 described above with respect to Figure 4. In many respects, the audio coding devices 510A to 510J can, as described below, operate in a substantially similar manner to the audio encoding device 20 albeit with slight deviations or modifications.

[0744] As shown in the examples of Figure 40A, the audio encoding device 510A comprises an audio compression unit 512, an audio encoding unit 514 and a bit stream generation unit 516. audio 512 may represent a unit that compresses spherical harmonic coefficients (SHC) 511 (“SHC 511”), which may also be denoted as higher order ambisonics coefficients (HO A) 511. In some cases, the audio compression unit 512 represents a unit that can either losslessly compress or perform lossy compression relative to the SHC 511. The SHC 511 may represent a plurality of SHCs, wherein at least one of the plurality of SHCs corresponds to a spherical base function having an order greater than one (1) (where SHC of this variety are termed as higher-order ambisonics (HOA) in order to distinguish from lower-order ambisonics of which an example is the so-called “format B”), as rme described in more detail above. Although the audio compression unit 512 can losslessly compress the SHC 511s, in some instances the audio compression unit 512 removes the SHC 511s that are not salient or relevant in describing the sound field when played back (wherein some may not be able to be heard by the human auditor system). In this sense, the lossy nature of this compression may not overly impact the perceived quality of the sound field when played back from the compressed version of the SHC 511.

[0745] In the example of Figure 40A, the audio compression unit includes a decomposition unit 518 and a sound field component extraction unit 520. The decomposition unit 518 may be similar to the linear invertible transform unit 30 of the device. coding unit 20. That is, decomposition unit 518 may represent a unit configured to perform a form of analysis called singular-value decomposition. Although described in relation to SVD, the techniques can be performed in relation to any transformation or similar decomposition that yields non-linearly corrected data sets. Also, reference to "sets" in this disclosure is intended to refer to "non-zero" sets unless specifically stated otherwise and is not intended to refer to the classical mathematical definition of sets which includes the so-called "empty set".

[0746] Anyway, the decomposition unit 518 performs a singular value decomposition (which, again, can be denoted by its acronym “SVD”) to transform the spherical harmonic coefficients 511 into two or more sets of transformed spherical harmonic coefficients . In the example of Figure 40, the decomposition unit 518 can perform SVD against the SHC 511 to generate a so-called V matrix 519, an S matrix 519B and a U matrix 519C. In the example of Figure 40, the decomposition unit 518 outputs each of the matrices separately rather than outputting the U.S.[k] vectors in combined form as discussed above with respect to the linear invertible transform unit 30.

[0747] As noted above, the matrix V* in the mathematical expression of SVD referenced above is denoted as the conjugate transpose of the matrix V to reflect that the SVD can be applied to matrices comprising complex numbers. When applied to matrices that comprise only real numbers, the complex conjugate of the matrix V (or, in other words, the matrix V*) can be assumed to be equal to the matrix V. Below it is assumed, for ease of illustration purposes, that the SHC 511 comprise real numbers with the result that matrix V is output via SVD instead of matrix V*. Although assumed to be the V matrix, the techniques can be applied in a similar way to SHC 511s that have complex coefficients, where the output of the SVD is the V* matrix. Accordingly, the techniques should not be limited in this regard to just providing application of SVD to generate a matrix V, but may include application of SVD to SHC 511 which have complex components to generate a matrix V*.

[0748] In any case, the decomposition unit 518 can perform a block shape of SVD with respect to each block (which may refer to a frame) of higher order ambisonics (HOA) audio data (where these audio data from ambisonics includes blocks or samples from the SHC 511 or any other form of multichannel audio data). A variable M can be used to denote the length of an audio frame in samples. For example, when an audio frame includes 1024 audio samples, M is equal to 1024. The decomposition unit 518 can therefore perform a block SVD with respect to a block, with the SHC 511 having SHC of M- by-(N+1)2, where N, again, denotes the order of the HOA audio data. The decomposition unit 518 can generate, through the realization of this SVD, the matrix V 519, the matrix S 519B and the matrix U 519C, where each of the matrices 519 to 519C ("matrices 519") can represent the respective matrices V , S and U described in more detail above. The decomposition unit 518 can pass or output these matrices 519A to a sound field component extraction unit 520. The V matrix 519A can be of size (N+1)2-by-(N+1)2, the matrix S 519B can be of size (N+1)2-by-(N+1)2 and the matrix U can be of size M-by-(N+1)2, where M refers to the number of samples in an audio frame. A typical value for M is 1024, although the techniques of this disclosure should not be limited to this typical diagonal matrix value.

[0749] Sound field component extraction unit 520 may represent a unit configured to determine and then extract discrete sound field components and background sound field components, effectively separating discrete sound field components from sound field components. background of the sound field. In this regard, the sound field component extraction unit 520 can perform many of the operations described above with respect to the sound field analysis unit 44, the background selection unit 48 and the foreground selection unit 36 of the device. of audio coding 20 shown in the example of Figure 4. Given that distinct components of the sound field, in some examples, require higher-order base functions (relating to background components of the sound field) (and therefore more SHC) to accurately represent the distinct nature of these components, separating the distinct components from the background components can enable more bits to be allocated to the distinct components and fewer bits to be (relatively speaking) allocated to the background components. Accordingly, through the application of such a transformation (in the form of SVD) or any other form of transformation, including PCA), the techniques described in this disclosure can facilitate the allocation of bits to multiple SHCs and thus the compression of the SHC 511s.

[0750] Furthermore, the techniques may also enable, as described in more detail below with respect to Figure 40B, an order reduction of the background components of the sound field given that higher order base functions are not, in some instances, needed to represent these background portions of the sound field given the diffuse or background nature of these components. The techniques can therefore enable compression of diffuse or background aspects of the sound field while preserving salient components or distinct aspects of the sound field by applying SVD to the SHC 511.

[0751] As further shown in the example of Figure 40, the sound field component extraction unit 520 includes a transpose unit 522, a salient component analysis unit 524 and a math unit 526. The transpose unit 522 represents a unit configured to transpose the V matrix 519A to generate a transpose of the V matrix 519, which is denoted as the "VT matrix 523." The transpose unit 522 can output this VT matrix 523 to the math unit 526. The VT matrix 523 can be of size (N+1)2- by-(N+1)2.

[0752] The salient component analysis unit 524 represents a unit configured to perform a salience analysis with respect to the S matrix 519B. The salient component analysis unit 524 may, in this regard, perform operations similar to those described above with respect to the sound field analysis unit 44 of the audio encoding device 20 shown in the example of Figure 4. The salient component analysis unit 524 can parse the diagonal values of matrix S 519B, select a variable number D from those components which has the largest value. In other words, the salient component analysis unit 524 can determine the value D, which separates the two subspaces (e.g., the foreground or predominant subspace and the background or ambient subspace) by analyzing the slope of the curve created by descending diagonal values of S, where large singular values represent foreground or distinct sounds and low singular values represent background components of the sound field. In some examples, the salient component analysis unit 524 may use a first and a second derivative of the singular value curve. The protruding component analysis unit 524 may also limit the number D to be between one and five. As another example, the salient component analysis unit 524 can limit the number D to be between one and (N+1)2. Alternatively, the salient component analysis unit 524 may pre-set the number D, such as a value of four. Anyway, once the number D is estimated, the salient component analysis unit 24 extracts the foreground and background subspace of the matrices U, V and S.

[0753] In some examples, the salient component analysis unit 524 can perform this analysis on all M samples, which can be restated as on a frame-by-frame basis. In this respect, D can vary from frame to frame. In other examples, the salient component analysis unit 24 may perform this analysis more than once per frame, analyzing two or more portions of the frame. Accordingly, the techniques should not be limited in this regard to the examples described in that disclosure.

[0754] In fact, the salient component analysis unit 524 can analyze the singular values of the diagonal matrix, which is detonated as the matrix S 519B in the example of Figure 40, identifying those values that have a greater relative value than the others. S 519B diagonal matrix values. The salient component analysis unit 524 can identify D values, extracting these values to generate the SDIST matrix 525A and the SBG matrix 525B. The SDIST matrix 525A can represent a diagonal matrix comprising D columns having (N+1)2 of the original S matrix 519B. In some cases, the SBG 525B matrix may represent a matrix that has (N+1)2-D columns, each of which includes (N+1)2 transformed spherical harmonic coefficients from the original S 519B matrix. Although described as an SDIST matrix representing a matrix comprising D columns having (N+1)2 values from the original S 519B matrix, the salient component parsing unit 524 can truncate this matrix to generate an SDIST matrix which has D columns which have D values from the original matrix S 519B, given that the matrix S 519B is a diagonal matrix and the (N+1)2 values of the D columns after the value in each column is many times a value of zero. Although described in relation to a full SDIST 525A matrix and a full SBG 525B matrix, the techniques can be deployed against truncated versions of this SDIST 525A matrix and a truncated version of this SBG 525B matrix. Accordingly, the techniques of such disclosure should not be limited in this regard.

, enquanto a matriz SBG 525B pode ser de um tamanho

. A matriz SDIST 525A pode incluir aqueles componentes principais ou, em outras palavras, valores singulares que são determinados como salientes em termos de serem componentes de áudio distintos (DIST) do campo sonoro, enquanto a matriz SBG 525B pode incluir aqueles valores singulares que são determinados como segundo plano (BG) ou, em outras palavras, ambiente ou componentes de áudio não distintos do campo sonoro. Embora mostradas como matrizes separadas 525A e 525B no exemplo da Figura 40, as matrizes 525A e 525B podem ser especificadas como uma única matriz com o uso da variável D para denotar o número de colunas (da esquerda para a direita) dessa única matriz que representa a matriz SDIST 525. Em alguns exemplos, a variável D pode ser definida como quatro.[0755] In other words, the SDIST 525A array can be of a size

, while the SBG 525B array can be of a size

. The SDIST 525A matrix can include those principal components or, in other words, singular values that are determined to be salient in terms of being distinct audio components (DIST) of the sound field, while the SBG 525B matrix can include those singular values that are determined as background (BG) or, in other words, environment or audio components not distinct from the sound field. Although shown as separate matrices 525A and 525B in the example in Figure 40, matrices 525A and 525B can be specified as a single matrix using the variable D to denote the number of columns (from left to right) of that single matrix that represents the array SDIST 525. In some examples, the variable D can be set to four.

[0756] The salient component analysis unit 524 can also analyze the U matrix 519C to generate the UDIST matrix 525C and the matrix UBG 525D. Often, the salient component analysis unit 524 can parse the matrix S 519B to identify the variable D, generating the matrix UDIST 525C and the matrix UBG 525B based on the variable D. That is, after identifying the columns D of the matrix S 519B that are protruded, the protruding component analysis unit 524 can split the U array 519C based on that determined variable D. In that case, the salient component analysis unit 524 can generate the UDIST matrix 525C to include the D columns (from left to right) of the (N+1)2 transformed spherical harmonic coefficients of the original U matrix 519C and the matrix UBG 525D to include the remaining (N+1)2-columns D of the (N+1)2 transformed spherical harmonic coefficients of the original U 519C matrix. The UDIST 525C array can be a size of M-by-D, while the UBG 525D array can be a size of M-by-(N+1)2-D. While shown as separate matrices 525C and 525D in the example of Figure 40, matrices 525C and 525D can be specified as a single matrix using the variable D to denote the number of columns (from left to right) of that single matrix that represents the UDIST 525B array.

[0757] The salient component analysis unit 524 can also analyze the VT matrix 523 to generate the VTDIST matrix 525E and the VTBG matrix 525F. Often, the salient component analysis unit 524 can parse matrix S 519B to identify variable D, generating matrix VTDIST 525E and matrix VBG 525F based on variable D. That is, after identifying the D columns of matrix S 519B that are protruded, the protruding component analysis unit 254 can split the V matrix 519A based on that determined variable D. In that case, the salient component analysis unit 524 can generate the VTDIST 525E matrix to include the (N+1)2 rows (from top to bottom) of the D-values of the original VT matrix 523 and the VTBG matrix 525F to include the ( N+1)2 rows remaining of the (N+1)2-D-values of the original VT 523 matrix. The VTDIST 525E matrix can be of a size of (N+1)2-by-D, while the matrix VTBG 525D can be of a size of (N+1)2-by- (N+1)2-D. Although shown as separate matrices 525E and 525F in the example in Figure 40, matrices 525E and 525F can be specified as a single matrix using the variable D to denote the number of columns (from left to right) of that single matrix that represents the VDIST 525e array. The salient component analysis unit 524 can output the SDIST matrix 525, the matrix SBG 525B, the matrix UDIST 525C, the matrix UBG 525D and the matrix VTBG 525F to the math unit 526, while also outputting the matrix VTDIST 525E to the bitstream generating unit 516.

[0758] Math unit 526 may represent a unit configured to perform matrix multiplications or any other math operation capable of being performed with respect to one or more matrices (or vectors). More specifically, as shown in the example in Figure 40, math unit 526 can represent a unit configured to perform matrix multiplication to multiply matrix UDIST 525C by matrix SDIST 525A to generate UDIST * SDIST 527 vectors of size M-per- D. Matrix math unit 526 can also represent a unit configured to perform matrix multiplication to multiply matrix UBG 525D by matrix SBG 525B and then matrix VTBG 525F to generate matrix UBG * SBG * VTBG 525F to generate spherical harmonic coefficients background 531 of size M-per-(N+1)2 (which may represent those of spherical harmonic coefficients 511 representative of background components of the sound field). The math unit 526 can output the UDIST*SDIST vectors 527 and the spherical background harmonic coefficients 531 to the audio encoding unit 514.

[0759] Audio encoding device 510 therefore differs from audio encoding device 20 in that audio encoding device 510 includes that math unit 526 configured to generate UDIST*SDIST 527 vectors and spherical harmonic coefficients background 531 through matrix multiplication at the end of the encoding process. The linear invertible transform unit 30 of the audio encoding device 20 performs the multiplication of the U and S matrices to output the US[k] 33 vectors at the relative start of the encoding process, which can facilitate future operations, such as rendering, not shown in the example of Figure 40. Furthermore, the audio encoding device 20, instead of retrieving the background SHCs 531 at the end of the encoding process, selects the background HOA coefficients 47 directly from the HOA coefficients 11, thus potentially avoiding array multiplications to retrieve the background SHC 531.

[0760] Audio encoding unit 514 may represent a unit that performs a form of encoding to further compress UDIST*SDIST vectors 527 and spherical background harmonic coefficients 531. Audio encoding unit 514 may operate in a substantially similar to the psychoacoustic audio encoder unit 40 of the audio encoding device 20 shown in the example of Figure 4. In some cases, this audio encoding unit 514 may represent one or more instances of an audio encoding encoding unit. Advanced (AAC). Audio encoding unit 514 can encode each column or row of UDIST * SDIST 527 vectors. Often, audio encoding unit 514 may invoke an example of an AAC encoding unit for each of the remaining order/suborder combinations. in the 531 background spherical harmonic coefficients. More information regarding how the 531 background spherical harmonic coefficients can be encoded using an AAC encoding unit can be found in a convention article by Eric Hellerud, et al. , titled “Encoding Higher Order Ambisonics with AAC,” presented at the 124th Convention, May 17-20, 2008 and available at: http://ro.uow.edu.au/cgi/viewcontent.cgi?article=8025&conte xt=engpapers . The audio coding unit 14 can output an encoded version of 527 UDIST* SDIST vectors (denoted “515 coded UDIST * SDIST vectors”) and an encoded version of 531 spherical background harmonic coefficients (denoted “background spherical harmonic coefficients”). coded 515B”) to bitstream generating unit 516. In some cases, audio encoding unit 514 may encode audio from background spherical harmonic coefficients 531 using a target bit rate lower than used to encode the UDIST * SDIST 527 vectors, thus potentially compressing the 531 background spherical harmonic coefficients more compared to the UDIST * SDIST 527 vectors.

[0761] The bitstream generation unit 516 represents a unit that formats data to conform to a known format (which may refer to a format known by a decoding device), thereby generating the stream of data. bits 517. Bitstream generating unit 42 may operate in a manner substantially similar to that described above with respect to bitstream generating unit 42 of audio encoding device 24 shown in the example of Figure 4. bitstream generation 516 may include a multiplexer that multiplexes encoded UDIST * SDIST vectors 515, encoded background spherical harmonic coefficients 515B, and VTDIST matrix 525E.

[0762] Figure 40B is a block diagram illustrating an exemplary audio encoding device 510B that can perform various aspects of the techniques described in this disclosure to compress spherical harmonic coefficients that describe two- or three-dimensional sound fields. Audio encoding device 510B may be similar to audio encoding device 510 in that audio encoding device 510B includes an audio compression unit 512, an audio encoding unit 514, and a current generating unit. bits 516. Furthermore, the audio compression unit 512 of the audio encoding device 510B may be similar to that of the audio encoding device 510 in that the audio compression unit 512 includes the decomposition unit 518. The compression unit The audio compression unit 512 of the audio encoding device 510B may be different from the audio compression unit 512 of the audio encoding device 510 wherein the sound field component extraction unit 520 includes an additional unit, denoted as a reduction unit. order 528A ("order 528 reduction unit"). For that reason, the sound field component extraction unit 520 of the audio encoding device 510B is denoted as the "sound field component extraction unit 520B."

[0763] The 528A order reduction unit represents a unit configured to perform additional order reduction of the 531 background spherical harmonic coefficients. In some cases, the 528A order reduction unit may rotate the sound field represented by the spherical harmonic coefficients. background 531 to reduce the number of background spherical harmonic coefficients 531 needed to represent the sound field. In some cases, since the spherical background harmonic coefficients 531 represent background components of the sound field, the order reduction unit 528A can remove, eliminate or otherwise delete (often zeroing out) those among the spherical harmonic coefficients of background 531 that correspond to higher order spherical basis functions. In this regard, the order reduction unit 528A can perform similar operations to the background selection unit 48 of the audio encoding device 20 shown in the example of Figure 4. The order reduction unit 528A can output a reduced version of the coefficients. 531 background spherical harmonics (denoted as “reduced background spherical harmonics 529”) to the audio encoding unit 514, which can perform audio coding in a manner described above to encode the reduced background spherical harmonics 529 and thereby generate the 515B encoded reduced spherical background harmonic coefficients.

[0764] The various clauses below may present various aspects of the techniques described in this disclosure.

[0765] Clause 132567-1. A device, such as audio encoding device 510 or audio encoding device 510B, comprising: one or more processors configured to perform singular value decomposition with respect to a plurality of spherical harmonic coefficients representative of a sound field to generate a matrix U representative of singular vectors to the left of the plurality of spherical harmonic coefficients, a matrix S representative of singular values of the plurality of spherical harmonic coefficients and a matrix V representative of singular vectors to the right of the plurality of spherical harmonic coefficients and to represent the plurality of spherical harmonic coefficients as a function of at least a portion of one or more of the U matrix, the S matrix and the V matrix.

[0766] Clause 132567-2. The device of clause 132567-1, wherein the one or more processors are further configured to generate a bit stream to include representing the plurality of spherical harmonic coefficients as one or more vectors of matrix U, matrix S, and matrix V which include combinations thereof or derivatives thereof.

[0767] Clause 132567-3. The device of clause 132567-1, wherein the one or more processors are further described to, when representing the plurality of spherical harmonic coefficients, determine one or more UDIST vectors included within the matrix U that describe distinct components of the sound field.

[0768] Clause 132567-4. The device of clause 132567-1, wherein the one or more processors are further configured to, when representing the plurality of spherical harmonic coefficients, determine one or more UDIST vectors included within the matrix U that describe distinct components of the sound field, determine a or more SDIST vectors included within the matrix S that also describe the distinct components of the sound field and multiply the one or more UDIST vectors and the one or more SDIST vectors to generate the UDIST * SDIST vectors.

[0769] Clause 132567-5. The device of clause 132567-1, wherein the one or more processors are further configured to, when representing the plurality of spherical harmonic coefficients, determine one or more UDIST vectors included within the matrix U that describe distinct components of the sound field, determine a or more SDIST vectors included within the matrix S that also describe the distinct components of the sound field and multiply the one or more UDIST vectors and the one or more SDIST vectors to generate one or more UDIST * SDIST vectors and where the one or more processors are additionally configured to audio encode the one or more UDIST * SDIST vectors to generate an audio encoded version of the one or more UDIST * SDIST vectors.

[0770] Clause 132567-6. The device of clause 132567-1, wherein the one or more processors are further described to, when representing the plurality of spherical harmonic coefficients, determine one or more UBG vectors included within the matrix U.

[0771] Clause 132567-7. The device of clause 132567-1, wherein the one or more processors are further configured to, when representing the plurality of spherical harmonic coefficients, analyze the matrix S to identify background and distinct components of the sound field.

[0772] Clause 132567-8. The device of clause 132567-1, wherein the one or more processors are further configured to, when representing the plurality of spherical harmonic coefficients, analyze the matrix S to identify background and distinct components of the sound field and determine, based on the analysis of the matrix S, one or more UDIST vectors of the matrix U that describe distinct components of the sound field and one or more UBG vectors of the matrix U that describe background components of the sound field.

[0773] Clause 132567-9. The device of clause 132567-1, wherein the one or more processors are further configured to, when representing the plurality of spherical harmonic coefficients, analyze the matrix S to identify background and distinct components of the sound field on an audio frame basis per audio frame and determine, based on the analysis of audio frame per audio frame of the matrix S, one or more UDIST vectors of the matrix U that describe distinct components of the sound field and one or more UBG vectors of the matrix U that describe components background of the sound field.

[0774] Clause 132567-10. The device of clause 132567-1, wherein the one or more processors are further configured to, when representing the plurality of spherical harmonic coefficients, analyze the matrix S to identify background and distinct components of the sound field, determine, based on the analysis of the matrix S, one or more UDIST vectors of the matrix U that describe distinct components of the sound field and one or more UBG vectors of the matrix U that describe background components of the sound field, determining, based on the analysis of the matrix S, a or more SDIST vectors and one or more SBG vectors of the matrix S that correspond to the one or more UDIST vectors and the one or more UBG vectors and determine, based on the analysis of the matrix S, one or more VTDIST vectors and one or more VTBG vectors of a transpose of the matrix V that correspond to one or more UDIST vectors and to one or more UBG vectors.

[0775] Clause 132567-11. The device of clause 132567-10, wherein the one or more processors are additionally configured to, when representing the plurality of spherical harmonic coefficients additionally, multiply the one or more UBG vectors by the one or more SBG vectors and then by one or more vectors VTBG to generate one or more UBG * SBG * VTBG vectors and where the one or more processors are additionally configured to audio encode the UBG * SBG * VTBG vectors to generate an audio encoded version of the UBG * SBG * VTBG vectors.

[0776] Clause 132567-12. The device of clause 132567-10, wherein the one or more processors are further configured to, when representing the plurality of spherical harmonic coefficients, multiply the one or more UBG vectors by the one or more SBG vectors and then by one or more VTBG vectors to generate one or more UBG * SBG * VTBG vectors and perform an order reduction process to eliminate those among the coefficients of the one or more UBG * SBG * VTBG vectors associated with one or more orders of spherical harmonic basis functions and generate, thus, a reduced-order version of one or more UBG * SBG * VTBG vectors.

[0777] Clause 132567-13. The device of clause 132567-10, wherein the one or more processors are further configured to, when representing the plurality of spherical harmonic coefficients, multiply the one or more UBG vectors by the one or more SBG vectors and then by one or more VTBG vectors to generate one or more UBG * SBG * VTBG vectors and perform an order reduction process to eliminate those among the coefficients of the one or more UBG * SBG * VTBG vectors associated with one or more orders of spherical harmonic basis functions and generate, thus, a reduced-order version of the one or more UBG * SBG * VTBG vectors and where the one or more processors are additionally configured to audio-encode the reduced-order version of the one or more UBG * SBG * VTBG vectors to generate an audio encoded version of one or more reduced-order UBG * SBG * VTBG vectors.

[0778] Clause 132567-14. The device of clause 132567-10, wherein the one or more processors are further configured to, when representing the plurality of spherical harmonic coefficients, multiply the one or more UBG vectors by the one or more SBG vectors and then by one or more VTBG vectors to generate one or more UBG * SBG * VTBG vectors and perform an order reduction process to eliminate those among the coefficients of one or more UBG * SBG * VTBG vectors associated with one or more orders greater than one (1) of functions harmonic basis and thereby generate a reduced-order version of the one or more UBG * SBG * VTBG vectors and audio-encode the reduced-order version of the one or more UBG * SBG * VTBG vectors to generate an encoded version of audio from one or more reduced-order UBG * SBG * VTBG vectors.

[0779] Clause 132567-15. The device of clause 132567-10, wherein the one or more processors are additionally configured to generate a stream of bits to include the one or more DIST vectors.

[0780] Clause 132567-16. The device of clause 132567-10, wherein the one or more processors are further configured to generate a bit stream to include the one or more VTDIST arrays without audio encoding the one or more VTDIST arrays.

[0781] Clause 132567-1F. A device, such as audio encoding device 510 or 510B, comprising one or more processors for performing a singular value decomposition with respect to multichannel audio data representative of at least a portion of the sound field to generate a matrix U representing the left singular vectors of the multichannel audio data, a matrix S representing the singular values of the multichannel audio data, and a matrix V representing the right singular vectors of the multichannel audio data and representing the multichannel audio data as a function of at least a portion of one or more of the U matrix, the S matrix, and the V matrix.

[0782] Clause 132567-2F. The device of clause 132567-1F, wherein the multichannel audio data comprises a plurality of spherical harmonic coefficients.

[0783] Clause 132567-3F. The device of clause 132567-2F, wherein the one or more processors are further configured to perform as recited by any combination of clauses 132567-2 through 132567-16.

[0784] From each of the various clauses described above, it is to be understood that any of the audio encoding devices 510A to 510J may perform a method or otherwise comprise means of performing each step of the method for which the audio encoding device 510A the 510J is configured to perform. In some cases, these means may comprise one or more processors. In some cases, the one or more processors may represent a special-purpose processor configured through instructions stored on a non-transient computer-readable storage medium. In other words, various aspects of the techniques in each of the sets of coding examples can provide a non-transient computer-readable storage medium that has instructions stored in it that, when executed, cause the one or more processors to perform the method for which the 510A to 510J audio encoding device has been configured to perform.

[0785] For example, a clause 132567-17 can be derived from the antecedent clause 132567-1 to be a method comprising performing a singular value decomposition with respect to a plurality of spherical harmonic coefficients representative of a sound field to generate a matrix U representative of singular vectors to the left of the plurality of spherical harmonic coefficients, a matrix S representative of singular values of the plurality of spherical harmonic coefficients and a matrix V representative of singular vectors to the right of the plurality of spherical harmonic coefficients and representing the plurality of coefficients spherical harmonics as a function of at least a portion of one or more of the U matrix, the S matrix, and the V matrix.

[0786] As another example, a clause 132567-18 can be derived from the preceding clause 1325671 to be a device, such as the audio encoding device 510B, comprising means for performing a singular value decomposition with respect to a plurality of spherical harmonic coefficients representative of a sound field to generate a matrix U representative of singular vectors to the left of the plurality of spherical harmonic coefficients, a matrix S representative of singular values of the plurality of spherical harmonic coefficients and a matrix V representative of singular vectors to the right of the plurality of spherical harmonic coefficients and means for representing the plurality of spherical harmonic coefficients as a function of at least a portion of one or more of the U matrix, the S matrix and the V matrix.

[0787] As yet another example, a clause 132567-18 can be derived from the antecedent clause 132567-1 to be a non-transient computer-readable storage medium that has instructions stored in it that, when executed, cause one or more processors to perform a singular value decomposition with respect to a plurality of spherical harmonic coefficients representative of a sound field to generate a matrix U representative of left singular vectors of the plurality of spherical harmonic coefficients, a matrix S representative of singular values of the plurality of spherical harmonic coefficients and a matrix V representative of singular vectors to the right of the plurality of spherical harmonic coefficients and representing the plurality of spherical harmonic coefficients as a function of at least a portion of one or more of the matrix U, matrix S and matrix V.

[0788] Various clauses may similarly be derived from clauses 132567-2 to 132567-16 for the various devices, methods, and non-transient computer-readable storage media derived as exemplified above. The same can be done for the various clauses listed throughout this revelation.

[0789] Figure 40C is a block diagram illustrating exemplary audio encoding devices 510C that can perform various aspects of the techniques described in this disclosure for compressing two- or three-dimensional sound. The audio encoding device 510C may be similar to the audio encoding device 510B in that the audio encoding device 510C includes an audio compression unit 512, an audio encoding unit 514 and a current generating unit. bits 516. Furthermore, the audio compression unit 512 of the audio encoding device 510C may be similar to that of the audio encoding device 510B in that the audio compression unit 512 includes the decomposition unit 518.

[0790] The audio compression unit 512 of the audio coding device 510C may, however, be different from the audio compression unit 512 of the audio coding device 510B in which the sound field component extraction unit 520 includes an additional unit, denoted vector reordering unit 532. For that reason, sound field component extraction unit 520 of audio encoding device 510C is denoted as "sound field component extraction unit 520C."

[0791] Vector reordering unit 532 may represent a unit configured to reorder UDIST * SDIST 527 vectors to generate one or more reordered UDIST * SDIST 533 vectors. In this regard, vector reordering unit 532 may operate in a similar to that described above with respect to the reordering unit 34 of the audio encoding device 20 shown in the example of Figure 4. The sound field component extraction unit 520C may invoke the vector reordering unit 532 to reorder the UDIST* vectors SDIST 527 due to the fact that the order of the UDIST * SDIST 527 vectors (where each vector of the UDIST * SDIST 527 vectors can represent one or more distinct mono audio objects present in the sound field) can vary from portions of the data from audio for the reason noted above. That is, given that the audio compression unit 512, in some instances, operates on those portions of the audio data generally called audio frames (which may have samples M of the spherical harmonic coefficients 511, where M is, in some instances, set to 1024), the position of vectors that correspond to these discrete mono audio objects as represented in the U 519C matrix from which the UDIST * SDIST 527 vectors are derived may vary from audio frame to audio frame.

[0792] Passing these UDIST*SDIST 527 vectors directly to the audio encoding unit 514 without reordering these UDIST*SDIST 527 vectors from audio frame to audio frame can reduce the extent of compression achievable for some compression schemes such as inheritance compression schemes that perform best when mono audio objects correlate (in channels, which is defined in this example by the order of UDIST * SDIST 527 vectors relative to each other) across audio frames. Furthermore, when not reordered, encoding of UDIST * SDIST 527 vectors can reduce the quality of audio data when retrieved. For example, AAC encoders, which can be represented in the example of Fig. 40C by the audio coding unit 514, can more efficiently compress the one or more reordered UDIST * SDIST vectors 533 from frame to frame compared to the compression achieved when the UDIST * SDIST 527 vectors are encoded directly from frame to frame. Although described above in relation to AAC encoders, the techniques can be performed in relation to any encoder that provides better compression when mono audio objects are specified through frames in a specific order or position (in channels).

[0793] As described in more detail below, the techniques can enable the 510C audio encoding device to reorder one or more vectors (i.e. UDIST * SDIST 527 vectors to generate one or more 533 reordered UDIST * SDIST vectors and facilitate , thereby compressing UDIST * SDIST vectors 527 by a legacy audio encoder, such as an audio encoding unit 514. Audio encoding device 510C may additionally perform the techniques described in this disclosure to audio encode one or more 533 reordered UDIST * SDIST vectors using the 514 audio encoding unit to generate a 515A encoded version of the 533 reordered UDIST * SDIST vectors.

[0794] For example, sound field component extraction unit 520C may invoke vector reordering unit 532 to reorder one or more first UDIST * SDIST vectors 527 of a first audio frame subsequent in time to the second frame to which one or more second UDIST * SDIST 527 vectors match. Although described in the context of a first audio frame being subsequent in time to the second audio frame, the first audio frame may precede the second audio frame in time. Accordingly, the techniques should not be limited to the example described in this disclosure.

[0795] The 532 vector reordering unit can first perform an energy analysis against each of the first UDIST * SDIST 527 vectors and the second UDIST * SDIST 527 vectors, computing a root mean square energy for at least a portion of (but often the total) first frame of audio and a portion of (but often the total) second frame of audio, and thus generate (assuming D is four) eight energies, one for each one of the first UDIST * SDIST 527 vectors of the first audio frame and one for each of the second UDIST * SDIST 527 vectors of the second audio frame. The vector reordering unit 532 can then compare each energy of the first UDIST * SDIST 527 vectors in turns with each of the second UDIST * SDIST 527 vectors as described above with respect to Tables 1 to 4.

[0796] In other words, when using frame-based SVD (or related methods such as KLT and PCA) decomposition into HoA signals, it may not be guaranteed that the order of vectors from frame to frame is consistent. For example, if there are two objects in the underlying sound field, the decomposition (which when properly performed can be termed “ideal decomposition”) may result in separating the two objects such that a vector represents an object in the matrix U. However, even when the decomposition can be denoted as an “ideal decomposition,” the vectors can alternate positions in the U matrix (and accordingly in the S and V matrix) from frame to frame. Additionally, there may well be phase differences, whereby the vector reordering unit 532 may invert the phase using phase inversion (dot-multiplying each element of the inverted vector by minus one or negative one). In order to feed these vectors, frame by frame in the same “AAC/Audio Encoding engine” may require the order to be identified (or, in other words, the signals to match), the phase to be rectified, and careful interpolation. on frame boundaries to be applied. Without this, the underlying audio codec can produce extremely jagged artifacts including those known as “time-smear” or “pre-echo”.

[0797] In accordance with various aspects of the techniques described in this disclosure, the audio encoding device 510C can apply multiple methodologies to identify/match vectors, using energy and cross-correlation at frame boundaries of the vectors. The 510C audio encoding device can also ensure that a 180 degree phase shift - which often appears at frame boundaries - is corrected. The vector reorder unit 532 can apply a fade/fade interpolation window shape between vectors to ensure a smooth transition between frames.

[0798] In this way, the 530C audio encoding device can reorder one or more vectors to generate one or more first reordered vectors and thus facilitates encoding by a legacy audio encoder, where the one or more vectors described represent distinct components of a sound field and audio encode the one or more reordered vectors using the legacy audio encoder to generate an encoded version of the one or more reordered vectors.

[0799] Various aspects of the techniques described in this disclosure may enable the 510C audio encoding device to operate in accordance with the following provisions.

[0800] Clause 133143-1 A. A device, such as the 510C audio encoding device, comprising: one or more processors configured to perform a power comparison between one or more first vectors and one or more second vectors to determine one or more first vectors reordered and to facilitate the extraction of one or both of the one or more first vectors and the one or more second vectors, where the one or more first vectors describe distinct components of a sound field in a first piece of data from audio and the one or more second vectors describe distinct components of the sound field in a second portion of the audio data.

[0801] Clause 133143-2A. The provision of clause 133143-1A, where the first one or more vectors do not represent background components of the sound field in the first portion of the audio data and where the one or more second vectors do not represent background components of the field sound in the second portion of the audio data.

[0802] Clause 133143-3A. The device of clause 133143-1 A, wherein the one or more processors are additionally configured to, after performing the power comparison, cross-correlate the one or more first vectors and the one or more second vectors to identify the one or more or more first vectors that correlated with the one or more second vectors.

[0803] Clause 133143-4A. The device of clause 133143-1A, wherein the one or more processors are further configured to discard one or more of the second vectors based on the power comparison to generate one or more reduced second vectors that have fewer vectors than the one or more more second vectors, perform a cross-correlation between at least one of the first one or more first vectors and the reduced second one or more vectors to identify one of the one or more reduced second vectors that correlates with at least one of the first one or more first vectors, and reorder at least one of the first one or more vectors based on cross correlation to generate the first one or more reordered vectors.

[0804] Clause 133143-5A. The device of clause 133143-1A, wherein the one or more processors are further configured to discard one or more of the second vectors based on the power comparison to generate one or more reduced second vectors that have fewer vectors than the one or more plus second vectors, cross-correlating at least one of the first one or more vectors and the reduced second one or more vectors to identify one of the one or more reduced second vectors that correlates with at least one of the first one or more first vectors, reorder at least one of the first one or more vectors based on cross-correlation and to generate the one or more first reorder vectors, encode the one or more first reorder vectors to generate the audio encoded version of the one or more first reorder vectors.

[0805] Clause 133143-6A. The device of clause 133143-1A, wherein the one or more processors are further configured to discard one or more of the second vectors based on the power comparison to generate one or more reduced second vectors that have fewer vectors than the one or more plus second vectors, cross-correlating at least one of the first one or more vectors and the reduced second one or more vectors to identify one of the one or more reduced second vectors that correlates with at least one of the first one or more first vectors, reordering at least one of the first one or more vectors based on cross-correlation and to generate the audio encoded version of the one or more first reordered vectors, generating a bit stream to include the encoded version of the one or more first reordered vectors.

[0806] Clause 133143-7A. The device of claims 3A to 6A, wherein the first portion of the audio data comprises a first audio frame having samples M, wherein the second portion of the audio data comprises a second audio frame having the same number, M , of samples, where the one or more processors are additionally configured to, when cross-correlation is performed, cross-correlate against the last values of M to Z of at least one of the first one or more vectors and the first values from M to Z of each of the one or more second reduced vectors to identify one of the one or more second reduced vectors that correlates with the at least one of the first one or more first vectors and where Z is less than M.

[0807] Clause 133143-8A. The device of claims 3A to 6A, wherein the first portion of the audio data comprises a first audio frame having samples M, wherein the second portion of the audio data comprises a second audio frame having the same number, M , of samples, where the one or more processors are additionally configured to, when cross-correlation is performed, cross-correlate against the last values of M through Y of at least one of the first one or more vectors and the first values from M to Z of each of the one or more second reduced vectors to identify one of the one or more second reduced vectors that correlates with the at least one of the first one or more first vectors and where both Z and Y are smaller than M.

[0808] Clause 133143-9A. The device of claims 3A to 6A, wherein the one or more processors are further configured to, when cross-correlation is performed, invert at least one of the one or more first vectors and the one or more second vectors.

[0809] Clause 133143-10A. The device of clause 133143-1A, wherein the one or more processors are further configured to perform a singular value decomposition with respect to a plurality of spherical harmonic coefficients representative of the sound field to generate the one or more first vectors and the one or more second vectors.

[0810] Clause 133143-11A. The device of clause 133143-1A, wherein the one or more processors are further configured to perform a singular value decomposition with respect to a plurality of spherical harmonic coefficients representative of the sound field to generate the matrix U representative of left singular vectors of the plurality of spherical harmonic coefficients, a matrix S representative of singular values of the plurality of spherical harmonic coefficients and a matrix V representative of singular vectors to the right of the plurality of spherical harmonic coefficients and generating the one or more first vectors and the one or more second ones vectors as a function of one or more of the matrix U, matrix S, and matrix V.

