RU2809646C1 - Multichannel signal generator, audio encoder and related methods based on mixing noise signal - Google Patents

Abstract

FIELD: computer engineering.

SUBSTANCE: processing audio data. The technical result is to provide comfortable stereo noise by modelling the spectral characteristics of background noise in both channels, as well as the degree of correlation between them while maintaining an average bit rate comparable to mono applications. The technical result is achieved by generating a first audio signal using a first audio source; generating a second audio signal using the second audio source; generating a mixing noise signal using a mixing noise source; and mixing the mixing noise signal and the first audio signal to obtain a first channel, and mixing the mixing noise signal and the second audio signal to obtain a second channel, using a first amplitude element affecting the amplitude of the first audio signal; using a first adder that sums the output signal of the first amplitude element and at least a portion of the noise mixing signal; using a second amplitude element affecting the amplitude of the second audio signal; using a second adder summing the output of the second amplitude element and at least a portion of the mixing noise signal.

EFFECT: comfortable stereo noise by modelling the spectral characteristics of background noise in both channels, as well as the degree of correlation between them while maintaining an average bit rate comparable to mono applications.

49 cl, 11 dwg

Description

The present invention relates, inter alia, to comfort noise generation (CNG) to provide discontinuous transmission (DTX) in stereo codecs. The invention also relates to a multi-channel signal generator, an audio encoder and related methods, for example based on a noise mixing signal. The invention may be implemented in a device, in a system, in a method, in a persistent storage module storing instructions that, when executed by a computer (processor, controller), instruct the computer (processor, controller) to carry out a particular method, and in an encoded multi-channel audio signal.

Introduction

Comfort noise generators are commonly used in discontinuous transmission (DTX) of audio signals, particularly audio signals containing speech. In this mode, the audio signal is first classified into active and inactive frames by a voice activity detector (VAD). Based on the VAD result, only active speech frames are encoded and transmitted at the nominal bit rate. During long pauses in which only background noise is present, the bit rate is reduced or zeroed and the background noise is encoded parametrically using silence insertion descriptor frames (SID frames). The average bit rate in this case is significantly reduced.

The noise is generated during inactive frames at the decoder side by means of a comfort noise generator (CNG). The SID frame size is very limited in practice. Therefore, the number of parameters describing background noise should be kept as small as possible. To this end, noise estimation is not applied directly to the output of spectral transforms. Instead, it is applied at a lower spectral resolution by averaging the input power spectrum between groups of frequency bands, for example, according to the bark scale. Averaging can be achieved using arithmetic or geometric means. Unfortunately, the limited number of parameters carried in SID frames prevents the precise spectral structure of background noise from being captured. Therefore, only the smoothed spectral envelope of noise can be reproduced by CNG. When the VAD initiates a CNG frame, the discrepancy between the smoothed spectrum of the reconstructed comfort noise and the spectrum of the actual background noise can become very audible during transitions between active frames (involving regular encoding and decoding of the noisy speech portion of the signal) and CNG frames.

Some example CNG technologies are contained in ITU-T Recommendations G.729B [1], G.729.1C [2], G.718 [3] or in the 3GPP specifications for AMR [4] and AMR-WB [5]. All these technologies generate comfort noise (CN) by using a linear prediction (LP) analysis/synthesis approach.

To further reduce the bit rate, the 3GPP communication codec for enhanced voice services (EVS) LTE [6] is equipped with a discontinuous transmission (DTX) mode that uses comfort noise generation (CNG) for inactive frames, i.e. for frames that are determined to consist only of background noise. For these frames, a low-speed parametric representation of the signal is transmitted via silent insertion descriptor (SID) frames at most every 8 frames (160 ms). This allows the CNG in the decoder to produce an artificial noise signal that resembles actual background noise. In EVS, CNG can be achieved using either a linear prediction scheme (LP-CNG) or a frequency domain scheme (FD-CNG), depending on the spectral characteristics of the background noise.

The LP-CNG approach in EVS [7] operates on the basis of band partitioning with coding consisting of an analytical/synthesis coding stage in the low band and in the high band. Unlike low-band coding, parametric modeling of the high-band noise spectrum is not performed on the high-band signal. Only the high-band signal energy is encoded and transferred to the decoder, and the high-band noise spectrum is generated exclusively at the decoder side. The CN low-pass bands and high-pass bands are synthesized by filtering the excitation through a synthesis filter. The low-band excitation is derived from the received low-band excitation energy and the frequency envelope of the low-band excitation. A low-pass synthesis filter is extracted from the received LP parameters in the form of line spectral frequency factors (LSFs). The high-band excitation is obtained using energy that is extrapolated from the low-band energy, and the high-band synthesis filter is derived from the LSF interpolation at the decoder side. The high-band synthesis is spectrally inverted and added to the low-band synthesis to form the final CN signal.

The FD-CNG approach [8],[9] uses a frequency domain noise estimation algorithm and then vector quantizes the smoothed spectral envelope of the background noise. The decoded envelope is detailed in the decoder by executing a second frequency domain noise estimator. Since a purely parametric representation is used during inactive frames, the noise signal is not available at the decoder in this case. In FD-CNG, noise estimation is performed in each frame (active and inactive) at the encoder and decoder sides based on a minimal statistical algorithm.