[0811] Clause 133143-12A. The device of clause 133143-1A, wherein the one or more processors are further configured to perform a singular value decomposition with respect to a plurality of spherical harmonic coefficients representative of the sound field to generate the matrix U representative of left singular vectors of the plurality of spherical harmonic coefficients, a matrix S representative of singular values of the plurality of spherical harmonic coefficients and a matrix V representative of singular vectors to the right of the plurality of spherical harmonic coefficients, perform a salience analysis with respect to the matrix S to identify a or more UDIST vectors from matrix U and one or more SDIST vectors from matrix S and determine the first one or more vectors and the second one or more vectors by multiplying, at least in part, the one or more UDIST vectors by the one or more SDIST vectors.

[0812] Clause 133143-13 A. The device of clause 133143-1 A, wherein the first portion of the audio data occurs in time before the second portion of the audio data.

[0813] Clause 133143-14A. The device of clause 133143-1A, wherein the first portion of the audio data occurs in time after the second portion of the audio data.

[0814] Clause 133143-15 A. The device of clause 133143-1 A, wherein the one or more processors are additionally configured to, when the power comparison is performed, compute a root mean square power for each of the one or more first vectors and the one or more second vectors and compare the root mean square energy computed for at least one of the first one or more vectors to the root mean square energy computed for each of the one or more plus second vectors.

[0815] Clause 133143-16A. The device of clause 133143-1A, wherein the one or more processors are further configured to reorder at least one of the first one or more vectors based on energy comparison to generate the one or more first reordered vectors and wherein the one or more processors are additionally configured to, when the first vectors are reordered, apply a fade-in interpolation window between the one or more first vectors to ensure a smooth transition when the first one or more reordered vectors are generated.

[0816] Clause 133143-17A. The device of clause 133143-1A, wherein the one or more processors are further configured to reorder the one or more first vectors based on at least the energy comparison to generate the one or more first reordered vectors, generate a stream of bits for include the first one or more reordered vectors or an encoded version of the one or more first reordered vectors and specify reordering information in the bitstream that describes how the first one or more reordered vectors were reordered.

[0817] Clause 133143-18 A. The provision of clause 133143-1 A, where the energy comparison facilitates the extraction of one or both of the one or more first vectors and the one or more second vectors in order to promote encoding of audio from one or both of the one or more first vectors and the one or more second vectors.

[0818] Clause 133143-1B. The device, such as the audio coding device 510C, comprising: one or more processors configured to cross-correlate against one or more first vectors and one or more second vectors to determine one or more reordered first vectors and facilitate the extraction of one or both of the one or more first vectors and the one or more second vectors, wherein the one or more first vectors describe distinct components of a sound field in a first piece of audio data and the one or more second vectors describe distinct components of the sound field in a second portion of the audio data.

[0819] Clause 133143-2B. The provision of clause 133143-1B, wherein the one or more first vectors do not represent background components of the sound field in the first portion of the audio data and wherein the one or more second vectors do not represent background components of the sound field in the second portion of the audio data.

[0820] Clause 133143-3B. The device of clause 133143-1B, wherein the one or more processors are further configured to, prior to performing cross-correlation, perform a power comparison between the one or more first vectors and the one or more second vectors to generate one or more more reduced second vectors that have fewer vectors than the one or more second vectors and where the one or more processors are additionally configured to, when cross-correlation is performed, cross-correlate between the first one or more vectors and one or more plus reduced second vectors to facilitate audio encoding of one or both of the one or more first vectors and the one or more second vectors.

[0821] Clause 133143-4B. Clause 133143-3A. The device of clause 133143-3B, wherein the one or more processors are additionally configured to, when the power comparison is performed, compute a root mean square power for each of the first one or more vectors and the one or more plus second vectors and compare the computed root mean square energy for at least one of the first one or more vectors to the computed root mean square energy for each of the one or more second vectors.

[0822] Clause 133143-5B. The device of clause 133143-3B, wherein the one or more processors are further configured to discard one or more of the second vectors based on power comparison to generate one or more reduced second vectors that have fewer vectors than the one or more second vectors, wherein the one or more processors are further configured to, when cross-correlation is performed, cross-correlate between at least one of the one or more first vectors and the one or more reduced second vectors to identify one of the one or more plus reduced second vectors that correlate with at least one of the first one or more vectors and wherein the one or more processors are further configured to reorder at least one of the first one or more first vectors based on cross correlation to generate the one or more first reordered vectors.

[0823] Clause 133143-6B. The device of clause 133143-3B, wherein the one or more processors are further configured to discard one or more of the second vectors based on power comparison to generate one or more reduced second vectors that have fewer vectors than the one or more second vectors, wherein the one or more processors are further configured to, when cross-correlation is performed, cross-correlate between at least one of the one or more first vectors and the one or more reduced second vectors to identify one of the one or more plus reduced second vectors that correlate with at least one of the first one or more vectors and wherein the one or more processors are further configured to reorder at least one of the first one or more first vectors based on cross correlation to generate the one or more first reordered vectors and encoding the one or more first reordered vectors to generate the audio encoded version of the one or more first vectors first reordered vectors.

[0824] Clause 133143-7B. The device of clause 133143-3B, wherein the one or more processors are further configured to discard one or more of the second vectors based on power comparison to generate one or more reduced second vectors that have fewer vectors than the one or more second vectors, wherein the one or more processors are further configured to, when cross-correlation is performed, cross-correlate between at least one of the one or more first vectors and the one or more reduced second vectors to identify one of the one or more plus reduced second vectors that correlate with at least one of the first one or more vectors and wherein the one or more processors are further configured to reorder at least one of the first one or more first vectors based on cross correlation to generate the one or more first reordered vectors, encoding the one or more first reordered vectors to generate the audio encoded version of the one or more primes three reordered vectors, generate a stream of bits to include the encoded version of the one or more reordered vectors.

[0825] Clause 133143-8B. The device of claims 3B to 7B, wherein the first portion of the audio data comprises a first audio frame having samples M, wherein the second portion of the audio data comprises a second audio frame having the same number, M , of samples, where the one or more processors are additionally configured to, when cross-correlation is performed, cross-correlate against the last values of M to Z of at least one of the first one or more vectors and the first values from M to Z of each of the one or more second reduced vectors to identify one of the one or more second reduced vectors that correlates with the at least one of the first one or more first vectors and where Z is less than M.

[0826] Clause 133143-9B. The device of claims 3B to 7B, wherein the first portion of the audio data comprises a first audio frame having samples M, wherein the second portion of the audio data comprises a second audio frame having the same number, M , of samples, where the one or more processors are additionally configured to, when cross-correlation is performed, cross-correlate against the last values of M through Y of at least one of the first one or more vectors and the first values from M to Z of each of the one or more second reduced vectors to identify one of the one or more second reduced vectors that correlates with the at least one of the first one or more first vectors and where both Z and Y are smaller than M.

[0827] Clause 133143-10B. The device of claim 1B, wherein the one or more processors are further configured to, when cross-correlation is performed, invert at least one of the one or more first vectors and the one or more second vectors.

[0828] Clause 133143-11B. The device of clause 133143-1B, wherein the one or more processors are further configured to perform a singular value decomposition with respect to a plurality of spherical harmonic coefficients representative of the sound field to generate the one or more first vectors and the one or more plus second vectors.

[0829] Clause 133143-12B. The device of clause 133143-1B, wherein the one or more processors are further configured to perform a singular value decomposition with respect to a plurality of spherical harmonic coefficients representative of the sound field to generate the matrix U representative of left singular vectors of the plurality of spherical harmonic coefficients, a matrix S representative of singular values of the plurality of spherical harmonic coefficients and a matrix V representative of singular vectors to the right of the plurality of spherical harmonic coefficients and generating the one or more first vectors and the one or more second vectors as a function of one or more of the matrix U, the matrix S, and the matrix V.

[0830] Clause 133143-13B. The device of clause 133143-1B, wherein the one or more processors are further configured to perform a singular value decomposition with respect to a plurality of spherical harmonic coefficients representative of the sound field to generate the matrix U representative of left singular vectors of the plurality of spherical harmonic coefficients, a matrix S representative of singular values of the plurality of spherical harmonic coefficients and a matrix V representative of singular vectors to the right of the plurality of spherical harmonic coefficients, perform a salience analysis with respect to the matrix S to identify one or more UDIST vectors from matrix U and one or more SDIST vectors from matrix S and determine the first one or more vectors and the second one or more vectors by multiplying, at least in part, the one or more UDIST vectors by the one or more vectors SDIST

[0831] Clause 133143-14B. The device of clause 133143-1B, wherein the one or more processors are further configured to perform a singular value decomposition with respect to a plurality of spherical harmonic coefficients representative of the sound field to generate the matrix U representative of left singular vectors of the plurality of spherical harmonic coefficients, a matrix S representative of singular values of the plurality of spherical harmonic coefficients and a matrix V representative of singular vectors to the right of the plurality of spherical harmonic coefficients and when the one or more first vectors and the one or more second vectors are determined, perform a salience analysis with respect to the matrix S to identify one or more VDIST vectors of the matrix V as at least one of the one or more first vectors and the one or more second vectors.

[0832] Clause 133143-15B. The device of clause 133143-1B, wherein the first portion of the audio data occurs in time before the second portion of the audio data.

[0833] Clause 133143-16B. The device of clause 133143-1B, wherein the first portion of the audio data occurs in time after the second portion of the audio data.

[0834] Clause 133143-17B. The device of clause 133143-1B, wherein the one or more processors are further configured to reorder at least one of the first one or more vectors based on cross-correlation to generate the first one or more reordered vectors and when the first vectors are reordered , apply a fade/fade interpolation window between the first one or more vectors to ensure a smooth transition when the first one or more reordered vectors are generated.

[0835] Clause 133143-18B. The device of clause 133143-1B, wherein the one or more processors are further configured to reorder the first one or more vectors based on at least cross-correlation to generate the first one or more reordered vectors, generate a bit stream to include the one or more first reordered vectors or an encoded version of the one or more first reordered vectors and specify in the bitstream how the first one or more reordered vectors.

[0836] Clause 133143-19B. The device of clause 133143-1B, where cross-correlation facilitates the extraction of one or both of the one or more first vectors and the one or more second vectors in order to promote audio encoding of the one or both of the one or more first vectors and the one or more second vectors.

[0837] Figure 40D is a block diagram illustrating an exemplary audio encoding device 510D that can perform various aspects of the techniques described in this disclosure to compress spherical harmonic coefficients that describe two- or three-dimensional sound fields. The audio encoding device 510D may be similar to the audio encoding device 510C in that the audio encoding device 510D includes an audio compression unit 512, an audio encoding unit 514 and a current generating unit. bits 516. Furthermore, the audio compression unit 512 of the audio encoding device 510D may be similar to that of the audio encoding device 510C in that the audio compression unit 512 includes the decomposition unit 518.

[0838] The audio compression unit 512 of the audio coding device 510D may, however, be different from the audio compression unit 512 of the audio coding device 510C in which the sound field component extraction unit 520 includes an additional unit, denoted as quantization unit 534 (“quant unit 534”). For that reason, the sound field component extraction unit 520 of the audio encoding device 510D is denoted as the "sound field component extraction unit 520D."

[0839] The quantization unit 534 represents a unit configured to quantize the one or more VTDIST 525E vectors and/or the one or more VTBG 525F vectors to generate one or more corresponding VTQ_DIST 525G vectors and/or one or more VTQ_BG 525H vectors. The quantization unit 534 can quantize (which is a signal processing term for mathematical rounding by eliminating bits used to represent a value) the one or more VTDIST vectors 525E in order to reduce the number of bits that are used to represent the one or more VTDIST 525E vectors in the bit stream 517. In some examples, the quantization unit 534 may quantize the 32-bit values of the one or more VTDIST 525E vectors, replacing those 32-bit values with 16-bit values rounded to generate one or more VTQ_DIST 525G vectors. In this regard, the quantization unit 534 may operate in a similar manner as described above with respect to the quantization unit 52 of the audio encoding device 20 shown in the example of Figure 4.

[0840] Quantization of this nature can introduce an error in the representation of the sound field that varies according to the coarseness of the quantization. In other words, using more bits to represent the one or more VTDIST 525E vectors can result in less quantization error. The quantization error due to the quantization of the VTDIST 525E vectors (which may be denoted “EDIST”) can be determined by subtracting the one or more VTDIST 525E vectors from the one or more VTQ_DIST 525G vectors.

[0841] In accordance with the techniques described in this disclosure, the 510D audio encoding device can compensate for one or more of the EDIST quantization errors by projecting the EDIST error onto or otherwise by modifying one or more of the UDIST vectors * SDIST 527 or the 531 spherical background harmonic coefficients generated by multiplying the one or more UBG 525D vectors by the one or more SBG 525B vectors and then by the one or more VTBG 525F vectors. In some examples, the 510D audio encoding device may only compensate for the EDIST error in the UDIST * SDIST 527 vectors. In other examples, the 510D audio encoding device may only compensate for the EBG error in the spherical background harmonic coefficients. In still other examples, the 510D audio encoding device can compensate for the EDIST error in both the UDIST * SDIST 527 vectors and the background spherical harmonic coefficients.

[0842] In operation, the salient component analysis unit 524 can be configured to output the one or more SDIST 525 vectors, the one or more SBG 525B vectors, the one or more UDIST 525C vectors, the one or more UBG 525D vectors , the one or more VTDIST 525E vectors and the one or more VTBG 525F vectors to the math unit 526. The salient component analysis unit 524 may also output the one or more VTDIST vectors 525E to the quantization unit 534. The unit 534 can quantize the one or more VTDIST 525E vectors to generate one or more VTQ_DIST 525G vectors. The quantization unit 534 may supply the one or more VTQ_DIST 525G vectors to the math unit 526, while also providing the one or more VTQ_DIST 525G vectors to the vector reorder unit 532 (as described above). The vector reordering unit 532 may operate with respect to the one or more VTQ_DIST 525G vectors in a similar manner as described above with respect to the VTDIST 525E vectors.

[0843] Upon receipt of these vectors 525 to 525G ("vectors 525"), the math unit 526 can first determine distinct spherical harmonic coefficients that describe distinct components of the sound field and spherical background harmonic coefficients that describe background components of the sound field. The matrix math unit 526 can be configured to determine the distinct spherical harmonic coefficients by multiplying the one or more UDIST 525C vectors by the one or more SDIST 525A vectors and then by the one or more VTDIST 525E vectors. The math unit 526 can be configured to determine the spherical background harmonic coefficients by multiplying the one or more UBG 525D vectors by the one or more SBG 525A vectors and then by the one or more VTBG 525E vectors.

[0844] The math unit 526 can then determine one or more 527' compensated UDIST * SDIST vectors (which may be similar to 527 UDIST * SDIST vectors except that these vectors include values to compensate for the EDIST error) by performing an pseudoinversion with respect to one or more VTQ_DIST 525G vectors and then multiplying the distinct spherical harmonics by the pseudoinversion of one or more VTQ_DIST 525G vectors. Vector reordering unit 532 may operate in the manner described above to generate reordered vectors 527', which are then audio encoded by audio coding unit 515A to generate audio coded reordered vectors 515', again as described above.

[0845] Math unit 526 can then project the EDIST error to the spherical background harmonic coefficients. The mathematics unit 526 may, in order to perform this projection, determine or otherwise recover the original spherical harmonic coefficients 511 by adding the distinct spherical harmonic coefficients to the background spherical harmonic coefficients. The math unit 526 can then subtract the quantized distinct spherical harmonic coefficients (which can be generated by multiplying the UDIST 525C vectors by the SDIST 525A vectors and then by the VTQ_DIST 525G vectors) and the spherical background harmonic coefficients from the spherical harmonic coefficients 511 to determine the remaining error due to quantization of the VTDIST vectors 519. The math unit 526 can then add this error to the quantized background spherical harmonic coefficients to generate compensated quantized spherical background harmonic coefficients 53.

[0846] In any case, the order reduction unit 528A can achieve an embodiment as described above to reduce compensated quantized background spherical harmonic coefficients 53 to reduced 529' background spherical harmonic coefficients, which can be audio encoded by the audio coding unit 514 in the manner described above to generate audio-encoded reduced spherical background harmonic coefficients 515B'.

[0847] In this way, the techniques can enable the 510D audio encoding device to quantize one or more first vectors, such as the 525E VTDIST vectors, representative of one or more components of a sound field and compensate for introduced error due to quantization. of the one or more first vectors and one or more second vectors, such as the UDIST * SDIST vectors 527 and/or the 531 background spherical harmonic coefficient vectors, which are also representative of the same one or more sound field components.

[0848] Furthermore, the techniques can provide such quantization error compensation in accordance with the following clauses.

[0849] Clause 133146-1B. A device, such as audio encoding device 510D, comprising: one or more processors configured to quantize one or more first vectors representative of one or more distinct components of a sound field and compensate for introduced error due to quantization of the one or more more first vectors in one or more second vectors that are also representative of the same one or more distinct components of the sound field.

[0850] Clause 133146-2B. The device of clause 133146-1B, wherein the one or more processors are configured to quantize one or more vectors of a transpose of a matrix V generated at least in part by performing a singular value decomposition with respect to a plurality of coefficients spherical harmonics that describe the sound field.

[0851] Clause 133146-3B. The device of clause 133146-1B, wherein the one or more processors are further configured to perform a singular value decomposition with respect to a plurality of spherical harmonic coefficients representative of a sound field to generate the matrix U representative of left singular vectors of the plurality of spherical harmonic coefficients, a matrix S representative of singular values of the plurality of spherical harmonic coefficients and a matrix V representative of singular vectors to the right of the plurality of spherical harmonic coefficients and wherein the one or more processors are configured to quantize one or more more vectors of a transpose of the matrix V.

[0852] Clause 133146-4B. The device of clause 133146-1B, wherein the one or more processors are configured to perform a singular value decomposition with respect to a plurality of spherical harmonic coefficients representative of a sound field to generate the matrix U representative of left singular vectors of the plurality of spherical harmonic coefficients, a matrix S representative of singular values of the plurality of spherical harmonic coefficients and a matrix V representative of singular vectors to the right of the plurality of spherical harmonic coefficients, wherein the one or more processors are configured to quantize one or more vectors of a transpose of the matrix V and where the one or more processors are configured to compensate for the error introduced due to quantization in one or more vectors U * S computed by multiplying one or more vectors U of the matrix U by one or more vectors S of the S matrix.

[0853] Clause 133146-5B. The device of clause 133146-1B, wherein the one or more processors are further configured to perform a singular value decomposition with respect to a plurality of spherical harmonic coefficients representative of a sound field to generate the matrix U representative of left singular vectors of the plurality of spherical harmonic coefficients, a matrix S representative of singular values of the plurality of spherical harmonic coefficients and a matrix V representative of singular vectors to the right of the plurality of spherical harmonic coefficients, determine one or more UDIST vectors of the matrix U, each one corresponds to one of the distinct components of the sound field, determine one or more SDIST vectors of the matrix S, each of which corresponds to the same of the distinct components of the sound field, and determine one or more VTDIST vectors of a transpose of the matrix V, where each corresponds to the same of the distinct components of the sound field o, where the one or more processors are configured to quantize the one or more VTDIST vectors to generate one or more VTQ_DIST vectors and where the one or more processors are configured to compensate for the error introduced due to quantization in one or more UDIST vectors * SDIST computed by multiplying one or more UDIST vectors from matrix U by one or more SDIST vectors from matrix S to generate one or more UDIST vectors * SDIST with error compensated.

[0854] Clause 133146-6B. The device of clause 133146-5B, wherein the one or more processors are configured to determine distinct spherical harmonic coefficients based on the one or more UDIST vectors, the one or more SDIST vectors and the one or more VTDIST vectors and perform a pseudo-inversion in relative to the VTQ_DIST vectors to divide the distinct spherical harmonic coefficients by the one or more VTQ_DIST vectors and thereby generate one or more error-compensated UC_DIST * SC_DIST vectors that compensate at least in part for the error introduced by quantizing the VTDIST vectors.

[0855] Clause 133146-7B. The device of clause 133146-5B, wherein the one or more processors are additionally configured to audio encode the one or more UDIST * SDIST vectors with error compensated.

[0856] Clause 133146-8B. The device of clause 133146-1B, wherein the one or more processors are further configured to perform a singular value decomposition with respect to a plurality of spherical harmonic coefficients representative of a sound field to generate the matrix U representative of left singular vectors of the plurality of spherical harmonic coefficients, a matrix S representative of singular values of the plurality of spherical harmonic coefficients and a matrix V representative of singular vectors to the right of the plurality of spherical harmonic coefficients, determine one or more vectors UBG of the matrix U that describe one or more background components of the sound field and one or more UDIST vectors of the matrix U that describe one or more distinct components of the sound field, determine one or more vectors SBG of the matrix S that describe the one or more background components of the sound field and one or more SDIST vectors of the matrix S that describe the one or more c components of the sound field and determine one or more VTDIST vectors and one or more VTBG vectors from a transpose of the matrix V, where the VTDIST vectors describe the one or more distinct components of the sound field and the VTBG describe the one or more components of sound field background, where the one or more processors are configured to quantize the one or more VTDIST vectors to generate one or more VTQ_DIST vectors and where the one or more processors are additionally configured to compensate for at least a portion of the error introduced due to quantization into spherical background harmonic coefficients formed by multiplying the one or more UBG vectors by the one or more SBG vectors and then by the one or more VTBG vectors in order to generate compensated spherical background harmonic coefficients.

[0857] Clause 133146-9B. The device of clause 133146-8B, where the one or more processors are configured to determine the error based on the VTDIST vectors and one or more UDIST*SDIST vectors formed by multiplying the UDIST vectors by the SDIST vectors and adding the determined error to the spherical background harmonic coefficients to generate the error-compensated spherical background harmonic coefficients.

[0858] Clause 133146-10B. The device of clause 133146-8B, wherein the one or more processors are further configured to audio-encode the error-compensated background spherical harmonic coefficients.

[0859] Clause 133146-1 IB. The provision of clause 133146-1B,

[0860] wherein the one or more processors are configured to compensate for the error introduced due to quantizing the one or more first vectors into one or more second vectors that are also representative of the same one or more sound field components to generate one or more second error compensated vectors, and wherein the one or more processors are further configured to generate a bit stream to include the one or more second error compensated vectors and the one or more first quantized vectors.

[0861] Clause 133146-12B. The device of clause 133146-1B, wherein the one or more processors are configured to compensate for the error introduced due to the quantization of one or more first vectors into one or more second vectors that are also representative of the same one or more sound field components to generate one or more error compensated second vectors, and wherein the one or more processors are further configured to audio encode the one or more error compensated second vectors and generate a bit stream to include the one or more second compensated vectors by audio-coded error and the first one or more quantized vectors.

[0862] Clause 133146-1C. A device, such as audio encoding device 510D, comprising: one or more processors configured to quantize one or more first vectors representative of one or more distinct components of a sound field and compensate for error introduced due to quantization of the one or more more first vectors in one or more second vectors that are representative of one or more background components of the sound field.

[0863] Clause 133146-2C. The device of clause 133146-1C, wherein the one or more processors are configured to quantize one or more vectors of a transpose of a generated matrix V by performing, at least in part, a singular value decomposition with respect to a plurality of spherical harmonic coefficients that describe the sound field.

[0864] Clause 133146-3C. The device of clause 133146-1C, wherein the one or more processors are further configured to perform a singular value decomposition with respect to a plurality of spherical harmonic coefficients representative of a sound field to generate a matrix U representative of left singular vectors of the plurality of spherical harmonic coefficients, a matrix S representative of singular values of the plurality of spherical harmonic coefficients and a matrix V representative of singular vectors to the right of the plurality of spherical harmonic coefficients, and wherein the one or more processors are configured to quantize a or more vectors of a transpose of the matrix V.

[0865] Clause 133146-4C. The device of clause 133146-1C, wherein the one or more processors are further configured to perform a singular value decomposition with respect to a plurality of spherical harmonic coefficients representative of a sound field to generate a matrix U representative of left singular vectors of the plurality of spherical harmonic coefficients, a matrix S representative of singular values of the plurality of spherical harmonic coefficients and a matrix V representative of singular vectors to the right of the plurality of spherical harmonic coefficients, determine one or more UDIST vectors of the matrix U, each of which corresponds to one of the distinct components of the sound field, determine one or more SDIST vectors of the matrix S, each of which corresponds to the same among the distinct components of the sound field, and determine one or more VTDIST vectors of a transpose of the matrix V, each one of which corresponds to the same among the distinct components of the sound field, wherein the one or more processors are configured to quantize the one or more VTDIST vectors to generate one or more VTQ_DIST vectors and compensate for at least a portion of the error introduced due to quantization on one or more UDIST * SDIST vectors computed by multiplying The one or more UDIST vectors of matrix U are combined with one or more SDIST vectors of matrix S so as to generate one or more error-compensated UDIST * SDIST vectors.

[0866] Clause 133146-5C. The device of clause 133146-4C, wherein the one or more processors are configured to determine distinct spherical harmonic coefficients based on the one or more UDIST vectors, the one or more SDIST vectors and the one or more VTDIST vectors and perform a pseudo-inversion in relative to the VTQ_DIST vectors to divide the distinct spherical harmonic coefficients by the one or more VTQ_DIST vectors and thus generate one or more UC_DIST * SC_DIST vectors that compensate at least in part for the error introduced through the quantization of the VTDIST vectors.

[0867] Clause 133146-6C. The device of clause 133146-4C, wherein the one or more processors are additionally configured to audio encode the one or more error-compensated UDIST * SDIST vectors.

[0868] Clause 133146-7C. The device of clause 133146-1C, wherein the one or more processors are further configured to perform a singular value decomposition with respect to a plurality of spherical harmonic coefficients representative of a sound field to generate a matrix U representative of the left singular vectors of the plurality of spherical harmonic coefficients, a matrix S representative of singular values of the plurality of spherical harmonic coefficients and a matrix V representative of the singular vectors to the right of the plurality of spherical harmonic coefficients, determine one or more vectors UBG of the matrix U that describe one or more background components of the sound field and one or more UDIST vectors of the matrix U that describe one or more distinct components of the sound fields, determine one or more SBG vectors of the matrix S that describe the one or more background components of the sound field and one or more SDIST vectors of the matrix S that describe the one or more distinct components of the sound field and determine one or more VTDIST vectors and one or more VTBG vectors from a transpose of the matrix V, where the VTDIST vectors describe the one or more distinct components of the sound field and the VTBG describe the one or more components sound field background, where the one or more processors are configured to quantize the one or more VTDIST vectors to generate one or more VTQ_DIST vectors, and where the one or more processors are configured to compensate for the error introduced due to quantization on the background spherical harmonic coefficients formed by multiplying the one or more UBG vectors by the one or more SBG vectors and then by the one or more VTBG vectors to generate error-compensated spherical background harmonic coefficients.

[0869] Clause 133146-8C. The device of clause 133146-7C, where the one or more processors are configured to determine the error based on the VTDIST vectors and one or more UDIST*SDIST vectors formed by multiplying the UDIST vectors by the SDIST vectors and adding the determined error to the spherical background harmonic coefficients to generate the error-compensated spherical background harmonic coefficients.

[0870] Clause 133146-9C. The device of clause 133146-7C, wherein the one or more processors are further configured to audio-encode the error-compensated background spherical harmonic coefficients.

[0871] Clause 133146-1OC. The device of clause 133146-1C, wherein the one or more processors are further configured to compensate for error introduced due to quantization of the one or more first vectors into one or more second vectors that are also representative of the same one or more field components sound to generate one or more second error compensated vectors and generate a bit stream to include the one or more second error compensated vectors and the one or more first quantized vectors.

[0872] Clause 133146-11C. The device of clause 133146-1B, wherein the one or more processors are further configured to compensate for error introduced due to quantization of the one or more first vectors into one or more second vectors that are also representative of the same one or more field components sound to generate one or more second error-compensated vectors and audio-encode the one or more second error-compensated vectors and generate a bitstream to include the one or more second audio-encoded error-compensated vectors and the first one or more quantized vectors.

[0873] In other words, when using frame-based SVD decomposition (or related methods such as KLT & PCA) on HoA signals for the purpose of bandwidth reduction, the techniques described in this disclosure may allow the 10D audio encoding quantize the first few vectors of the matrix U (multiplied by the corresponding singular values of the matrix S) as well as the corresponding vectors of the vector V. This will comprise the 'foreground' or 'distinct' components of the sound field. The techniques may then allow the audio encoding device 510D to encode the U*S vectors using a 'black box' audio encoding mechanism, such as an AAC encoder. The vector V can be either scalar or vector quantized.

[0874] Additionally, some of the remaining vectors in the U matrix can be multiplied with the corresponding singular values of the S matrix and the V matrix and also encoded using a 'black box' audio coding mechanism. They will comprise the 'background' components of the sound field. A simple 16-bit scalar quantization of the V vectors can result in overhead of approximately 80 kbps for 4th order (25 coefficients) and 160 kbps for 6th order (49 coefficients). Coarse quantization can result in larger quantization errors. The techniques described in this disclosure can compensate for the quantization error of the vectors V - by 'projecting' the quantization error of the vector V onto the foreground and background components.

[0875] The techniques in this description may include calculating a quantized version of the actual V vector. This quantized vector V can be called V (where V - V+e). The underlying HoA signal - for the foreground components - that the techniques are trying to recreate is given by H_f=USV, where U, S and V contain only the foreground elements. For the purpose of this discussion, U.S. will be replaced by a single set of vectors U. Thus, H_f = UV. Given that you have an erroneous V, techniques are trying to recreate H_f as closely as possible. Thus, the techniques can allow the audio encoding device 10D to find U such that H_f=U'V. The 10D audio encoding device can use a pseudo-inversion methodology that allows U = H_f [V']A(- 1). By using the so-called 'black box' audio coding mechanism to encode U, the techniques can minimize the error in H, caused by what can be referred to as the erroneous V vector.

[0876] In a similar way, the techniques can also allow the audio encoding device to project the error due to quantization of V in the background elements. The audio encoding device 510D can be configured to recreate the total HoA signal which is a combination of the foreground and background HoA signals, that is, H=H_f + H_b. This can again be modeled as H = H_f +e + H_b, due to the quantization error in V. Thus, instead of putting the H_b through the 'black box audio encoder', one puts (e+ H_b) through the audio encoder, in the effect compensation of the error in V. In practice, this compensates the error only up to the order determined by the audio encoding device 510D to send to the background elements.

[0877] Figure 40E is a block diagram illustrating an exemplary audio encoding device 510E that can perform various aspects of the techniques described in this disclosure to compress spherical harmonic coefficients that describe two or three dimensional sound fields. The audio encoding device 510E may be similar to the audio encoding device 510D in that the audio encoding device 510E includes an audio compression unit 512, an audio encoding unit 514, and a stream generation unit. bits 516. Further, the audio compression unit 512 of the audio encoding device 510E may be similar to that of the audio encoding device 510D wherein the audio compression unit 512 includes a decomposition unit 518.

[0878] The audio compression unit 512 of the audio encoding device 510E may, however, differ from the audio compression unit 512 of the audio encoding device 510D in that the math unit 526 of the component extraction unit 520 performs additional aspects of the techniques described in this disclosure to further reduce the V matrix 519A before including the reduced transpose version of the V matrix 519A in the bit stream 517. For that reason, the sound field component extraction unit 520 of the audio encoding device 510E is denoted as the “520E sound field component extraction unit.”

[0879] In the example of Figure 40E, the order reduction unit 528, instead of forwarding the reduced background spherical harmonics 529' to the audio encoding unit 514, returns the reduced background spherical harmonics 529 ' to math unit 526. As noted above, these reduced background spherical harmonic coefficients 529' may have been reduced by removing those of the coefficients that correspond to spherical basis functions that have one or more orders and/or suborders identified . The reduced order of reduced background spherical harmonic coefficients 529' can be denoted by the variable NBG.

[0880] Given that the sound field component extraction unit 520E may not perform order reduction with respect to one or more 533' reordered UDIST * SDIST vectors, the order of this decomposition of the spherical harmonic coefficients that describe the distinct components of the sound field (which can be denoted by the NDIST variable) can be greater than the background order, NBG. In other words, NBG can be commonly smaller than NDIST. One possible reason that NBG may be smaller than NDIST is that the background components are assumed not to have much directionality so that higher order spherical basis functions are not required, thus allowing for order reduction and resulting in NBG being smaller than NDIST.

[0881] Given that the one or more reordered VTQ_DIST vectors 539 were previously sent overtly, without audio encoding those vectors 539 in the bitstream 517, as shown in the examples of Figures 40A to 40D, the one or more reordered VTQ_DIST vectors 539 may consume considerable bandwidth. As an example, each of the one or more 539 reordered VTQ_DIST vectors, when quantized to 16-bit scalar values, can consume approximately 20 Kbps for fourth-order Ambisonics audio data (where each vector has 25 coefficients) and 40 Kbps for sixth-order Ambisonics audio data (where each vector has 49 coefficients).

[0882] In accordance with various aspects of the techniques described in this disclosure, the sound field component extraction unit 520E can reduce the amount of bits that need to be specified for the spherical harmonic coefficients or decompositions thereof, such as the one or more reordered VTQ_DIST vectors 539. In some examples, the math unit 526 can determine, based on the reduced-order spherical harmonic coefficients 529', which of the reordered VTQ_DIST vectors 539 should be removed and recombined with the reduced-order spherical harmonic coefficients 529' and those among the reordered VTQ_DIST vectors 539 that should form the VTSMALL vectors 521. That is, the math unit 526 can determine an order of reduced-order spherical harmonic coefficients 529', where that order can be denoted as NBG. The reordered VTQ_DIST arrays 539 can be of an order denoted by the variable NDIST, where NDIST is greater than the order NBG.

[0883] The math unit 526 can then parse the first NBG orders of the reordered VTQ_DIST vectors 539, removing those vectors that specify decomposed spherical harmonic coefficients that correspond to spherical base functions that have an order less than or equal to NBG. These removed reordered VTQ_DIST vectors 539 can then be used to form intermediate spherical harmonic coefficients by multiplying those among the reordered UDIST * SDIST vectors 533' representative of the decomposed versions of the spherical harmonic coefficients 511 that correspond to spherical basis functions that have a order less than or equal to NBG by the reordered VTQ_DIST vectors removed 539 to form the intermediate distinct spherical harmonic coefficients. The math unit 526 can then generate modified background spherical harmonic coefficients 537 by adding the intermediate distinct spherical harmonic coefficients to the reduced order spherical harmonic coefficients 529'. The math unit 526 can then pass these modified background spherical harmonic coefficients 537 to the audio encoding unit 514, which audio-encodes these coefficients 537 to form the modified audio-encoded spherical background harmonic coefficients 515B' .

[0884] The math unit 526 can then pass the one or more VTSMALL vectors 521, which can represent those vectors 539 representative of a decomposed form of spherical harmonic coefficients 511 that correspond to spherical base functions that have a higher order than that NBG is less than or equal to NDIST. In that respect, the math unit 526 can perform similar operations to the coefficient reduction unit 46 of the audio coding device 20 shown in the example of Figure 4. The math unit 526 can pass the one or more VTSMALL vectors 521 to the bitstream generating unit 516, which can generate the bitstream 517 to include the VTSMALL vectors 521 often in their original non-audio encoded form. Since the VTSMALL arrays 521 include fewer arrays than the reordered VTQ_DIST arrays 539, the techniques can facilitate allocating fewer bits to the reordered VTQ_DIST arrays 539 just by specifying the VTSMALL arrays 521 in the bitstream 517.

[0885] Although shown as not being quantized, in some instances the 510E audio encoding device can quantize the 525F VTBG vectors. In some instances, such as when the audio coding unit 514 is not used to compress the spherical background harmonic coefficients, the audio coding device 510E may quantize the VTBG vectors 525F.

[0886] Thus, the techniques can allow the 510E audio encoding device to determine at least one of one or more vectors decomposed of spherical harmonic coefficients to be recombined with the spherical background harmonic coefficients to reduce an amount of bits required to be allocated to one or more vectors in a bit stream, where the spherical harmonic coefficients describe a sound field, and where the spherical background harmonic coefficients describe one or more background components of the same sound field.

[0887] That is, the techniques may allow the 510E audio encoding device to be configured in a manner indicated by the following clauses.

[0888] Clause 133149-1 A. A device, such as the 510E audio encoding device comprising: one or more processors configured to determine at least one of the one or more decomposed vectors of spherical harmonic coefficients to be recombined with the background spherical harmonic coefficients to reduce the amount of bits required to be allocated to one or more vectors in a bit stream, where spherical harmonic coefficients describe a sound field, and where spherical background harmonic coefficients describe a or more background components of the same sound field.

[0889] Clause 133149-2A. The device of clause 133149-1A, wherein the one or more processors are further configured to generate a reduced set of the one or more vectors by removing the at least one particular of the one or more vectors from the one or more vectors.

[0890] Clause 133149-3A. The device of clause 133149-1 A, wherein the one or more processors are further configured to generate a reduced set of the one or more vectors by removing the at least one particular of the one or more vectors from the one or more vectors, recombining the at least one removed from among the one or more vectors with the spherical background harmonic coefficients to generate the modified background spherical harmonic coefficients and generate the bit stream to include the reduced set of the one or more vectors and the spherical harmonic coefficients backgrounds modified.

[0891] Clause 133149-4A. The device of clause 133149-3A, wherein the reduced set of one or more vectors is included in the bit stream without the first being audio encoded.

[0892] Clause 133149-5A. The device of clause 133149-1 A, wherein the one or more processors are further configured to generate a reduced set of the one or more vectors by removing the at least one particular of the one or more vectors from the one or more vectors, recombining the at least one removed from among the one or more vectors with the spherical background harmonic coefficients to generate the modified spherical background harmonic coefficients, audio-encode the modified spherical background harmonic coefficients, and generate the bit stream to include the reduced set of the one or more vectors and the audio-encoded modified spherical background harmonic coefficients.

[0893] Clause 133149-6A. The device of clause 133149-1A, wherein the one or more vectors comprise vectors representative of at least some aspects of the one or more distinct components of the sound field.

[0894] Clause 133149-7A. The device of clause 133149-1A, wherein the one or more vectors comprise one or more vectors of a transpose of a matrix V generated by performing, at least in part, a singular value decomposition with respect to the plurality of spherical harmonic coefficients that describe the sound field.

[0895] Clause 133149-8A. The device of clause 133149-1A, wherein the one or more processors are further configured to perform a singular value decomposition with respect to the plurality of spherical harmonic coefficients to generate a matrix U representative of left singular vectors of the plurality of spherical harmonic coefficients to generate a matrix U representative of left singular vectors of the plurality of spherical harmonic coefficients , a matrix S representative of singular values of the plurality of spherical harmonic coefficients and a matrix V representative of singular vectors to the right of the plurality of spherical harmonic coefficients, and wherein the one or more vectors comprise one or more vectors of a transpose of the matrix V .

[0896] Clause 133149-9A. The device of clause 133149-1 A, wherein the one or more processors are additionally configured to perform an order reduction with respect to the spherical background harmonic coefficients in order to remove those of the spherical background harmonic coefficients that correspond to the spherical basis functions that have an identified order and/or suborder, where the spherical background harmonic coefficients correspond to an NBG order.

[0897] Clause 133149-10A. The device of clause 133149-1 A, wherein the one or more processors are additionally configured to perform an order reduction with respect to the spherical background harmonic coefficients in order to remove those of the spherical background harmonic coefficients that correspond to the spherical basis functions that have an identified order and/or suborder, where the background spherical harmonic coefficients correspond to an NBG order that is less than the order of the distinct spherical harmonic coefficients, NDIST, and where the spherical harmonic coefficients distinct represent the distinct components of the sound field.

[0898] Clause 133149-11A. The device of clause 133149-1 A, wherein the one or more processors are additionally configured to perform an order reduction with respect to the spherical background harmonic coefficients in order to remove those of the spherical background harmonic coefficients that correspond to the spherical basis functions that have an identified order and/or suborder, where the background spherical harmonic coefficients correspond to an NBG order that is less than the order of the distinct spherical harmonic coefficients, NDIST, and where the spherical harmonic coefficients distinct components represent the distinct components of the sound field and are not subjected to order reduction.

[0899] Clause 133149-12A. The device of clause 133149-1A, wherein the one or more processors are further configured to perform a singular value decomposition with respect to the plurality of spherical harmonic coefficients to generate a matrix U representative of the left singular vectors of the plurality of harmonic coefficients spherical, a matrix S representative of singular values of the plurality of spherical harmonic coefficients and a matrix V representative of the singular vectors to the right of the plurality of spherical harmonic coefficients and determining one or more vectors VTDIST and one or more VTBG from a transpose of the matrix V, the one or more VTDIST vectors describe one or more distinct components of the sound field and the one or more VTBG vectors describe one or more background components of the sound field, and wherein the one or more vectors include the one or more VTDIST vectors.

[0900] Clause 133149-13 A. The device of clause 133149-1 A, wherein the one or more processors are further configured to perform a singular value decomposition with respect to the plurality of spherical harmonic coefficients to generate a U matrix representative of the singular vectors to the left of the plurality of spherical harmonic coefficients, a matrix S representative of singular values of the plurality of spherical harmonic coefficients and a matrix V representative of the singular vectors to the right of the plurality of spherical harmonic coefficients, determine one or more vectors VTDIST and one or more VTBG of a transpose of the matrix V, the one or more VTDIST vectors describe one or more distinct components of the sound field and the one or more VBG vectors describe one or more background components of the sound field and quantize the one or more VTDIST vectors to generate one or more VTQ_DISTs, and where the one or more vectors include the one or more VTQ_DIST vectors .

[0901] Clause 133149-14A. The device of clause 133149-12A or clause 133149-13A, wherein the one or more processors are further configured to determine one or more UDIST vectors and one or more UBG vectors of the matrix U, the one or more UDIST vectors describe the one or more distinct components of the sound field and the one or more UBG vectors describe the one or more background components of the sound field and determine the one or more SDIST vectors and the one or more SBG vectors of the matrix S, the one or more SDIST vectors describe the one or more distinct components of the sound field and the one or more SBG vectors describe the one or more background components of the sound field.

[0902] Clause 133149-15A. The device of clause 133149-14A, wherein the one or more processors are further configured to determine the spherical background harmonic coefficients as a function of the one or more UBG vectors, the one or more SBG vectors and the one or more VTBG vectors , perform order reduction with respect to the spherical background harmonic coefficients to generate the reduced spherical background harmonic coefficients that have an order equal to NBG, multiply the one or more UDISTs by the one or more SDIST vectors to generate one or more UDIST * SDIST vectors, remove the at least one determined among the one or more vectors from the one or more vectors to generate a reduced set of the one or more vectors, multiply the one or more UDIST * SDIST vectors by at least one removed from the one or more more VTDIST vectors or the one or more VTQ_DIST vectors to generate the intermediate spherical distinct harmonic coefficients and add the spherical distinct harmonic coefficients to the background spherical harmonic coefficient to recombine the at least one removed from among the one or more VTDIST vectors or the one or more VTQ_DIST vectors with the spherical background harmonic coefficients.

[0903] Clause 133149-16A. The device of clause 133149-14A, wherein the one or more processors are further configured to determine the spherical background harmonic coefficients as a function of the one or more UBG vectors, the one or more SBG vectors and the one or more VTBG vectors , perform order reduction with respect to the spherical background harmonic coefficients to generate the reduced spherical background harmonic coefficients that have an order equal to NBG, multiply the one or more UDISTs by the one or more SDIST vectors to generate one or more UDIST * SDIST vectors, reorder the one or more UDIST * SDIST vectors to generate one or more reordered UDIST * SDIST vectors, remove the at least one particular among the one or more vectors from the one or more vectors to generate a reduced set of the one or more more vectors, multiply the one or more UDIST * SDIST vectors reordered by at least one removed from among the one or more VTDIST vectors or the one or more VTQ_DIST vectors to generate the co efficient intermediate distinct spherical harmonics and adding the intermediate distinct spherical harmonics to the background spherical harmonics to recombine the at least one removed from among the one or more VTDIST vectors or the one or more VTQ_DIST vectors with the background spherical harmonics.

[0904] Clause 133149-17A. The clause 133149-15A or clause 133149-16A device, wherein the one or more processors are additionally configured to audio-encode the background spherical harmonic coefficients after adding the intermediate distinct spherical harmonic coefficients to the spherical background harmonic coefficients and generating the bit stream to include the audio encoded spherical background harmonic coefficients.

[0905] Clause 133149-18 A. The device of clause 133149-1 A, wherein the one or more processors are further configured to perform a singular value decomposition with respect to the plurality of spherical harmonic coefficients to generate a matrix U representative of the singular vectors to the left of the plurality of spherical harmonic coefficients, a matrix S representative of the singular values of the plurality of spherical harmonic coefficients and a matrix V representative of the singular vectors to the right of the plurality of spherical harmonic coefficients, determine one or more VTDIST vectors and one or more VTBG of a transpose of matrix V, the one or more VDIST vectors describe one or more distinct components of the sound field and the one or more VBG vectors describe one or more background components of the sound field, quantize the one or more VTDIST vectors to generate one or more VTQ_DIST vectors and reorder the one or more VTQDIST vectors to generate one or more VTQ_DIST vectors reordered, and where the one or more vectors include the one or more reordered VTQ_DIST vectors.

[0906] Figure 40F is a block diagram illustrating the exemplary audio encoding device 510F that can perform various aspects of the techniques described in this disclosure to compress spherical harmonic coefficients that describe two or three dimensional sound fields. Audio encoding device 510F may be similar to audio encoding device 510C in that audio encoding device 510F includes an audio compression unit 512, an audio encoding unit 514, and a stream generation unit 510F. bits 516. In addition, the audio compression unit 512 of the audio encoding device 510F may be similar to that of the audio encoding device 510C wherein the audio compression unit 512 includes a decomposition unit 518 and a reordering device 532, which can operate similarly to similar units of the audio coding device 510C. In some examples, the audio encoding device 510F may include a quantization unit 534, as described with respect to Figures 40D and 40E, for quantizing one or more vectors of any of the UDIST 525C vectors, the 525D UBG vectors, the 525D vectors, the VTDIST 525E and the VTBG 525 J vectors.

[0907] The audio compression unit 512 of the audio coding device 510F may, however, differ from the audio compression unit 512 of the audio coding device 510C in that the protruding component analysis unit 524 of the sound field component extraction 520 may perform a content analysis to select the number of foreground components, denoted as D in the context of Figures 40A to 40J. In other words, the salient component analysis unit 524 can operate against the U, S, and V matrices 519 in the manner described above to identify whether the decomposed versions of the spherical harmonic coefficients were generated from the synthetic audio objects or from of a natural recording with a microphone. The salient component analysis unit 524 can then determine D based on its synthetic determination.

[0908] In addition, the audio compression unit 512 of the audio encoding device 510F may differ from the audio compression unit 512 of the audio encoding device 510C in that the sound field component extraction unit 520 may include an additional unit, a 528F order power saving and reduction unit (pictured as “528F order power saving and power saving unit”). For these reasons, the sound field component extraction unit 520 of the audio encoding device 510F is denoted as the "sound field component extraction unit 510F."

[0909] The energy preserving and order reduction unit 528F represents a unit configured to perform order reduction of the background components of the VBG matrix 525H representative of the right singular vectors of the plurality of spherical harmonic coefficients 511 while preserving energy (and concomitant sound pressure) of the sound field described in part by the full VBG matrix 525H. In this regard, the order reduction and energy preservation unit 528F can perform operations similar to those described above with respect to the background selection unit 48 and the power compensation unit 38 of the audio encoding device 20 shown in the example of the Figure 4.

, em que D representa um número de componentes principais ou, em outras palavras, valores singulares que são determinados para serem salientes em termos de serem componentes de áudio distintos do campo sonoro. Isto é, a matriz VBG completa 525H inclui aqueles valores singulares que são determinados como sendo segundo plano (BG) ou, em outras palavras, componentes ambientes ou de áudio não distinto do campo sonoro.[0910] Complete VBG matrix 525H has dimensionality

, where D represents a number of principal components or, in other words, singular values that are determined to be salient in terms of being audio components distinct from the sound field. That is, the complete VBG matrix 525H includes those singular values that are determined to be background (BG) or, in other words, ambient or audio components not distinct from the sound field.

. A unidade de transposta 522 aplica uma operação de transposta à matriz VBG’ reduzida 5251 para gerar e emitir uma matriz VTBG’ reduzida transposta 525J à unidade de matemática 526, que pode operar para reconstruir os componentes de som de segundo plano do campo sonoro computando-se UBG*SBG*VTBG com o uso da matriz UBG 525D, da matriz SBG 525B e da matriz VTBG’ reduzida transposta 525J.[0911] As described above with respect to, for example, the order reduction unit 524 of Figures 40B to 40E, the order reduction and energy preservation unit 528F may remove, eliminate or otherwise erase (often by zeroing ) those among the background singular values of the matrix VBG 525H that correspond to higher-order spherical basis functions. The order reduction and energy preservation unit 528F can output a reduced version of the VBG matrix 525H (denoted as matrix "VBG'5251" and referred to hereinafter as "reduced VBG' matrix 5251") to the transpose unit 522. VBG' reduced 5251 can have dimensionality

. The transpose unit 522 applies a transpose operation to the reduced VTBG' matrix 5251 to generate and output a reduced VTBG' matrix transpose 525J to the math unit 526, which can operate to reconstruct the background sound components of the sound field by computing it. if UBG*SBG*VTBG using matrix UBG 525D, matrix SBG 525B and reduced transposed matrix VTBG' 525J.

[0912] Pursuant to the techniques described in this document, the 528F energy preservation and order reduction unit is additionally configured to compensate for possible reductions in the overall energy of the sound field's background sound components caused by order reduction of the full VBG matrix 525H to generate the reduced VBG' matrix 5251. In some examples, the energy preserving and order reduction unit 528F compensates by determining a compensation gain in the form of amplification values to apply to each of the (N+1)2-D columns of the reduced matrix VBG' 5251 in order to increase the root mean square (RMS) energy of the reduced matrix VBG' 5251 to equal or at least closer to the RMS of the full VBG matrix 525H, before of outputting the reduced matrix VBG' 5251 to the transpose unit 522.

[0913] In some instances, the order reduction and energy preservation unit 528F can determine the RMS energy of each column of the full VBG matrix 525H and the RMS energy of each column of the reduced VBG' matrix 5251, then determine the value of amplification for the column as the ratio of the first to the last, as indicated in the following equation:

[0914] where a is the amplification value for a column, BG represents a single column of matrix VBG 525H, and vBG' represents the corresponding single column of matrix VBG' 5251. This can be represented in matrix notation as:

[0915] where VBGRMS is an RMS vector that has elements that denote the RMS of each column of the matrix VBG 525H, VBG'RMS is an RMS vector that has elements that denote the RMS of each column of the reduced matrix VBG' 5251, and A is a amplification value vector that has elements for each column of the matrix VBG 525H. The order reduction and energy preservation unit 528F applies a scalar multiplication to each column of the reduced VBG matrix 5251 using the corresponding amplification value, α, or in vector form:

.[0916] where VgG represents a reduced matrix VBG' 5251 including power offset. The order reduction and energy preservation unit 528F may output the reduced VBG' matrix 5251 including energy compensation to the transpose unit 522 to equalize (or nearly equalize) the RMS of the reduced VBG' matrix 5251 to this of the complete VBG matrix 525H. The output dimensionality of the reduced VBG' matrix 5251 including energy compensation can be (f| +1)2 x (N+1)2 - Energy give can be

.

[0917] In some examples, to determine each RMS of the respective columns of the reduced VBG' matrix 5251 and of the full VBG matrix 525H, the energy preserving and order reduction unit 528F can first apply a spherical harmonic coefficient (SHC) renderer reference to columns. The application of the reference SHC renderer by the energy preservation and order reduction unit 528F allows the determination of the RMS in the SHC domain to determine the energy of the overall sound field described by each column of the frame represented by the reduced matrix VBG' 5251 and by the full VBG matrix 525H. Thus, in such examples, the order reduction and energy preservation unit 528F can apply the reference SHC renderer to each column of the full VBG matrix 525H and to each reduced column of the reduced VBG' matrix 5251, determine the respective values of RMS for the column and the reduced column and determine the amplification value for the column as the ratio between the RMS value for the column and the RMS value for the reduced column. In some examples, order reduction for the reduced matrix VBG' 5251 proceeds in the direction of the coincident column to preserve energy. This can be expressed in pseudocode as follows:

[0918] R = ReferenceRenderer;

[0919] for m = nuniDist+1: numChannels

[0920] fullV = V(:,m); //take a column of V => fullV

[0921] reducedV =[fullV(l:numBG); zeros(numChannels-numBG,l)];

[0922] alpha=sqrt(sum( (fullV**R).A2/sum((reducedV**R) .A2));

[0923] if isnan(alpha) || isinf(alpha), alpha = 1; end;

[0924] V_out(:,m) = reducedV * alpha;

[0925] end

e a dimensionalidade de V_out pode ser

.[0926] In the above pseudocode, numChannels can represent (N+1)2 - D, numBG can represent ((+1)2, V can represent the matrix VBG 525H and V_out can represent the reduced matrix VBG' 5251 and R can represent the reference SHC renderer of the energy-preserving and order-reduction unit 528F. The dimensionality of V can be

and the dimensionality of V_out can be

.

[0927] As a result, the audio encoding device 510F can, by representing the plurality of spherical harmonic coefficients 511, reconstruct the background sound components using a reduced-order VBG' matrix 5251 that includes compensation of the energy that can be lost as a result of the order reduction process.

[0928] Figure 40G is a block diagram illustrating the exemplary audio encoding device 510G that can perform various aspects of the techniques described in this disclosure to compress spherical harmonic coefficients that describe two or three dimensional sound fields. In the example of Figure 40G, the audio encoding device 510G includes a sound field component extraction unit 520F. In turn, the sound field component extraction unit 520F includes a salient component analysis unit 524G.

[0929] The audio compression unit 512 of the audio encoding device 510G may, however, differ from the audio compression unit 512 of the audio encoding device 10F in that the audio compression unit 512 of the encoding device 510G audio system includes a 524G protruding component analysis unit. The 524G salient component analysis unit may represent a unit configured to determine the salience or distinctness of audio data representing a sound field, using the directionality-based information associated with the audio data.

[0930] While energy-based determinations can improve the rendering of an SVD decomposed sound field to identify the audio components distinct from the sound field, energy-based determinations can also cause a device to misidentify the components of a sound field. background audio as distinct audio components, in cases where the background audio components exhibit a high power level. That is, a power-only separation of distinct and background audio components may not be robust, as energetic (e.g. louder) background audio components may be incorrectly identified as being 320/ 500 distinct audio components. To more robustly distinguish between discrete and background audio components of the sound field, various aspects of the techniques described in this disclosure may allow the 524G salient component analysis unit to perform directionality-based analysis of the SHC 511 to separate the distinct and background audio components of the decomposed versions of the SHC 511.

[0931] The 524G salient component analysis unit may, in the example of Figure 40H, represent a unit configured or otherwise operable to separate distinct (or foreground) elements from background elements included in one or more of the V array 519, the S array 519B and the U array 519C, similar to the salient component analysis units 524 of the previously described audio coding devices 510-510F. According to some SVD-based techniques, most energy components (for example, the first few vectors of one or more of the V, S, and U matrices 519-519C or a matrix derived from them) can be treated as components. distinct. However, most of the energy components (which are represented by vectors) of one or more of the 519-519C matrices may not, in all scenarios, represent the components/signals that are the most directional.

[0932] Unlike the previously described salient component analysis units 524, the salient component analysis unit 524G may implement one or more aspects of the techniques described herein to identify foreground elements based on the directionality of the vectors of a or more of matrices 519-519C or a matrix derived therefrom. In some 321/500 examples, the 524G salient component analysis unit can identify or select as distinct audio components (where the components may also be referred to as “objects”), one or more vectors based on both energy and vector directionality. For example, the 524G salient component analysis unit can identify those vectors from one or more of the 519-519C matrices (or a matrix derived therefrom) that exhibit both high energy and high directionality (e.g., represented as a quotient of directionality) as the distinct audio components. As a result, if the salient component analysis unit 524G determines that a particular vector is relatively less directional compared to other vectors in one or more of the matrices 519-519C (or a matrix derived therefrom), then regardless of the energy level associated with the particular vector, the salient component analysis unit 524G can determine that the particular vector represents the background (or ambient) audio components of the sound field represented by the SHC 511. protruding component 524G can perform operations similar to those described above with respect to the sound field analysis unit 44 of the audio encoding device 20 shown in the example of Figure 4.

[0933] In some deployments, the 524G salient component analysis unit can identify distinct audio objects (which, as noted above, may also be referred to as “components”) based on directionality by performing the following operations. The salient component analysis unit 524G may multiply (eg, using one or more matrix multiplication processes) the V matrix 519A by the S matrix 322/500 519B. By multiplying the V matrix 519A and the S matrix 519B, the salient component analysis unit 524G can obtain a VS matrix. Additionally, the salient component analysis unit 524G can square (ie, exponentiate by a power of two) at least some of the entries of each of the vectors (which can be in a row) of the VS matrix. In some instances, the 524G salient component analysis unit may sum those squared entries of each vector that are associated with an order greater than 1. As an example, if each matrix vector includes 25 entries, the analysis unit 524G salient component can, with respect to each vector, square the inputs of each vector starting at the fifth input and ending at the twenty-fifth input, summing the squared inputs to determine a directionality quotient (or a directionality indicator) . Each summation operation can result in a directionality quotient for a corresponding vector. In this example, the 524G salient component analysis unit can determine that those inputs of each row that are associated with an order less than or equal to 1, i.e., the first to fourth input, are more generally directed to the amount of energy and less to the directionality of those inputs. That is, the lower-order ambisonics associated with an order of zero or one correspond to spherical basis functions which, as illustrated in Figure 1 and Figure 2, do not give much in terms of the pressure wave direction, but rather In addition, they provide some volume (which is representative of energy).

[0934] The operations described in the example above can also be expressed according to the following pseudocode. The pseudocode below includes annotations, in the form of 323/500 comment statements that are enclosed within consecutive occurrences of the strings “/*” and “*/” (without quotes).

[0935] /* The next line is directed to parse each row independently and sum the values in the first (as an example) row from the fifth input to the twenty-fifth input to determine the directionality quotient or directionality metric for a corresponding vector. Square the entries before the sum. Entries in each row that are associated with an order greater than 1 are associated with higher-order ambisonics and are thus more likely to be directional. */

[0936] /* The next line is directed to sorting the sum of squares for the generated VS matrix and selecting a set of the highest values (e.g. three or four of the highest values) */

[0937] In other words, according to the above pseudocode, the salient component analysis unit 524G can select entries of each vector of the decomposed VS matrix from those of the SHC 511 that correspond to a spherical basis function that has an order greater than one. The salient component analysis unit 524G can then square these entries for each vector of the VS matrix, summing the squared entries to identify, compute, or otherwise determine a directionality metric or quotient for each vector of the VS matrix. Next, the 524G salient component analysis unit can sort the vectors from the VS matrix based on the respective directionality metrics of each of the vectors. The 524G salient component analysis unit can sort these vectors in descending order of directionality metrics, so that those vectors with the highest corresponding directionality are first and those vectors with the lowest corresponding directionality are last. The 524G salient component analysis unit can then select the non-zero subset of vectors that have the highest relative directionality metric.

[0938] In accordance with some aspects of the techniques described herein, the 510G audio encoding device or one or more components thereof may identify or otherwise use a predetermined number of the vectors of the VS matrix as the distinct audio components . For example, after selecting entries 5 through 25 of each row of the VS matrix and squaring and summing the selected entries to determine the relative directionality metric for each respective vector, the 524G salient component analysis unit can implement the additional selection among the vectors to identify the vectors that represent the distinct audio components. In some examples, the salient component analysis unit 524G may select a predetermined number of vectors from the VS matrix by comparing the directionality quotients of the vectors. As an example, the 524G salient component analysis unit can select the four vectors represented in the VS matrix that have the four largest directionality quotients (and which are the first four vectors of the assorted VS matrix). In turn, the 524G salient component analysis unit can determine that the four selected vectors represent the four most distinct audio objects associated with the corresponding SHC representation of the sound field.

[0939] In some examples, the 524G salient component analysis unit may reorder the vectors derived from the VS matrix to reflect the distinction of the four selected vectors as described above. In one example, the 524G salient component analysis unit can reorder the vectors so that the four selected entries are relocated to the top of the VS matrix. For example, the 524G salient component analysis unit can modify the VS matrix so that all four selected entries are placed in a first (or topmost) row of the resulting reordered VS matrix. Although described herein with respect to the protruding component parsing unit 524G, in various implementations, other components of the audio encoding device 510G, such as the vector reordering unit 532, may perform reordering.

[0940] The salient component analysis unit 524G can communicate the resulting matrix (i.e. the VS matrix, reordered or not, as the case may be) to the bit stream generation unit 516. bitstream generation 516 can use VS matrix 525K to generate bitstream 517. For example, if salient component analysis unit 524G has reordered VS matrix 525K, bitstream generation unit 516 can use the top row of the reordered version of the VS 525K matrix as the distinct audio objects, such as quantizing or discarding the remaining vectors of the reordered version of the VS 525K matrix. By quantizing the remaining vectors of the reordered version of the VS 525K matrix, the bitstream generation unit 16 can treat the remaining vectors as ambient or background audio data.

[0941] In the examples where the salient component analysis unit 524G did not order the VS 525K matrix, the bitstream generation unit 516 can distinguish the distinct data from the background audio data, based on the particular inputs ( for example, the 5th to 25th entries) of each row of the VS 525K array, as selected by the 524G protruding component analysis unit. For example, the bitstream generating unit 516 can generate the bitstream 517 by quantizing or discarding the first four entries of each row of the VS array 525K.

[0942] In this manner, the 510G audio coding device and/or components thereof, such as the 524G salient component analysis unit, can deploy the techniques of this disclosure to determine or otherwise utilize the ratios of greater or lesser coefficients of the audio data in order to distinguish between distinct audio objects and background audio data representative of the sound field. For example, as described, the salient component analysis unit 524G may use the power ratios based on the values of the various entries of the VS matrix 525K generated by the salient component analysis unit 524H. By combining the data provided by the V matrix 519A and the matrix S 519B, the salient component analysis unit 524G can generate the VS matrix 525K to provide the information on both the directionality and the overall energy of the various components of the audio data. , in the form of vectors and related data (eg directionality quotients). More specifically, the V matrix 519A can provide information related to directionality determinations, while the S matrix 519B can provide information related to general energy determinations for the audio data components.

[0943] In other examples, the salient component analysis unit 524G can generate the VS 525K matrix using the reordered VTDIST vectors 539. In these examples, the salient component analysis unit 524G can determine the distinction based on the V matrix 519, before any modification based on matrix S 519B. In other words, according to these examples, the salient component analysis unit 524G can determine directionality using only the V matrix 519, without performing the step of generating the matrix VS 525K. More specifically, the V matrix 519A can provide information about the manner in which the components (e.g., vectors of the V matrix 519) of the audio data are mixed, and potentially information about the various synergistic effects of the data transmitted by the vectors. For example, the V matrix 519A can provide the information about the “arrival direction” of the various audio components represented by the vectors, such as the arrival direction of each audio component, as relayed to the audio encoding device 510G by a EigenMike®. As used herein, the term "audio data component" may be used interchangeably with the "input" of any of matrices 519 or any matrices derived therefrom.

[0944] In accordance with some implementations of the techniques of this disclosure, the 524G salient component analysis unit may supplement or augment the SHC representations with extra information to make various determinations described herein. As an example, the salient component analysis unit 524G can augment the SHC with extra information in order to determine the salience of various audio components represented in the arrays 519-519C. As another example, the salient component analysis unit 524G and/or the vector reordering unit 532 can augment the HOA with extra data to distinguish between discrete audio objects and background audio data.

[0945] In some instances, the 524G salient component analysis unit can detect that portions (eg distinct audio objects) of the audio data exhibit Keynesian energy. An example of such distinct objects can be associated with a human voice that modulates. In the case of voice-based audio data that modulates, the salient component analysis unit 524G can determine that the energy of the modulation data, as the ratio for the energies of the remaining components, remains approximately constant (e.g., constant within threshold range) or approximately stationary over time. Traditionally, if the energy characteristics of discrete audio components with Keynesian energy (e.g., those with the modulation voice) change from one audio frame to another, a device may not have the ability to identify the series of audio components. as a single signal. However, the 524G salient component analysis unit can implement techniques of this revelation to determine a directionality or an aperture of the distance object represented as a vector in the various matrices.

[0946] More specifically, the 524G salient component analysis unit can determine that characteristics such as directionality and/or aperture are unlikely to change substantially across audio frames. As used herein, the aperture represents a ratio of higher order coefficients to lower order coefficients within the audio data. Each row of the V matrix 519A may include vectors that correspond to particular SHCs. The 524G salient component analysis unit can determine that lower-order SHCs (eg, associated with an order less than or equal to) tend to represent ambient data, while higher-order inputs tend to represent discrete data. Additionally, the 524G salient component analysis unit can determine that, in many instances, higher-order SHCs (e.g., associated with an order greater than 1) exhibit greater energy and that the energy ratio of higher-order SHCs for lower order remains substantially similar (or approximately constant) from audio frame to audio frame.

[0947] One or more components of the 524G salient component analysis unit can determine the characteristics of the audio data such as directionality and aperture, using the V matrix 519. In this way, the components of the 510G audio encoding device , such as the 524G salient component analysis unit, can implement the techniques described in this document to determine salience and/or distinguish distinct audio objects from background audio, using the information based on directionality. Using directionality to determine salience and/or distinction, the 524G salient component analysis unit can arrive at more robust determinations than in cases of a device configured to determine salience and/or distinction using data based on energy only. While described above with respect to determinations based on salience directionality and/or distinctiveness, the 524G salient component analysis unit can deploy the techniques of this disclosure to use directionality in addition to other characteristics, such as energy, to determine salience and /or distinguishing particular components of the audio data, as represented by vectors of one or more of the matrices 519-519C (or any matrix derived therefrom).

[0948] In some examples the method includes identifying one or more audio objects distinct from one or more spherical harmonic coefficients (SHC) associated with the audio objects based on a given directionality for one or more of the audio objects. In one example, the method additionally includes determining the directionality of one or more audio objects based on the spherical harmonic coefficients associated with the audio objects. In some examples, the method additionally includes performing a singular value decomposition with respect to spherical harmonic coefficients to generate a matrix U representative of left singular vectors of the plurality of spherical harmonic coefficients, a matrix S representative of singular values of the plurality of harmonic coefficients spherical and a matrix V representative of singular vectors to the right of the plurality of spherical harmonic coefficients; and representing the plurality of spherical harmonic coefficients as a function of at least a portion of one or more of the matrix U, the matrix S and the matrix V, wherein the determination of the respective directionality of the one or more audio objects is based on at least in part in the V matrix.

[0949] In one example, the method additionally includes reordering one or more vectors of the matrix V so that vectors that have a higher directionality quotient are positioned above vectors that have a lower directionality quotient in the reordered matrix V. In one example, the method additionally includes determining that vectors that have the higher directionality quotient include greater directional information than vectors that have the lower directionality quotient. In one example, the method additionally includes multiplying the matrix V by the matrix S to generate a matrix VS, where the matrix VS includes one or more vectors. In one example, the method additionally includes selecting entries from each row of the VS matrix that are associated with an order greater than 1, frame raising each of the selected entries to form corresponding squared entries, and for each row of the VS matrix , sum all inputs raised to the frame to determine a directionality quotient for a corresponding vector.

[0950] In some examples, each row of the VS matrix includes 25 entries. In one example, selecting entries from each row of the VS matrix associated with order greater than 1 includes selecting all entries starting with a 5th entry of each row of the VS matrix and ending with a 25th entry of each row of the VS matrix. In one example, the method additionally includes selecting a subset of the vectors from the VS matrix to represent the distinct audio objects. In some examples, subset selection includes selecting four vectors from the VS matrix, and the four selected vectors have the four highest directionality quotients of all vectors from the VS matrix. In one example, the determination that the selected subset of the vectors represents the distinct audio objects is based on both the directionality and an energy of each vector.

[0951] In some examples the method includes identifying one or more audio objects distinct from one or more spherical harmonic coefficients associated with the audio objects based on a directionality and energy determined for one or more of the audio objects. In one example, the method additionally includes determining one or both of the directionality and energy of the one or more audio objects based on the spherical harmonic coefficients associated with the audio objects. In some examples, the method additionally includes performing a singular value decomposition with respect to spherical harmonic coefficients representative of the sound field to generate a matrix U representative of singular vectors to the left of the plurality of spherical harmonic coefficients, a matrix S representative of singular values of the plurality of spherical harmonic coefficients and a matrix V representative of singular vectors to the right of the plurality of spherical harmonic coefficients and representing the plurality of spherical harmonic coefficients as a function of at least a portion of one or more of the matrix U, the matrix S and the matrix V, where the determination of the respective directionality of the one or more audio objects is based at least in part on the matrix V and where the determination of the respective energy of the one or more audio objects is based at least on part, in matrix S.

[0952] In one example, the method additionally includes multiplying the matrix V by the matrix S to generate a matrix VS, where the matrix VS includes one or more vectors. In some examples, the method additionally includes selecting VS matrix entries that are associated with an order greater than 1, framing each of the selected entries to form corresponding squared entries, and, for each row of the VS matrix, summing all the inputs raised to the frame to generate a directionality quotient for a corresponding vector of the VS matrix. In some examples, each row of the VS matrix includes 25 entries. In one example, selecting entries from each row of the VS matrix associated with order greater than 1 comprises selecting all entries starting with a 5th entry of each row of the VS matrix and ending with a 25th entry of each row of the VS matrix. In some examples, the method additionally includes selecting a subset of the vectors to represent the distinct audio objects. In one example, subset selection comprises selecting four vectors from the VS matrix, and the four selected vectors have the four highest directionality quotients of all vectors from the VS matrix. In some examples, the determination that the selected subset of the vectors represents the distinct audio objects is based on both the directionality and an energy of each vector.

[0953] In some examples, a method includes determining, using information based on directionality, one or more first vectors that describe the distinct components of the sound field and one or more second vectors that describe the background components of the field. sound, both the one or more first vectors and the one or more second vectors generated by at least performing a transformation with respect to the plurality of spherical harmonic coefficients. In one example, the transformation comprises a singular-value decomposition that generates a matrix U representative of the left-hand singular vectors of the plurality of spherical harmonic coefficients, a matrix S representative of the singular values of the plurality of spherical harmonic coefficients, and a matrix V representative of the vectors singulars to the right of the plurality of spherical harmonic coefficients. In one example, the transformation comprises principal component analysis to identify the distinct components of the sound field and the background components of the sound field.

[0954] In some examples, a device is configured or otherwise operable to perform any of the techniques described herein or any combination of the techniques. In some examples, a computer-readable storage medium is encoded with instructions that, when executed, cause one or more processors to perform any of the techniques described herein or any combination of the techniques. In some examples, a device includes means for performing any of the techniques described herein or any combination of the techniques.

[0955] That is, prior aspects of the techniques may allow the 510G audio encoding device to be configured to operate in accordance with the following clauses.

[0956] Clause 134954-1B. A device, such as the 510G audio encoding device, comprising: one or more processors configured to identify one or more audio objects distinct from one or more spherical harmonic coefficients associated with the audio objects, based on a directionality and an energy determined for one or more of the audio objects.

[0957] Clause 134954-2B. The device of clause 134954-IB, wherein the one or more processors are further configured to determine one or both of the directionality and energy of the one or more audio objects based on the spherical harmonic coefficients associated with the audio objects.

[0958] Clause 134954-3B. The device of any one of claims 1B or 2B or combinations thereof, wherein the one or more processors are further configured to perform a singular value decomposition with respect to spherical harmonic coefficients representative of the sound field to generate a matrix U representative of the vectors left singular vectors of the plurality of spherical harmonic coefficients, a matrix S representative of the singular values of the plurality of spherical harmonic coefficients and a matrix V representative of the right singular vectors of the plurality of spherical harmonic coefficients and representing the plurality of spherical harmonic coefficients as a function of at least a portion of one or more of the U matrix, the S matrix and the V matrix, wherein the one or more processors are configured to determine the respective directionality of the one or more audio objects based at least in part , in matrix V, and where the one or more processors are configured to determine the respective energy of one or more audio objects based, at least in part, on the matrix S.

[0959] Clause 134954-4B. The device of clause 134954-3B, wherein the one or more processors are further configured to multiply the matrix V by the matrix S to generate a matrix VS, the matrix VS including one or more vectors.

[0960] Clause 134954-5B. The device of clause 134954-4B, wherein the one or more processors are further configured to select entries from each row of the VS matrix that are associated with an order greater than 1, frame each of the selected entries to form high entries squared and, for each row of the VS matrix, sum all entries raised to the frame to generate a directionality quotient for a corresponding vector of the VS matrix.

[0961] Clause 134954-6B. The device of any one of claims 5B and 6B or a combination thereof, wherein each row of the VS matrix includes 25 inputs.

[0962] Clause 134954-7B. The device of clause 134954-6B, wherein the one or more processors are configured to select all entries starting at a 5th entry in each row of the VS array and ending at a 25th entry in each row of the VS array.

[0963] Clause 134954-8B. The device of either clause 134954-6B and clause 134954-7B or a combination thereof, wherein the one or more processors are further configured to select a subset of the vectors to represent distinct audio objects.

[0964] Clause 134954-9B. The device of clause 134954-8B, where the one or more processors are configured to select four vectors from the VS matrix, and where the four selected vectors have the four highest directionality quotients of all vectors from the VS matrix.

[0965] Clause 134954-10B. The device of either clause 134954-6B and clause 134954-7B or a combination thereof, wherein the one or more processors are further configured to determine that the selected subset of vectors representing distinct audio objects is based on both in directionality and in an energy of each vector.

[0966] Clause 134954-1C. A device, such as the 510G audio encoding device, comprising: one or more processors configured to determine, using directionality-based information, one or more first vectors that describe the distinct components of the sound field, and one or more second vectors describing the background components of the sound field, both the one or more first vectors and the one or more second vectors generated by at least performing a transformation with respect to the plurality of spherical harmonic coefficients.

[0967] Clause 134954-2C. The method of clause 134954-1C, wherein the transformation comprises a singular value decomposition that generates a matrix U representative of the left singular vectors of the plurality of spherical harmonic coefficients, a matrix S representative of the singular values of the plurality of spherical harmonic coefficients, and a matrix V representative of the singular vectors to the right of the plurality of spherical harmonic coefficients.

[0968] Clause 134954-3C. The method of clause 134954-2C, which further comprises the operations recited by any combination of clause 134954-1A to clause 134954-12A and clause 134954-1B to clause 134954-9B.

[0969] Clause 134954-4C. The method of clause 134954-1C, wherein the transformation comprises principal component analysis to identify the distinct components of the sound field and the background components of the sound field.

[0970] Figure 40H is a block diagram illustrating the exemplary audio encoding device 510H that can perform various aspects of the techniques described in this disclosure to compress spherical harmonic coefficients that describe two or three dimensional sound fields. The audio encoding device 510H may be similar to the audio encoding device 510G in that the audio encoding device 510H includes an audio compression unit 512, an audio encoding unit 514, and a stream generation unit. bits 516. Furthermore, the audio compression unit 512 of the audio encoding device 510H may be similar to that of the audio encoding device 510G wherein the audio compression unit 512 includes a decomposition unit 518 and a 520G sound field component extraction unit, which can operate similarly to similar units of 510G audio encoding device. In some examples, the audio encoding device 510H may include a quantization unit 534, as described with respect to Figures 40D and 40E, to quantize one or more vectors of any of the UDIST 525C vectors, the UBG 525D vectors, the 525D vectors, the VTDIST 525E and the VTBG 525J vectors.

[0971] The audio compression unit 512 of the audio encoding device 510H may, however, differ from the audio compression unit 512 of the audio encoding device 510G in that the audio compression unit 512 of the encoding device audio unit 510H includes an additional unit denoted as interpolation unit 550. Interpolation unit 550 may represent a unit that interpolates subframes of a first audio frame from subframes of the first audio frame and a temporarily subsequent or preceding audio frame, as described in more detail below with respect to Figures 45 and 45B. The interpolation unit 550 can, in performing this interpolation, reduce computational complexity (in terms of processing cycles and/or memory consumption) by potentially reducing the extent to which the decomposition unit 518 is required to decompose the SHC 511. In this regard, the interpolation unit 550 can perform operations similar to those described above with respect to the spatiotemporal interpolation unit 50 of the audio encoding device 24 shown in the example of Figure 4.

[0972] That is, the singular value decomposition performed by the decomposition unit 518 potentially consumes a lot of memory and/or processor, while also, in some examples, it takes extensive amounts of time to decompose the SHC 511, especially as the order of SHC 511 increases. In order to reduce the amount of time and make SHC 511 compression more efficient (in terms of processing cycles and/or memory consumption), the techniques described in this disclosure may provide for interpolation of one or more subframes of the first frame of data. audio, where each of the subframes may represent decomposed versions of the SHC 511. Instead of performing SVD over the entire frame, the techniques may allow the decomposition unit 518 to decompose a first subframe from a first audio frame, generating a V matrix 519'.