The method for generating comfortable noise in the case of two (or more) channels is described in [10]. [10] describes a system for stereo DTX and CNG that combines mono-SID with a per-band coherence metric calculated for the two stereo input channels in the encoder. At the decoder, mono CNG information and coherence values are decoded from the bit stream, and the target coherence in a number of frequency bands is synthesized. To reduce the bit rate of the resulting stereo SID frame, the coherence values are encoded using a prediction circuit followed by variable bit rate entropy encoding. Comfort noise is generated for each channel using the methods described in the previous paragraphs, and then the two CNs are mixed for each frequency band using a weighted formula based on the transmitted coherence values of the frequency bands included in the SID frame.

Background/disadvantages of the state of the art

In a stereo system, separately shaping the background noise results in a completely decorrelated noise that sounds unpleasant and significantly different from the actual background noise, causing abrupt audible transitions when switching to/from the active mode background to the DTX mode backgrounds. In addition, it is not possible to maintain a stereo image of the background using only two fully decorrelated noise sources. Finally, if there is a source of background noise, and the speaker moves with a handheld device around the source, then the spatial representation of the background noise must change over time, something that cannot be replicated when reconstructing the background noise independently for each channel. Therefore, it is necessary to create a new approach to solve this problem for stereo signals.

This is also solved in [10]; however, in embodiments, inserting a common noise source for the two channels to simulate correlated noise to generate the final comfort noise plays an important role in simulating a stereo background noise recording.

Existing speech communication codecs typically encode only mono signals. Consequently, most existing DTX systems are designed for mono-CNG. Simply using DTX mode to operate independently on both channels of a stereo signal seems straightforward, but does present several challenges. First, this approach requires passing two sets of parameters describing two background noise signals in two channels. This should increase the data rate required to transmit SID frames, which reduces the benefit of reducing network load. Another problematic aspect is the VAD solution, which must be synchronized between channels to avoid weirdness and distortion in the spatial image of the stereo signal, as well as to optimize the system bit rate reduction. Additionally, when applying CNG to the receiver side independently for both channels, two independent CNG algorithms typically must produce two random noise signals with zero or very low coherence. This should result in a very wide stereo image in the generated comfort noise. On the other hand, applying only to the noise generator and using the same comfort noise signal in both channels results in very high coherence and a very narrow stereo image. However, for most stereo signals, the stereo image and its spatial impression should fall somewhere between these two extremes. Switching to or from active frames in DTX mode should therefore introduce sharp audible transitions. Additionally, if there is a source of background noise, and the speaker moves with a handheld device around the source, then the spatial representation of the background noise must change over time, something that cannot be replicated when reconstructing the background noise independently for each channel. Therefore, a new approach is needed to solve this problem for stereo signals.

The system described in [10] resolves these issues by passing information to the mono-CNG along with parameter values that are used to re-synthesize a stereo image of the background noise at the decoder. This type of DTX system is optimally suited for parametric stereo encoders that apply downmixing to two input channels before encoding and transmitting, from which mono CNG parameters can be extracted. However, in a discrete stereo encoding scheme, the two channels are typically still encoded together, and upmixing parameters such as a fine-grained coherence metric are typically not extracted. Thus, for similar stereo encoders, a different approach is required.

Aspects of the Present Invention

The present examples provide efficient transmission of stereo speech signals. Stereo transmission can improve user experience and speech intelligibility compared to (mono) transmission of only one channel of audio, particularly in situations with overlapping background noise or other sounds. Stereo signals can be encoded parametrically, whereby mono downmixing of two stereo channels is applied and this one downmix channel is encoded and sent to the receiving device along with auxiliary information that is used to approximate the original stereo signal at the decoder. Another approach is to use discrete stereo encoding, which aims to remove redundancy between channels to achieve a more compact two-channel representation of the original signal through some signal pre-processing. The two processed channels are then encoded and transmitted. At the decoder, inverse processing is applied. However, auxiliary information relevant for stereo processing can be transmitted along two channels. The main difference between parametric and discrete stereo coding methods therefore lies in the number of transmitted channels.

Typically, there are periods in a conversation during which not all speakers are actively speaking. The input signal to the speech encoder during these periods therefore consists mainly of background noise or (virtually) silence. To reduce the data rate and load on the transmission network, speech encoders attempt to distinguish between frames that contain speech (active frames) and frames that contain mostly background noise or silence (inactive frames). For inactive frames, the data rate can be significantly reduced by not encoding the audio signal as in active frames, but instead extracting a low bit rate parametric description of the current background noise in the form of a silent insertion descriptor (SID) frame. This SID frame is periodically sent to the decoder to update the parameters describing the background noise, while for inactive frames in the interim the bit rate is reduced or even no information is transmitted at all. At the decoder, the background noise is remodeled using the parameters carried in the SID frame through a comfort noise generation (CNG) algorithm. Thus, the transmission rate can be reduced or even reset to zero for inactive frames without the user interpreting this as an interruption or end of the connection.

A DTX system for discretely encoded stereo signals is described, consisting of a stereo SID and a CNG method that generates comfortable stereo noise by modeling the spectral characteristics of the background noise in both channels, as well as the degree of correlation between them, while maintaining an average bit rate comparable to mono applications.

Disclosure of the Invention

In accordance with an aspect, a multi-channel signal generator is provided for generating a multi-channel signal having a first channel and a second channel comprising:

- a first audio source for generating a first audio signal;

- a second audio source for generating a second audio signal;

- noise source during mixing to generate a mixing noise signal; And

- a mixer for mixing the mixing noise signal and the first audio signal to obtain a first channel, and for mixing the mixing noise signal and the second audio signal to obtain a second channel.