[0973] Decomposing unit 518 may also decompose a second subframe of a second audio frame, wherein that second audio frame may be temporarily subsequent or temporarily preceding the first audio frame. The decomposition unit 518 may output a V matrix 519' for that subframe of the second audio frame. The interpolation unit 550 can then interpolate the remaining subframes of the first audio frame based on the V matrices 519' decomposed from the first and second subframes, outputting the V matrix 519, the S matrix 519B and the U matrix 519C, wherein the decompositions for the remaining subframes can be computed based on the SHC 511, the V matrix 519A for the first audio frame, and the interpolated V matrices 519 for the remaining subframes of the first audio frame. Interpolation can therefore avoid computing the decays for the remaining subframes of the first audio frame.

[0974] Also, as noted above, the U 519C matrix may not be continuous from frame to frame, whereby the discrete components of the U 519C matrix decomposed from a first audio frame of the SHC 511s may be specified in rows and /or different columns than in the U 519C matrix decomposed from a second audio frame from the SHC 511. By performing this interpolation, the discontinuity can be reduced as a linear interpolation can have a smoothing effect that can reduce any artifacts introduced due to frame limits (or, in other words, the segmentation of SHC 511s into frames). Using the V matrix 519' to perform this interpolation and then retrieving the U matrices 519C based on the V interpolated matrix 519' of the SHC 511 can smooth out any effects of reordering the U matrix 519C.

[0975] In operation, the interpolation unit 550 may interpolate one or more subframes of a first audio frame of a first decomposition, e.g., the V matrix 519', from a portion of a first plurality of spherical harmonic coefficients 511 included in the first frame and a second decomposition, for example, matrix V 519' of a portion of a second plurality of spherical harmonic coefficients 511 included in a second frame to generate the decomposed interpolated spherical harmonic coefficients for the one or more subframes.

[0976] In some examples, the first decomposition comprises the first matrix V 519' representative of the singular vectors to the right of the portion of the first plurality of spherical harmonic coefficients 511. Likewise, in some examples, the second decomposition comprises the second matrix V 519' representative of the singular vectors to the right of the portion of the second plurality of spherical harmonic coefficients.

[0977] The interpolation unit 550 may perform temporal interpolation with respect to the one or more subframes based on the first V matrix 519' and the second V matrix 519'. That is, the interpolation unit 550 can temporarily interpolate, for example, the second, third and fourth subframes out of four total subframes for the first audio frame based on a V matrix 519' decomposed from the first subframe of the first audio frame and the V matrix 519' decomposed from the first subframe of the second audio frame. In some examples, this temporal interpolation is a linear temporal interpolation, where the matrix V 519' decomposed from the first subframe of the first audio frame is weighted heavier when interpolating the second subframe of the first audio frame than when interpolating the fourth subframe of the first audio frame. By interpolating the third subframe, the V arrays 519' can be weighted equally. By interpolating the fourth subframe, the V matrix 519' decomposed from the first subframe of the second audio frame can be more heavily weighted than the V matrix 519' decomposed from the first subframe of the first audio frame.

[0978] In other words, linear temporal interpolation can weight V 519' matrices given the proximity of one of the subframes of the first audio frame to be interpolated. For the second subframe to be interpolated, the V matrix 519' decomposed from the first subframe of the first audio frame is weighted more heavily given its proximity to the second subframe to be interpolated than the V matrix 519' decomposed from the first subframe of the second audio frame. The weights can be equivalent for that reason when interpolating the third subframe based on the V 519' matrices. The weight applied to the V matrix 519' decomposed from the first subframe of the second audio frame can be greater than that applied to the V matrix 519' decomposed from the first subframe of the first audio frame as the fourth subframe is to be interpolated is closer to the first subframe of the second audio frame than to the first subframe of the first audio frame.

[0979] Although, in some examples, only a first subframe of each audio frame is used to perform the interpolation, the portion of the first plurality of spherical harmonic coefficients may comprise two of four subframes of the first plurality of spherical harmonic coefficients 511. In such and in other examples, the portion of the second plurality of spherical harmonic coefficients 511 comprises two of four subframes of the second plurality of spherical harmonic coefficients 511.

[0980] As noted above, a single device, for example the audio encoding device 510H, can perform the interpolation while also decomposing the portion of the first plurality of spherical harmonic coefficients to generate the first decompositions of the portion of the first plurality of coefficients spherical harmonics. In these and other examples, the decomposition unit 518 may decompose the portion of the second plurality of spherical harmonic coefficients to generate the second decompositions of the portion of the second plurality of spherical harmonic coefficients 511. Although described with respect to a single device, two or more devices can perform the techniques described in this disclosure, wherein one of the two devices performs the decomposition and the other of the devices performs the interpolation in accordance with the techniques described in this disclosure.

[0981] In other words, 3D audio based on spherical harmonics can be a parametric representation of the 3D pressure field in terms of orthogonal basis functions on a sphere. The higher the N order of the representation, the potentially higher the spatial resolution and often the greater the number of spherical harmonic (SH) coefficients (for a total of (N+1)2 coefficients). For many applications, bandwidth compression of the coefficients may be required to be able to efficiently transmit and store the coefficients. These techniques addressed in this disclosure can provide a frame-based dimensionality reduction process that uses Singular Value Decomposition (SVD). SVD analysis can decompose each frame of coefficients into three matrices U, S, and V. In some examples, the techniques can manipulate some of the U-vectors as directional components of the underlying sound field. However, when manipulated in this way, these (U-shaped) vectors are discontinuous from frame to frame - even though they represent the same distinct audio component. These discontinuities can lead to significant artifacts when components are fed through transform audio encoders.

[0982] The techniques described in this disclosure can address this discontinuity. That is, the techniques can be based on the observation that the matrix V can be interpreted as orthogonal spatial geometric axes in the domain of Spherical Harmonics. The matrix U can represent a projection of the Spherical Harmonics (HOA) data in terms of those basis functions, where the discontinuity can be attributed to the basis functions (V) that change every frame - and are therefore discontinuous. This is different from similar decomposition, such as the Fourier Transform, where the basis functions are, in some instances, constants from frame to frame. In these terms, SVD can be considered as a compatibility search algorithm. The techniques described in this disclosure may allow the interpolation unit 550 to maintain continuity between the base (V) functions from frame to frame - by interpolation therebetween.

[0983] In some examples, the techniques allow the interpolation unit 550 to divide the frame of the SH data into four subframes, as described above and further described below with respect to the Figures. 45 and 45B. The interpolation unit 550 can then compute the SVD for the first subframe. Similarly, the SVD is computed for the first subframe of the second frame. For each of the first frame and second frame, the interpolation unit 550 can convert the V vectors into a spatial map by projecting the vectors onto a sphere (using a projection matrix such as a design matrix). T). The interpolation unit 550 can then interpret the vectors in V as formed on a sphere. To interpolate the V matrices for the three subframes between the first subframe of the first frame and the first subframe of the next frame, the interpolation unit 550 can then interpolate these spatial formats - and then transform them back to vectors of SH through the inversion of the projection matrix. The techniques of this disclosure can thus provide a smooth transition between V matrices.

[0984] In this way, the 510H audio encoding device can be configured to perform various aspects of the techniques set forth below in relation to the following clauses.

[0985] Clause 135054-1A. A device, such as audio encoding device 510H, comprising: one or more processors configured to interpolate one or more subframes of a first frame from a first decomposition of a portion of a first plurality of spherical harmonic coefficients included in the first frame and a second decomposition of a portion of a second plurality of spherical harmonic coefficients included in a second frame to generate the decomposed interpolated spherical harmonic coefficients for the one or more subframes.

[0986] Clause 135054-2A. The device of clause 135054-1A, wherein the first decomposition comprises a first matrix V representative of the singular vectors to the right of the portion of the first plurality of spherical harmonic coefficients.

[0987] Clause 135054-3A. The device of clause 135054-1A, wherein the second decomposition comprises a second matrix V representative of the singular vectors to the right of the portion of the second plurality of spherical harmonic coefficients.

[0988] Clause 135054-4A. The device of clause 135054-1A, wherein the first decomposition comprises a first matrix V representative of the singular vectors to the right of the portion of the first plurality of spherical harmonic coefficients, and wherein the second decomposition comprises a second matrix V representative of the singular vectors to the right. right of the portion of the second plurality of spherical harmonic coefficients.

[0989] Clause 135054-5A. The device of clause 135054-1A, wherein the one or more processors are further configured to, when interpolating the one or more subframes, temporarily interpolate the one or more subframes based on the first decomposition and the second decomposition.

[0990] Clause 135054-6A. The provision of clause 135054-1A, wherein the one or more processors are additionally configured to, by interpolating the one or more subframes, project the first decomposition into a spatial domain to generate the first projected decompositions, project the second decomposition into the spatial domain to generate the second projected decompositions, spatially interpolate the first projected decompositions and the second projected decompositions to generate a spatially interpolated first projected decomposition and a second spatially interpolated projected decomposition, and temporarily interpolate the one or more subframes based on the first spatially interpolated projected decomposition, and the second spatially interpolated projected decomposition.

[0991] Clause 135054-7A. The device of clause 135054-6A, wherein the one or more processors are further configured to project the temporarily interpolated spherical harmonic coefficients that result from interpolating the one or more subframes back into a spherical harmonic domain.

[0992] Clause 135054-8A. The device of clause 135054-1A, wherein the portion of the first plurality of spherical harmonic coefficients comprises a single subframe of the first plurality of spherical harmonic coefficients.

[0993] Clause 135054-9A. The device of clause 135054-1A, wherein the portion of the second plurality of spherical harmonic coefficients comprises a single subframe of the second plurality of spherical harmonic coefficients.

[0994] Clause 135054-10A. The provision of clause 135054-1A,

[0995] where the first frame is divided into four subframes, and

[0996] wherein the portion of the first plurality of spherical harmonic coefficients comprises only the first subframe of the first plurality of spherical harmonic coefficients.

[0997] Clause 135054-11A. The provision of clause 135054-1A,

[0998] where the second frame is divided into four subframes, and

[0999] wherein the portion of the second plurality of spherical harmonic coefficients comprises only the first subframe of the second plurality of spherical harmonic coefficients.

[1000] Clause 135054-12A. The device of clause 135054-1A, wherein the portion of the first plurality of spherical harmonic coefficients comprises two of four subframes of the first plurality of spherical harmonic coefficients.

[1001] Clause 135054-13A. The device of clause 135054-1A, wherein the portion of the second plurality of spherical harmonic coefficients comprises two of four subframes of the second plurality of spherical harmonic coefficients.

[1002] Clause 135054-14A. The device of clause 135054-1A, wherein the one or more processors are further configured to decompose the portion of the first plurality of spherical harmonic coefficients to generate first decompositions of the portion of the first plurality of spherical harmonic coefficients.

[1003] Clause 135054-15A. The device of clause 135054-1A, wherein the one or more processors are further configured to decompose the portion of the second plurality of spherical harmonic coefficients to generate the second decompositions of the portion of the second plurality of spherical harmonic coefficients.

[1004] Clause 135054-16A. The device of clause 135054-1A, wherein the one or more processors are further configured to perform a singular value decomposition with respect to the portion of the first plurality of spherical harmonic coefficients to generate a matrix U representative of left singular vectors of the first plurality of spherical harmonic coefficients, a matrix S representative of singular values of the first plurality of spherical harmonic coefficients and a matrix V representative of singular vectors to the right of the first plurality of spherical harmonic coefficients.

[1005] Clause 135054-17A. The clause device 135054-1A, wherein the one or more processors are further configured to perform a singular value decomposition with respect to the portion of the second plurality of spherical harmonic coefficients to generate a matrix U representing left singular vectors of the second plurality of spherical harmonic coefficients, a matrix S representing singular values of the second plurality of spherical harmonic coefficients and a matrix V representing singular vectors to the right of the second plurality of spherical harmonic coefficients.

[1006] Clause 135054-18A. The device of clause 135054-1A, wherein each of the first and second plurality of spherical harmonic coefficients represents a plane wave representation of the sound field.

[1007] Clause 135054-19A. The clause device 135054-1A, wherein each of the first and second plurality of spherical harmonic coefficients represents one or more mono audio objects mixed together.

[1008] Clause 135054-20A. The device of clause 135054-1A, wherein each of the first and second plurality of spherical harmonic coefficients comprises respective first and second spherical harmonic coefficients representing a three-dimensional sound field. [0952] clause 135054-21A. The device of clause 135054-1A, wherein each of the first and second plurality of spherical harmonic coefficients is associated with at least one spherical basis function that has an order greater than one.

[1009] Clause 135054-22A. The device of clause 135054-1A, wherein each of the first and second plurality of spherical harmonic coefficients is associated with at least one spherical basis function that has an order of four.

[1010] While described above as being performed by the exemplary audio encoding device 510H, the various audio decoding devices 24 and 540 may also perform any of various aspects of the techniques set forth in connection with clauses 135054-1A through 135054-22A .

[1011] Fig. 40I is a block diagram illustrating an exemplary audio encoding device 510I that can perform various aspects of the techniques described in this disclosure to compress spherical harmonic coefficients that describe two or more three-dimensional sound fields. The exemplary audio encoding device 510I may be similar to the exemplary audio encoding device 510H wherein the exemplary audio encoding device 510I includes an audio compression unit 512, an audio encoding unit 514, and a generation unit. 516. In addition, the audio compression unit 512 of the exemplary audio encoding device 510I may be similar to that of the exemplary audio encoding device 510H wherein the audio compression unit 512 includes a decomposition unit. 518 and a sound field component extraction unit 520, which may operate similarly to similar units of the exemplary audio encoding device 510H. In some examples, the exemplary audio coding device 10I may include a quantizing unit 351/500 34, as described in connection with Figures 3D and 3E, for quantizing one or more vectors of any one of UDIST 25C, UBG 25D, VTDIST 25E and VTBG 25J.

[1012] However, although both the audio compression unit 512 of the exemplary audio coding device 510I and the audio compression unit 512 of the exemplary audio coding device 10H include a sound field component extraction unit, the Sound field component extraction unit 520I of exemplary audio encoding device 510I may include an additional module referred to as V compression unit 552. V compression unit 552 may represent a unit configured to compress a spatial component of the sound field, that is, one or more of the VTDIST 539 vectors in this example. That is, the singular value decomposition performed in relation to the SHC can decompose the SHC (which is representative of the sound field) into energy components represented by vectors of the matrix S, time components represented by the matrix U and special components represented by the matrix V The V compression unit 552 can perform operations similar to those described above with respect to the quantization unit 52.

[1013] For example purposes, the VTDIST 539 vectors are assumed to comprise two row vectors that have 25 elements each (implying a fourth-order HOA representation of the sound field). Although described in relation to the two vectors in a row, any number of vectors can be included in the VTDIST vectors 539 through (N+1)2, where n denotes the order of the HOA representation of the sound field.

[1014] The compression unit V can receive the VTDIST 539 vectors and perform a 352/500 compression scheme to generate VTDIST 539' compressed vector representations. This compression scheme can involve any conceivable compression scheme for compressing the elements of an array or generally data, and should not be limited to the example described below in more detail.

[1015] The VTDIST 552 compression unit can perform, as an example, a compression scheme that includes one or more of transforming floating-point representations of each element of the VTDIST 539 vectors to integer representations of each element of the VTDIST 539 vectors, standardizing the quantization of the integer representations of the VTDIST 539 vectors; and the categorization and encoding of the quantized integer representations of the VTDIST 539 vectors. Several of the one or more processes of this compression scheme can be dynamically controlled by parameters to reach or almost reach, as an example, a rate of target bits for the resulting bitstream 517.

[1016] Since each of the VTDIST 539 vectors are orthonormal to each other, each of the VTDIST 539 vectors can be encoded independently. In some examples, as described in more detail below, each element of each VTDIST 539 array can be encoded using the same encoding mode (defined by various sub-modes).

[1017] In any case, as noted above, this encoding scheme may involve first transforming the floating-point representations of each element (which is, in some examples, a 32-bit floating-point number) of each of the vectors VTDIST 539 for a 16-bit integer representation. The V 552 compression unit can perform these floating-point-to-integer transformations by multiplying each element of a given one of the VTDIST 539 vectors by 215, which is, in some examples, performed by a right shift 353/500 by 15.

[1018] The compression unit V 552 can then perform uniform quantization with respect to all elements of the data one of the vectors VTDIST 539. The compression unit V 552 can identify a quantization step size based on a value , which can be denoted as an nbits parameter. The V compression unit 552 can dynamically determine this nbit parameter based on a target bit rate. The V compression unit 552 can determine the quantization step size as a function of this nbit parameter. As an example, the V 552 compression unit can set the quantization step size (denoted as “delta” or “ ” in this disclosure) to be equal to 2l6-nbits. In this example, if nbits equals six, delta equals 210 and there are 26 levels of quantization. In this respect, for a vector element , the quantized vector element q is equal to

[1019] The V compression unit 552 can then perform categorization and residual encoding of the quantized vector elements. As an example, the compression unit V 552 can, for a given quantized vector element q identify a category (by determining a category identifier cid) to which that element corresponds using the following equation:

[1020] The V compression unit 552 can then Huffman-encode this category index cid, while also identifying a sign bit that indicates whether q is a positive value or a negative value. The V 552 compression unit can next identify a residue in this category. As an example, the 354/500 compression unit V 552 can determine this residual according to the following equation:

[1021] The V compression unit 552 can then block encode this residue with cid-l bits.

[1022] The following example illustrates a simplified example of this categorization and residual encoding process. First, nbits is assumed to be six so that q [-31,31]. The following is assumed:

[1023] In addition, the following is assumed:

[1024] So for a q = [6, -17, 0, 0, 3], the following can be determined:

[1025] » cid = 3,5,0,0,2

[1026] » sign = 1,0,x,x,l

[1027] » residual = 2,1,x,x, 1

[1028] » Bits for 6 = '0010' + ' + '10'

[1029] » Bits for -17 = '00111 ' + '0' + '0001 '

[1030] » Bits for 0 = ‘0’

[1031] » Bits for 0 = '0'

[1032] » Bits for 3 = '000 ' + '1' + '1 '

[1033] » Total Bits = 7+ 10+ 1 + 1 +5 = 24

[1034] » Average bits = 24/5 = 4.8

[1035] Although not shown in the above simplified example, the V 552 compression unit can select different Huffman codebooks for different nbit values during cid encoding. In some examples, the compression unit V 552 may provide a different Huffman codebook for values of nbits 6,..., 15. In addition, the compression unit V 552 may include five 356/500 codebooks. different Huffman values for each of the different nbit values in the range of 6, 15 for a total of 50 Huffman codebooks. In this regard, the V compression unit 552 may include a plurality of different Huffman codebooks to accommodate cid encoding in a number of different statistical contexts.

[1036] To illustrate, the V compression unit 552 may, for each of the nbit values, include a first Huffman codebook for encoding vector elements one through four, a second Huffman codebook for encoding vector elements five through nine, a third Huffman codebook for coding vector elements nine and up. These first three Huffman codebooks can be used when the one of the VTDIST 539 vectors to be compressed is not predicted from a temporally subsequent one that corresponds to one of the VTDIST 539 vectors and is not representative of spatial information from a data object. synthetic audio (that defined, for example, originally by a pulse code modulated (PCM) audio object). The V-compression unit 552 may additionally include, for each of the nbit values, a fourth Huffman codebook to encode the one of the VTDIST vectors 539 when that one of the VTDIST vectors 539 is predicted from a temporally subsequent one that corresponds to one of the VTDIST 539 vectors. The VTDIST 552 compression unit may also include, for each of the nbit values, a fifth Huffman codebook to encode the one of the VTDIST 539 vectors when that one of the VTDIST 539 vectors is representative of a synthetic audio object. Huffman's various codebooks can be developed for each of these different statistical contexts, that is, the unanticipated and non-synthetic context, the predicted context, and the synthetic context in this example.

[1037] The following table illustrates the Huffman table selection and the bits to be specified in the bitstream to allow the decompression unit to select the appropriate Huffman table:

[1038] In the above table, the prediction mode (“Pred mode”) indicates whether the prediction was performed for the current vector, while the Huffman Table (“HT info”) indicates additional Huffman codebook information (or table) used to select one of Huffman tables one through five.

[1039] The following table further illustrates this Huffman table selection process given various statistical contexts or scenarios.

[1040] In the above table, the column “Record” indicates the encoding context when the 358/500 vector is representative of an audio object that has been registered while the column “Synthetic” indicates an encoding context for when the vector is representative of a synthetic audio object. The row “Without Prediction” indicates encoding context when prediction is not performed against vector elements, while row “With prediction” indicates encoding context when prediction is performed against vector elements. As shown in this table, the compression unit V 552 selects HT{1, 2, 3} when the vector is representative of a recorded audio object and prediction is not performed with respect to vector elements. The V 552 compression unit selects HT5 when the audio object is represented from a synthetic audio object and prediction is not performed with respect to vector elements. The V compression unit 552 selects HT4 when the vector is representative of a recorded audio object and prediction is performed with respect to vector elements. The V 552 compression unit selects HT5 when the audio object is representative of a synthetic audio object and the prediction is performed with respect to vector elements.

[1041] In this way, the techniques can allow an audio compression device to compress a spatial component of a sound field, where the spatial component is generated by performing a synthesis based on a plurality of spherical harmonic coefficients.

[1042] Figure 43 is a diagram illustrating the V compression unit 552 shown in Figure 40I in more detail. In the example of Fig. 43, the V compression unit 552 includes a uniform quantization unit 600, an nbit unit 602, a prediction unit 604, a prediction mode unit 606 ("Pred Mode Unit 606"), a 359 /500 residual and category coding unit 608 and a Huffman table selection unit 610. The uniform quantization unit 600 represents a unit configured to perform the quantization described above with respect to one of the special components denoted as in the example in Fig. 43 (which can represent any of the VTDIST 539 vectors). Unit of nbits 602 represents a unit configured to determine the parameter or value of nbits.

[1043] The prediction unit 604 represents a unit configured to perform the prediction with respect to the quantized spatial component denoted as q in the example of Fig. 43. The prediction unit 604 can perform the prediction by performing an element subtraction from one of the current VTDIST vectors 539 by a temporally subsequent one that corresponds to one of the VTDIST vectors 539. The result of this prediction can be referred to as a predicted spatial component.

[1044] Prediction mode unit 606 may represent a unit configured to select prediction mode. The Huffman table selection unit 610 may represent a unit configured to select an appropriate Huffman table to encode the cid. The prediction mode unit 606 and the Huffman table selection unit 610 may operate, as an example, according to the following pseudocode:

[1045] For a given nbits, retrieve all Huffman Tables that have nbits

[1046] B00 = 0; B01 = 0; B10 = 0; B11 = 0; // initialize to compute expected bits per encoding mode

[1047] for m = 1:(# elements in vector)

[1048] // calculate expected number of bits for a vector element v(m)

[1049] // without prediction and using Huffman's Table 5

[1050] B00 = B00 + calculate_bits(v(m) , HT5);

[1051] // without prediction and using Huffman's Table {1,2,3}

[1052] B01 = B01 + calculate_bits(v(m) , HTq); q in {1,2,3}

[1053] // calculate expected number of bits for residual prediction

[1054] e(m e(m) ) = v(m) - vp(m); // vp(m) : frame vector element

[1055] previous // with prediction and using Huffman's Table 4

[1056] B10 = B10 + calculate_bits(e(m) , HT4);

[1057] // with prediction and using Huffman's Table 5

[1058] B11 = B11 + calculate_bits(e(m) , HT5);

[1059] end

[1060] // find a better prediction mode and Huffman table that yield minimum bits

[1061] // best prediction mode and Huffman table are flagged by pflag and Htflag respectively

[1062] [Be, id] = min([BOO B01 B10 B11]);

[1063] Switch Case 1: id

[1064] Case 1:pflag = 0; HTflag = 0;

[1065] Case 2: pflag = 0; HTflag = 1;

[1066] Case 3: pflag = 1; HTflag = 0;

[1067] Case 4: pflag = 1; HTflag = 1;

[1068] end Residual and category encoding unit 608 may represent a unit configured to perform categorization and residual encoding of a predicted spatial component or the quantized spatial component (when prediction is turned off) in the manner described in more detail below .

[1069] As shown in the example of Figure 43, the V compression unit 552 can output various parameters or values for inclusion in the bit stream 517 or side information (which itself can be a separate bit stream from the bit stream 517). bits 517). Assuming the information is specified in the bitstream 517, the V compression unit 552 can output the nbits value, prediction mode and Huffman table information to the bitstream generating unit 516 along with the compressed version of the spatial component (shown as compressed spatial component 539' in the example of Fig. 40I), which in this example may refer to the Huffman code selected to encode the cid, the sign bit, and the block-coded residue. The value of nbits can be specified once in the 517 bit stream for all VTDIST 539 arrays, while the prediction mode and Huffman table information can be specified for each of the VTDIST 539 arrays. The bit stream portion that specifies the compressed version of the spatial component is shown in the example of Figures 10B and 10c.

[1070] In this manner, the exemplary audio encoding device 510H can perform various aspects of the techniques set forth in relation to the following clauses.

[1071] Clause 141541-1A. A device, such as exemplary audio encoding device 510H, comprising: one or more processors configured to obtain a bit stream comprising a compressed version of a spatial component of a sound field, wherein the spatial component is generated by performing A vector-based synthesis is performed with respect to a plurality of spherical harmonic coefficients.

[1072] Clause 141541-2A. The clause device 141541-1A, wherein the compressed version of the spatial component is represented in the bit stream using, at least in part, a field that specifies a prediction mode used during compression of the spatial component.

[1073] Clause 141541-3A. The device of any combination of clause 141541-1A and clause 141541-2A, where the compressed version of the spatial component is represented in the bitstream using, at least in part, Huffman table information that specifies a table of Huffman used during compression of the spatial component.

[1074] Clause 141541-4A. The device of any combination of clause 141541-1A through clause 141541-3A, where the compressed version of the spatial component is represented in the bit stream using, at least in part, a field that indicates a value that expresses a size of quantization step or a variable of the same used during the compression of the spatial component.

[1075] Clause 141541-5A. The clause device 141541-4A, wherein the value comprises an nbit value.

[1076] Clause 141541-6A. The device of any combination of clause 141541-4A and clause 141541-5A, wherein the bit stream comprises a compressed version of a plurality of special components of the sound field of which the compressed version of the spatial component is included, and wherein the value expresses the quantization step size or a variable thereof used during compression of the plurality of special components.

[1077] Clause 141541-7A. The device of any combination of clause 141541-1A through clause 141541-6A, where the compressed version of the spatial component is represented in the bitstream using, at least in part, a Huffman code to represent a category identifier which identifies a compression category to which the spatial component corresponds.

[1078] Clause 141541-8A. The device of any combination of clause 141541-1A through clause 141541-7A, wherein the compressed version of the spatial component is represented in the bit stream using, at least in part, a sign bit that identifies whether the spatial component is a positive value or a negative value.

[1079] Clause 141541-9A. The device of any combination of clause 141541-1A through clause 141541-8A, wherein the compressed version of the spatial component is presented in the bit stream using, at least in part, a Huffman code to represent a residual value of the spatial component.

[1080] Clause 141541-10A. The device of any combination of clause 141541-1A through clause 141541-9A, wherein the device comprises an exemplary audio encoding device and a bit stream generating device.

[1081] Clause 141541-12A. The device of any combination of clause 141541-1A through clause 141541-11A, wherein the vector-based synthesis comprises a single-valued decomposition.

[1082] While described as being performed by the 510H audio encoding device, the techniques can also be performed by any of the 24 and/or 540 audio decoding devices.

[1083] In this manner, the exemplary audio encoding device 510H can additionally perform various aspects of the techniques set forth in connection with the following clauses.

[1084] Clause 141541-1D. A device, such as exemplary audio encoding device 510H, comprising: one or more processors configured to generate a bit stream comprising a compressed version of a spatial component of a sound field, the spatial component generated by performing a vector-based synthesis with respect to a plurality of spherical harmonic coefficients.

[1085] Clause 141541-2D. The clause device 141541-1D, wherein the one or more processors are additionally configured to, during bitstream generation, generate the bitstream to include a field specifying a prediction mode used during spatial component compression .

[1086] Clause 141541-3D. The device of any combination of clause 141541-1D and clause 141541-2D, wherein the one or more processors are additionally configured to, during bitstream generation, generate the bitstream to include Huffman table information that specifies a Huffman table used during spatial component compression.

[1087] Clause 141541-4D. The device of any combination of clause 141541-1D through clause 141541-3D, wherein the one or more processors are additionally configured to, during bitstream generation, generate the bitstream to include a field indicating a value which expresses a quantization step size or a variable thereof used during the compression of the spatial component.

[1088] Clause 141541-5D. The clause device 141541-4D, wherein the value comprises an nbit value.

[1089] Clause 141541-6D. The device of any combination of clause 141541-4D and clause 141541-5D, wherein the one or more processors are further configured to, during bitstream generation, generate the bitstream to include a compressed version of a plurality of special components of the sound field from which the compressed version of the spatial component is included, and where the value expresses the quantization step size or a variable thereof used during compression of the plurality of special components.

[1090] Clause 141541-7D. The device of any combination of clause 141541-1D through clause 141541-6D, wherein the one or more processors are additionally configured to, during bitstream generation, generate the bitstream to include a Huffman code to represent a category identifier that identifies a compression category to which the spatial component corresponds.

[1091] Clause 141541-8D. The device of any combination of clause 141541-1D through clause 141541-7D, wherein the one or more processors are additionally configured to, during bit stream generation, generate the bit stream to include a sign bit that identifies whether the spatial component is a positive value or a negative value.

[1092] Clause 141541-9D. The device of any combination of clause 141541-1D through clause 141541-8D, wherein the one or more processors are additionally configured to, during bitstream generation, generate the bitstream to include a Huffman code to represent a residual value of the spatial component.

[1093] Clause 141541-10D. The device of any combination of clause 141541-1D through clause 141541-10D, wherein the vector-based synthesis comprises a single-valued decomposition.

[1094] The exemplary audio encoding device 510H may further be configured to implement various aspects of the techniques as set out in the following clauses.

[1095] Clause 141541-1E. A device, such as exemplary audio encoding device 510H, comprising: one or more processors configured to compress a spatial component of a sound field, wherein the spatial component is generated by performing vector-based synthesis with respect to to a plurality of spherical harmonic coefficients.

[1096] Clause 141541-2E. The clause device 141541-1E, wherein the one or more processors are additionally configured to, during compression of the spatial component, convert the spatial component from a floating point representation to an integer representation.

[1097] Clause 141541-3E. The device of any combination of clause 141541-1E and clause 141541-2E, wherein the one or more processors are additionally configured to, during spatial component compression, dynamically determine a value indicative of a quantization step size, and quantize the spatial component based on the value to generate a quantized spatial component.

[1098] Clause 141541-4E. The device of any combination of claims 1E to 3E, wherein the one or more processors are further configured to, during compression of the spatial component, identify a category to which the spatial component corresponds.

[1099] Clause 141541-5E. The device of any combination of clause 141541-1E through clause 141541-4E, wherein the one or more processors are further configured to, during compression of the spatial component, identify a residual value for the spatial component.

[1100] Clause 141541-6E. The device of any combination of clause 141541-1E through clause 141541-5E, wherein the one or more processors are additionally configured to, during spatial component compression, perform a prediction with respect to the spatial component and a subsequent spatial component to generate a predicted spatial component.

[1101] Clause 141541-7E. The device of any combination of clause 141541-1E, wherein the one or more processors are additionally configured to, during compression of the spatial component, convert the spatial component from a floating-point representation to an integer representation, dynamically determine a indicative value of a quantization step size, quantize the entire representation of the spatial component based on the value to generate a quantized spatial component, identify a category to which the spatial component corresponds based on the quantized spatial component to generate a category identifier, determine a signal from the spatial component, identify a residual value for the spatial component based on the quantized spatial component and the category identifier, and generate a compressed version of the spatial component based on the category identifier, signal, and residual value.

[1102] Clause 141541-8E. The device of any combination of clause 141541-1E, wherein the one or more processors are additionally configured to, during compression of the spatial component, convert the spatial component from a floating-point representation to an integer representation, dynamically determine a indicative value of a quantization step size, quantize the entire representation of the spatial component based on the value to generate a quantized spatial component, perform a prediction against the spatial component and a subsequent spatial component to generate a predicted spatial component, identify a category to which the predicted spatial component corresponds based on the quantized spatial component to generate a category identifier, determine a signal from the spatial component, identify a residual value for the spatial component based on the quantized spatial component and the category identifier, and generate a compressed version of the spatial component based on the category identifier, the sign and the residual value.

[1103] Clause 141541-9E. The device of any combination of clause 141541-1E through clause 141541-8E, wherein the vector-based synthesis comprises a single-valued decomposition.

[1104] Various aspects of the techniques may further allow the 510H audio encoding device to be configured to operate as set out in the following clauses.

[1105] Clause 141541-1F. A device, such as exemplary audio encoding device 510H, comprising: one or more processors configured to identify a Huffman codebook for use in compressing a current spatial component from a plurality of special components based on a order of the current spatial component with respect to the remainder of the plurality of special components, wherein the spatial component is generated by performing vector-based synthesis with respect to a plurality of spherical harmonic coefficients.

[1106] Clause 141541-2F. The clause 141541-3F device, wherein the one or more processors are additionally configured to perform any combination of the steps enumerated in clause 141541-1A through clause 141541-12A, clause 141541-1B through clause 141541-10B and clause 141541 -1C through clause 141541-9C.

[1107] Various aspects of the techniques may further allow the 510H audio encoding device to be configured to operate as set out in the following clauses.

[1108] Clause 141541-1H. A device, such as audio encoding device 510H, comprising: one or more processors configured to determine a quantization step size to be used when compressing a spatial component of a sound field, wherein the spatial component is generated by performing vector-based synthesis against a plurality of spherical harmonic coefficients.

[1109] Clause 141541-2H. The clause 141541-1H device, wherein the one or more processors are additionally configured to, during quantization step size determination, determine the quantization step size based on a target bit rate.

[1110] Clause 141541-3H. The device of clause 141541-1H, wherein the one or more processors are further configured to, during selection of one of a plurality of quantization step sizes, determine an estimate of a number of bits used to represent the spatial component, and determining the quantization step size based on a difference between the estimate and a target bitrate.

[1111] Clause 141541-4H. The device of clause 141541-1H, wherein the one or more processors are further configured to, during selection of one of a plurality of quantization step sizes, determine an estimate of a number of bits used to represent the spatial component, determine a difference between the estimate and a target bitrate and determine the quantization step size by adding the difference to the target bitrate.

[1112] Clause 141541-5H. The device of clause 141541-3H or clause 141541-4H, wherein the one or more processors are additionally configured to, during the determination of the number of bits estimate, calculate the estimate of the number of bits that should be generated for the spatial component given a codebook that corresponds to the target bitrate.

[1113] Clause 141541-6H. The device of clause 141541-3H or clause 141541-4H, wherein the one or more processors are additionally configured to, during the determination of the number of bits estimate, calculate the estimate of the number of bits that should be generated for the spatial component given an encoding mode used during compression of the spatial component.

[1114] Clause 141541-7H. The device of clause 141541-3H or clause 141541-4H, wherein the one or more processors are additionally configured to, during the determination of the number of bits estimate, calculate a first estimate of the number of bits that should be generated for the component given a first encoding mode to be used during compression of the spatial component, calculate a second estimate of the number of bits that should be generated for the spatial component given a second encoding mode to be used during compression of the spatial component, select the one of the first estimate and the second estimate that has the smallest number of bits to use as the given estimate of the number of bits.

[1115] Clause 141541-8H. The device of clause 141541-3H or clause 141541-4H, wherein the one or more processors are further configured to, during the determination of the number of bits estimate, identify a category identifier that identifies a category to which the spatial component corresponds , identify a bit length of a residual value for the spatial component that would result during compression of the spatial component that corresponds to the category, and determine the estimate of the number of bits by, at least in part, adding a number of bits used to represent the category identifier for the residual value bit length.

[1116] Clause 141541-9H. The device of any combination of clause 141541-1H through clause 141541-8H, wherein the vector-based synthesis comprises a single-valued decomposition.

[1117] While described as being performed by the audio encoding device 510H, the techniques set forth in the above clauses, clause 141541-1H through clause 141541-9H, may also be performed by the audio decoding device 540D.

[1118] Additionally, various aspects of the techniques may allow the 510H audio encoding device to be configured to operate as set forth in the following clauses.

[1119] Clause 141541-1J. A device, such as audio encoding device 510J, comprising: one or more processors configured to select one of a plurality of codebooks to be used during compression of a spatial component of a sound field, wherein the component space is generated by performing vector-based synthesis against a plurality of spherical harmonic coefficients.

[1120] Clause 141541-2J. The device of clause 141541-1J, wherein the one or more processors are additionally configured to, during the selection of one of the plurality of books and codes, determine an estimate of a number of bits used to represent the spatial component using from each of the plurality of codebooks, and selecting that of the plurality of codebooks that resulted in the determined estimate having the fewest bits.

[1121] Clause 141541-3J. The device of clause 141541-1J, wherein the one or more processors are further configured to, during the selection of one of the plurality of codebooks, determine an estimate of a number of bits used to represent the spatial component using one or more of the plurality of codebooks, wherein the one or more of the plurality of codebooks is selected based on an order of spatial component elements to be compressed relative to other spatial component elements.

[1122] Clause 141541-4J. The device of clause 141541-1J, wherein the one or more processors are further configured to, during the selection of one of the plurality of codebooks, determine an estimate of a number of bits used to represent the spatial component using one of a plurality of codebooks designed to be used when the spatial component is not predicted from a subsequent spatial component.

[1123] Clause 141541-5J. The device of clause 141541-1J, wherein the one or more processors are further configured to, during the selection of one of the plurality of codebooks, determine an estimate of a number of bits used to represent the spatial component using one of a plurality of codebooks designed to be used when the spatial component is predicted from a subsequent spatial component.

[1124] Clause 141541-6J. The device of clause 141541-1J, wherein the one or more processors are further configured to, during the selection of one of the plurality of codebooks, determine an estimate of a number of bits used to represent the spatial component using one of a plurality of codebooks designed for use when the spatial component is representative of a synthetic audio object in the sound field.

[1125] Clause 141541-7J. The clause 141541-1J device, wherein the synthetic audio object comprises a pulse code modulated (PCM) audio object.

[1126] Clause 141541-8J. The device of clause 141541-1J, wherein the one or more processors are additionally configured to, during the selection of one of the plurality of codebooks, determine an estimate of a number of bits used to represent the spatial component using of one of a plurality of codebooks designed to be used when the spatial component is representative of an audio object registered in the sound field.

[1127] Clause 141541-9J. The device of any combination of claims 1J to 8J, wherein the vector-based synthesis comprises a singular value decomposition.

[1128] In each of the various examples described above, it is to be understood that the audio encoding device 510 may perform a method or otherwise comprise means for each step of the method for which the audio encoding device 510 is configured to perform. In some examples, these means may comprise one or more processors. In some examples, the one or more processors may represent a special-purpose processor configured through instructions stored on non-transient computer-readable storage media. In other words, the various aspects of the techniques in each of the sets of encoding examples can provide a non-transient computer-readable storage medium that has stored on it instructions that, when executed, cause the one or more processors to perform the method for which the audio encoding device 510 was configured to perform.

[1129] Figure 40J is a block diagram illustrating an exemplary audio encoding device 510J that can perform various aspects of the techniques described in this disclosure to compress spherical harmonic coefficients that describe two or more three-dimensional sound fields. The exemplary audio encoding device 510J may be similar to the exemplary audio encoding device 510G in that the exemplary audio encoding device 510J includes an audio compression unit 512, an audio encoding unit 514 and a generation unit. 516. In addition, the audio compression unit 512 of the exemplary audio coding device 510J may be similar to that of the exemplary audio coding device 510G wherein the audio compression unit 512 includes a decomposition unit. 518 and a sound field component extraction unit 520, which may operate similarly to similar units of the exemplary audio encoding device 510I. In some examples, the exemplary audio encoding device 510J may include a quantization unit 534, as described with respect to Figures 40D through 40E, for quantizing one or more vectors of any of the UDIST vectors 525C, the UBG vectors 525D, the VTDIST 525E vectors and the VTBG 525J vectors.

[1130] The audio compression unit 512 of the audio encoding device 510J may, however, differ from the audio compression unit 512 of the audio encoding device 510G in that the audio compression unit 512 of the audio encoding device 512 audio 510J includes an additional unit denoted interpolation unit 550. Interpolation unit 550 may represent a unit that interpolates subframes of a first audio frame from subframes of the first audio frame and a temporally subsequent second or preceding audio frame , as described in more detail below with respect to Figures 45 and 45B. The interpolation unit 550 can, by performing this interpolation, reduce computational complexity (in terms of processing cycles and/or memory consumption) by potentially reducing the extent to which the decomposition unit 518 is required to decompose the SHC 511. Interpolation unit 550 may operate in a similar manner as described above with respect to interpolation unit 550 of audio coding devices 510H and 510I shown in the examples of Figures 40H and 40I.

[1131] In operation, the interpolation unit 200 may interpolate one or more subframes of a first audio frame from a first decomposition, e.g., the V matrix 19', of a portion of a first plurality of spherical harmonic coefficients 11 included in the first frame and a second decomposition, e.g. matrix V 19', of a portion of a second plurality of spherical harmonic coefficients 11 included in a second frame to generate interpolated spherical harmonic coefficients decomposed for the one or more subframes.

[1132] The interpolation unit 550 can obtain interpolated spherical harmonic coefficients decomposed by a time segment by, at least in part, performing an interpolation with respect to a first decomposition of a first plurality of spherical harmonic coefficients and a decomposition of a second plurality of spherical harmonic coefficients. Smoothing unit 554 may apply the decomposed interpolated spherical harmonic coefficients to smooth at least one of special components and time components of the first plurality of spherical harmonic coefficients and the second plurality of spherical harmonic coefficients. Smoothing unit 554 can generate smoothed UDIST matrices 525C' as described above with respect to Figures 37 to 39. The first and second decompositions can refer to V1T 556, V2T 556B in Figure 40J.

[1133] In some cases, VT or other V vectors or V matrices may be output in a quantized version for interpolation. In this way, the V vectors for the interpolation can be identical to the V vectors in the decoder, which also performs the V vector interpolation, for example, to recover the multidimensional signal.

[1134] In some examples, the first decomposition comprises the first matrix V 519' representative of singular vectors to the right of the portion of the first plurality of spherical harmonic coefficients 511. Similarly, in some examples, the second decomposition comprises the second matrix V 519' representative of singular vectors to the right of the portion of the second plurality of spherical harmonic coefficients.

[1135] Interpolation unit 550 may perform temporal interpolation with respect to the one or more subframes based on the first V 519' and the second V matrix 19'. That is, the interpolation unit 550 can temporally interpolate, for example, the second, third and fourth subframes of a total of four subframes for the first audio frame based on a V matrix 519' decomposed from the first subframe of the first audio frame and in the V matrix 519' decomposed from the first subframe of the second audio frame. In some examples, this temporal interpolation is linear temporal interpolation, where the matrix V 519' decomposed from the first subframe of the first audio frame is weighted more heavily during interpolation of the second subframe of the first audio frame than during interpolation of the fourth subframe of the first audio frame. During the interpolation of the third subframe, the V arrays 519' may be weighted equally. During the interpolation of the fourth subframe, the V matrix 519' decomposed from the first subframe of the second audio frame can be more heavily weighted than the V matrix 519' decomposed from the first subframe of the first audio frame.

[1136] In other words, linear temporal interpolation can weight V matrices 519' given the proximity of one of the subframes of the first audio frame to be interpolated. For the second subframe to be interpolated, the V matrix 519' decomposed from the first subframe of the first audio frame is weighted more heavily given its proximity to the second subframe to be interpolated than the V matrix 519' decomposed from the first subframe of the second audio frame. The weights may be equivalent for that reason during the interpolation of the third subframe based on the V matrices 519'. The weight applied to the V matrix 519' decomposed from the first subframe of the second audio frame can be greater than that applied to the V matrix 519' decomposed from the first subframe of the first audio frame as the fourth subframe is to be interpolated is closer to the first subframe of the second audio frame than the first subframe of the first audio frame.

[1137] In some examples, the interpolation unit 550 can project the first V matrix 519' decomposed from the first subframe of the first audio frame into a spatial domain to generate the first projected decompositions. In some examples, this projection includes a projection onto a sphere (for example, using a projection matrix, such as a T-design matrix). The interpolation unit 550 can then project the second decomposed V matrix 519' from the first subframe of the second audio frame into the spatial domain to generate the second projected decompositions. The interpolation unit 550 can then spatially interpolate (which again can be a linear interpolation) the first projected decompositions and the second projected decompositions to generate a first spatially interpolated projected decomposition and a second spatially interpolated projected decomposition. The interpolation unit 550 can then temporally interpolate the one or more subframes based on the first spatially interpolated projected decomposition and the second spatially interpolated projected decomposition.

[1138] In this example where the interpolation unit 550 spatially and then temporally projects the V arrays 519', the interpolation unit 550 can project the temporally interpolated spherical harmonic coefficients that result from interpolation of the one or more subframes back to a spherical harmonic domain, thereby generating the V matrix 519, the S matrix 519B and the U matrix 519C.

[1139] In some examples, the portion of the first plurality of spherical harmonic coefficients comprises a single subframe of the first plurality of spherical harmonic coefficients 511. In some examples, the portion of the second plurality of spherical harmonic coefficients comprises a single subframe of the second plurality of spherical harmonic coefficients. spherical harmonic coefficients 511. In some examples, that single subframe from which the V matrices 19' are decomposed is the first subframe.

[1140] In some examples, the first frame is divided into four subframes. In these and other examples, the portion of the first plurality of spherical harmonic coefficients comprises only the first subframe of the plurality of spherical harmonic coefficients 511. In these and other examples, the second frame is divided into four subframes, and the portion of the second plurality of spherical harmonic coefficients spherical harmonics 511 comprises only the first subframe of the second plurality of spherical harmonic coefficients 511.

[1141] Although, in some examples, only a first subframe of each audio frame is used to perform interpolation, the portion of the first plurality of spherical harmonic coefficients may comprise two of four subframes of the first plurality of spherical harmonic coefficients 511. In such and other examples, the portion of the second plurality of spherical harmonic coefficients 511 comprises two of four subframes of the second plurality of spherical harmonic coefficients 511.

[1142] As noted above, a single device, for example, audio encoding device 510J, can perform the interpolation while also decomposing the portion of the first plurality of spherical harmonic coefficients to generate the first decompositions of the portion of the first plurality of spherical harmonic coefficients spherical. In these and other examples, the decomposition unit 518 may decompose the portion of the second plurality of spherical harmonic coefficients to generate second decompositions of the portion of the second plurality of spherical harmonic coefficients. While described with respect to a single device, two or more devices can perform the techniques described in this disclosure, wherein one of the two devices performs decomposition and the other of the devices performs interpolation in accordance with the techniques described in this disclosure.

[1143] In some examples, the decomposition unit 518 may perform a singular value decomposition with respect to the portion of the first plurality of spherical harmonic coefficients 511 to generate a V matrix 519' (as well as an S matrix 519B' and a U matrix 519C', which are not shown for ease of illustration purposes) representative of singular vectors to the right of the first plurality of spherical harmonic coefficients 511. In these and other examples, the decomposition unit 518 can perform singular value decomposition with respect to the portion of the second plurality of spherical harmonic coefficients 511 to generate a V matrix 519' (as well as an S matrix 519B' and a U matrix 519C', which are not shown for ease of illustration purposes) representative of right-hand singular vectors of the second plurality of spherical harmonic coefficients.

[1144] In some examples, as noted above, each of the first and second pluralities of spherical harmonic coefficients represents a plane wave representation of the sound field. In these and other examples, each of the first and second pluralities of spherical harmonic coefficients 511 represents one or more mono audio objects mixed together.

[1145] In other words, 3D audio based on spherical harmonics can be a parametric representation of the 3D pressure field in terms of orthogonal basis functions on a sphere. The higher the N order of the representation, the potentially higher the spatial resolution, and often the greater the number of spherical harmonic (SH) coefficients (for a total of (N+1)2 coefficients). For many applications, bandwidth compression of the coefficients may be required to be able to efficiently transmit and store the coefficients. These techniques addressed in this disclosure can provide a frame-based dimensionality reduction process using Singular Value Decomposition (SVD). SVD analysis can decompose each frame of coefficients into three matrices U, S and V. In some examples, the techniques can deal with some of the U-vectors as directional components of the underlying sound field. However, when treated this way, these (U-shaped) vectors are discontinuous from frame to frame - even though they represent the same distinct audio component. These discontinuities can lead to significant artifacts when components are fed through transform audio encoders.

[1146] The techniques described in this disclosure can address this discontinuity. That is, the techniques can be based on the observation that the matrix V can be interpreted as orthogonal spatial axes in the domain of Spherical Harmonics. The matrix U can represent a projection of the Spherical Harmonics (HOA) data in terms of these basis functions, where the discontinuity can be attributed to the basis functions (V) that change every frame - and they are therefore discontinuous. This is different from similar decomposition, such as the Fourier transform, where the basis functions are, in some instances, constants from frame to frame. In these terms, SVD can be considered as a compatibility search algorithm. The techniques described in this disclosure may allow the interpolation unit 550 to maintain continuity between the base (V) functions from frame to frame - by interpolating between them.

[1147] In some examples, the techniques allow the interpolation unit 550 to divide the SH data frame into four subframes as described above and described further below with respect to Figures 45 and 45B. The interpolation unit 550 can then compute the SVD for the first subframe. Similarly, we compute the SVD for the first subframe of the second frame. For each of the first and second frames, the interpolation unit 550 can convert the V vectors to a spatial map by projecting the vectors onto a sphere (using a projection matrix such as a project matrix in T). The interpolation unit 550 can then interpret the vectors in V as shapes on a sphere. To interpolate the V matrices for the three subframes between the first subframe of the first frame and the first subframe of the next frame, the interpolation unit 550 can then interpolate these spatial formats - and then transform them back to vectors SH through the inverse of the projection matrix. The techniques of this disclosure can thus provide a smooth transition between V matrices.

[1148] Figures 41 and 41D are block diagrams, each of which illustrates an exemplary audio decoding device 540A-540D that can perform various aspects of the techniques described in this disclosure for decoding spherical harmonic coefficients that describe sound fields of two or more three dimensions. The 540A audio decoding device can represent any device that has the ability to decode audio data, such as a desktop computer, a laptop computer, a workstation, a tablet or slate type computer, a audio recording device, a cell phone (which includes so-called “smart phones”), a personal media player device, a personal gaming device, or any other type of device that has the ability to decode audio data.

[1149] In some examples, the audio decoding device 540A performs an audio decoding process that is reciprocal to the audio encoding process performed by any of the audio encoding devices 510 or 510B with the exception of performing reduction of order (as described above with respect to the examples of Figures 40B to 40J), which is, in some examples, used by audio encoding devices 510B to 510J to facilitate the removal of extraneous extraneous data.

[1150] Although shown as a single device, i.e., device 540A in the example of Figure 41, the various components or units referenced below as being included within device 540A may form separate devices that are external to device 540. In other words , although described in this disclosure as being performed by a single device, i.e., the device 540A in the example of Figure 41, the techniques may be implanted or otherwise performed by a system comprising multiple devices, where each such device may each include one or more of the various components or units described in more detail below. Consequently, the techniques should not be limited in this regard to the example in Figure 41.

[1151] As shown in the example of Figure 41, the audio decoding device 540A comprises an extraction unit 542, an audio decoding unit 544, a math unit 546 and an audio rendering unit 548. The extraction unit 542 represents a unit configured to extract 515B coded reduced background spherical harmonic coefficients, 515A coded UDIST * SDIST vectors, and 525E VTDIST vectors from bit stream 517. Extractor unit 542 outputs spherical second harmonic coefficients. 515B encoded reduced planes and 515A encoded UDIST * SDIST vectors to the audio decoding unit 544, while also outputting the VTDIST matrix 525E to the math unit 546. In this regard, the extraction unit 542 may operate in a similar manner to the unit extraction 72 of the audio decoding device 24 shown in the example of Figure 5.

[1152] Audio decoding unit 544 represents a unit for decoding the encoded audio data (often according to a reciprocal audio decoding scheme, such as an AAC decoding scheme) in order to recover the UDIST* vectors SDIST 527 and reduced background spherical harmonic coefficients 529. The audio decoding unit 544 outputs the UDIST * SDIST vectors 527 and reduced background spherical harmonic coefficients 529 to the math unit 546. In this regard, the decoding unit 544 can operate in a similar manner to the psychoacoustic decoding unit 80 of the audio decoding device 24 shown in the example of Figure 5.

[1153] Math unit 546 can represent a unit configured to perform matrix addition and multiplication (as well as, in some examples, any other matrix math operation). The math unit 546 may first perform a matrix multiplication of the UDIST * SDIST vectors 527 by the VTDIST matrix 525E. The math unit 546 can then add the result of multiplying the UDIST * SDIST vectors 527 by the VTDIST matrix 525E by the reduced background spherical harmonic coefficients 529 (which, again, can refer to the result of multiplying the UBG matrix 525D by the matrix SBG 525B and then matrix VTBG 525F) to the result of matrix multiplication of vectors UDIST * SDIST 527 by matrix VTDIST 525E to generate the reduced version of the original spherical harmonic coefficients 11, which are denoted as recovered spherical harmonic coefficients 547 The math unit 546 may output the recovered spherical harmonic coefficients 547 to the audio rendering unit 548. In this regard, the math unit 546 may operate in a similar manner to the foreground formulation unit 78 and the coefficient formulation unit 546. of HOA 82 of the audio decoding device 24 shown in the example of Figure 5.

[1154] The 548 audio rendering unit represents a unit configured to render the 549A-549N channels (the “549 channels”, which may also be generically referred to as the “549 multi-channel audio data” or “high -speakers 549”). The audio rendering unit 548 can apply a transform (often expressed as a matrix) to the recovered spherical harmonic coefficients 547. Since the recovered spherical harmonic coefficients 547 describe the sound field in three dimensions, the recovered spherical harmonic coefficients 547 represent an audio format that facilitates the rendering of 549A multichannel audio data in a way that has the ability to accommodate most local decoder speaker geometries (which may refer to the geometry of the speakers that will play data multichannel audio stream 549). More information on rendering the 549A multichannel audio data is described above with respect to Figure 48.

[1155] Although described in the context of the multichannel audio data 549A being multichannel surround sound audio data 549, the audio rendering unit 48 can also perform a form of binauralization to binauralize the recovered spherical harmonic coefficients 549A and thereby , generate two binaurally rendered channels 549. Consequently, the techniques should not be limited to surround sound forms of multichannel audio data, but may include binauralized multichannel audio data.

[1156] The various clauses listed below may present various aspects of the techniques described in this disclosure.

[1157] Clause 132567-1B. A device, such as audio decoding device 540, comprising: one or more processors configured to determine one or more first vectors that describe distinct components of the sound field and one or more second vectors that describe background components of the sound field , both the one or more first vectors and the one or more second vectors generated at least by performing a singular value decomposition with respect to the plurality of spherical harmonic coefficients.

[1158] Clause 132567-2B. The device of clause 132567-1B, wherein the first one or more vectors comprise one or more audio-encoded UDIST * SDIST vectors that, prior to audio encoding, were generated by multiplying one or more audio-encoded UDIST vectors from a matrix U by one or more SDIST vectors of a matrix S, wherein the matrix U and the matrix S are generated at least by performing singular-value decomposition with respect to the plurality of spherical harmonic coefficients, and wherein the one or more processors are additionally configured to audio decode the one or more audio encoded UDIST * SDIST arrays to generate an audio decoded version of the one or more audio encoded UDIST * SDIST arrays.

[1159] Clause 132567-3B. The device of clause 132567-1B, wherein the first one or more vectors comprise one or more audio-encoded UDIST * SDIST vectors that, prior to audio encoding, were generated by multiplying one or more audio-encoded UDIST vectors from a matrix U by one or more SDIST vectors from a matrix S, and one or more vectors VTDIST from a transpose of a matrix V, where the matrix U, matrix S, and matrix V are generated at least by performing the decomposition of singular value with respect to the plurality of spherical harmonic coefficients, and wherein the one or more processors are additionally configured to audio decode the one or more audio encoded UDIST * SDIST vectors to generate an audio decoded version of the one or more vectors UDIST*SDIST audio encoded.

[1160] Clause 132567-4B. The clause device 132567-3B, wherein the one or more processors are further configured to multiply the UDIST * SDIST vectors by the VTDIST vectors to recover those of the plurality of spherical harmonics representing the distinct components of the sound field.

[1161] Clause 132567-5B. The device of clause 132567-1B, wherein the one or more second vectors comprise one or more audio-encoded UBG * SBG * VTBG vectors that, prior to audio encoding, were generated by multiplying the UBG vectors included within a matrix U by vectors SBG enclosed within a matrix S and then by vectors VTBG enclosed within a transpose of a matrix V, and wherein the matrix S, the matrix U and the matrix V were each generated by at least performing the singular value decomposition with respect to the plurality of spherical harmonic coefficients.

[1162] Clause 132567-6B. The device of clause 132567-1B, wherein the one or more second vectors comprise one or more audio-encoded UBG * SBG * VTBG vectors that, prior to audio encoding, were generated by multiplying UBG vectors included within a matrix U by SBG vectors included within a matrix S and then by vectors VTBG included within a transpose of a matrix V, and where the matrix S, the matrix U and the matrix V were generated at least by performing the decomposition of singular value with respect to the plurality of spherical harmonic coefficients, and wherein the one or more processors are additionally configured to audio decode the one or more audio-encoded UBG * SBG * VTBG vectors to generate one or more UBG * SBG * vectors VTBG decoded by audio.

[1163] Clause 132567-7B. The device of clause 132567-1B, wherein the first one or more vectors comprise one or more audio-encoded UDIST * SDIST vectors that, prior to audio-encoding, were generated by multiplying one or more audio-encoded UDIST vectors from a matrix U by one or more SDIST vectors from a matrix S, and one or more vectors VTDIST from a transpose of a matrix V, where the matrix U, matrix S, and matrix V were generated at least by performing the decomposition of singular value with respect to the plurality of spherical harmonic coefficients, and wherein the one or more processors are additionally configured to audio decode the one or more audio-encoded UDIST * SDIST vectors to generate the one or more UDIST * SDIST vectors, and multiply the UDIST * SDIST vectors by the VTDIST vectors to recover those of the plurality of spherical harmonic coefficients that describe the distinct components of the sound field, where the plus one is The second vectors comprise one or more audio-encoded UBG * SBG * VTBG vectors that, prior to audio coding, were generated by multiplying UBG vectors included within matrix U by SBG vectors included within matrix S and then by vectors VTBG included within the transpose matrix V, and wherein the one or more processors are additionally configured to audio decode the one or more audio encoded UBG * SBG * VTBG vectors to recover at least a portion of the plurality of spherical harmonic coefficients that describe background components of the sound field, and adds the plurality of spherical harmonic coefficients that describe the distinct components of the sound field to at least a portion of the plurality of spherical harmonic coefficients that describe background components of the sound field to generate a reconstructed version of the sound field. plurality of spherical harmonic coefficients.

[1164] Clause 132567-8B. The device of clause 132567-1B, wherein the first one or more vectors comprise one or more UDIST*SDIST vectors which, prior to audio encoding, were generated by multiplying one or more audio-encoded UDIST vectors from a matrix U by one or more SDIST vectors from a matrix S, and one or more VTDIST vectors from a transpose of a matrix V, where the matrix U, the matrix S, and the matrix V were generated at least by performing the single-valued decomposition into with respect to the plurality of spherical harmonic coefficients, and wherein the one or more processors are additionally configured to determine a D value that indicates the number of vectors to be extracted from a bit stream to form the one or more UDIST*SDIST vectors and the one or more VTDIST vectors.

[1165] Clause 132567-9B. The device of clause 132567-10B, wherein the first one or more vectors comprise one or more UDIST * SDIST vectors that, prior to audio encoding, were generated by multiplying one or more audio-encoded UDIST vectors of a matrix U by one or more SDIST vectors from a matrix S, and one or more VTDIST vectors from a transpose of a matrix V, where the matrix U, the matrix S, and the matrix V were generated at least by performing the single-valued decomposition into with respect to the plurality of spherical harmonic coefficients, and wherein the one or more processors are additionally configured to determine a D value in an audio frame, on an audio frame basis, which indicates the number of vectors to be extracted from a stream of bits to form the one or more UDIST * SDIST vectors and the one or more VTDIST vectors.

[1166] Clause 132567-1G. A device, such as the audio decoding device 540, comprising: one or more processors configured to determine one or more first vectors that describe distinct components of a sound field and one or more second vectors that describe background components of the sound field , both the one or more first vectors and the one or more second vectors generated by at least performing a single-value decomposition against multichannel audio data representing at least a portion of the sound field.

[1167] Clause 132567-2G. The device of clause 132567-1G, wherein the multichannel audio data comprises a plurality of spherical harmonic coefficients.

[1168] Clause 132567-3G. The clause 132567-2G device, wherein the one or more processors are further configured to perform any combination of clause 132567-2B to clause 132567-9B.

[1169] From each of the various clauses described above, it is to be understood that any one of the audio decoding devices 540A to 540D may perform a method or may otherwise comprise means for performing each step of the method that the 540A to 540D audio decoding devices are configured to perform. In some examples, these means may comprise one or more processors. In some examples, the one or more processors may represent a special-purpose processor configured through instructions stored on non-transient computer-readable storage media. In other words, various aspects of the techniques in each of these sets of encoding examples can provide non-transient computer-readable storage media that have instructions stored on them that, when executed, cause the one or more processors to perform the method. that the 540A to 540D audio decoding devices have been configured to perform.

[1170] For example, a clause 132567-10B can be derived from the aforementioned clause 132567-1B to be a method comprising: determining one or more first vectors that describe distinct components of a sound field and one or more second vectors that describe components background of the sound field, both the one or more first vectors and the one or more second vectors generated by at least performing a single-value decomposition with respect to a plurality of spherical harmonic coefficients representing the sound field.

[1171] As another example, a clause 132567-1 IB can be derived from the aforementioned clause 132567-1B to be a device, such as the audio decoding device 540, which comprises means for determining one or more first vectors that describe distinct components of the sound field and one or more second vectors describing background components of the sound field, wherein both the one or more first vectors and the one or more second vectors are generated by at least performing a single-valued decomposition with respect to the plurality of spherical harmonic coefficients; and means for storing the one or more first vectors and the one or more second vectors.

[1172] As yet another example, a clause 132567-12B can be derived from the aforementioned clause 132567-1B to be a non-transient computer-readable storage medium that has instructions stored on it that, when executed, cause a or more processors determine one or more first vectors that describe distinct components of a sound field and one or more second vectors that describe background components of the sound field, where both the one or more first vectors and the one or more second vectors are generated at least by performing a single-value decomposition with respect to a plurality of spherical harmonic coefficients included in higher-order ambisonic audio data describing the sound field.

[1173] Various clauses may be derived similarly from clauses 132567-2B to 132567-9B for the various non-transient computer readable storage devices, methods and media derived as exemplified above. The same may be done for the various other clauses listed throughout the present disclosure.

[1174] Fig. 41B is a block diagram illustrating an exemplary audio decoding device 540B that can perform various aspects of the techniques described in the present disclosure to decode spherical harmonic coefficients that describe two-dimensional or three-dimensional sound fields. The audio decoding device 540B may be similar to the audio decoding device 540, except that in some instances, the extractor unit 542 may extract reordered VTDIST vectors 539 instead of VTDIST vectors 525E. In other examples, the extraction unit 542 can extract the VTDIST 525E vectors and then reorder these VTDIST 525E vectors based on reordering information specified in the bitstream or inferred (through analysis of other vectors) to determine the vectors Reordered VTDISTs 539. In this respect, the extraction unit 542 can operate similarly to the extraction unit 72 of the audio decoding device 24 shown in the example of Figure 5. In any case, the extraction unit 542 can output the reordered VTDIST vectors 539 to the mathematics unit 546, wherein the process described above in relation to the recovery of spherical harmonic coefficients can be performed in relation to these reordered VTDIST vectors 539.

[1175] In this way, the techniques may enable the 540B audio decoding device to decode audio from one or more reordered vectors representing distinct components of a sound field, the one or more reordered vectors being reordered to facilitate compression of one or more more vectors. In these or other examples, the audio decoding device 540B may recombine the one or more reorder vectors with one or more additional reorder vectors to recover spherical harmonic coefficients representative of distinct components of the sound field. In these and other examples, the audio decoding device 540B can then retrieve a plurality of spherical harmonic coefficients based on spherical harmonic coefficients representative of distinct components of the sound field and spherical harmonic coefficients representative of background components of the field. sound.

[1176] That is, various aspects of the techniques may provide for configuring the 540B audio decoding device to decode one or more reordered vectors in accordance with the following clauses.

[1177] Clause 133146-1F. A device, such as audio encoding device 540B, comprising: one or more processors configured to determine a number of vectors corresponding to components in the sound field.

[1178] Clause 133146-2F. The device of clause 133146-1F, wherein the one or more processors are configured to determine the number of vectors after performing order reduction in accordance with any combination of the examples described above.

[1179] Clause 133146-3F. The device of clause 133146-1F, wherein the one or more processors are further configured to perform order reduction in accordance with any combination of the examples described above.

[1180] Clause 133146-4F. The device of clause 133146-1F, where the one or more processors are configured to determine the number of vectors from a specified value in a bit stream, and where the one or more processors are additionally configured to parse the stream of bits based on the given number of vectors to identify one or more vectors in the bit stream that represent distinct components of the sound field.

[1181] Clause 133146-5F. The device of clause 133146-1F, where the one or more processors are configured to determine the number of vectors from a specified value in a bit stream, and where the one or more processors are additionally configured to parse the stream of bits based on the given number of vectors to identify one or more in the stream of bits that represent background components of the sound field.

[1182] Clause 133143-1C. A device, such as audio decoding device 540B, comprising: one or more processors configured to reorder one or more reordered vectors representative of distinct components of a sound field.

[1183] Clause 133143-2C. The provision of clause 133143-1C, wherein the one or more processors are further configured to determine the one or more reordered vectors and to determine reorder information that describes how the one or more reordered vectors were reordered, wherein the one or more processors are further configured to reorder, during reordering of the one or more reordered vectors, the one or more reordered vectors based on the determined reordering information.

[1184] Clause 133143-3C. The device of 1C, wherein the one or more reordered vectors comprises the one or more first reordered vectors recited by any combination of claims 1A to 18A or any combination of claims 1B to 19B, and wherein the first one or more vectors are determined according to the method recited by any combination of claims 1A to 18A or any combination of claims 1B to 19B.

[1185] Clause 133143-4D. A device, such as the audio decoding device 540B, comprising: one or more processors configured to decode an audio one or more reordered vectors representing distinct components of a sound field, the one or more reordered vectors being reordered to facilitate the compression of one or more vectors.

[1186] Clause 133143-5D. The device of clause 133143-4D, wherein the one or more processors are further configured to recombine the one or more reorder vectors with one or more additional reorder vectors to recover spherical harmonic coefficients representative of distinct components of the sound field.

[1187] Clause 133143-6D. The device of clause 133143-5D, wherein the one or more processors are further configured to retrieve a plurality of spherical harmonic coefficients based on spherical harmonic coefficients representative of distinct components of the sound field and spherical harmonic coefficients representative of background components of the sound field.

[1188] Clause 133143-1E. A device, such as audio decoding device 540B, comprising: one or more processors configured to reorder one or more vectors to generate one or more first reorder vectors and thereby facilitate encoding via a legacy audio encoder, wherein the one or more vectors describe represent distinct components of a sound field, and audio encode the one or more reordered vectors using the legacy audio encoder to generate an encoded version of the one or more reordered vectors.

[1189] Clause 133143-2E. The device of 1E, wherein the one or more reordered vectors comprises the one or more first reordered vectors recited by any combination of claims 1A to 18A or any combination of claims 1B to 19B, and wherein the first one or more vectors are determined according to the method recited by any combination of claims 1A-18A or by any combination of claims 1B-19B.

[1190] Fig. 41C is a block diagram illustrating another exemplary audio encoding device 540C. The 540C audio decoding device can represent any device capable of decoding audio data, for example, a desktop computer, a laptop-type computer, a workstation, a tablet or slate-type computer, a recording device audio device, a cell phone (including so-called "smart phones"), a personal media playback device, a personal gaming device, or any other type of device capable of decoding audio data.

[1191] In the example of Fig. 41C, the audio decoding device 540C performs an audio decoding process that is reciprocal to the audio encoding process performed by any of the audio encoding devices 510B to 510E with the exception of performing of order reduction (those described above with respect to the examples of Figures 40B to 40J), which, in some examples, is used by the audio encoding device 510B to 510J to facilitate the removal of extraneous extraneous data.

[1192] Although shown as a single device, i.e., device 540C in the example of Figure 41C, the various components or units indicated below as included in device 540C may form separate devices that are external to device 540C. In other words, although described in the present disclosure as being performed by a single device, i.e., the device 540C in the example of Figure 41C, the techniques may be implemented or otherwise performed by a system comprising multiple devices, wherein each such device may each include one or more of the various components or units described in more detail below. Consequently, in this respect, the techniques should not be limited to the example in Figure 41 C.

[1193] Furthermore, the 540C audio encoding device may be similar to the 540B audio encoding device. However, the extraction unit 542 may determine the one or more VTSMALL arrays 521 from the bit stream 517 instead of the reordered VTQ_DIST arrays 539 or VTDIST arrays 525E (as in the case described with respect to the audio encoding device 510 of the Figure 40). As a result, the extraction unit 542 can pass the VTSMALL vectors 521 to the math unit 546.

[1194] In addition, extraction unit 542 can determine audio-encoded modification spherical harmonic coefficients 515B' from bitstream 517, passing these coefficients 515B' to audio decoding unit 544, which can decode audio the coded modified background spherical harmonic coefficients 515B to retrieve the modified background spherical harmonic coefficients 537. The audio decoding unit 544 can pass these modified background spherical harmonic coefficients 537 to the math unit 546.

[1195] The math unit 546 can then multiply the audio-decoded (and possibly unordered) UDIST * SDIST vectors 527' by the one or more VTSMALL vectors 521 to retrieve the higher order distinct spherical harmonic coefficients. Then, the math unit 546 may sum the distinct higher-order spherical harmonic coefficients to the modified background spherical harmonic coefficients 537 to retrieve the plurality of the spherical harmonic coefficients 511 or some derivative thereof (which may be a derivative due to the order performed in the encoder unit 510E).

[1196] In this way, the techniques can enable the 540C audio decoding device to determine, from a bit stream, at least one of one or more vectors decomposed from spherical harmonic coefficients that have been recombined with harmonic coefficients background spherical harmonics to reduce the amount of bits that are required to be allocated to the one or more vectors in the bit stream, where spherical harmonic coefficients describe a sound field, and where spherical background harmonic coefficients describe one or more background components of the same sound field.

[1197] In this regard, various aspects of the techniques may enable the 540C audio decoding device, in some examples, to be configured to determine, from a stream of bits, at least one of one or more vectors decomposed from of spherical harmonic coefficients that have been recombined with spherical background harmonic coefficients, where the spherical harmonic coefficients describe a sound field, and where the spherical background harmonic coefficients describe one or more background components of the same sound field.

[1198] In these and other examples, the audio decoding device 540C is configured to obtain from the bit stream a first portion of spherical harmonic coefficients that have an order equal to NBG. In these and other examples, the audio decoding device 540C is further configured to obtain, from the bit stream, an audio-encoded first portion of spherical harmonic coefficients having an order equal to NBG, and to audio-decode the first audio encoded portion of the spherical harmonic coefficients to generate a first portion of the spherical harmonic coefficients.

[1199] In these and other examples, the at least one or more vectors comprise one or more VTSMALL vectors, with the one or more VTSMALL vectors being determined from a transpose of a matrix V generated by performing a value decomposition unique with respect to the plurality of spherical harmonic coefficients.

[1200] In these and other examples, at least one of the one or more vectors comprises one or more VTSMALL vectors, the one or more VTSMALL vectors being determined from a transpose of a matrix V generated by performing a single-value decomposition with respect to the plurality of spherical harmonic coefficients, and the audio decoding device 540C is further configured to obtain, from the bit stream, one or more UDIST*SDIST vectors that have been derived from a matrix U and from a matrix S, in which both of them were generated by performing the single-value decomposition with respect to the plurality of spherical harmonic coefficients, and to multiply the UDIST * SDIST vectors by the VTSMALL vectors.

[1201] In these and other examples, the at least one of the one or more vectors comprises one or more VTSMALL vectors, the one or more VTSMALL vectors being determined from a transpose of a matrix V generated by performing a single-value decomposition with respect to the plurality of spherical harmonic coefficients, and the audio decoding device 540C is further configured to obtain from the bit stream one or more UDIST*SDIST vectors that were derived from a matrix U and a matrix S, both of which are generated by performing the single-value decomposition with respect to the plurality of spherical harmonic coefficients, and to multiply the UDIST * SDIST vectors by the VTSMALL vectors in order to recover different spherical harmonic coefficients of higher order background and to sum the background spherical harmonic coefficients that include the distinct background spherical harmonic coefficients lower order values to distinct higher order spherical harmonic coefficients of the background in order to recover, at least partially, the plurality of spherical harmonic coefficients.

[1202] In these and other examples, the at least one of the one or more vectors comprises one or more VTSMALL vectors, the one or more VTSMALL vectors being determined from a transpose of a matrix V generated by performing a single-value decomposition with respect to the plurality of spherical harmonic coefficients, and the audio decoding device 540C is further configured to obtain from the bit stream one or more UDIST*SDIST vectors that were derived from a matrix U and a matrix S, both of which are generated by performing the single-value decomposition with respect to the plurality of spherical harmonic coefficients, and to multiply the UDIST * SDIST vectors by the VTSMALL vectors in order to recover different spherical harmonic coefficients of higher order background, to sum the background spherical harmonic coefficients that include the distinct background spherical harmonic coefficients those of lower order to distinct higher order spherical harmonic coefficients of the background in order to at least partially recover the plurality of spherical harmonic coefficients and render the recovered plurality of spherical harmonic coefficients

[1203] Fig. 41D is a block diagram illustrating another exemplary audio encoding device 540D. The 540D audio decoding device can represent any device capable of decoding audio data, e.g. a desktop computer, a laptop-type computer, a workstation, a tablet or slate-type computer, a recording device audio device, a cell phone (including so-called "smart phones"), a personal media playback device, a personal gaming device, or any other type of device capable of decoding audio data.

[1204] In the example of Fig. 41D, the audio decoding device 540D performs an audio decoding process that is reciprocal to the audio encoding process performed by any of the audio encoding devices 510B to 510J with the exception of performing of order reduction (those described above with respect to the examples of Figures 40B to 40J), which, in some examples, is used by audio encoding devices 510B to 510J to facilitate the removal of extraneous extraneous data.

[1205] Although shown as a single device, i.e., device 540D in the example of Figure 41D, the various components or units indicated below as included in device 540D may form separate devices that are external to device 54DC. In other words, although described in the present disclosure as being performed by a single device, i.e., the device 540D in the example of Figure 41D, the techniques may be implanted or otherwise performed by a system comprising multiple devices, in that each such device may each include one or more of the various components or units described in more detail below. Consequently, in this respect, the techniques should not be limited to the example of Figure 41D.

[1206] Furthermore, the audio decoding device 540D can be similar to the audio decoding device 540B, except that the audio decoding device 540D performs additional V-decompression that is, in general, reciprocal to the compression performed by the V compression unit 552 described above with respect to Figure 401. In the example of Figure 41D, extraction unit 542 includes a V decompression unit 555 that performs such V decompression of compressed spatial components 539' included in bit stream 517 (and , specified generally in accordance with the example shown in one of Figures 10B and 10c). The V 555 decompression unit can decompress the VTDIST 539 vectors based on the following equation:

[1207] In other words, the V decompression unit 555 can first parse the nbits value from the bit stream 517 and identify the appropriate set of five Huffman code tables to use when decoding the Huffman code. representative of the cid. Based on the prediction mode and Huffman encoding information specified in bitstream 517 and possibly the order of the spatial component element relative to the other spatial component elements, the V decompression unit 555 can identify the correct table among the five Huffman tables defined for the value of analyzed nbits. Using this Huffman table, the V 555 decompression unit can decode the cid value from the Huffman code. Then, the V decompression unit 555 can parse the signal bit and the residual block code, with the decoding of the residual block code in order to identify the residual. In accordance with the above equation, the V 555 decompression unit can decode one of the VTDIST 539 vectors.

[1208] The above can be summarized in the

[1209] In the above syntax table, the first switching command with the four cases (case 0 to 3) provides a way to determine the length of the VTDIST vector in terms of the number of coefficients. The first case, case 0, indicates that all coefficients for the VTDIST vectors are specified. The second case, case 1, indicates that only those coefficients of the VTDIST vector corresponding to an order greater than a MmNumOfCoeffsForAmbHOA are specified, which can denote what is denoted as (NDIST +1) - (NBG + 1) above. The third case, case 2, is similar to the second case, but additionally subtracts coefficients identified by NumOfAddAmbHoaChan, which denotes a variable to specify additional channels (where "channels" indicates a specific coefficient corresponding to a given order, suborder combination) corresponding to an order that exceeds the NBG order The fourth case, case 3, indicates that only those coefficients from the VTDIST vector left after removing the coefficients identified by NumOfAddAmbHoaChan are specified.