According to an aspect, the first audio source is a first noise source and the first audio signal is a first noise signal, or the second audio source is a second noise source and the second audio signal is a second noise signal,

- wherein the first noise source or the second noise source is configured to generate the first noise signal or the second noise signal in such a way that the first noise signal or the second noise signal is decorrelated with respect to the mixing noise signal.

According to an aspect, the mixer is configured to generate the first channel and the second channel such that the magnitude of the mixing noise signal in the first channel is equal to the magnitude of the mixing noise signal in the second channel or is within a range of 80-120 percent relative to the magnitude of the mixing noise signal in the second channel .

According to an aspect, the mixer includes a control input for receiving a control parameter, and wherein the mixer is configured to control the amount of a mixing noise signal in a first channel and a second channel in response to the control parameter.

According to an aspect, each of the first audio source, the second audio source, and the mixing noise source is a Gaussian noise source.

According to an aspect, the first audio source includes a first noise generator for generating a first audio signal as a first noise signal, wherein the second audio source includes a decorrelator for decorrelating the first noise signal to generate a second audio signal as a second noise signal, and wherein the mixing noise source comprises a second generator noise, or

- wherein the first audio source contains a first noise generator for generating a first audio signal as a first noise signal, wherein the second audio source contains a second noise generator for generating a second audio signal as a second noise signal, and wherein the mixing noise source contains a decorrelator for decorrelating the first noise signal signal or a second noise signal to form a mixing noise signal, or

- wherein one of the first audio source, the second audio source and the mixing noise source contains a noise generator for generating a noise signal, and the other of the first audio source, the second audio source and the mixing noise source contains a first decorrelator for decorrelating the noise signal, and furthermore one of the first audio source, the second audio source, and the mixing noise source comprises a second decorrelator for decorrelating the noise signal, wherein the first decorrelator and the second decorrelator are different from each other such that the output signals of the first decorrelator and the second decorrelator are decorrelated from each other, or

- wherein the first audio source contains a first noise generator, wherein the second audio source contains a second noise generator, and wherein the mixing noise source contains a third noise generator, wherein the first noise generator, the second noise generator and the third noise generator are configured to form mutually decorrelated noise signals.

According to an aspect, one of the first audio source, the second audio source, and the mixing noise source comprises a pseudo-random number sequence generator configured to generate a pseudo-random number sequence in response to a seed, and wherein at least two of the first audio source, the second audio source, and the noise source during mixing, they are configured to initialize the generator of pseudo-random number sequences using different initial numbers.

In an aspect, at least one of the first audio source, the second audio source, and the mixing noise source is configured to operate using a previously stored noise table, or

- wherein at least one of the first audio source, the second audio source and the mixing noise source is configured to generate a complex spectrum for the frame using the first noise value for the real part and the second noise value for the imaginary part,

- wherein, if necessary, at least one noise generator is configured to generate a complex spectral noise value for the frequency element k of the resolution using, for one of the real part and the imaginary part, the first random value with index k, and using, for the other of a real part and an imaginary part, a second random value with index (k+M), wherein the first noise value and the second noise value are included in a noise array, for example, extracted from a random number sequence generator or from a noise table, or from a noise process , in the range from the start index to the end index, wherein the start index is less than M, and where the end index is equal to or less than 2M, wherein M and k are integers.

According to the aspect, the mixer contains:

- a first amplitude element for influencing the amplitude of the first audio signal;

- a first adder for summing the output signal of the first amplitude element and at least part of the noise mixing signal;

- a second amplitude element for influencing the amplitude of the second audio signal;

- a second adder for summing the output of the second amplitude element and at least a portion of the mixing noise signal,

- in this case, the magnitude of the impact performed through the first amplitude element and the magnitude of the impact performed through the second amplitude element are equal to each other, or the magnitude of the impact performed through the second amplitude element differs by less than 20 percent relative to the value performed through the first amplitude element .

According to an aspect, the mixer includes a third amplitude element for influencing the amplitude of the mixing noise signal,

- in this case, the magnitude of the impact performed by means of the third amplitude element depends on the magnitude of the impact performed by means of the first amplitude element or the second amplitude element in such a way that the magnitude of the impact performed by means of the third amplitude element becomes greater when the magnitude of the impact performed by means of the first amplitude element element, or the magnitude of the effect performed by the second amplitude element becomes smaller.

According to an aspect, the amount of action performed by the third amplitude element is the square root of the value of c _q , and the amount of action performed by the first amplitude element and the amount of action performed by the second amplitude element is the square root of the difference between one and c _q .

According to an aspect, an input interface for receiving encoded audio data in a sequence of frames comprising an active frame and an inactive frame after the active frame; And

- an audio decoder for decoding encoded audio data for the active frame to generate a decoded multi-channel signal for the active frame,

- wherein the first audio source, the second audio source, the mixing noise source and the mixer are active in the inactive frame to generate a multi-channel signal for the inactive frame.

According to an aspect, the encoded audio signal for an active frame has a first set of coefficients describing a first number of frequency bins; And

- the encoded audio signal for the inactive frame has a second set of coefficients describing the second number of frequency bins,

- in this case, the first number of frequency bins is greater than the second number of frequency bins.

According to an aspect, the encoded audio data for an inactive frame comprises silence insert descriptor data comprising comfort noise data indicating signal energies for each channel of the two channels or for each of a first linear combination of the first and second channels and a second linear combination of the first and second channels for the inactive frame, and indicating coherence between the first channel and the second channel in the inactive frame, and

- wherein the mixer is configured to mix the mixing noise signal and the first audio signal or the second audio signal based on the comfort noise data indicating coherence, and

- wherein the multi-channel signal generator further comprises a signal modification module for modifying the first channel and the second channel of either the first audio signal, or the second audio signal, or the mixing noise signal, wherein the signal modification module is configured to be controlled by comfort noise data indicating signal energies for the first an audio channel and a second audio channel, or indicating signal energies for a first linear combination of the first and second channels and a second linear combination of the first and second channels.

According to an aspect, the audio data for the inactive frame comprises:

- a first frame of a silence insert descriptor for the first channel and a second frame of a silence insert descriptor for the second channel, wherein the first frame of the silence insert descriptor contains:

- comfort noise parameter data for the first channel and/or for the first linear combination of the first and second channels, and

- comfort noise generation support information for the first channel and the second channel, and

- wherein the second frame of the silence insertion descriptor contains:

- comfort noise parameter data for the second channel and/or for the second linear combination of the first and second channels, and

- coherence information indicating coherence between the first channel and the second channel in the inactive frame, and

- wherein the multi-channel signal generator contains a controller for controlling the generation of a multi-channel signal in an inactive frame using the comfort noise generation auxiliary information for the first frame of the silence insert descriptor to determine the comfort noise generation mode for the first channel and the second channel and/or for the first linear combination of the first and the second channels and a second linear combination of the first and second channels, using the coherence information in the second frame of the silence insert descriptor to specify the coherence between the first channel and the second channel in the inactive frame, and using the comfort noise parameter data from the first frame of the silence insert descriptor, and with using the comfort noise parameter data from the second frame of the silence insert descriptor to define the power situation of the first channel and the power situation of the second channel.

According to an aspect, the audio data for the inactive frame comprises:

- at least one silent insert descriptor frame for a first linear combination of the first and second channels and a second linear combination of the first and second channels,

- wherein at least one frame of the silence insertion descriptor contains:

- data (p_noise) of comfort noise parameters for the first linear combination of the first and second channels, and

- auxiliary information on the formation of comfortable noise for the second linear combination of the first and second channels,

- wherein the multi-channel signal generator comprises a controller for controlling the generation of a multi-channel signal in an inactive frame using the comfort noise generation auxiliary information for the first linear combination of the first and second channels and the second linear combination of the first and second channels, using coherence information in the second frame of the silence insert descriptor for defining coherence between the first channel and the second channel in the inactive frame, and using the comfort noise parameters data from at least one silence insertion descriptor frame, and using the comfort noise parameters data from at least one silence insertion descriptor frame to define a power situation the first channel and the energy situation of the second channel.

According to an aspect, a spectral-time converter for converting the resulting first channel and the resulting second channel, spectrally adjustable and coherently adjustable, into corresponding time domain representations that are to be combined or concatenated with the time domain representations of the corresponding channels of the decoded multi-channel signal for the active frame.

According to an aspect, the audio data for the inactive frame comprises:

- a silence insertion descriptor frame, wherein the silence insertion descriptor frame contains data on comfort noise parameters for the first and second channel and auxiliary information on the generation of comfort noise for the first channel and the second channel and/or for the first linear combination of the first and second channels and the second linear combination of the first and second channels and coherence information indicating coherence between the first channel and the second channel in the inactive frame, and

- wherein the multi-channel signal generator contains a controller for controlling the generation of a multi-channel signal in an inactive frame using the auxiliary information of comfort noise generation for the silence insertion descriptor frame to determine the comfort noise generation mode for the first channel and the second channel, using coherence information in the silence insertion descriptor frame for specifying coherence between the first channel and the second channel in the inactive frame, and using the comfort noise parameter data from the silence insert descriptor frame to specify the power situation of the first channel and the power situation of the second channel.

According to an aspect, the encoded audio data for an inactive frame comprises silence insert descriptor data comprising comfort noise data indicating the energy of signals for each channel in the middle/side view, and coherence data indicating the coherence between the first channel and the second channel in the left/right view, with wherein the multi-channel signal generator is configured to convert the middle/side representation of the signal energy into a left/right representation of the signal energy in the first channel and the second channel,

- wherein the mixer is configured to mix the mixing noise signal into the first audio signal and the second audio signal based on the coherence data to obtain the first channel and the second channel, and

- wherein the multi-channel signal generator further comprises a signal modification module configured to modify the first and second channels by generating the first and second channels based on the energy of the signals in the left/right region.

According to an aspect, the multi-channel signal generator is configured to, in the case that the audio data includes signaling signaling indicating that the energy in a side channel is less than a predetermined threshold value, zero the side channel coefficients.

According to an aspect, the audio data for the inactive frame comprises:

- at least one silent insertion descriptor frame, wherein the at least one silence insertion descriptor frame contains comfort noise parameter data for the middle and side channel and comfort noise generation auxiliary information for the middle and side channel and coherence information indicating coherence between the first channel and the second channel in the inactive frame, and

- wherein the multi-channel signal generator contains a controller for controlling the generation of a multi-channel signal in an inactive frame using the auxiliary information of comfort noise generation for the silence insertion descriptor frame to determine the comfort noise generation mode for the first channel and the second channel, using coherence information in the silence insertion descriptor frame to specify coherence between the first channel and the second channel in the inactive frame, and using these comfort noise parameters or a processed version thereof from the silence insert descriptor frame to define the power situation of the first channel and the power situation of the second channel.

According to an aspect, the multi-channel signal generator is configured to scale the energy coefficients of the signals for the first and second channel by means of gain information encoded using the comfort noise parameter data for the first and second channel.