[1210] After this switching command, the decision to perform uniform dequantization is controlled by NbitsQ (or nbits, as denoted above), which results, if not equal to 5, in the application of Huffman decoding. The value of cid, indicated above, is equal to the two least significant bits of the value of NbitsQ. The prediction mode discussed above is denoted as the PFlag in the syntax table above, although the HT info bit is denoted as the CbFlag in the syntax table above. The remaining syntax specifies how decoding takes place substantially similar to that described above.

[1211] In this manner, the techniques of the present disclosure can enable the audio decoding device 540D to obtain a bit stream comprising a compressed version of a spatial component of a sound field, the spatial component being generated by performing a vector-based synthesis against a plurality of spherical harmonic coefficients, and decompress the compressed version of the spatial component to obtain the spatial component.

[1212] Other techniques may enable the 540D audio decoding device to decompress a compressed version of a spatial component of a sound field, with the spatial component being generated by performing vector-based synthesis with respect to a plurality of spherical harmonic coefficients.

[1213] In this way, the 540D audio encoding device can perform various aspects of the techniques presented below in relation to the following clauses.

[1214] Clause 141541-1B. A device comprising:

[1215] one or more processors configured to obtain a bit stream comprising a compressed version of a spatial component of a sound field, the spatial component being generated by performing vector-based synthesis against a plurality of spherical harmonic coefficients, and to decompress the compressed version of the spatial component to obtain the spatial component.

[1216] Clause 141541-2B. The provision of clause 141541-1B, where the compressed version of the spatial component is represented in the bitstream using, at least partially, a field that specifies a prediction mode used during compression of the spatial component, and where the one or more processors are further configured to decompress, during decompression of the compressed version of the spatial component, a compressed version of the spatial component based at least partially on the prediction mode in order to obtain the spatial component.

[1217] Clause 141541-3B. The device of any combination of clause 141541-1B and clause 141541-2B, where the compressed version of the spatial component is represented in the bitstream using, at least partially, Huffman table information that specifies a table of Huffman used during compression of the spatial component, and where the one or more processors are additionally configured to, during decompression of the compressed version of the spatial component, decompress the compressed version of the spatial component based at least partially on the table information of Huffman.

[1218] Clause 141541-4B. The device of any combination of clause 141541-1B to clause 141541-3B, wherein the compressed version of the spatial component is represented in the bit stream using, at least partially, a field that indicates a value that expresses a size of quantization step or a variable thereof used during compression of the spatial component, and wherein the one or more processors are additionally configured to, during decompression of the compressed version of the spatial component, decompress the compressed version of the spatial component based on, at least, least partially, in value.

[1219] Clause 141541-5B. The device of clause 141541-4B, wherein the value comprises an nbit value.

[1220] Clause 141541-6B. The device of any combination of clause 141541-4B and clause 141541-5B, wherein the bit stream comprises a compressed version of a plurality of spatial components of the sound field whose compressed version of the spatial component is included, wherein the value expressed the quantization step size or a variable thereof used during compression of the plurality of spatial components and the one or more processors are further configured to, during decompression of the compressed version of the spatial component, decompress the plurality of compressed version of the spatial component based, at least partially, on value.

[1221] Clause 141541-7B. The provision of any combination of clause 141541-1B to clause 141541-6B, wherein the compressed version of the spatial component is represented in the bitstream using, at least partially, a Huffman code to represent a category identifier that identifies a compression category to which the spatial component corresponds, and in which the one or more processors are further configured to decompress, during decompression of the compressed version of the spatial component, the compressed version of the spatial component based at least partially on the Huffman code.

[1222] Clause 141541-8B. The device of any combination of clause 141541-1B to clause 141541-7B, wherein the compressed version of the spatial component is represented in the bit stream using, at least partially, a sign bit that identifies whether the spatial component is a positive value or a negative value, and wherein the one or more processors are further configured to decompress, during decompression of the compressed version of the spatial component, the compressed version of the spatial component based at least partially on the sign bit.

[1223] Clause 141541-9B. The device of any combination of clause 141541-1B to clause 141541-8B, wherein the compressed version of the spatial component represented in the bitstream using, at least partially, a Huffman code to represent a residual value of the spatial component , and wherein the one or more processors are further configured to decompress, during decompression of the compressed version of the spatial component, the compressed version of the spatial component based at least partially on the Huffman code.

[1224] Clause 141541-10B. The device of any combination of clause 141541-1B to clause 141541-10B, wherein the vector-based synthesis comprises a single-valued decomposition.

[1225] Additionally, the 540D audio decoding device can be configured to perform various aspects of the techniques presented below in relation to the following clauses.

[1226] Clause 141541-1C. A device, such as the audio decoding device 540D, comprising: one or more processors configured to decompress a compressed version of a spatial component of a sound field, the spatial component being generated by performing vector-based synthesis with respect to a plurality of spherical harmonic coefficients.

[1227] Clause 141541-2C. The device of any combination of clause 141541-1C and clause 141541-2C, wherein the one or more processors are further configured to obtain, during decompression of the compressed version of the spatial component, a category identifier that identifies a category to which the spatial component was categorized when compressed, obtain a signal that identifies whether the spatial component is a positive or negative value, obtain a residual value associated with the compressed version of the spatial component, and decompress the compressed version of the spatial component based on the category identifier, in sign and residual value.

[1228] Clause 141541-3C. The device of clause 141541-2C, wherein the one or more processors are further configured to obtain, during the retrieval of the category identifier, a Huffman code representative of the category identifier and to decode the Huffman code to obtain the category identifier category.

[1229] Clause 141541-4C. The device of clause 141541-3C, where the one or more processors are additionally configured to identify, during the decoding of the Huffman code, a Huffman table used to decode the Huffman code based on, at least partially, a position relative of the spatial component in a vector that specifies a plurality of spatial components.

[1230] Clause 141541-5C. The device of any combination of clause 141541-3C and clause 141541-4C, wherein the one or more processors are additionally configured to identify, during Huffman code decoding, a Huffman table used to decode the Huffman code with based on, at least partially, a prediction mode used during the compression of the spatial component.

[1231] Clause 141541-6C. The device of any combination of clause 141541-3C to clause 141541-5C, wherein the one or more processors are additionally configured to identify, during Huffman code decoding, a Huffman table used to decode the Huffman code based on , at least partially, on the Huffman table information associated with the compressed version of the spatial component.

[1232] Clause 141541-7C. The device of clause 141541-3C, wherein the one or more processors are additionally configured to identify, during the decoding of the Huffman code, a Huffman table used to decode the Huffman code based at least partially on a position relative of the spatial component into a vector that specifies a plurality of spatial components, in a prediction mode used during compression of the spatial component and Huffman table information associated with the compressed version of the spatial component.

[1233] Clause 141541-8C. The device of clause 141541-2C, wherein the one or more processors are further configured to decode, during residual value acquisition, a block code representative of the residual value in order to obtain the residual.

[1234] Clause 141541-9C. The device of any combination of clause 141541-1C to clause 141541-8C, wherein the vector-based synthesis comprises a single-valued decomposition.

[1235] Additionally, the 540D audio decoding device can be configured to perform various aspects of the techniques presented below in relation to the following clauses. Clause 141541-1G. A device, such as the audio decoding device 540D, comprising: one or more processors configured to identify a Huffman codebook for use in decompressing a compressed version of a current spatial component from a plurality of compressed spatial components based on in an order of the compressed version of the current spatial component with respect to the remaining components among the plurality of compressed spatial components, the spatial component being generated by performing a vector-based synthesis with respect to a plurality of spherical harmonic coefficients.

[1236] Clause 141541-2G. The device of clause 141541-1G, wherein the one or more processors are additionally configured to perform any combination of the steps recited in clause 141541-1D to clause 141541-10D, and in clause 141541-1E to clause 141541-9E.

[1237] Figures 42 to 42C are each block diagrams illustrating in more detail the order reduction unit 528A shown in the examples of Figures 40B to 40J. Figure 42 is a block diagram illustrating an order reduction unit 528, which may represent an example of the order reduction unit 528A of Figures 40B to 40J. Order reduction unit 528A may receive or otherwise determine a target bit rate 535 and perform order reduction with respect to background spherical harmonic coefficients 531 based only on that target bit rate 535. In some examples, the order reduction unit 528A may access a table or other data structures that use the target bit rate 535 to identify those orders and/or suborders that must be removed from the spherical background harmonic coefficients 531 to in order to generate the reduced background spherical harmonic coefficients 529.

[1238] In this way, the techniques can enable an audio encoding device, such as audio encoding devices 510B to 410J, to perform, based on a target bit rate 535, order reduction relative to a plurality of spherical harmonic coefficients or decompositions thereof, such as background spherical harmonic coefficients 531, in order to generate reduced spherical harmonic coefficients 529 and reduced decompositions thereof, wherein the plurality of spherical harmonic coefficients represents a sound field.

[1239] In each of the various examples described above, it is to be understood that the audio decoding device 540 may perform a method or otherwise comprises means for performing each step of the method that the audio decoding device 540 is configured to perform, in some instances, such means may comprise one or more processors. In some examples, the one or more processors may represent a special-purpose processor configured through instructions stored on non-transient computer-readable storage media. In other words, various aspects of the techniques in each of these sets of encoding examples can provide a non-transient computer-readable storage medium that has instructions stored on it that, when executed, cause the one or more processors to perform the method. that the 540 audio decoding device has been configured to perform.

[1240] Figure 42B is a block diagram illustrating an order reduction unit 528B, which may represent an example of the order reduction unit 528A of Figures 40B to 40J. In the example of Fig. 42B, instead of performing order reduction based only on a target bit rate 535, the order reduction unit 528B can perform order reduction based on a content analysis of the spherical harmonic coefficients of background 531. Order reduction unit 528B may include a content analysis unit 536A that performs such content analysis.

[1241] In some examples, the content analysis unit 536A may include a spatial analysis unit 536A that performs a form of content analysis called spatial analysis. Spatial analysis may involve analyzing the spherical background harmonic coefficients 531 to identify spatial information that describes the shape or other spatial properties of the background components of the sound field. Based on this spatial information, the order reduction unit 528B can identify those orders and/or suborders that must be removed from the spherical background harmonic coefficients 531 in order to generate reduced spherical background harmonic coefficients 529.

[1242] In some examples, the content analysis unit 536A may include a diffusion analysis unit 536B that performs a form of content analysis called diffusion analysis. Diffusion analysis can analyze the spherical background harmonic coefficients 531 to identify the diffusion information that describes the diffusivity of the background components of the sound field. Based on this diffusion information, the order reduction unit 528B can identify those orders and/or suborders that must be removed from the spherical background harmonic coefficients 531 in order to generate reduced spherical background harmonic coefficients 529.

[1243] While shown to be inclusive of both the spatial analysis unit 536A and the scatter analysis unit 36B, the content analysis unit 536A may include only the spatial analysis unit 536, only the scatter analysis unit 536B, or both the spatial analysis unit 536A and the diffusion analysis unit 536B. In some instances, the content analysis unit 536A may perform other forms of content analysis in addition to, or as alternatives to, spatial analysis and diffusion analysis. Accordingly, the techniques described in the present disclosure should not be limited in this regard.

[1244] In this way, the techniques can enable an audio coding device, such as the 510B to 510J audio coding devices, to perform, based on a content analysis, a plurality of spherical harmonic coefficients or decompositions thereof that describes a sound field, order reduction with respect to the plurality of spherical harmonic coefficients or decompositions thereof in order to generate reduced spherical harmonic coefficients or reduced decompositions thereof.

[1245] In other words, the techniques may enable a device, such as the 510B to 510J audio encoding devices, to be configured in accordance with the following clauses.

[1246] Clause 133146-1E. A device, such as any of audio encoding devices 510B to 510J, comprising one or more processors configured to perform, based on a content analysis, a plurality of spherical harmonic coefficients or decompositions thereof that describe a sound field , an order reduction with respect to the plurality of spherical harmonic coefficients or decompositions thereof in order to generate reduced spherical harmonic coefficients or reduced decompositions thereof.

[1247] Clause 133146-2E. The device of clause 133146-1E, wherein the one or more processors are further configured to perform, prior to performing order reduction, a single-value decomposition with respect to the plurality of spherical harmonic coefficients in order to identify one or more first vectors that describe distinct components of the sound field and one or more second vectors that identify background components of the sound field, and where the one or more processors are configured to perform order reduction with respect to the one or more first vectors, the one or more second vectors or both the one or more first vectors and the one or more second vectors.

[1248] Clause 133146-3E. The device of clause 133146-1E, wherein the one or more processors are further configured to perform content analysis with respect to the plurality of spherical harmonic coefficients or decompositions thereof.

[1249] Clause 133146-4E. The device of clause 133146-3E, wherein the one or more processors are configured to perform spatial analysis with respect to the plurality of spherical harmonic coefficients or decompositions thereof.

[1250] Clause 133146-5E. The device of clause 133146-3E, wherein performing the content analysis comprises performing a diffusion analysis with respect to the plurality of spherical harmonic coefficients or decompositions thereof.

[1251] Clause 133146-6E. The device of clause 133146-3E, wherein the one or more processors are configured to perform analysis and diffusion analysis with respect to the plurality of spherical harmonic coefficients or decompositions thereof. Clause 133146-7E. The device of claim 1, wherein the one or more processors are configured to perform, based on content analysis of the plurality of spherical harmonic coefficients or decompositions thereof and a target bit rate, order reduction with respect to to the plurality of spherical harmonic coefficients or in the decompositions thereof in order to generate the reduced spherical harmonic coefficients or the reduced decompositions thereof.

[1252] Clause 133146-8E. The device of clause 133146-1E, wherein the one or more processors are further configured to audio encode the reduced spherical harmonic coefficients or decompositions thereof.

[1253] Clause 133146-9E. The device of clause 133146-1E, wherein the one or more processors are further configured to audio encode the reduced spherical harmonic coefficients or reduced decompositions thereof and to generate a bit stream to include the reduced spherical harmonic coefficients or their reduced decompositions.

[1254] Clause 133146-10E. The device of clause 133146-1E, wherein the one or more processors are further configured to specify one or more orders and/or one or more suborders of spherical basis functions to which these reduced spherical harmonic coefficients or reduced decompositions thereof correspond in a bit stream that includes the reduced spherical harmonic coefficients or the reduced decompositions thereof.

[1255] Clause 133146-1 IE. The device of clause 133146-1E, wherein reduced spherical harmonic coefficients or reduced decompositions thereof have lesser values than the plurality of spherical harmonic coefficients or decompositions thereof.

[1256] Clause 133146-12E. The device of clause 133146-1E, wherein the one or more processors are further configured to remove those of a plurality of spherical harmonic coefficients or vectors from the decompositions thereof that have a specified order and/or suborder to generate the reduced spherical harmonic coefficients or their reduced decompositions.

[1257] Clause 133146-13E. The device of clause 133146-1E, wherein the one or more processors are configured to zero out those among the plurality of spherical harmonic coefficients or those vectors of the decomposition thereof which have a specified order and/or suborder to generate the spherical harmonic coefficients reduced or reduced decomposition thereof.

[1258] Figure 42C is a block diagram illustrating an order reduction unit 528C, which may represent an example of the order reduction unit 528A of Figures 40B to 40J. The order reduction unit 528C of Fig. 42B is substantially the same as the order reduction unit 528B, but may receive or otherwise determine a target bit rate 535 in the manner described above with respect to the reduction unit. 528A of Fig. 42, while also performing the content analysis in the manner described above with respect to the order reduction unit 528B of Fig. 42B. Then, order reduction unit 528C can perform order reduction with respect to spherical background harmonic coefficients 531 based on that target bit rate 535 and content analysis.

[1259] In this way, the techniques can 422/500 enable an audio encoding device, for example, audio encoding devices 510B to 510J, to perform a content analysis with respect to the plurality of spherical harmonic coefficients or decompositions of the same. While performing order reduction, audio encoding devices 510B to 510J can perform, based on target bit rate 535 and content analysis, order reduction with respect to the plurality of spherical harmonic coefficients or decompositions of the same to generate the reduced spherical harmonic coefficients or the reduced decompositions thereof.

[1260] As one or more vectors are removed, audio encoding devices 510B to 510J can specify the number of vectors in the bit stream as control data. Audio encoding devices 510B to 510J can specify this number of vectors in the bit stream to facilitate the extraction of vectors from the bit stream by the audio decoding device.

[1261] Figure 44 is an exemplary diagrammatic illustration of operations performed by the audio encoding device 410D to compensate for quantization error in accordance with various aspects of the techniques described in the present disclosure. In the example of Figure 44, math unit 526 of audio encoding device 510D is shown as a dashed block to denote that math operations can be performed by math unit 526 of audio decoding device 510D.

[1262] As shown in the example of Figure 44, the math unit 526 can first multiply the UDIST * SDIST 527 vectors by the VTDIST 525E vectors in order to generate distinct 423/500 spherical harmonic coefficients (denoted as "HDIST 630 vectors"). Then the math unit 526 can divide the HDIST vectors 630 by the quantized version of the VTDIST 525E vectors (which are denoted, again, as "VTQ_DIST 525G vectors"). The math unit 526 can perform the division by determining a pseudoinverse of the VTQ_DIST 525G vectors and then multiplying the HDIST vectors by the pseudoinverse of the VTQ_DIST 525G vectors, issuing an error compensated version of UDIST * SDIST (which can be abbreviated as "USDIST" or "USDIST" vectors). The error compensated version of USDIST can be denoted as USDIST 527' vectors in the example in Figure 44. In this way, the techniques can effectively project the quantization error, at least partially, to the USDIST 527 vectors, generating the USDIST vectors 527'.

[1263] Then the math unit 526 can subtract the USDIST 527' vectors from the UDIST * SDIST 527 vectors to determine the USERR 634 vectors (which can account for at least a portion of the error due to quantization projected onto the UDIST * SDIST 527 vectors ). Then the math unit 526 can multiply the USERR 634 vectors by the VTQ_DIST 525G vectors in order to determine the HERR 636 vectors. Mathematically, the HERR 636 vectors can be equivalent to the USDIST 527 vectors - USDIST 527 vectors, whose result is, then multiplied by the VTDIST 525E vectors. Then the math unit 526 can sum the HERR vectors 636 to the spherical background harmonic coefficients 531 (denoted as ¾G vectors 531 in the example in Figure 44) computed by multiplying the UBG vectors 525D by the SBG 525B vectors and then , by the VTBG 525F vectors. The math unit 526 can sum the HERR 636 vectors to the 531 HBG vectors, which effectively projects at 424/500 minus a portion of the quantization error into the HBG 531 vectors to generate compensated HBG 53 vectors. In this way, the techniques can project at least a portion of the quantization error onto the HBG 531 vectors.

[1264] Figures 45 and 45B are diagrams illustrating subframe interpolation of portions of two frames in accordance with various aspects of the techniques described in the present disclosure. In the example of Figure 45, a first frame 650 and a second frame 652 are shown. The first frame 650 can include spherical harmonic coefficients (“SH[1]”) that can be decomposed into matrices U[1], S[1] and V'[1]. The second frame 652 may include spherical harmonic coefficients (“SH[2]”). These SH[1] and SH[2] can identify different SHC 511 frames described above.

[1265] In the example of Figure 45B, the decomposition unit 518 of the audio encoding device 510H shown in the example of Figure 40H can separate each of the frames 650 and 652 into four respective subframes 651A to 651D and 653A to 653D. Then, the decomposition unit 518 can decompose the first subframe 651A (denoted as "SH[1,1]") of frame 650 into matrices U[1, 1], S[1, 1] and V[1, 1 ], outputting the matrix V[1, 1] 519' to the interpolation unit 550. Then, the decomposition unit 518 may decompose the second subframe 653A (denoted as "SH[2,1]") of the frame 652 into matrices U[1, 1], S[1, 1] and V[1, 1], outputting the matrix V[2, 1] 519' to the interpolation unit 550. The decomposition unit 518 can also output SH[1, 1], SH[1, 2], SH[1, 3] and SH[1, 4] of SHC 11 and SH[2, 1], SH[2, 2], SH[2, 3] and SH[ 2, 4] from the SHC 511 to the interpolation unit 550.

[1266] The interpolation unit 425/500 550 can then perform the interpolations identified at the bottom of the illustration shown in the example of Figure 45B. That is, the interpolation unit 550 can interpolate V'[1, 2] based on V'[1, 1] and V'[2, 1]. The interpolation unit 550 can also interpolate V'[1, 3] based on V'[1, 1] and V'[2, 1]. In addition, the interpolation unit 550 can also interpolate V'[1, 4] based on V'[1, 1] and V'[2, 1]. These interpolations may involve a projection of V'[1, 1] and V'[2, 1] onto the spatial domain, as shown in the example of Figures 46 to 46E, followed by a temporal interpolation and then a projection back into the spherical harmonic domain.

[1267] The interpolation unit 550 can then derive U[1, 2]S[1, 2] by multiplying SH[1, 2] by (V'[1, 2])-1, U[ 1, 3]S[1, 3] multiplying SH[1, 3] by (V'[1, 3])-1 and U[1, 4]S[1, 4] multiplying SH[1 , 4] by (V'[1, 4])-1. Thereafter, the interpolation unit 550 can reform the frame into decomposed form by outputting the V matrix 519, the S matrix 519B and the U matrix 519C.

[1268] Figures 46A to 46E are diagrams illustrating a cross-section of a projection of one or more vectors of a decomposed version of a plurality of spherical harmonic coefficients that have been interpolated in accordance with the techniques described in the present disclosure. Fig. 46A illustrates a cross-section of a projection of one or more first vectors of a first V matrix 19' that has been decomposed from the SHC 511 of a first subframe from a first frame through an SVD process. Figure 46B illustrates a cross-section of a projection of one or more first vectors of a second V matrix 519' that has been decomposed from the SHC 511 of a second subframe from a second frame through an SVD process. 426/500

[1269] Figure 46C illustrates a cross-section of a projection of one or more interpolated vectors into a V matrix 519A representative of a second subframe of the first frame, such vectors having been interpolated in accordance with the techniques described in the present matrix disclosure. V 519' decomposed from the first subframe of the first frame of the SHC 511 (i.e., the one or more vectors of the matrix V 519' shown in the example of Figure 46 in that example) and the first subframe of the second frame of the SHC 511 (i.e. that is, the one or more vectors of the matrix V 519' shown in the example of Figure 46B in that example).

[1270] Figure 46D illustrates a cross-section of a projection of one or more interpolated vectors into a V matrix 519A representative of a third subframe of the first frame, such vectors having been interpolated in accordance with the techniques described in the present matrix disclosure V 519' decomposed from the first subframe of the first frame of the SHC 511 (i.e., the one or more vectors of the matrix V 519' shown in the example of Figure 46 in that example) and the first subframe of the second frame of the SHC 511 (i.e. that is, the one or more vectors of the matrix V 519' shown in the example of Figure 46B in that example).

[1271] Figure 46E illustrates a cross-section of a projection of one or more interpolated vectors into a V matrix 519A representative of a fourth subframe of the first frame, these vectors having been interpolated in accordance with the techniques described in the present matrix disclosure V 519' decomposed from the first subframe of the first frame of the SHC 511 (i.e., the one or more vectors of the matrix V 519' shown in the example of Figure 46 in that example) and the first subframe of the second frame of the SHC 511 (that is, the one or more vectors of matrix V 519' shown in the example of Fig. 46B in that example).

[1272] Figure 47 is a block diagram illustrating in more detail the extraction unit 542 of the audio decoding devices 540A to 540D shown in the examples of Figures 41 to 41D. In some examples, the extraction unit 542 may represent a front-end which may be termed an "integrated decoder", which can perform two or more decoding schemes (where, by carrying out these two or more schemes, the decoder can be considered to "integrate" the two or more schemes). As shown in the example of Figure 44, extraction unit 542 includes a multiplexer 620 and extraction subunits 622A and 622B ("extraction subunits 622"). The multiplexer 620 identifies those of the frame-coded SCH arrays 547 to 547N to be sent to the extraction subunit 622A and the extraction subunit 622B based on the corresponding indication of whether the associated frame-coded SCH arrays 547 to 547N are generated from a synthetic audio object or from a recording. Each of the 622A extraction subunits can perform a different decoding scheme (which may be called "decompression") which, in some examples, is adapted both for SHC generated from a synthetic audio object and for a SHC generated from a recording. Each of the extraction subunits 622A can perform a respective scheme among these decompression schemes in order to generate the frames from the SHC 547, which are sent to the SHC 547.

[1273] For example, the extraction unit 622A can perform a decompression scheme to reconstruct 428/500 of the SA from a predominant signal (PS) using the following formula:

[1274] where DirV is a directional vector (representative of various directions and widths), which can be transmitted over an auxiliary channel. In this example, the 622B extraction unit can perform a decompression scheme that reconstructs the PS HOA matrix using the following formula:

[1275] where Ynm is the spherical harmonic function and theta and fi information can be sent through the auxiliary channel.

[1276] In this regard, the techniques enable the extraction unit 538 to select one of a plurality of decompression schemes based on the indication of whether a compressed version of spherical harmonic coefficients representative of a sound field is generated from an object of synthetic audio and decompress the compressed version of the spherical harmonic coefficients using the scheme selected from the plurality of decompression schemes. In some examples, the device comprises an integrated decoder.

[1277] Figure 48 is a block diagram illustrating the audio rendering unit 48 of the audio decoding device 540A to 540D shown in the examples of Figures 41A to 41D in more detail. Figure 48 illustrates a conversion of recovered spherical harmonic coefficients 547 into multi-channel audio data 549A that is compatible with a decoder local speaker geometry. For some 429/500 local speaker geometries (which, again, can refer to a speaker geometry at the decoder), some transforms that guarantee invertibility can result in less than desirable audio-image quality. That is, sound reproduction may not always result in the correct location of sounds when compared to the audio that is captured. In order to correct for this less-than-desirable image quality, the techniques can be further improved so as to introduce a concept that can be termed “virtual speakers”.

[1278] Rather than requiring one or more speakers to be relocated or positioned in particular or defined regions of space that have certain tolerances specified by a standard such as the ITU-R BS.775-1 verified above, the above frame can be modified to include some form of panning, for example, vector-based amplitude panning (VBAP), distance-based amplitude panning, or other forms of panning. Focusing on VBAP, for purposes of illustration, VBAP can effectively introduce what can be characterized as "virtual speakers." VBAP can modify a feed to one or more speakers so that those or more speakers speakers effectively emit sound that appears to originate from a virtual speaker at one or more of a location and angle different from at least one of the location and/or angle of the one or more speakers supporting the virtual speaker.

[1279] For purposes of illustration, the following equation for determining speaker feeds in terms of SHC might be as follows:

[1280] In the above equation, the VBAP matrix is the size of M rows by N columns, where M denotes the number of speakers (and equals five in the above equation) and N denotes the number of virtual speakers. VBAP matrix A can be computed as a function of vectors from the defined listener location for each of the speaker positions and the vectors from the defined listener location for each of the virtual speaker positions. The matrix AD in the above equation can be N rows in size by (order+1)2 columns, where the order can refer to the order of the SH functions. The matrix D can represent the following:

[1281] The matrix g (or vector, since there is only a single column) can represent the gain for speaker feeds for the speakers arranged in the local decoder geometry. In the equation, matrix g is of size M. Matrix A (or vector, since there is only a single column) can denote SHC 520, and is of size (Order+1)(Order+1), which can also be denoted as (Order+1)2.

[1282] In effect, the VBAP matrix is an MxN matrix that provides what might be termed a "gain adjustment" that factors in the location of the speakers and the portion of the virtual speakers. Introducing panning in this way can result in in better 431/500 reproduction of multichannel audio that results in a better image quality when reproduced by local speaker geometry. Also, by incorporating VBAP into this equation, the techniques can overcome poor speaker geometries that do not align with those specified in various patterns.

[1283] In practice, the equation can be inverted and employed to transform the SHC back to multichannel feeds for a particular speaker geometry or configuration, which, again, may be termed local decoder geometry in the present disclosure. That is, the equation can be inverted in order to solve the matrix g. The inverted equation can be as follows]:

[1284] The matrix g can represent the speaker gain, in this example, for each of the five speakers in a 5.1 speaker configuration. The virtual speaker locations used in this configuration can match the locations defined in a 5.1 multi-channel format specification or standard. The location of speakers that can support each of these virtual speakers can be determined using any known audio localization technique, many of which involve playing a tone that has a particular frequency to determine a location of each speaker. speaker in relation to a headend unit (such as an audio/video receiver (A/V receiver), television, gaming system, 432/500 digital video disc system, or other types of audio systems). headend). Alternatively, a user of the headend unit can manually specify the location of each of the speakers. Either way, given these locations and possible angles, the headend unit can resolve the gains by assuming an optimal virtual speaker setup via VBAP.

[1285] In this regard, a device or apparatus may perform vector base amplitude panning or other form of panning on the plurality of virtual channels to produce a plurality of channels that drive the speakers and, a local decoder geometry for emit sounds that appear to originate from virtual speakers configured in a different local geometry. Therefore, the techniques may enable the audio decoding device 40 to perform a transform on the plurality of spherical harmonic coefficients, such as the recovered spherical harmonic coefficients 47, in order to produce a plurality of channels. Each of the plurality of channels may be associated with a corresponding different region of space. Furthermore, each of the plurality of channels may comprise a plurality of virtual channels, wherein the plurality of virtual channels may be associated with the corresponding different region of space. Therefore, a device can perform vector base amplitude panning on the virtual channels to produce the plurality of channels of the multichannel audio data 49.

[1286] Figures 49A through 49E(ii) are diagrams illustrating receptive audio coding systems 560 A 560C, 567D, 569D, 571E and 573E that can implement various aspects of the techniques described in the present disclosure. As shown in the example of Figure 49A, the audio encoding system 560A may include an audio encoding device 562 and an audio decoding device 564. The audio encoding device 562 may be similar to any of these devices. coding codes 20 and 510A to 510D shown in the example of Figures 4 and 40A to 40D, respectively. Audio decoding device 564 may be similar to audio decoding device 24 and 40 shown in the example of Figures 5 and 41.

[1287] As described above, higher-order ambisonics (HO A) is a way of describing all directional information of a sound field based on a spatial Fourier transform. In some examples, the higher the order of ambisonics, N, the greater the spatial resolution and the greater the number of spherical harmonic coefficients (SH) (N+1)2. Thus, the higher the order of ambisonics N, in some examples, results in higher bandwidth requirements to transmit and store the coefficients. Due to the fact that the bandwidth requirements of HO A are quite high compared to, for example, 5.1 or 7.1 surround system audio data, a bandwidth reduction may be desired for many applications.

[1288] In accordance with the techniques described in the present disclosure, the 560A audio coding system can perform a method based on separating discrete (foreground) from non-distinguished from non-distinguished (background or ambient) elements in a spatial sound scene. This separation can allow the 560A audio encoding system to process foreground and background elements independently of each other. In this example, the 560 A audio 434/500 encoding system exploits the property that foreground elements can attract more attention (by the listener) and can be easier to locate (again, by the listener) compared to background elements. . As a result, the 560A audio encoding system can store or stream HOA content more effectively.

[1289] In some examples, the 560A audio encoding system can accomplish this separation using the Singular Value Decomposition (SVD) process. The SVD process can separate a table of HOA coefficients into 3 matrices (U, S, V). The matrix U contains the singular left vectors and the matrix V contains the right singular vectors. The matrix S Diagonal S contains the non-negative ranked singular values on its diagonal. A generally satisfactory (or, in some examples, perfect reconstruction that assumes unlimited precision in representing the HOA coefficients) of the HOA coefficients is given by U*S*V. By reconstructing only the subspace with the largest D:U(:,1:D)*S(1:D,:)*V subspace, the 560A audio coding system can extract the most salient spatial information from this frame. of HOA, i.e. foreground sound elements (and perhaps some strong previous room reflections). The remaining U(:,D+1:end)*S(D+1:end,:)*V' can reconstruct background elements and content reverb.

[1290] The 560 A audio coding system can determine the D value, which separates the two subspaces, by analyzing the slope of the curve created by decreasing diagonal values of S: large singular values represent foreground sounds, singular values low ones represent background values. The 560A audio coding system can use a 435/500 first and second derivative of the singular value curve. The 560A audio coding system can also limit the D number between one and five. Alternatively, the audio coding system 560 A may preset the number D, for example, to a value of four. In any case, once the number D is estimated, the audio coding system 560A extracts the foreground and background subspace of the U and S matrices.

[1291] The 560 A audio coding system can then reconstruct the HO A coefficients of the background scenario via U(:,D+l:end)*S(D+l:end,:) *V, which results in (N+1)2 channels of HOA coefficients. Since it is known that background elements are, in some instances, not as salient and not as locatable relative to foreground elements, the 560A audio coding system may truncate the channel order. of HOA. Additionally, the 560A audio coding system can compress these channels with lossy or lossless audio codecs, for example AAC, or optionally a more aggressive audio codec compared to the codec used to compress the protruding foreground elements. In some examples, in order to save bandwidth, the 560A audio encoding system may transmit foreground elements differently. That is, the audio coding system can transmit the left singular vectors U(:,1:D) after being compressed with lossy or lossless audio codecs (e.g. AAC) and can transmit these left singular values compressed together of the reconstruction matrix R=S(1:D,:)*V. R can represent a D x (N+1)2 matrix, which can differ across frames.

[1292] On the receiving side of the 560 audio coding system, the 436/500 audio coding system can multiply these two matrices in order to reconstruct a (N+1)2 channel HOA frame. Since the foreground and background HOA channels are summed together, the 560A audio encoding system can render to any speaker array using any Ambisonic renderer. Since the techniques provide for the separation of foreground elements (direct or distinct sound) from background elements, a hearing impaired person can control the mixing of foreground and background elements to increase intelligibility. In addition, other audio effects can also be applied, for example a dynamic compressor only on foreground elements.

[1293] Figure 49B is a block diagram illustrating the 560B audio coding system in more detail. As shown in the example of Figure 49B , audio encoding system 560B may include an audio encoding device 566 and an audio decoding device 568. Audio encoding device 566 may be similar to audio encoding devices 24 and 510E shown in the examples of Figures 4 and 40E. Audio decoding device 568 may be similar to audio decoding device 540B and 40 shown in the example of Figures 5 and 41B.

[1294] In accordance with the techniques described in the present disclosure, while using frame based SVD (or related methods, e.g. KLT & PCA) decomposition into HoA signals, for the purpose of bandwidth reduction, the audio coding device 66 can quantize the first vectors of the matrix U (multiple by the corresponding singular values of the matrix S) as well as the corresponding vectors of the VT vector. This will comprise the 437/500 foreground components of the sound field. The techniques can make it possible for the 566 audio encoding device to encode the UDIST*SDIST vector using a "black box" audio encoding mechanism. The vector V can be either a scalar or a quantized vector. Furthermore, some or all of the remaining vectors in the U matrix can be multiplied by the corresponding singular values of the S matrix and the V matrix and also encoded using a "black box" audio coding mechanism. They will understand the 'background' components of the sound field.

[1295] Since the louder hearing components are decomposed into the 'foreground components', 566 audio encoding devices can reduce the Ambisonics order of the 'background' components before using a "black box" audio, due to the fact that the background is assumed not to contain important findable content. Depending on the ambisonics order of the foreground components, an audio encoding unit 566 can transmit the corresponding V-vector(s), which can be quite large. For example, a simple 16-bit scalar quantization of the V vectors will result in an overhead of approximately 20 kbps for the 4th order (25 coefficients) and 40 kbps for the 6th order (49 coefficients) per foreground component. The techniques described in the present disclosure may provide a method of reducing this V-Vector overhead.

podem ser usados para reconstruir os componentes de primeiro plano ou distintos até a ordem NBG. Com o uso das técnicas descritas na presente revelação, o dispositivo de codificação de áudio 566 pode reconstruir o primeiro plano até a ordem NBG e mesclar os

canais resultantes com os canais de segundo plano, o que resulta em um campo sonoro completo até a ordem NBG. O dispositivo de codificação de áudio 566 pode, em seguida, reduzir o vetor-V para aqueles coeficientes com o índice maior do que (NBG+1)2 para transmissão, (em que esses vetores podem ser denominados de "VTSMALL”). No lado receptor, a unidade de decodificação de áudio 568 pode reconstruir os canais de áudio de primeiro plano para a ordem de ambisonics maior do que NBG multiplicando-se os elementos de primeiro plano pelos vetores VTSMALL.[1296] For purposes of illustration, it is assumed that the ambisonics order of the foreground elements is NDIST and the ambisonics order of the NBG background elements, as described above. Since audio encoding device 566 can reduce the Ambisonics order of background elements, 438/500 as described above, NBG can be smaller than NDIST. The length of the foreground V-vector that needs to be transmitted to reconstruct the foreground elements on the receiving side is the length of (NDIST+1)2 per foreground element, whereas the first coefficients

can be used to reconstruct foreground or distinct components up to NBG order. Using the techniques described in the present disclosure, the audio encoding device 566 can reconstruct the foreground to NBG order and merge the

resulting channels with the background channels, which results in a full sound field up to the NBG order. The audio encoding device 566 can then reduce the V-vector to those coefficients with the index greater than (NBG+1)2 for transmission, (where these vectors may be referred to as "VTSMALL"). receiver side, the audio decoding unit 568 can reconstruct the foreground audio channels to the order of ambisonics greater than NBG by multiplying the foreground elements by the VTSMALL vectors.

[1297] Figure 49C is a block diagram illustrating the 560C audio coding system in more detail. As shown in the example of Figure 49C, audio encoding system 560B may include an audio encoding device 567 and an audio decoding device 569. Audio encoding device 567 may be similar to audio encoding devices 20 and 510F shown in the examples of Figures 4 and 40F. Audio decoding device 569 may be similar to audio decoding devices 540B and 40 shown in the example of Figures 5 and 41B.

[1298] In accordance with the 439/500 techniques described in the present disclosure, while using frame based SVD (or related methods, e.g. KLT & PCA) decomposition into HoA signals, for the purpose of width reduction bandwidth, the audio coding device 567 can quantize the first vectors of the matrix U (multiple by the corresponding singular values of the matrix S) as well as the corresponding vectors of the VT vector. This will comprise the 'foreground' components of the sound field. The techniques can make it possible for the 567 audio encoding device to encode the UDIST*SDIST vector using a "black box" audio encoding mechanism. The vector V can be either a scalar or a quantized vector. Furthermore, some or all of the remaining vectors in the U matrix can be multiplied by the corresponding singular values of the S matrix and the V matrix and also encoded using a "black box" audio coding mechanism. They will understand the 'background' components of the sound field.

[1299] Since the loudest hearing components are decomposed into the 'foreground components', 567 audio encoding devices can reduce the Ambisonics order of the 'background' components before using a "black box" audio, due to the fact that it is assumed) that the background does not contain important findable content. The audio encoding device 567 can reduce the order in order to preserve the overall energy of the sound field in accordance with techniques described herein. Depending on the Ambisonics order of the foreground components, a 567 audio encoding unit can transmit the corresponding V-vector(s), which can be quite large. For example, a simple 16-bit scalar quantization of the 440/500 V vectors will result in an overhead of approximately 20 kbps for the 4th order (25 coefficients) and 40 kbps for the 6th order (49 coefficients) per foreground component. The techniques described in the present disclosure may provide a method for reducing this V-Vector(s) overhead.

[1300] For purposes of illustration, it is assumed that the Ambisonics order of the foreground elements and background elements is N. The audio encoding device 567 can reduce the Ambisonics order of the background elements of( s) V-vector(s) of N a so that < N. Audio encoding device 67 additionally applies compensation to increase the values of the background elements of the V-vector(s) to preserve energy of the sound field described by SHCs. Exemplary techniques for applying compensation are described above with respect to Figure 40F. On the receiving side, the audio decoding unit 569 can reconstruct the background audio channels to the order of ambisonics. Figures 49D(i) and 49D(ii) illustrate an audio encoding device 567D and an audio decoding device 569D respectively. The audio encoding device 567D and the audio decoding device 569D may be configured to perform one or more directionality-based distinguishing determinations, in accordance with aspects of the present disclosure. Higher Order Ambisonics (HO A) is a method for describing all directional information of a sound field based on the spatial Fourier transform. The higher the order of Ambisonics N, the greater the spatial resolution, the greater the number of spherical harmonic (SH) coefficients (N+1)^2, the greater the 441/500 bandwidth required to transmit and store the data. Due to the fact that the bandwidth requirements of HO A are quite high, for many applications a bandwidth reduction is desired.

[1301] Previous descriptions have described how SVD (single value decomposition) or related processes can be used for spatial audio compression. The techniques described in this document present an improved algorithm for selecting the salient elements also known as the foreground elements. After an SVD-based decomposition of an HOA audio frame into its U, S, V matrices, the techniques base the selection of salient K elements exclusively on the first K channels of the U matrix [ U(:,1:K)* S(1:K,1:K)]. This results in selecting the audio elements with the highest energy. However, these elements are not guaranteed to be directional as well. Therefore, the techniques are aimed at finding sound elements that have high energy and are also directional. This is potentially achieved by weighting the matrix V against the matrix S. Then, for each row of this resulting matrix, the highest indexed elements (which are associated with the higher-order coefficients HOA) are squared and summed, which results in a value per row [sumVS in the pseudocode described with respect to Figure 40H]. In accordance with the workflow represented in the pseudocode, higher-order Ambisonic coefficients starting at the 5th index are considered. These values are sorted according to their size and the assortment index is used to rearrange the original U, S, and V matrices accordingly. The SVD-based compression algorithm described 442/500 earlier in the present disclosure can then be applied without further modification.

[1302] Figures 49E(i) and 49E(ii) are block diagrams illustrating an audio encoding device 571E and an audio decoding device 573E respectively. Audio encoding device 571E and audio decoding device 573E can perform various aspects of the techniques described above with respect to the examples in Figures 49 to 49D(ii), with the exception that audio encoding device 57 IE can perform the single-valued decomposition against a power spectral density (PDS) matrix of the HOA coefficients to generate a matrix S2 and a matrix V. The matrix S2 may denote a squared matrix S, and consequently, the matrix S2 can undergo a square root operation to obtain the matrix S. The audio encoding device 57 IE can perform, in some examples, quantization with respect to the matrix V to obtain a quantized matrix V (which may be denoted as matrix V).

[1303] Audio encoding device 571E can obtain matrix U by first multiplying matrix S by quantized matrix V to generate matrix SV. The 57 IE audio encoding device can then obtain the pseudoinverse of the SV matrix and then multiply the HOA coefficients by the pseudoinverse of the SV matrix to obtain the U matrix. Performing SVD with respect to spectral density of power of the HOA coefficients rather than the coefficients themselves, the 57 IE audio encoding device can potentially reduce the computational complexity of performing the SVD in terms of one or more of processor cycles and storage space, while at the same time realizes the same 443/500 source audio coding efficiency as if SVD were applied directly to the HOA coefficients.

[1304] The audio decoding device 573E can be similar to those audio decoding devices described above, except that the audio decoding device 573 can reconstruct the HOA coefficients from the decompositions of the HOA coefficients achieved by applying the SVD and in relation to the power spectral density of the HOA coefficients instead of the HOA coefficients directly.

[1305] Figures 50A and 50B are block diagrams each illustrating one of two different approaches to potentially reducing background content ordering in accordance with the techniques described in the present disclosure. As shown in the example in Figure 50, the first approach can employ order reduction with respect to vectors UBG * SBG * VT to reduce the order from N to fj, and which is less than (<) N. That is, the order reduction unit 528A shown in the examples of Figure 40B to 40J can perform order reduction to truncate or otherwise reduce the N order of vectors UBG * SBG * VT to fj, where is less than (< ) No.

em que essa matriz VBG é, então, usada no lugar da matriz VT durante a geração 444/500 dos vetores UBG * SBG * VT, que realizam de maneira eficaz o truncamento a fim de gerar os vetores UBG * SBG * VT de tamanho M x (fj + l)2.[1306] As an alternative approach, the order reduction unit 528A can, as shown in the example in Figure 50B, perform this truncation with respect to the VT which eliminates the rows so that they are ((+1)2, which is not is illustrated in the example of Figure 40B for ease of illustration. In other words, the order reduction unit 528A can remove one or more orders from the VT matrix to efficiently generate a matrix of VBG.

where this VBG matrix is then used in place of the VT matrix during the 444/500 generation of the UBG * SBG * VT vectors, which effectively truncate to generate the UBG * SBG * VT vectors of size M x (fj + l)2.

[1307] Figure 51 is a block diagram illustrating examples of a distinct component compression path of an audio encoding device 700A that can implement various aspects of the techniques described in the present disclosure for compressing spherical harmonic coefficients 701. For example in Figure 51, the distinct component compression path may refer to a processing path of the audio encoding device 700A that compresses the distinct components of the sound field represented by the SHC 701. Background component compression, may represent a processing path of the 700A audio encoding device that compresses the SHC 701's background components.

[1308] Although not shown for purposes of ease of illustration, the background compression path can operate against the SHC 701 directly instead of the SHC 701 decompositions. This is similar to what was described above with respect to Figures 49 to 49C, with the exception that instead of recomposing the background components from the UBG, SBG, and VBG matrices and then performing some of the pseudoacoustic encoding (e.g. using an AAC encoder) on those components background components, the background component processing path can operate against the SHC 701 directly (as described above with respect to the audio encoding device 20 shown in the example in Figure 4), by compressing these background components with the use of the 445/500 piscoacoustic encoder. By performing pseudo-acoustic coding against the SHC 701 directly, discontinuities can be reduced while also reducing computational complexity (in terms of operations required to compress the background components) compared to performing pseudo-acoustic coding. against the recomposed background components. Although indicated in terms of distinguished and background, the term "prominent" may be used in place of "distinguished" and the term "environment" may be used in place of "background" in the present revelation.

[1309] In any case, the spherical harmonic coefficients 701 (“SHC 701”) may comprise a matrix of coefficients that have a size of M x (N+1)2 where M denotes the number of samples (and in some examples, is 1024) in an audio frame and N denotes the highest order of the basis function to which the coefficients correspond. As noted above, N is commonly defined as four (4) for a total of 1024X25 coefficients. Each of the SHC 701 corresponding to a particular order, the suborder combination can be referred to as a channel. For example, all M sample coefficients corresponding to a first-order, zero-order base function can represent one channel, while the coefficients corresponding to zero-order, zero-suborder base function can represent another channel, etc. . SHC 701 may also be referred to in the present disclosure as higher order ambisonics content (HO A) 701 or as a SH 701 signal.

[1310] As shown in the example of Figure 51, the audio encoding device 700A includes an parsing unit 702, a vector-based synthesis unit 704, a vector reduction unit 706, an encoding unit 446/500 pseudoacoustics 708, a coefficient reduction unit 710 and a compression unit 712 ("length unit 712"). Analysis unit 702 may represent a unit configured to perform analysis against SHC 701 in order to identify distinct sound field components (D) 703 and a total number of background components (BGTOT) 705. Compared to devices of audio coding described above, the audio coding device 700A does not make this determination in relation to the decompositions of the SHC 701, but directly in relation to the SHC 701.

e uma matriz [V] 709 que tem um tamanho de

. A matriz [U.S.] 707 pode representar uma matriz que resulta de uma multiplicação de matriz da matriz [U] e da matriz [S] gerada através de uma aplicação de SVD aos SHC 701.[1311] Vector-based synthesis unit 704 represents a unit configured to perform some form of vector-based synthesis with respect to SHC 701, e.g. SVD, KLT, PCA or any other vector-based synthesis, to generate, in the SVD examples, an array [US] 707 that has a size of

and an array [V] 709 that has a size of

. The matrix [US] 707 can represent a matrix that results from a matrix multiplication of the matrix [U] and the matrix [S] generated by applying SVD to the SHC 701.

Embora não mostrada para propósitos de facilitar ilustração, a interpolação da matriz [V] 709 pode ocorrer antes da redução da matriz [V] 709 de maneira semelhante àquela descrita mais detalhadamente acima. Ademais, embora não mostrado para propósitos de facilitar ilustração, reordenação da matriz [U.S.] reduzida 711 e/ou da matriz [V] reduzida 712 da maneira descrita mais detalhadamente acima. Consequentemente, as técnicas não devem ser limitadas a esses aspectos (tais como, projeção de erro ou qualquer outro aspecto das técnicas supracitadas descritas acima, porém não mostrados no exemplo da Figura 51). A unidade de codificação psicoacústica 708 representa uma unidade configurada para realizar codificação pseudoacústica em relação à matriz [U.S.] 711 para gerar um fluxo de bits 715. A unidade de redução de coeficiente 710 pode representar uma unidade configurada para reduzir o número de canais da matriz [V] reduzida 713. Em outras palavras, a unidade de redução de coeficiente 710 pode representar uma unidade configurada para eliminar aqueles coeficientes dos vetores V distintos (que formam a matriz [V] reduzida 713) que têm poucas informações direcionais ou não têm informações direcionais. Conforme descrito acima, em alguns exemplos, aqueles coeficientes dos vetores V distintos correspondentes a uma primeira função e a uma função de base de ordem zero (denotada como NBG acima) fornecem poucas informações direcionais, portanto, podem ser removidos dos vetores V distintos (através do quê é denominado de "redução de ordem” acima). Nesse exemplo, uma melhor flexibilidade pode ser fornecida não apenas para identificar esses coeficientes que correspondem a NBG, porém para identificar canais de HOA adicionais (que podem ser denotados pelo 448/500 TotalOfAddAmbHOAChan variável) a partir de conjunto de [(NBG +1)2+1, (N+1)2]. A unidade de análise 702 pode analisar os SHC 701 para determinar BGTOT, o que pode identificar não apenas

porém o TotalOfAddAmbHOAChan. A unidade de redução de coeficiente 710 pode, então, remover esses coeficientes correspondentes ao (NBG+1)2 e ao TotalOfAddAmbHOAChan a partir da matriz [V] reduzida 713 para gerar uma matriz [V] pequena 717 do tamanho

[1312] The vector reduction unit 706 may represent a unit configured to reduce the number of vectors in the matrix [US] 707 and matrix [V] 709 so that each of the remaining vectors in the matrix [US] 707 and the matrix matrix [V] 709 identifies a distinct or prominent component of the sound field. The vector reduction unit 706 may perform this based on the number of distinct components D 703. The number of distinct components D 703 may, in effect, represent a matrix of numbers, where each number identifies different distinct vectors from the matrices 707 and 709. The vector reduction unit 706 can output a reduced [US] matrix 711 of size M x D and 447/500 a reduced [V] matrix 713 of size

Although not shown for purposes of ease of illustration, matrix [V] 709 interpolation can occur before matrix [V] 709 reduction in a manner similar to that described in more detail above. Further, although not shown for purposes of ease of illustration, reordering of the reduced [US] matrix 711 and/or the reduced [V] matrix 712 in the manner described in more detail above. Consequently, the techniques should not be limited to these aspects (such as error projection or any other aspect of the aforementioned techniques described above, but not shown in the example in Figure 51). The psychoacoustic encoding unit 708 represents a unit configured to perform pseudoacoustic encoding against the [US] array 711 to generate a bit stream 715. The coefficient reduction unit 710 may represent a unit configured to reduce the number of channels in the array. Reduced [V] 713. In other words, the coefficient reduction unit 710 may represent a unit configured to eliminate those coefficients from the distinct V vectors (which form the reduced [V] matrix 713) that have little or no directional information. directional. As described above, in some examples, those coefficients of the distinct V vectors corresponding to a first function and a zero-order base function (denoted as NBG above) provide little directional information, so they can be removed from the distinct V vectors (via of what is termed "order reduction" above.) In this example, better flexibility can be provided not only to identify those coefficients that correspond to NBG, but to identify additional HOA channels (which can be denoted by the 448/500 TotalOfAddAmbHOAChan variable) from the set of [(NBG+1)2+1, (N+1)2]. The parsing unit 702 can analyze the SHC 701 to determine BGTOT, which can identify not only

however the TotalOfAddAmbHOAChan. The coefficient reduction unit 710 can then remove those coefficients corresponding to (NBG+1)2 and TotalOfAddAmbHOAChan from the reduced [V] matrix 713 to generate a small [V] matrix 717 of size

[1313] The compression unit 712 can then perform the scalar quantization check above and/or Huffman coding to compress the small [V] matrix 717, output the compressed small [V] matrix 717 as auxiliary channel information 719 (“aux channel info 719”). The compression unit 712 can output the auxiliary channel information 719 in a manner similar to that shown in the example of Figures 10 to 100(ii). In some examples, a bitstream generating unit similar to that described above may incorporate auxiliary channel information 719 into bitstream 715. Furthermore, although referred to as bitstream 715, audio encoding device 700A may include, as noted above, a background component processing path that results in another bitstream, wherein a bitstream generating unit similar to those described above can generate a bitstream similar to the bitstream 17 described above that includes bitstream 715 and bitstream output through the background component processing path.

[1314] In accordance with the techniques described in the present disclosure, the analysis unit 702 may be configured to determine a first non-zero set of coefficients of a vector, i.e., the 449/500 vectors of the reduced [V] matrix 713 in this example, to be used to represent the distinct component of the sound field. In some examples, the parsing unit 702 may determine that all coefficients of each vector forming the reduced [V] matrix 713 should be included in the auxiliary channel information 719. Therefore, the parsing unit 702 may set the BGTOT equal to zero.

[1315] Then, the audio encoding device 700A can effectively act reciprocally to that described in relation to the table denoted as "decoded vectors". In addition, the audio encoding device 700A can specify a syntax element in a header of an access unit (which may include one or more frames from among a plurality of frames) which among a plurality of configuration modes has been selected. Although described as being specified on a per-access unit basis, parsing unit 702 may specify this syntax element on a per-frame basis in any other periodic basis or non-periodic basis (such as once for the entire bit stream). configuration modes were selected to specify the non-zero set of coefficients of the reduced [V] matrix 713 to represent the directional aspects of this c distinct component. The syntax element can be denoted as "codedVVecLength". In this way, the 700A audio encoding device can signal or otherwise specify in the bitstream which of the four configuration modes was used to specify the small array [ V] 717 in the bitstream. Although described in relation to four configuration modes, the techniques should not be limited to four configuration modes, but to any 450/500 number of configuration modes, including a single configuration mode or a plurality of configuration modes.

[1316] Therefore, various aspects of the techniques may enable the 700A audio encoding device to be configured to operate in accordance with the following clauses.

[1317] Clause 133149-1F. A device comprising: one or more processors configured to select one or more of a plurality of configuration modes by which a non-zero set of coefficients of a vector is to be specified, the vector being decomposed from the plurality of spherical harmonic coefficients describing a sound field and representing a distinct component of the sound field, and specifying the non-zero set of vector coefficients based on the mode selected from among the plurality of configuration modes.

[1318] Clause 133149-2F. The provision of clause 133149-1F, wherein the mode among the plurality of configuration modes indicates that the non-zero set of coefficients includes all coefficients.

[1319] Clause 133149-3F. The provision of clause 133149-1F, wherein the mode among the plurality of configuration modes indicates that the non-zero set of coefficients includes those among the coefficients corresponding to an order greater than an order of a basis function to which a or more of the plurality of spherical harmonic coefficients correspond.

[1320] Clause 133149-4F. The provision of clause 133149-1F, wherein the mode among the plurality of configuration modes indicates that the non-zero set of coefficients includes those among the coefficients corresponding to an order greater than an order of a base function to which one of the plurality of spherical harmonic coefficients corresponds to and excludes at least one of the corresponding coefficients at an order greater than the order of the basis function to which the one or more of the plurality of spherical harmonic coefficients correspond.

[1321] Clause 133149-5F. The provision of clause 133149-1F, wherein the mode among the plurality of configuration modes indicates that the non-zero set of coefficients includes all coefficients with the exception of at least one of the coefficients.

[1322] Clause 133149-6F. The device of clause 133149-1F, wherein the one or more processors are further configured to specify the mode selected from among the plurality of configuration modes in a bit stream.

[1323] Clause 133149-1G. A device comprising: one or more processors configured to determine one of a plurality of configuration modes by which to extract a non-zero set of coefficients from a vector in accordance with one of a plurality of configuration modes, wherein the vector was decomposed from plurality of spherical harmonic coefficients that describe a sound field and that represent a distinct component of the sound field, and extracted the non-zero set of vector coefficients based on the mode obtained from among the plurality of configuration modes .

[1324] Clause 133149-2G. The provision of clause 133149-1G, wherein the mode among the plurality of configuration modes indicates that the non-zero set of coefficients includes all coefficients.

[1325] Clause 133149-3G. The provision of clause 133149-1G, wherein the mode among the plurality of configuration modes indicates that the non-zero set of coefficients includes those among the coefficients corresponding to an order greater than an order of a basis function to which one or more of the plurality of spherical harmonic coefficients correspond.

[1326] Clause 133149-4G. The provision of clause 133149-1F, wherein the mode among the plurality of configuration modes indicates that the non-zero set of coefficients includes those among the coefficients corresponding to an order greater than an order of a basis function to which a among the plurality of spherical harmonic coefficients corresponds to and excludes at least one of the coefficients corresponding to an order greater than the order of the basis function to which the one or more of the plurality of spherical harmonic coefficients correspond.

[1327] Clause 133149-5G. The provision of clause 133149-1G, wherein the mode among the plurality of configuration modes indicates that the non-zero set of coefficients includes all coefficients with the exception of at least one of the coefficients.

[1328] Clause 133149-6G. The device of clause 133149-1G, wherein the one or more processors are further configured to determine, during mode determination among the plurality of configuration modes, the mode among the plurality of configuration modes based on a signaled value in a stream of bits.

[1329] Figure 52 is a block diagram illustrating another example of an audio decoding device 750A that can implement various aspects of the techniques described in the present disclosure in order to reconstruct, or quasi-reconstruct, the SHC 701s. In the example of Figure 52, the audio decoding device 750A is similar to the audio decoding device 540D shown in the example of Figure 4 ID, except that the extraction unit 542 receives bitstream 715' (which is similar to the bitstream 715 described above with respect to the example in Figure 51, except that bitstream 715' also includes audio encoded version of SHCBG 752) and auxiliary channel information 719. For this reason, the extraction unit is denoted as "extraction unit 542'".

[1330] Furthermore, extraction unit 542' differs from extraction unit 542 in that extraction unit 542' includes a modified form of V decompression unit 555 (which is shown as "V 555 decompression unit"' in the example of Figure 52.) The decompression unit V 555' receives the auxiliary channel information 719 and the syntax element denoted codedVVecLength 754. The extraction unit 542' parses the codedVVecLength 754 from the bit stream 715' (and , in one example, from the header access unit included in the bit stream 715'). analysis module 758 configurable to operate in any one of the configuration modes described above 760.

[1331] Mode configuration unit 756 receives syntax element 754 and selects one of the 760 configuration modes. Mode configuration unit 756 then configures parsing unit 758 with the mode selected from among the modes of configuration 760. Parsing unit 758 represents a unit configured to operate in accordance with any of the 454/500 modes of configuration 760 to parse a compressed form of small vectors [V] 717 of auxiliary channel information 719. The unit 758 analyzer can operate in accordance with the switching command shown in the following table. TABLE — DECODED VECTORS

[1332] In the syntax table above, the first switching command with the four cases (case 0 to 3) provides a way to determine the lengths of each vector of the small [V] matrix 717 in terms of the number of coefficients. The first case, case 0, indicates that all coefficients for the VTDIST vectors are specified. The second case, case 1, indicates that only those coefficients of the VTDIST vector corresponding to an order greater than a MinNumOfCoeffsForAmbHOA are specified, which can denote what is denoted as (NDIST +1) - (NBG + 1) above. The third case, case 2, is similar to the second case, but it additionally subtracts coefficients identified by NumOfAddAmbHoaChan, which denotes a variable to specify additional channels (where "channels" indicates a specific coefficient corresponding to a given combination of order, suborder ) corresponding to an order that exceeds the NBG order The fourth case, case 3, indicates that only those coefficients from the VTDIST vector left after removing the coefficients identified by NumOfAddAmbHoaChan are specified.

[1333] In this regard, the audio decoding device 750A may operate in accordance with the techniques described in the present disclosure to determine a first non-zero set of coefficients from a vector representing a distinct component of the sound field, the vector being was decomposed from a plurality of spherical harmonic coefficients that describe a sound field.

[1334] In addition, the audio decoding device 457/500 750A may be configured to operate in accordance with the techniques described in the present disclosure to determine one of a plurality of configuration modes through which a non-zero set is to be extracted. of coefficients of a vector in accordance with a plurality of configuration modes, the vector being decomposed from a plurality of spherical harmonic coefficients that describe a sound field and represent a distinct component of the sound field, and the non-zero set extracted of the vector coefficients based on the mode obtained from among the plurality of configuration modes.

[1335] Fig. 53 is a block diagram illustrating another example of an audio encoding device 570 that can perform various aspects of the techniques described in the present disclosure. In the example of Figure 53, the audio encoding device 570 may resemble one or more of the audio encoding devices 510A to 510J (where it is assumed that the order reduction unit 528A is to be included in the sound field component extraction 20, but not shown for purposes of illustration). However, audio encoding device 570 may include a more general transform unit 572 which may comprise decomposition unit 518 in some examples.

[1336] Figure 54 is a block diagram illustrating in more detail an exemplary implementation of the audio encoding device 570 shown in the example of Figure 53. As illustrated in the example of Figure 54, the transform unit 572 of the audio encoding device 570 is shown in the example of Figure 53. audio encoding 570 includes a rotation unit 654. The 458/500 sound field component extraction unit 520 of the audio encoding device 570 includes a spatial analysis unit 650, a content feature analysis unit 652, a coherent component extraction unit 656, and a fuzzy component extraction unit 658. The audio encoding unit 514 of the audio encoding device 570 includes an AAC encoding mechanism 660 and an AAC encoding mechanism 162. bit stream generation unit 516 of audio coding device 570 includes a multiplexer (MUX) 164.

[1337] The bandwidth - in terms of bits/second - required to represent 3D audio data in the form of SHC can make mesos prohibitive in terms of consumer usage. For example, when using a sampling rate of 48 kHz, and with the same resolution of 32 bits - a fourth order SHC representation represents a bandwidth of 36 Mbits/second (25x48000x32 bps). When compared to prior art audio coding for stereo signals, which is typically around 100 kbits/second, this is quite a figure. The techniques deployed in the example of Figure 54 can reduce the bandwidth of 3D audio representations.

[1338] The spatial analysis unit 650, the content feature analysis unit 652 and the rotation unit 654 may receive the SHC 511. As described in another paragraph in the present disclosure, the SHC 511 may be representative of a sound field . In the example of Figure 54, the spatial analysis unit 650, the content feature analysis unit 652 and the rotation unit 654 can receive twenty-five SHC for a fourth order fourth (n=4) representation of the sound field. 459/500

[1339] The spatial analysis unit 650 can analyze the sound field represented by the SHC 511 to identify distinct sound field components and diffuse sound field components. Distinct components of the sound field are sounds that are perceived to originate from an identifiable direction or that are otherwise distinct from background or diffuse components of the sound field. For example, the sound generated by an individual musical instrument may be perceived as originating from an identifiable direction. In contrast, diffuse background components of the sound field are perceived to originate from an identifiable direction. For example, the sound of wind blowing through a forest can be a diffuse component of a sound field.

[1340] The spatial analysis unit 650 can identify one or more distinct components that attempt to identify an ideal angle through which the sound field should be rotated in order to align those distinct components that have the most energy with the vertical geometric axes and/or horizontal (relative to an assumed microphone that recorded this sound field). The spatial analysis unit 650 can identify this ideal angle so that the sound field can be rotated so that these distinct components better align with the underlying spherical base functions shown in the examples in Figures 1 and 2.

[1341] In some examples, the spatial analysis unit 650 may represent a unit configured to perform a form of diffusion analysis to identify a percentage of the sound field represented by the SHC 511 that includes diffuse sounds (which may refer to sounds that have low steering levels or lower-order SHC, meaning those SHC 511 460/500 that have an order less than or equal to one). As an example, a spatial analysis unit 650 can perform diffusion analysis in a manner similar to that described in a dissertation by Ville Pulkki entitled "Spatial Sound Reproduction with Directional Audio Coding", published in J. Audio Eng. Soc, Volume 55, #6, dated June 2007. In some examples, the spatial analysis unit 650 can analyze only a non-zero subset of the HO A coefficients, for example, the zero and first order coefficients among the SHC 511, during the performing diffusion analysis to determine the percentage of diffusion.

[1342] The content characteristic analysis unit 652 can determine, based at least partially on the SHC 511, whether the SHC 511 was generated through a natural recording of a sound field or produced artificially (i.e. synthetically), for example from an audio object, for example a PCM object. Additionally, the content characteristic analysis unit 652 can then determine, based at least partially on whether the SHC 511's were generated through an actual recording of a sound field or from an artificial audio object. , the total number of channels to be included in the bitstream 517. For example, the content feature analysis unit 652 can determine, based at least partially on whether the SHC 511 was generated from recording an actual field sound or from an artificial audio object, that bitstream 517 must be included in sixteen channels. Each of the channels can be a mono channel. The content feature analysis unit 652 may further perform the determination of the total number of channels to be included in the 461/500 bit stream 517 based on an output bit rate of the bit stream 517, e.g. 1, 2 Mbps.

[1343] In addition, the content feature analysis unit 652 can determine, based at least partially on whether the SHC 511s were generated from a recording of an actual sound field or from an artificial audio object, how many of the channels must be allocated to coherent components or, in other words, to distinct components of the sound field and how many of the channels must be allocated to diffuse components, or in other words, background components of the sound field. For example, when SHC 511s were generated from recording an actual sound field using, for example, an Eigenmic, the content characteristic analysis unit 652 can allocate three of the channels to coherent components of the sound field. and can allocate the remaining channels to diffuse components of the sound field. In this example, when the SHC 511 were generated from an artificial audio object, the content characteristic analysis unit 652 can allocate five of the channels to the coherent components of the sound field and can allocate the remaining channels to the diffuse components of the field. sound. In this way, the content analysis block (i.e., content feature analysis unit 652) can determine the type of sound field (e.g., diffuse/directional, etc.) and, in turn, determine the number of coherent/fuzzy to be extracted.

[1344] The target bit rate can influence the number of components and the bit rate of individual AAC encoding engines (eg AAC 660, 662 encoding engines). In other words, the content feature analysis unit 652 may 462/500 additionally perform the determination of how many channels should be allocated to coherent components and how many should be allocated to fuzzy components based on an output bit rate of the stream. 517 bits, for example 1.2 Mbps.

[1345] As noted above, the content feature analysis unit 652 can determine whether the SHC 511 was generated from a recording of an actual sound field or from an artificial audio object. The content characteristic analysis unit 652 can make this determination in several ways. For example, audio encoding device 570 can use 4th order SHC. In this example, the content feature analysis unit 652 can encode 24 channels and predict a 25th channel (which can be represented as a vector). The content feature analysis unit 652 can apply scalars to at least some of the 24 channels and sum the resulting values to determine the 25th vector. Additionally, in that example, the content feature analysis unit 652 can determine an accuracy of the predicted 25th channel. In this example, if the accuracy of the 25th channel is relatively high (eg accuracy exceeds a particular threshold), SHC 511s are likely to be generated from a synthetic audio object. In contrast, if the accuracy of the predicted 25th channel is relatively low (eg, the accuracy is below the particular threshold), SHC 511s are more likely to represent a recorded sound field. In this example, if the signal-to-noise ratio (SNR) of the 25th channel is greater than 100 decibels (dbs), the SHC 511s are more likely to represent a sound field generated from a synthetic audio object. In contrast, the SNR of a sound field recorded using an eigen microphone can be 5 to 20 dBs. Thus, there may be an apparent demarcation in SNR relationships between the sound field represented by the SHC 511 generated from a real direct recording and from a synthetic audio object.

[1346] Additionally, the content feature analysis unit 652 can select, based at least partially on whether the SHC 511s were generated from a recording of an actual sound field or from an artificial audio object, books of codes to quantize the vector V. In other words, the content feature analysis unit 652 can select different codebooks for use in quantizing the vector V, depending on whether the sound field represented by the HO coefficients A is recorded or synthetic.

[1347] In some instances, the content feature analysis unit 652 can determine, on a recurring basis, whether SHC 511s were generated from a recording of an actual sound field or from an artificial audio object. In some of these examples, the recurring basis may be every frame. In other examples, the content characteristic analysis unit 652 may perform this determine once. Additionally, the content feature analysis unit 652 can determine, on a recurring basis, the total number of channels and the allocation of coherent component channels and fuzzy component channels. In some of these examples, the recurring basis may be every frame. In other examples, the content characteristic analysis unit 652 may perform this determine once. In some examples, the content feature analysis unit 652 may select, on a recurring basis, codebooks for use in quantizing the vector V. In some of these examples, the recurring basis may be on a per-frame basis. In other examples, the content characteristic analysis unit 652 may perform this determine once.

[1348] Rotation unit 654 can perform a rotation operation of the HOA coefficients. As discussed in another paragraph in the present disclosure (e.g., with respect to Figures 55 and 55B), performing the rotation operation can reduce the number of bits required to represent the SHC 511s. In some examples, the rotation analysis performed by the Rotation unit 652 is an example of a single-value decomposition ("SVD") analysis. Principal component analysis (“PCA”), independent component analysis (“ICA”), and the Karhunen-Loeve Transform (“KLT”) are related techniques that may be applicable.

[1349] In the example of Fig. 54, the coherent component extraction unit 656 receives rotated SHCs 511 from the rotation unit 654. Additionally, the coherent component extraction unit 656 extracts from the rotated SHCs 511 the rotated SHCs 511 associated with the coherent components of the sound field.

[1350] In the example of Fig. 54, coherent component extraction unit 656 receives rotated SHC 511 from rotation unit 654. Additionally, coherent component extraction unit 656 extracts from the rotated SHC 511 the rotated SHC 511 associated with the coherent components of the sound field.

[1351] In addition, the coherent component extraction unit 656 generates one or more coherent component channels. Each of the coherent component channels may include a different subset of the rotated SHC 511 associated with the sound field coherent coefficients. In the example of Fig. 54 , the coherent component extraction unit 656 can generate from one to 16 coherent component channels. The number of coherent component channels generated by the coherent component extraction unit 656 can be determined by the number of channels allocated by the content characteristic analysis unit 652 to the coherent components of the sound field. The bit rates of the coherent component channels generated by the coherent component extraction unit 656 can be determined by the content feature analysis unit 652.

[1352] Similarly, in the example of Figure 54, fuzzy component extraction unit 658 receives rotated SHC 511 from rotation unit 654. Additionally, diffuse component extraction unit 658 extracts from the rotated SHC 511 the Rotated SHC 511 associated with the diffuse components of the sound field.

[1353] In addition, the fuzzy component extraction unit 658 generates one or more fuzzy component channels. Each of the diffuse component channels may include a different subset of the rotated SHC 511 associated with the diffuse sound field coefficients. In the example of Fig. 54, the fuzzy component extraction unit 658 can generate from one to 9 fuzzy component channels. The number of fuzzy component channels generated by the fuzzy component extraction unit 658 can be determined by the number of channels allocated by the content characteristic analysis unit 652 to the coherent components of the sound field. The bit rates of the fuzzy component channels generated by the fuzzy component extraction unit 658 can be determined by the content feature analysis unit 652.

[1354] In the example of Fig. 54, the ACC encoding unit 660 can use an AAC codec to encode the component coherent channels generated by the component coherent extraction unit 656. Similarly, the ACC encoding unit 662 can use an AAC codec to encode the fuzzy component channels generated by the fuzzy component extraction unit 658. The multiplexer 664 ("MUX 664") can multiplex the coded coherent component channels and the coded fuzzy component channels, together with auxiliary data (e.g., an ideal angle determined by the spatial analysis unit 650), to generate the bit stream 517.

[1355] In this way, the techniques can enable the audio encoding device 570 to determine whether representative spherical harmonic coefficients of a sound field are generated from a synthetic audio object.

[1356] In some examples, the audio encoding device 570 can determine, based on whether the spherical harmonic coefficients are generated from a synthetic audio object, a subset of the spherical harmonic coefficients representative of distinct components of the sound field . In this and other examples, the audio encoding device 570 may generate a bit stream to include the subset of the spherical harmonic coefficients. The audio encoding device 570 may, in some instances, audio encode the subset of the spherical harmonic coefficients and generate a bit stream to include the audio encoded subset of the spherical harmonic coefficients.

[1357] In some examples, the audio encoding device 570 may determine, based on whether the spherical harmonic coefficients are generated from a synthetic audio object, a subset of the spherical harmonic coefficients representative of background components of the sound field. In this and other examples, the audio encoding device 570 may generate a bit stream to include the subset of the spherical harmonic coefficients. In this and other examples, the audio encoding device 570 can audio encode the subset of the spherical harmonic coefficients, and generate a bit stream to include the audio encoded subset of the spherical harmonic coefficients.

[1358] In some examples, audio encoding devices 570 can perform a spatial analysis against the spherical harmonic coefficients to identify an angle through which the sound field represented by the spherical harmonic coefficients must be rotated and performed a rotation operation to rotating the sound field by the identified angle in order to generate rotated spherical harmonic coefficients.

[1359] In some examples, the audio encoding device 570 can determine, based on whether the spherical harmonic coefficients are generated from a synthetic audio object, a first subset of the spherical harmonic coefficients representative of distinct components of the field sound, and determine, on a 468/500 basis, whether the spherical harmonic coefficients are generated from a synthetic audio object, from a second subset of the spherical harmonic coefficients representative of background components of the sound field. In this and other examples, audio encoding devices 570 can audio encode the first subset of the spherical harmonic coefficients that have a higher target bit rate than that used to audio encode the second subset of the spherical harmonic coefficients.

[1360] In this way, various aspects of the techniques can enable the audio encoding device 570 to determine whether SCH 511s are generated from a synthetic audio object in accordance with the following clauses.

[1361] Clause 132512-1. A device, such as audio encoding device 570, comprising: wherein the one or more processors are further configured to determine whether representative spherical harmonic coefficients of a sound field are generated from a synthetic audio object.

[1362] Clause 132512-2. The device of clause 132512-1, wherein the one or more processors are additionally configured to, when determining whether representative spherical harmonic coefficients of the sound field are generated from the synthetic audio object, exclude a first vector from a matrix spherical harmonic coefficient with frame that stores at least a portion of the spherical harmonic coefficients representative of the sound field to obtain a reduced matrix of spherical harmonic coefficient with frame.

[1363] Clause 132512-3. The device of clause 132512-1, wherein the one or more processors are 469/500 additionally configured to, when determining whether representative spherical harmonic coefficients of the sound field are generated from the synthetic audio object, exclude a first vector of a framed spherical harmonic coefficient matrix that stores at least a portion of the spherical harmonic coefficients representative of the sound field to obtain a framed spherical harmonic coefficient reduced matrix, and predicting a vector of the framed spherical harmonic coefficient reduced matrix based on in remaining vectors of the reduced spherical harmonic coefficient matrix with frame.

[1364] Clause 132512-4. The device of clause 132512-1, wherein the one or more processors are additionally configured to, when determining whether representative spherical harmonic coefficients of the sound field are generated from the synthetic audio object, exclude a first vector from a matrix spherical harmonic coefficient with frame that stores at least a portion of the spherical harmonic coefficients representative of the sound field to obtain a reduced matrix of spherical harmonic coefficient with frame, and predict a vector of the reduced spherical harmonic coefficient with frame matrix based on at least least partially on a sum of the remaining vectors of the framed spherical harmonic coefficient reduced matrix.

[1365] Clause 132512-5. The device of clause 132512-1, wherein the one or more processors are additionally configured to, during the determination of whether representative spherical harmonic coefficients of the sound field are generated from the synthetic audio object, predict a vector from a matrix of framed spherical harmonic coefficient that stores at least a 470/500 portion of the spherical harmonic coefficients based, at least in part, on a sum of remaining vectors of the framed spherical harmonic coefficient matrix.

[1366] Clause 132512-6. The device of clause 132512-1, wherein the one or more processors are additionally configured to, during the determination of whether representative spherical harmonic coefficients of the sound field are generated from the synthetic audio object, predict a vector from a matrix of framed spherical harmonic coefficient that stores at least a portion of the spherical harmonic coefficients based at least partially on a sum of remaining vectors of the framed spherical harmonic coefficient matrix, and computing an error based on the predicted vector.

[1367] Clause 132512-7. The device of clause 132512-1, wherein the one or more processors are additionally configured to, when determining whether representative spherical harmonic coefficients of the sound field are generated from the synthetic audio object, predicts a vector from a matrix of framed spherical harmonic coefficient that stores at least a portion of the spherical harmonic coefficients based at least partially on a sum of remaining vectors of the framed spherical harmonic coefficient matrix, and computing an error based on the predicted vector and the corresponding vector of spherical harmonic coefficient matrix with frame.

[1368] Clause 132512-8. The device of clause 132512-1, wherein the one or more processors are additionally configured to, during the determination of whether representative spherical harmonic coefficients of the sound field are generated from the synthetic audio object, predict a vector from a matrix of 471/500 spherical harmonic coefficient with frame that stores at least a portion of the spherical harmonic coefficients based at least partially on a sum of remaining vectors of the spherical harmonic coefficient matrix with frame, and computer an error with a sum of the absolute value of the difference of the predicted vector and the corresponding vector of the spherical harmonic coefficient matrix with frame.