According to an aspect, the multi-channel signal generator is configured to convert the generated multi-channel signal from a frequency domain version to a time domain version.

According to an aspect, the first audio source is a first noise source and the first audio signal is a first noise signal, or the second audio source is a second noise source and the second audio signal is a second noise signal,

- wherein the first noise source or the second noise source is configured to generate the first noise signal or the second noise signal such that the first noise signal or the second noise signal is at least partially correlated, and

- the mixing noise source is configured to generate a mixing noise signal with a first part of the mixing noise and a second part of the mixing noise, wherein the second part of the mixing noise is at least partially decorrelated with respect to the first part of the mixing noise; And

- the mixer is used for mixing the first part of the noise when mixing the mixing noise signal and the first audio signal to obtain the first channel, and for mixing the second part of the noise when mixing the mixing noise signal and the second audio signal to obtain the second channel.

According to an aspect, there is provided a method for generating a multi-channel signal having a first channel and a second channel comprising:

- generating a first audio signal using the first audio source;

- generating a second audio signal using a second audio source;

- generation of a mixing noise signal using a noise source during mixing; And

- mixing the mixing noise signal and the first audio signal to obtain a first channel, and mixing the mixing noise signal and the second audio signal to obtain a second channel.

In accordance with an aspect, an audio encoder is provided for generating an encoded multi-channel audio signal for a sequence of frames comprising an active frame and an inactive frame, the audio encoder comprising:

- an activity detector for analyzing the multi-channel signal to determine a frame of the sequence of frames as representing an inactive frame;

- a noise parameter calculation module for calculating first noise parameter data for a first channel of the multi-channel signal and for calculating second noise parameter data for a second channel of the multi-channel signal;

a coherence calculation module for calculating coherence data indicating a coherence situation between the first channel and the second channel in the inactive frame; And

- an output interface for generating an encoded multi-channel audio signal having encoded audio data for an active frame and, for an inactive frame, first noise parametric data, second noise parametric data, or a first linear combination of the first noise parametric data and second noise parametric data, and a second linear combination of the first parametric data noise and second parametric noise data, and coherence data.

According to an aspect, the coherence calculation module is configured to calculate a coherence value and quantize the coherence value to obtain a quantized coherence value, wherein the output interface is configured to use the quantized coherence value as coherence data in an encoded multi-channel signal.

According to an aspect, the coherence calculation module is configured to:

- calculating the real intermediate value and the imaginary intermediate value from the complex spectral values for the first channel and the second channel in the inactive frame;

- calculating a first energy value for the first channel and a second energy value for the second channel in the inactive frame; And

- calculating coherence data using a real intermediate value, an imaginary intermediate value, a first energy value and a second energy value, or

- smoothing at least one of the real intermediate value, the imaginary intermediate value, the first energy value and the second energy value and calculating the coherence data using the at least one smoothed value.

According to an aspect, the coherence calculation module is configured to calculate a real intermediate value as a sum over the real parts of products of complex spectral values for corresponding frequency bins of the first channel and the second channel in the inactive frame, or

- calculating the imaginary intermediate value as a sum over the imaginary parts of the products of complex spectral values for the corresponding frequency bins of the first channel and the second channel in the inactive frame.

According to an aspect, the coherence calculation module is configured to square the smoothed real intermediate value and square the smoothed imaginary intermediate value and sum the squared values to obtain a first component number,

- wherein the coherence calculation module is configured to multiply the smoothed first and second energy values to obtain a second component number, and combine the first and second component numbers to obtain a resultant number for the coherence value on which the coherence data is based.

According to an aspect, the coherence calculation module is configured to calculate the square root of the resultant number to obtain a coherence value on which the coherence data is based.

According to an aspect, the coherence calculation module is configured to quantize the coherence value using a uniform quantizer to obtain the quantized coherence value as n bits as coherence data.

According to an aspect, the output interface is configured to generate a first silence insert descriptor frame for a first channel and a second silence insert descriptor frame for a second channel, wherein the first silence insert descriptor frame contains comfort noise parameter data for the first channel and comfort noise generating auxiliary information for the first channel. channel and a second channel, and wherein the second frame of the silent insert descriptor contains comfort noise parameter data for the second channel and coherence information indicating coherence between the first channel and the second channel in the inactive frame, or

- wherein the output interface is configured to generate a silence insertion descriptor frame, wherein the silence insertion descriptor frame contains comfort noise parameter data for the first and second channel and auxiliary comfort noise generation information for the first channel and the second channel and coherence information indicating coherence between the first channel and the second channel in the inactive frame,

- or wherein the output interface is configured to generate a first frame of a silence insert descriptor for the first channel and a second channel and a second frame of a silence insert descriptor for the first channel and a second channel, wherein the first frame of the silence insert descriptor contains comfort noise parameter data for the first channel and a second channel and comfort noise generating auxiliary information for the first channel and a second channel, and wherein the second frame of the silence insert descriptor contains comfort noise parameter data for the first channel and the second channel and coherence information indicating coherence between the first channel and the second channel in the inactive frame.

According to an aspect, the uniform quantizer is configured to calculate n bits such that the value for n is equal to the value of the bits occupied by the comfort noise generation auxiliary information for the first frame of the silence insertion descriptor.

According to an aspect, the activity detector is configured to:

- analyzing the first channel of the multi-channel signal to classify the first channel as active or inactive, and

- analyzing the second channel of the multi-channel signal to classify the second channel as active or inactive, and

- determining a frame of the frame sequence to be an inactive frame if both the first channel and the second channel are classified as inactive.

According to an aspect, the noise parameter calculation module is configured to calculate first gain information for the first channel and second gain information for the second channel, and provide noise parameter data as the first gain information for the first channel and the second gain information.

According to an aspect, the noise parameter calculation module is configured to convert at least some of the first noise parameter data and the second noise parameter data from a left/right view to a middle/side view with a middle channel and a side channel.

According to an aspect, the noise parameter calculation module is configured to reconvert the middle/side representation of at least some of the first noise parametric data and the second noise parametric data into a left/right representation,

- wherein the noise parameter calculation module is configured to calculate, from the re-converted left/right representation, first gain information for the first channel and second gain information for the second channel, and provide first gain information for the first channel included in the first noise parameter data and the second information gain included in the second noise parameter data.

According to an aspect, the noise parameter calculation module is configured to calculate:

- first gain information by comparison:

- a version of the first parametric noise data for the first channel, re-converted from the middle/side view to the left/right view; With

- a version of the first parametric noise data for the first channel before conversion from the middle/side view to the left/right view; and/or

- second gain information by comparison:

- a version of the second parametric noise data for the second channel re-converted from the middle/side view to the left/right view; With

- a version of the second parametric noise data for the second channel before conversion from the middle/side view to the left/right view.

According to an aspect, the noise parameter calculation module is configured to compare the energy of a second linear combination between the first noise parameter data and the second noise parameter data with a given energy threshold value and:

- in the event that the energy of the second linear combination between the first noise parametric data and the second noise parametric data is greater than a predetermined energy threshold, the coefficients of the side channel noise shape vector are set to zero; And

- in the event that the energy of the second linear combination between the first noise parametric data and the second noise parametric data is less than a predetermined energy threshold, the coefficients of the side channel noise shape vector are stored.

According to an aspect, the audio encoder is configured to encode a second linear combination between the first noise parametric data and the second noise parametric data with fewer bits than the number of bits through which the first linear combination between the first noise parametric data and the second noise parametric data is encoded.

According to an aspect, the output interface is configured to:

- generating a coded multi-channel audio signal having coded audio data for the active frame using a first set of coefficients for a first number of frequency bins; And

- generating first parametric noise data, second parametric noise data, or a first linear combination of the first parametric noise data and second parametric noise data and a second linear combination of the first parametric noise data and second parametric noise data using a second set of coefficients describing a second number of frequency bins,

- in this case, the first number of frequency bins is greater than the second number of frequency bins.

According to an aspect, there is provided an audio encoding method for generating an encoded multi-channel audio signal for a sequence of frames comprising an active frame and an inactive frame, the method comprising:

- analyzing the multi-channel signal to determine the frame of the frame sequence as representing an inactive frame;

- calculating the first parametric noise data for the first channel of the multi-channel signal and/or for the first linear combination of the first and second channels of the multi-channel signal and calculating the second parametric noise data for the second channel of the multi-channel signal and/or for the second linear combination of the first and second channels of the multi-channel signal;

- calculating coherence data indicating a coherence situation between the first channel and the second channel in the inactive frame; And

- generating an encoded multi-channel audio signal having encoded audio data for the active frame and, for the inactive frame, first noise parametric data, second noise parametric data, and coherence data.

According to an aspect, a computer program is provided for implementing, when executed on a computer or processor, the above or below method.

According to an aspect, there is provided an encoded multi-channel audio signal organized in a sequence of frames, wherein the sequence of frames comprises an active frame and an inactive frame, wherein the encoded multi-channel audio signal comprises:

- encoded audio data for the active frame;

- first parametric noise data for the first channel in the inactive frame;

- second parametric noise data for the second channel in the inactive frame; And

- coherence data indicating the coherence situation between the first channel and the second channel in the inactive frame.

According to an aspect, the first audio source is a first noise source and the first audio signal is a first noise signal, or the second audio source is a second noise source and the second audio signal is a second noise signal,

- wherein the first noise source or the second noise source is configured to generate the first noise signal or the second noise signal in such a way that the first noise signal or the second noise signal is decorrelated with respect to the mixing noise signal.

According to an aspect, the mixer is configured to generate the first channel and the second channel such that the magnitude of the mixing noise signal in the first channel is equal to the magnitude of the mixing noise signal in the second channel or is within a range of 80-120 percent relative to the magnitude of the mixing noise signal in the second channel .

According to an aspect, the mixer includes a control input for receiving a control parameter, and wherein the mixer is configured to control the amount of a mixing noise signal in a first channel and a second channel in response to the control parameter.

According to an aspect, each of the first audio source, the second audio source, and the mixing noise source is a Gaussian noise source.

According to an aspect, the first audio source includes a first noise generator for generating a first audio signal as a first noise signal, wherein the second audio source includes a decorrelator for decorrelating the first noise signal to generate a second audio signal as a second noise signal, and wherein the mixing noise source comprises a second generator noise, or

- wherein the first audio source contains a first noise generator for generating a first audio signal as a first noise signal, wherein the second audio source contains a second noise generator for generating a second audio signal as a second noise signal, and wherein the mixing noise source contains a decorrelator for decorrelating the first noise signal signal or a second noise signal to form a mixing noise signal, or

- wherein one of the first audio source, the second audio source and the mixing noise source contains a noise generator for generating a noise signal, and the other of the first audio source, the second audio source and the mixing noise source contains a first decorrelator for decorrelating the noise signal, and furthermore one of the first audio source, the second audio source, and the mixing noise source comprises a second decorrelator for decorrelating the noise signal, wherein the first decorrelator and the second decorrelator are different from each other such that the output signals of the first decorrelator and the second decorrelator are decorrelated from each other, or

- wherein the first audio source contains a first noise generator, wherein the second audio source contains a second noise generator, and wherein the mixing noise source contains a third noise generator, wherein the first noise generator, the second noise generator and the third noise generator are configured to form mutually decorrelated noise signals.