[1369] Clause 132512-9. The device of clause 132512-1, wherein the one or more processors are further configured to predict, during the determination of whether representative spherical harmonic coefficients of the sound field are generated from the synthetic audio object, a vector of a matrix of framed spherical harmonic coefficient that stores at least a portion of the spherical harmonic coefficients based at least partially on a sum of remaining vectors of the framed spherical harmonic coefficient matrix, compute an error based on the predicted vector and the corresponding vector of the framed spherical harmonic coefficient matrix, compute a ratio based on a corresponding vector energy of the framed spherical harmonic coefficient matrix and on error, and compare the ratio to a threshold to determine whether representative spherical harmonic coefficients of the sound field are generated from synthetic audio object.

[1370] Clause 132512-10. The device of any one of claims 4 to 9, wherein the one or more processors are further configured to predict, during vector prediction, a first non-zero vector of the framed spherical harmonic coefficient matrix that stores at least the portion of the spherical harmonic coefficients.

[1371] Clause 132512-11. The 472/500 device of any one of claims 1 to 10, wherein the one or more processors are further configured to specify an indication of whether spherical harmonic coefficients are generated from the synthetic audio object in a bitstream that stores a compressed version of the spherical harmonic coefficients.

[1372] Clause 132512-12. The device of clause 132512-11, where the indication is a single bit.

[1373] Clause 132512-13. The device of clause 132512-1, wherein the one or more processors are additionally configured to determine, based on whether the spherical harmonic coefficients are generated from a synthetic audio object, a subset of the spherical harmonic coefficients representative of distinct components of the sound field.

[1374] Clause 132512-14. The device of clause 132512-13, wherein the one or more processors are further configured to generate a bit stream to include the subset of the spherical harmonic coefficients.

[1375] Clause 132512-15. The device of clause 132512-13, wherein the one or more processors are further configured to audio encode the subset of the spherical harmonic coefficients, and generate a bit stream to include the audio encoded subset of the spherical harmonic coefficients.

[1376] Clause 132512-16. The provision of clause 132512-1, wherein the one or more processors are additionally configured to determine, based on whether the spherical harmonic coefficients are generated from a synthetic audio object, a subset of the spherical harmonic coefficients representative of components of background of the sound field. 473/500

[1377] Clause 132512-17. The device of clause 132512-16, wherein the one or more processors are further configured to generate a bit stream to include the subset of the spherical harmonic coefficients.

[1378] Clause 132512-18. The device of clause 132512-15, wherein the one or more processors are further configured to audio encode the subset of the spherical harmonic coefficients, and generate a bit stream to include the audio encoded subset of the spherical harmonic coefficients.

[1379] Clause 132512-18. The device of clause 132512-1, wherein the one or more processors are additionally configured to perform spatial analysis in relation to spherical harmonic coefficients to identify an angle through which the sound field represented by the spherical harmonic coefficients must be rotated, and perform an operation of rotation to rotate the sound field by the identified angle to generate rotated spherical harmonic coefficients.

[1380] Clause 132512-20. The device of clause 132512-1, wherein the one or more processors are further configured to determine, based on whether the spherical harmonic coefficients are generated from a synthetic audio object, a first subset of the spherical harmonic coefficients representative of components distinct from the sound field, and to determine, based on whether the spherical harmonic coefficients are generated from a synthetic audio object, a second subset of the spherical harmonic coefficients representative of background components of the sound field.

[1381] Clause 132512-21. The device of 474/500 clause 132512-20, wherein the one or more processors are additionally configured to audio encode the first subset of spherical harmonic coefficients that have a higher target bit rate than that used to audio encode the second subset of spherical harmonic coefficients.

[1382] Clause 132512-22. The device of clause 132512-1, wherein the one or more processors are further configured to perform a single-valued decomposition with respect to spherical harmonic coefficients to generate a matrix U representative of singular left vectors of the plurality of spherical harmonic coefficients, a matrix S representative of singular values of the plurality of spherical harmonic coefficients and a matrix V representative of singular straight vectors of the plurality of spherical harmonic coefficients.

[1383] Clause 132512-23. The device of clause 132512-22, wherein the one or more processors are further configured to determine, based on whether spherical harmonic coefficients are generated from a synthetic audio object, those portions of one or more of the matrix U , the matrix S and the matrix V representative of distinct components of the sound field.

[1384] Clause 132512-24. The device of clause 132512-22, wherein the one or more processors are further configured to determine, based on whether spherical harmonic coefficients are generated from a synthetic audio object, those portions of one or more of the matrix U , the matrix S and the matrix V representing background components of the sound field.

[1385] Clause 132512-1C. A device, such as the audio encoding device 570, comprising: one or more processors configured to determine whether representative spherical harmonic coefficients of a sound field are generated from a synthetic audio object based on a ratio computed as a function of at least an energy of a vector of spherical harmonic coefficients and an error derived based on a predicted version of the vector of spherical harmonic coefficients and the vector of spherical harmonic coefficients.

[1386] In each of the various examples described above, it is to be understood that the audio encoding device 570 may perform a method or otherwise comprises means for performing each step of the method that the audio decoding device 570 is configured to perform, in some instances, such means may comprise one or more processors. In some examples, the one or more processors may represent a special-purpose processor configured through instructions stored on non-transient computer-readable storage media. In other words, various aspects of the techniques in each of these sets of encoding examples can provide a non-transient computer-readable storage medium that has instructions stored on it that, when executed, cause the one or more processors to perform the method. that the audio encoding device 570 has been configured to perform.

[1387] Figures 55 and 55B are diagrams illustrating an exemplary embodiment of various aspects of the techniques described in the present disclosure for rotating a sound field 640. Figure 55 is a diagram illustrating a sound field 640 prior to rotation in accordance with the various aspects of the techniques described in the present disclosure. In the example of Figure 55, sound field 640 includes two high pressure locations, denoted locations 642A and 642B. These locations 642A and 642B ("locations 642") lie along a line 644 that has a non-zero slope (which is another way of calling a line that is not horizontal, since horizontal lines have a slope equal to zero). Since locations 642 have a z coordinate in addition to the x and y coordinates, higher order spherical basis functions may be necessary to correctly represent that sound field 640 (since these higher order spherical basis functions describe the upper non-horizontal portions Instead of reducing the sound field 640 directly to the SHCs 511, the audio encoding device 570 may rotate the sound field 640 until the line 644 connecting the locations 642 is horizontal.

[1388] Figure 55B is a diagram illustrating the sound field 640 after being rotated until the line connecting the locations 642 is horizontal. As a result of rotating the sound field 640 in this way, the SHC 511 can be derived so that the higher order coefficients within the SHC 511 are specified as zero, since the rotated sound field 640 no longer has any pressure locations (or energy) with z coordinates. In this way, audio encoding device 570 can rotate, translate, or more generally adjust sound field 640 so as to reduce the number of SHC 511 that have non-zero values. In combination with various other aspects of the techniques, the audio encoding device 570 may then, instead of signaling a 32-bit signaled number that identifies that these higher-order coefficients among the SHC 511 have zero values, signal in a field of the bit stream 517 that these higher order coefficients among the SHC 511 are not signaled. Audio encoding device 570 may also specify rotation information in bitstream 517 which indicates how often sound field 640 has been rotated, for example expressing an azimuth and elevation in the manner described above. An extractor device, such as the audio encoding device, can then imply that these unsigned coefficients within the SHC 511 have a value of zero and perform, during reproduction of the sound field 640 based on SHC 511, rotation in order to rotate the sound field 640 so that the sound field 640 resembles the sound field 640 shown in the example of Figure 55. In this way, the audio encoding device 570 can reduce the number of SHC 511 whose specification is required in bitstream 517, in accordance with the techniques described in the present disclosure.

[1389] A 'spatial compression' algorithm can be used to determine the optimal rotation of the sound field. In one embodiment, the audio encoding device 570 may apply the algorithm to iterate through all possible combinations of azimuth and elevation (i.e., 1024x512 combinations in the example above), rotating the sound field for each combination and calculating the number of SHC 511s that are above the threshold value. The candidate azimuth/elevation combination that produces the lowest number of SHC 511s above the threshold value can be considered as what might be termed the “optimal rotation.” In this rotated form, the sound field may require the lowest number of SHC 511s. to represent the sound field and can then be considered compressed. In some examples, the tuning may comprise this ideal rotation, and the tuning information described above may include this rotation information (which may be referred to as "optimal rotation" ) (in terms of azimuthal and elevation angles).

[1390] In some examples, instead of specifying just the azimuthal angle and elevation angle, the audio encoding device 570 can specify additional angles in the form of, for example, Euler angles. Euler angles specify the angle of rotation about the z-axis, the anterior x-axis, and the anterior z-axis. While described in the present disclosure with respect to combinations of azimuthal and elevation angles, the techniques of the present disclosure should not be limited to specifying only azimuthal and elevation angles, but may include specifying any number of angles, including all three angles. of Euler verified above. In that sense, the audio coding device 570 can rotate the sound field so as to reduce a number of the plurality of hierarchical elements that provide relevant information in the description of the sound field and specify Euler angles as rotation information in the bit stream. Euler angles, as seen above, can describe how the sound field has been rotated. During the use of Euler angles, the bit stream extractor device can analyze the bit stream to determine rotation information that includes Euler angles, and during reproduction of the sound field based on the elements among the plurality of hierarchical elements which provide relevant information in the description of the sound field, the rotation of the sound field based on Euler angles.

[1391] Furthermore, in some instances, instead of 479/500 explicitly specifying these angles in the bitstream 517, the audio encoding device 570 may specify an index (which may be called the "rotation index") associated with predefined combinations of the one or more angles that specify the rotation. In other words, the rotation information may, in some examples, include the rotation index. In these examples, a certain value of the rotation index, such as a value of zero, may indicate that no rotation was performed. This rotation index can be used in relation to a rotation table. That is, the audio encoding device 570 may include a rotation table that comprises an entry for each of the angle combinations azimuth and elevation angle.

[1392] Alternatively, a rotation table may include an entry for each matrix transform representative of each combination of azimuth angle and elevation angle. That is, the audio encoding device 570 can store a rotation table that has an entry for each matrix transformation to rotate the sound field by each of the combinations of azimuth and elevation angles. Typically, audio encoding device 570 receives SHC 511 and derives SHC 511', when rotation is performed, according to the following equation:

[1393] In the above equation, the SHC 511' are computed as a function of an encoding matrix to encode a sound field in terms of a second frame of reference {EncMat2), an inversion matrix to revert the SHC 511 back to a sound field in terms of a first reference frame (InvMat1), and SHC 511. 480/500 EncMat2 has a size of 25x32, whereas InvMat2 has a size of 32x25. Both SHC 511' and SHC 511 are size 25, where SHC 511' can be further reduced by removing those that do not specify audio protruding information. EncMat2 can vary for each combination of azimuth and elevation angle, whereas InvMat1 can remain static for each combination of azimuth and elevation angle. The rotation table can include an entry that results from multiplying each EncMat2 other than InvMat1.

[1394] Figure 56 is a diagram illustrating an example sound field captured in accordance with a first frame of reference which is then rotated in accordance with the techniques described in the present disclosure to express the sound field in terms of a second frame of reference. In the example of Figure 56, the sound field surrounding an Eigen 646 microphone is captured assuming a first frame of reference, which is denoted by the geometric axes X1, Y1, and in the example of Figure 56. The SHC 511s describe the sound field in terms this first frame of reference. The InvMat1 transforms the SHC 511s back into the sound field, which allows the sound field to be rotated relative to the second frame of reference denoted by the geometric axes X2, Y2, and Z2 in the example in Figure 56. The EncMat2 described above can rotate the sound field and generate the SHC 511' that describe that rotated sound field in terms of the second frame of reference.

[1395] Anyway, the above equation can be derived as follows. Since the sound field is recorded with a certain coordinate system, so that the front is considered as the x-axis direction, the 32 microphone positions of an Eigen microphone (or other microphone configurations) are defined 481/500 from that coordinate system reference. Then the rotation of the sound field can be thought of as a rotation of that frame of reference. For the assumed frame of reference, the SHC 511 can be calculated as follows:

[1396] In the above equation, o represents the spherical basis functions at the position (Posj) of the ith microphone (where i can be 1 to 32 in this example). The vector mici denotes the microphone signal to the ith microphone during a time t. The positions (Posi) indicate the position of the microphone in the first reference frame (that is, the reference frame before the rotation in this example).

[1397] The above equation can alternatively be expressed in terms of the mathematical expressions denoted above as:

[1398] In order to rotate the sound field (or in the second reference frame), the position (Posi) is calculated in the second reference frame. As long as the original microphone signals are present, the sound field can be rotated arbitrarily. However, the original microphone signals (mici(t)) are often not available. The problem then may be how to retrieve the microphone signals (mici(t)) from the SHC 511. If a T-model is used (coo on an Eigen microphone 32 microphone), the solution to this problem can be achieved by solving the following equation:

conforme verificado acima.[1399] This InvMat1 can specify the spherical harmonic base functions computed according to the position of the microphones as specified with respect to the first reference frame. This equation can also be expressed as

as verified above.

A EncMat2 especifica as funções de base harmônica esférica de uma posição girada (Posi'). Dessa maneira, a EncMat2 pode especificar de manei eficaz uma combinação do ângulo azimutal e de elevação. Desse modo, quando a tabela de rotação armazena o resultado de

para cada combinação dos ângulos azimutal e de elevação, a tabela de rotação especifica, de maneira eficaz, cada combinação dos ângulos azimutal e de elevação. A equação acima também pode ser expressa como:

[1400] Once the microphone signals (mici(t)) are retrieved in accordance with the above equation, the microphone signals (mici(t)) describing the sound field can be rotated to compute the corresponding SHC 511' to the second frame of reference, which results in the following equation:

EncMat2 specifies the spherical harmonic basis functions of a rotated position (Posi'). In this way, EncMat2 can effectively specify a combination of azimuthal and elevation angle. Thus, when the rotation table stores the result of

for each combination of azimuth and elevation angles, the rotation table effectively specifies each combination of azimuth and elevation angles. The above equation can also be expressed as:

ao passo que EncMat2 pode corresponder a

[1401] where 2, 2 represents a second azimuthal angle and a second elevation angle different 483/500 from the first azimuth angle and elevation angle represented by 1, 1. The 1, 1 corresponds to the first frame of reference whereas the 2, 2 correspond to the second frame of reference. Therefore, InvMat1 can correspond to

whereas EncMat2 can correspond to

[1402] The above may represent a more simplified version of the computation that does not consider the filtering operation, represented above in several equations denoting the derivation of the SHC 511 in the frequency domain by the function jn(•), which refers to the spherical function of Bessel of order n. In the time domain, this function jn(•) represents a filtering operation that is specific to a particular order, n. With filtering, rotation can be performed in order. For purposes of illustration, consider the following equations:

[1403] From these equations, the SHC rotated 511' for orders are done separately, as the bn(t) is different for each order. As a result, the above equation can be altered as follows to compute the first-order coefficients among the 511' rotated SHCs:

[1404] Since there are three first-order coefficients among the SHC 511, each of the vectors of SHC 511' and 511 is of size three in the above equation. Likewise, for the second order, the following equation can be applied:

[1405] Again, since there are five second-order coefficients among the SHC 511, each of the vectors of SHC 511' and 511 is of size five in the above equation. The remaining equations for the other orders, i.e. the third and fourth orders, can be similar to the one described above, following the same pattern regarding matrix sizes (where the number of rows of EncMat2, the number of columns of InvMat1 and the sizes of the third and fourth order SHC 511 and SHC 511' vectors are equal to the number of suborders (m times two plus 1) of each of the third and fourth order spherical harmonic basis functions.

[1406] Therefore, the audio encoding device 570 can perform this rotation operation with respect to each combination of azimuthal and elevation angle in an attempt to identify the so-called ideal rotation. The audio encoding device 570 may, after performing this rotation operation, compute the number of SHC 511' above the threshold value. In some examples, the audio encoding device 570 may perform this rotation to derive a series SHC 511' representing the sound field over a duration of time, for example an audio frame. By performing this rotation to derive the series of SHC 511' representing the sound field over that length of time, the audio encoding device 570 can reduce the number of rotation operations that have to be performed compared to accomplishing this. for each set of SHC 511s describing the sound field for durations of time less than a 485/500 frame or other length. In any case, the audio coding device 570 can store, throughout this process, the coefficients among the SHC 511' that have the smallest number of the SHC 511' greater than the threshold value.

[1407] However, performing this rotation operation with respect to each processor-intensive elevation and azimuth angle combination is time consuming. As a result, the audio encoding device 570 may not perform what might be characterized as this "brute force" deployment of the rotation algorithm. Instead, the audio encoding device 570 may provide rotations with respect to a subset of possibly known (statistically speaking) azimuthal and elevation angle combinations that provide generally satisfactory compression, performing additional rotations relative to combinations around those in that subset that provide better compression compared to other combinations in the subset.

[1408] Alternatively, the audio encoding device 570 may perform this rotation with respect to only the known subset of combinations. Alternatively, the audio encoding device 570 may follow a path (spatially) of combinations, performing rotations relative to that path of combinations. Alternatively, audio encoding device 570 may specify a compression threshold that defines a maximum number of SHC 511' that have non-zero values above the threshold value. This compression threshold can effectively set a stopping point for the search so that when the audio encoding device 570 performs a rotation and determines that the number of SHC 511' that has a value above the threshold set is less than or equal to (or less in some instances) the 486/500 compression threshold, the 570 audio encoding device will stop performing any additional rotation operations on the remaining combinations. 'Still alternatively, audio encoding devices 570 may traverse a hierarchically arranged tree (or other data structures) of combinations, performing rotation operations with respect to the current combination, and traversing the tree to the right or left (e.g. example with respect to binary trees) depending on the number of SHC 511' that have a non-zero value greater than the threshold value.

[1409] In this sense, each of these alternatives involves performing a first and second rotation operation and comparing the result of performing the first and second rotation operations in order to identify one of the first and second rotation operations that results in the smallest number of the SHC 511' which has a non-zero value greater than the threshold value. Accordingly, the audio encoding device 570 may perform a first rotation operation in the sound field to rotate the sound field in accordance with a first azimuthal angle and a first elevation angle and determine a first number of the plurality of hierarchical elements representative of the field. sound field rotated in accordance with the first azimuth angle and the first elevation angle that provide relevant information in describing the sound field. Audio encoding device 570 may also perform a second sound field rotation operation to rotate the sound field in accordance with a second azimuth angle and a second elevation angle and determine a second number from the plurality of hierarchical elements representative of the sound field. rotated in accordance with the 487/500 second azimuth angle and the second elevation angle that provides relevant information in describing the sound field. Additionally, audio encoding device 570 may select the first rotation operation or the second rotation operation based on a comparison of the first number of the plurality of hierarchical elements and the second number of the plurality of hierarchy elements.

[1410] In some examples, the rotation algorithm can be applied against a duration of time, where subsequent invocations of the rotation algorithm can perform rotation operations based on past invocations of the rotation algorithm. In other words, the rotation algorithm can be adaptive based on past rotation information determined during the rotation of the sound field for an earlier duration of time. For example, audio encoding device 570 may rotate the sound field over a first time duration, e.g., an audio frame, to identify SHC 511' for that first time duration. Audio encoding devices 570 may specify rotation information and SHC 511' in bitstream 517 in any of the ways described above. This rotation information can be termed as first rotation information in that it describes the rotation of the sound field for the first time duration. Audio encoding device 570 can then, based on this first rotation information, rotate the sound field over a second time duration, e.g., a second audio frame, to identify SHC 511' for that second time. time duration. The audio encoding device 570 can use this first rotation information while performing the second rotation operation over the second 488/500 duration of time to initiate a search for an "optimal" combination of azimuth and elevation angles, for example. example. Audio encoding device 570 can then specify the SHC 511' and corresponding rotation information for the second time duration (which may be termed "second rotation information") in bitstream 517.

[1411] While described above in relation to different ways to implement the rotation algorithm in order to reduce processing time and/or consumption, the techniques can be performed in relation to any algorithm that can reduce or otherwise increase to expedite the identification of what can be termed the “optimal rotation.” Furthermore, the techniques can be performed against any algorithm that identifies non-ideal rotations, but which can improve performance in other respects, often measured in terms of speed or processor or other resource usage.

[1412] Figures 57 to 57E are each a diagram illustrating bit streams 517A to 517E formed in accordance with the techniques described in the present disclosure. In the example of Figure 57A, the bit stream 517A may represent an example of the bit stream 517 shown in Figure 53 above. Bit stream 517A includes a present field of SHC 670 and a field that stores SHC 511' (where the field is denoted "SHC 511'"). The present field of SHC 670 may include a bit corresponding to each of the SHC 511. The SHC 511' may represent the coefficients among the SHC 511's that are specified in the bit stream, which may be smaller in number than the number of the SHC 511's. Typically, each of the SHC 511' are the coefficients among SHC 511s that have non-zero values. As noted above, for a 489/500 fourth-order representation of any given sound field, (1+4)2 or 25 SHC are required. Eliminating one or more of these SHC and replacing those zero-valued SHCs with a single bit can store 31 bits, which can be allocated to express other portions of the sound field in more detail or otherwise removed to facilitate efficient bandwidth utilization.

[1413] In the example of Fig. 57B, bitstream 517B may represent an example of bitstream 517 shown in Fig. 53 above. Bitstream 517B includes a transform information field 672 ("transform information 672") and a field that stores SHC 511' (where the field is denoted "SHC 511 "'). above, may comprise translation information, rotation information, and/or any other form of information denoting an adjustment in reaction to a sound field. In some examples, 672 transform information may also specify a higher order of SHC 511 which are specified in the bit stream 517B as SHC 511'. That is, the transform information 672 may indicate an order of three, which the extraction device may understand as an indication that SHC 511' includes the coefficients among SHC 511 or more and that includes the coefficients among SHC 511s that have an order of three. The extraction device can then be configured to set SHC 511s that have an order of four or higher to zero, thereby removing potentially making explicit SHC 511 signaling of order four or higher in the bit stream.

[1414] In the example of Fig. 57C, the bit stream 517C may represent an example of the bit stream 517 shown in Fig. 53 above. Bitstream 517C includes the 490/500 transform information field 672 ("transform information 672"), the present field of SHC 670, and a field that stores SHC 511' (wherein the field is denoted "SHC 51" ). Instead of being configured to understand which order of SHC 511s are not signaled as described above with respect to Figure 57B, the SHC 670 present field can explicitly signal which of the SHC 511s are specified in the stream. from bits 517C to SHC 511'.

[1415] In the example of Fig. 57D, the bit stream 517D may represent an example of the bit stream 517 shown in Fig. 53 above. Bitstream 517D includes a 674th order field ("60th order"), the SHC present field 670, an azimuth flag 676 ("AZF 676"), an elevation flag 678 ("ELF 678"), an azimuth angle field 680 ("azimuth 680"), an elevation angle field 682 ("elevation 682"), and a field that stores SHC 511' (where, again, the field is denoted "SHC 511"). of order 674 specifies the order of SHC 511', that is, the order denoted by n above to the highest order of the spherical basis function used to represent the sound field. The field of order 674 is shown to be a field of 8 bits, but can be of various other bit sizes, such as three (which is the number of bits required to specify fourth order). The present field of SHC 670 is shown as a 25-bit field. However, again, the present field of SHC 670 can be of various other sizes in bits. The present field of SHC 670 is shown as 25 bits to indicate that the present field of SHC 670 can include a bit for each of the spherical harmonic coefficients corresponding to a fourth-order representation of the sound field.

[1416] Azimuth flag 676 represents a one-bit flag that specifies whether the azimuth field 491/500 680 is present in bitstream 517D. When azimuth flag 676 is set to one, azimuth field 680 for SHC 511' is present in bitstream 517D. When azimuth flag 676 is set to zero, azimuth field 680 for SHC 511' is not present or otherwise specified in bitstream 517D. Likewise, elevation flag 678 represents a one-bit flag that specifies whether elevation field 682 is present in bit stream 517D. When elevation flag 678 is set to one, elevation field 682 for SHC 511' is present in bitstream 517D. When elevation flag 678 is set to zero, elevation field 682 for SHC 511' is not present or otherwise specified in bitstream 517D. Although one is described that the corresponding field is present and zero signals that the corresponding field is not present, the convention can be reversed so that a zero specifies that the corresponding field is specified in bitstream 517D and one specifies that the corresponding field is not specified in bitstream 517D. Therefore, the techniques described in the present disclosure should not be limited in this regard.

[1417] Azimuthal field 680 represents a 10-bit field that specifies, when present in bitstream 517D, the azimuthal angle. Although shown as a 10-bit field, the azimuth field 680 can have other bit sizes. Elevation field 682 represents a 9-bit field that specifies, when present in bitstream 517D, the elevation angle. The azimuth angle and elevation angle specified in fields 680 and 682, respectively, can represent, in combination with flags 676 and 678, the rotation information described above. This rotation information can be used 492/500 to rotate the sound field to bring the SHC 511s back to the original reference frame.

[1418] The field of SHC 511' is shown as a variable field that has a size of X. The field of SHC 511' may vary due to the number of SHC 511' specified in the bit stream as denoted by the present field of SHC 670. The size X can be derived as a function of the number of ones in the present field of SHC 670 times 32 bits (which is the size of each SHC 511').

[1419] In the example of Figure 57E, bitstream 517E may represent another example of bitstream 517 shown in Figure 53 above. Bitstream 517E includes a field of order 674 ("order 60"), a present field of SHC 670 and a rotation index field 684, and a field that stores SHC 511' (wherein, again, the field is denoted "SHC 511'“). The order field 674, the present field of SHC 670 and the field of SHC 511' may be substantially similar to those described above. The rotation index field 684 may represent a 20-bit field used to specify one of 1024x512 combinations (or, in other words, 524288) of the azimuth and elevation angles. In some examples, only 19 bits can be used to specify this rotation index field 684, and the audio encoding device 570 can specify an additional flag in the bitstream to indicate whether a rotation operation has been performed (thus if the rotation index field 684 is present in the bitstream.) This rotation index field 684 specifies the rotation index checked above, which may refer to one and entered in a rotation table common to both the audio encoding device 570 and the bitstream extractor device. This rotation table can store, in some examples, the 493/500 different combinations of azimuth and elevation angles. Alternatively, the rotation table can store the matrix described above, which effectively stores the different combinations of azimuth and elevation angles in matrix form.

[1420] Figure 58 is a flowchart illustrating an exemplary operation of the audio encoding device 570 shown in the example of Figure 53 in implementing rotating aspects of the techniques described in the present disclosure. Initially, the audio encoding device 570 may select a combination of azimuthal angle and an elevation angle in accordance with one or more of the various rotation algorithms described above (800). The audio encoding device 570 can then rotate the sound field according to the selected azimuth and elevation angle (802). As described above, the audio encoding device 570 can first derive the sound field from the SHC 511 using the InvMat1 verified above. The audio encoding device 570 can also determine the SHC 511' representing the rotated sound field (804). Although described as being separate steps or operations, the audio encoding device 570 may apply a transform (which may represent the [EncMat2][//?v at;]) that represents the selection of the combination of azimuth angle and angle of view. elevation, deriving the sound field from the SHC 511, rotating the sound field and determining the SHC 511' representing the rotated sound field.

[1421] In any case, the audio encoding device 570 may then compute a number of the determined SHC 511' that are greater than a threshold value, compared to a number shifted to a previous iteration relative to the threshold value. to a previous combination of azimuth angle and elevation angle (806, 808). In the first iteration against the first combination of azimuthal angle and elevation angle, this comparison can be to a predetermined previous number (which can be set to zero). In any case, if the given number of SHC 511' is less than the previous numbers ("SIM" 808), the audio encoding device 570 stores the SHC 51, the azimuth angle and the elevation angle, which frequently replace the previous SHC 511', azimuthal angle and elevation angle stored from a previous iteration of the rotation algorithm (810).

[1422] If the determined number of SHC 511' is not less than the previous number ("NO" 808) or after storing the SHC 511', the azimuthal angle and elevation angle in place of the previously stored SHC 51, the azimuth angle and elevation angle, the audio encoding device 570 can determine whether the rotation algorithm has ended. That is, the audio encoding device 570 can, for example, determine whether the entire available combination of azimuth angle and elevation angle has been evaluated. In other examples, the audio coding device 570 can determine if other criteria are satisfied (e.g., if all of a defined subset of matching were performed, if a certain one was traversed, if a hierarchical tree was traversed to a leaf node , etc.) so that the audio encoding device 570 has finished applying the rotation algorithm. If not finished (“No” 812), the audio encoding device 570 can perform the above process against another selected combination (800 to 812). If done ("YES" 812), the audio encoding device 570 can specify 495/500 the stored SHC 51, the azimuth angle, and the elevation angle in the bitstream 517 in one of several ways described above (814) .

[1423] Figure 59 is a flowchart illustrating an exemplary operation of the audio encoding device 570 shown in the example of Figure 53 in performing the transforming aspects of the techniques described in the present disclosure. Initially, the audio encoding device 570 may select a matrix that represents a linear invertible transform (820). An example of a matrix representing a linear invertible transform might be the matrix shown above which is the result of [EncMatf\[IncMatf\. The audio encoding device 570 can then apply the matrix to the sound field to transform the sound field (822). The audio encoding device 570 can also determine the SHC 511' representing the rotated sound field (824). Although described as having separate steps or operations, the audio encoding device 570 can apply a transform (which represents the result of [EncMat2][//?v at;]), deriving the sound field from the SHC 511, can transform the sound field and determine the SHC 511' which represents the transform sound field.

[1424] In any case, the audio encoding device 570 may then compute a number of the determined SHC 511' that are greater than a threshold value, which comprises that number to a number computed for a previous iteration with respect to to an earlier application of a transform matrix (826, 828). If the given number of SHC 511' is less than the previous number ("Yes" 828), the audio encoding device 570 stores the SHC 511' and the matrix (or some derivative thereof, e.g. an index 496 /500 associated with the matrix), often replacing the previous SHC 511' and the matrix (or a derivative thereof) stored from a previous iteration of the rotation algorithm (830).

[1425] If the number of SHC 511' is not less than the previous number ("No" 828) or after storing the SHC 511' and matrix in place of the previously stored SHC 511' and matrix, the encoding device of audio 570 can determine if the transform algorithm is finished (832). That is, the audio encoding device 570 can determine, for example, whether all transform matrices have been evaluated. In other examples, the audio coding device 570 can determine if other criteria are satisfied (e.g., if all of a defined subset of matching were performed, if a certain one was traversed, if a hierarchical tree was traversed to a leaf node , etc.) so that the audio encoding device 570 has finished applying the transform algorithm. If not terminated (“No” 832), the audio encoding device 570 can perform the above process with respect to another selected transform matrix (820 to 832). If terminated ("Yes" 832), audio encoding device 570 can specify the SHC 511' and array stored in bitstream 517 in one of several ways described above (834).

[1426] In some examples, the transform algorithm may perform a single iteration, evaluating a single transform. That is, the transform matrix may comprise any matrix that represents a linear invertible transform. In some examples, the linear invertible transform can transform 497/500 the sound field from the spatial domain to the frequency domain. Examples of such a linear invertible transform may include a discrete Fourier transform (DFT). Application of the DFT may only involve a single iteration, so it does not necessarily include the steps of determining whether the transform algorithm is determined. Consequently, the techniques should not be limited to the example in Figure 59.

[1427] In other words, an example of a linear invertible transform is a discrete Fourier transform (DFT). The twenty-five SHC 511' can be operated by the DFT to form a set of twenty-five complex coefficients. The audio encoding device 570 may also zero-fill the twenty-five SHC 511' to be an interior number multiple of 2, so as to potentially increase the resolution of the DFT bin size and potentially have a more efficient method of DFT, for example by applying a fast Fourier transform (FFT). In some examples, increasing the DFT resolution beyond 25 points is not necessarily required. In the transform domain, audio encoding device 570 may apply a threshold to determine whether there is any spectral energy in a particular compartment. Audio encoding device 570, in this context, may then discard or zero spectral coefficient energy that is below this threshold, and audio encoding device 570 may apply an inverse transform to recover SHC 511' that have one or more of the discarded or zeroed SHC 511's. That is, after the inverse transform is applied, subthreshold coefficients are not present and, as a result, fewer bits can be used to encode the sound field. 498/500

[1428] In one or more examples, the functions described may be implemented in hardware, software, firmware, or any combination thereof. If implemented in software, functions may be stored in, or transmitted over, as one or more instructions or code on a computer-readable medium and executed by a hardware-based processing unit. Computer readable media may include computer readable storage media, which correspond to tangible media such as data storage media or communication media which include any media that facilitates the transfer of a computer program from one place to another, for example, according to a communication protocol. In this way, computer-readable media may correspond, in general, to (1) tangible computer-readable storage media that are non-transient (2) a communication medium, eg, a signal or carrier wave. Data storage media can be any available media that can be accessed by one or more computers or one or more processors to retrieve instructions, code structures and/or data for implementing the techniques described in the present disclosure. A computer program product may include computer readable media.

[1429] By way of example, and not limitation, such computer readable storage media may comprise RAM, ROM, EEPROM, CD-ROM or other optical disk storage, magnetic disk storage or other magnetic storage devices, or any other medium that can be used to store the desired program code in the form of instructions or data structures and that can be accessed 499/500 by a computer. Also, any connection is properly termed a computer readable media. For example, if instructions are transmitted from a website, server, or other remote source using coaxial cable, fiber optic cable, twisted pair, digital subscription line (DSL), or wireless technologies such as infrared , radio and microwave, then coaxial cable, fiber optic cable, twisted pair, DSL or wireless technologies such as infrared, radio and microwave are included in the definition of media. It should be understood, however, that computer-readable storage media and data storage media do not include connections, carrier waves, signals, or other transient media, but are instead directed toward non-transient, tangible media. Magnetic disc and optical disc, as used herein, include compact disc (CD), laser disc, optical disc, digital versatile disc (DVD), floppy disk and blu-ray disc, where magnetic discs normally reproduce data in a magnetic, while optical discs reproduce the data optically with lasers. Combinations of the above should also fall within the scope of computer readable media.

[1430] Instructions may be executed by one or more processors, such as one or more digital signal processors (DSPs), general-purpose microprocessors, application-specific integrated circuits (ASICs), field-programmable logic arrays (FPGAs), or other set of equivalent integrated or discrete logic circuits. Therefore, the term "processor," as used herein, may refer to any of the aforementioned structures or any other structure suitable for implementing the 500/500 techniques described in this document. Additionally, in some respects, the functionality described in this document may be provided within dedicated hardware and/or software modules configured to encrypt and decode or incorporated into a combined codec. Also, the techniques can be fully implemented in one or more circuits or logic elements.

[1431] The techniques of the present disclosure can be implemented in a wide variety of devices or apparatus, including a handset, an integrated circuit (IC), or a set of ICs (eg, a chip set). Various components, modules, or units are described in the present disclosure to emphasize the functional aspects of devices configured to perform the disclosed techniques, but do not necessarily require realization by different hardware units. Instead, as described above, multiple units may be combined into one codec hardware unit or provided by a collection of interoperable hardware units, including one or more processors as described above, together with appropriate software and/or firmware.

[1432] Various modalities of the techniques have been described. These and other aspects of the techniques are within the scope of the claims that follow.

Claims (15)

1. Method characterized by comprising: obtaining a stream of bits (21) comprising a compressed version of a spatial component of a sound field (57), wherein the spatial component is defined in a spherical harmonic domain and is generated by performing a linear decomposition with respect to a plurality of spherical harmonic coefficients, wherein the compressed version of the spatial component is represented in the bit stream using, at least in part, a field (262) that specifies a prediction mode (237 ) used when compressing the spatial component. Method according to claim 1, characterized in that the compressed version of the spatial component is still represented in the bit stream using, at least in part, one of: Huffman table information that specifies a Huffman table used when compressing the spatial component; a Huffman code to represent a category identifier that identifies a compression category to which the spatial component corresponds; a sign bit that identifies whether the spatial component has a positive value or a negative value; and a Huffman code to represent a residual value of the spatial component. Method according to claim 1, characterized in that the decomposition comprises a singular value decomposition. 4. Device (16, 24) characterized in that it comprises one or more processors configured to: obtain a bit stream (16, 24) comprising a compressed version of a spatial component of a sound field (57), the spatial component is defined in a spherical harmonic domain and is generated by performing a linear decomposition with respect to a plurality of spherical harmonic coefficients, where the compressed version of the spatial component is represented in the bit stream using, at least in part, a field (262) that specifies a prediction mode (237) used when compressing the spatial component. Device according to claim 4, characterized in that the compressed version of the spatial component is subsequently represented in the bit stream using, at least in part, Huffman table information that specifies a Huffman table used when compressing the spatial component. Device according to claim 4, characterized in that the compressed version of the spatial component is further represented in the bit stream using, at least in part, a Huffman code to represent a category identifier that identifies a category of compression to which the spatial component corresponds. Device according to claim 4, characterized in that the compressed version of the spatial component is further represented in the bit stream using, at least in part, a sign bit that identifies whether the spatial component has a positive or negative value. a negative value. Device according to claim 4, characterized in that the compressed version of the spatial component is further represented in the bit stream using, at least in part, a Huffman code to represent a residual value of the spatial component. Device according to claim 4, characterized in that the decomposition comprises a singular value decomposition. 10. Method characterized by comprising: generating a stream of bits (21) that comprises a compressed version of a spatial component of a sound field (57), wherein the spatial component is defined in a spherical harmonic domain and is generated by performing a linear decomposition with respect to a plurality of spherical harmonic coefficients, wherein the compressed version of the spatial component is represented in the bit stream using, at least in part, a field (262) that specifies a prediction mode (237 ) used when compressing the spatial component. The method of claim 10, characterized in that generating the bit stream comprises generating the bit stream including one of: Huffman table information specifying a Huffman table used when compressing the spatial component; a Huffman code to represent a category identifier that identifies a compression category to which the spatial component corresponds; a sign bit that identifies whether the spatial component has a positive value or a negative value; and a Huffman code to represent a residual value of the spatial component. 12. Device (16, 24) characterized in that it comprises one or more processors configured to: generate a bit stream (21) comprising a compressed version of a spatial component of a sound field (57), the spatial component being defined into a spherical harmonic domain and is generated by performing a linear decomposition with respect to a plurality of spherical harmonic coefficients, where the compressed version of the spatial component is represented in the bit stream using, at least in part, a field (262) which specifies a prediction mode (237) used when compressing the spatial component. Device according to claim 12, characterized in that one or more processors are configured to generate the bit stream including one of: Huffman table information specifying a Huffman table used when compressing the spatial component; a Huffman code to represent a category identifier that identifies a compression category to which the spatial component corresponds; a sign bit that identifies whether the spatial component has a positive value or a negative value; and a Huffman code to represent a residual value of the spatial component. Device according to claim 12, characterized in that the decomposition comprises a singular value decomposition. 15. Memory characterized by comprising instructions stored therein which, when executed, cause one or more processors to perform a method as defined in any one of claims 1 to 3, 10 or 11.

Images