According to an aspect, one of the first audio source, the second audio source, and the mixing noise source comprises a pseudo-random number sequence generator configured to generate a pseudo-random number sequence in response to a seed, and

- wherein at least two of the first audio source, the second audio source and the mixing noise source are configured to initialize the pseudo-random number sequence generator using different initial numbers.

In an aspect, at least one of the first audio source, the second audio source, and the mixing noise source is configured to operate using a previously stored noise table, or

- wherein at least one of the first audio source, the second audio source and the mixing noise source is configured to generate a complex spectrum for the frame using the first noise value for the real part and the second noise value for the imaginary part,

- in this case, if necessary, at least one noise generator is configured to generate a complex spectral noise value for the frequency element k of the resolution using for one of the real part and the imaginary part the first random value with index k, and using for the other of the real part and imaginary part of the second random value with index (k+M),

- wherein the first noise value and the second noise value are included in a noise array, for example, extracted from a random number sequence generator or from a noise table, or from a noise process, in the range from the initial index to the final index, and wherein the initial index is less than M, and wherein the final index is equal to or less than 2M, with M and k being integers.

According to the aspect, the mixer contains:

- a first amplitude element for influencing the amplitude of the first audio signal;

- a first adder for summing the output signal of the first amplitude element and at least part of the noise mixing signal;

- a second amplitude element for influencing the amplitude of the second audio signal;

- a second adder for summing the output of the second amplitude element and at least a portion of the mixing noise signal,

- in this case, the magnitude of the impact performed through the first amplitude element and the magnitude of the impact performed through the second amplitude element are equal to each other or differ by less than 20 percent relative to the value performed through the first amplitude element.

According to an aspect, the mixer includes a third amplitude element for influencing the amplitude of the mixing noise signal, wherein the amount of influence performed by the third amplitude element depends on the amount of influence performed by the first amplitude element or the second amplitude element such that the amount of influence performed by of the third amplitude element becomes larger when the amount of action performed by the first amplitude element or the amount of action performed by the second amplitude element becomes smaller.

According to an aspect, the multi-channel signal generator further comprises:

- an input interface for receiving encoded audio data in a sequence of frames containing an active frame and an inactive frame after the active frame; And

- an audio decoder for decoding encoded audio data for the active frame to generate a decoded multi-channel signal for the active frame,

- wherein the first audio source, the second audio source, the mixing noise source and the mixer are active in the inactive frame to generate a multi-channel signal for the inactive frame.

According to an aspect, the encoded audio data for the inactive frame comprises silence insert descriptor data comprising comfort noise data indicating signal energies for each channel of two channels for the inactive frame and indicating coherence between the first channel and the second channel in the inactive frame, and

- wherein the mixer is configured to mix the mixing noise signal and the first audio signal or the second audio signal based on the comfort noise data indicating coherence, and wherein the multi-channel signal generator further comprises a signal modification module for modifying the first channel and the second channel of either the first audio signal or the second audio signal, or noise mixing signal,

- wherein the signal modification module is configured to be controlled by comfort noise data indicating signal energies for the first audio channel and the second audio channel.

According to an aspect, the audio data for the inactive frame comprises:

- a first frame of a silence insert descriptor for the first channel and a second frame of a silence insert descriptor for the second channel, wherein the first frame of the silence insert descriptor contains comfort noise parameter data for the first channel and auxiliary comfort noise generation information for the first channel and the second channel, and wherein a second frame of the silent insert descriptor contains comfort noise parameter data for the second channel and coherence information indicating coherence between the first channel and the second channel in the inactive frame, and

- wherein the multi-channel signal generator contains a controller for controlling the generation of a multi-channel signal in an inactive frame using the auxiliary information of comfort noise generation for the first frame of the silence insert descriptor for determining the mode of comfort noise generation for the first channel and the second channel, using coherence information in the second frame of the descriptor silence inserts to set the coherence between the first channel and the second channel in the inactive frame, and using comfort noise generation data from the first frame of the silence insert descriptor, and using comfort noise generation data from the second frame of the silence insert descriptor to specify the energy situation of the first channel, and energy situation of the second channel.

According to an aspect, further comprising a spectral-time converter for converting the resulting first channel and the resulting second channel, spectrally adjustable and coherently adjustable, into corresponding time domain representations that are to be combined or concatenated with the time domain representations of the corresponding channels of the decoded multi-channel signal for the active frame .

According to an aspect, the audio data for the inactive frame comprises:

- a silent insertion descriptor frame, wherein the silence insertion descriptor frame contains comfort noise parameter data for the first and second channel and comfort noise generation auxiliary information for the first channel and the second channel and coherence information indicating coherence between the first channel and the second channel in the inactive frame, And

- wherein the multi-channel signal generator contains a controller for controlling the generation of a multi-channel signal in an inactive frame using the auxiliary information of comfort noise generation for the silence insertion descriptor frame to determine the comfort noise generation mode for the first channel and the second channel, using coherence information in the second insertion descriptor frame silence to set the coherence between the first channel and the second channel in the inactive frame, and using the comfort noise generation data from the silence insert descriptor frame to set the power situation of the first channel and the power situation of the second channel.

According to an aspect, the first audio source is a first noise source and the first audio signal is a first noise signal, or the second audio source is a second noise source and the second audio signal is a second noise signal,

- wherein the first noise source or the second noise source is configured to generate the first noise signal or the second noise signal such that the first noise signal or the second noise signal is at least partially correlated, and

- wherein the mixing noise source is configured to generate a mixing noise signal with a first part of the mixing noise and a second part of the mixing noise, wherein the second part of the mixing noise is at least partially decorrelated with respect to the first part of the mixing noise; And

- wherein the mixer is configured to mix the first part of the noise when mixing the mixing noise signal and the first audio signal to obtain the first channel, and mixing the second part of the noise when mixing the mixing noise signal and the second audio signal to obtain the second channel.

According to an aspect, a method for generating a multi-channel signal having a first channel and a second channel comprises:

- generating a first audio signal using the first audio source;

- generating a second audio signal using a second audio source;

- generation of a mixing noise signal using a noise source during mixing; And

- mixing the mixing noise signal and the first audio signal to obtain a first channel, and mixing the mixing noise signal and the second audio signal to obtain a second channel.

According to an aspect, an audio encoder is provided for generating an encoded multi-channel audio signal for a sequence of frames comprising an active frame and an inactive frame, the audio encoder comprising:

- an activity detector for analyzing the multi-channel signal to determine a frame of the sequence of frames as representing an inactive frame;

- a noise parameter calculation module for calculating first noise parameter data for a first channel of the multi-channel signal and for calculating second noise parameter data for a second channel of the multi-channel signal;

a coherence calculation module for calculating coherence data indicating a coherence situation between the first channel and the second channel in the inactive frame; And

- an output interface for generating an encoded multi-channel audio signal having encoded audio data for an active frame and, for an inactive frame, first noise parametric data, second noise parametric data, and coherence data.

According to an aspect, the coherence calculation module is configured to calculate a coherence value and quantize the coherence value to obtain a quantized coherence value, wherein the output interface is configured to use the quantized coherence value as coherence data in an encoded multi-channel signal.

According to an aspect, the coherence calculation module is configured to:

- calculating the real intermediate value and the imaginary intermediate value from the complex spectral values for the first channel and the second channel in the inactive frame;

- calculating a first energy value for the first channel and a second energy value for the second channel in the inactive frame; And

- calculating coherence data using a real intermediate value, an imaginary intermediate value, a first energy value and a second energy value, or

- smoothing at least one of the real intermediate value, the imaginary intermediate value, the first energy value and the second energy value and calculating the coherence data using the at least one smoothed value.

According to an aspect, the coherence calculation module is configured to calculate a real intermediate value as a sum over the real parts of products of complex spectral values for corresponding frequency bins of the first channel and the second channel in the inactive frame, or

- calculating the imaginary intermediate value as a sum over the imaginary parts of the products of complex spectral values for the corresponding frequency bins of the first channel and the second channel in the inactive frame.

According to an aspect, the coherence calculation module is configured to square the smoothed real intermediate value and square the smoothed imaginary intermediate value and sum the squared values to obtain a first component number,

- wherein the coherence calculation module is configured to multiply the smoothed first and second energy values to obtain a second component number, and combine the first and second component numbers to obtain a resultant number for the coherence value on which the coherence data is based.

According to an aspect, an audio encoder is provided, wherein the coherence calculation module is configured to calculate the square root of the resultant number to obtain a coherence value on which the coherence data is based.

According to an aspect, the coherence calculation module is configured to quantize the coherence value using a uniform quantizer to obtain the quantized coherence value as N bits as coherence data.

According to the aspect, an audio encoder is provided,

- wherein the output interface is configured to generate a first frame of a silence insertion descriptor for the first channel and a second frame of a silence insertion descriptor for the second channel, wherein the first frame of the silence insertion descriptor contains data on comfortable noise parameters for the first channel and auxiliary information on the formation of comfortable noise for the first channel and a second channel, and wherein the second frame of the silent insert descriptor contains comfort noise parameter data for the second channel and coherence information indicating coherence between the first channel and the second channel in the inactive frame, or

- wherein the output interface is configured to generate a silence insertion descriptor frame, wherein the silence insertion descriptor frame contains comfort noise parameter data for the first and second channel and auxiliary comfort noise generation information for the first channel and the second channel and coherence information indicating coherence between the first channel and the second channel in the inactive frame.

According to an aspect, the uniform quantizer is configured to calculate N bits such that the value for N is equal to the value of the bits occupied by the comfort noise generation auxiliary information for the first frame of the silence insert descriptor.

According to an aspect, an audio encoding method for generating an encoded multi-channel audio signal for a sequence of frames comprising an active frame and an inactive frame, the method comprising:

- analyzing the multi-channel signal to determine the frame of the frame sequence as representing an inactive frame;

- calculating first parametric noise data for a first channel of the multi-channel signal and calculating second parametric noise data for a second channel of the multi-channel signal;

- calculating coherence data indicating a coherence situation between the first channel and the second channel in the inactive frame; And

- generating an encoded multi-channel audio signal having encoded audio data for the active frame and, for the inactive frame, first noise parametric data, second noise parametric data, and coherence data.

According to an aspect, an encoded multi-channel audio signal arranged in a sequence of frames, wherein the sequence of frames comprises an active frame and an inactive frame, wherein the encoded multi-channel audio signal comprises:

- encoded audio data for the active frame;

- first parametric noise data for the first channel in the inactive frame;

- second parametric noise data for the second channel in the inactive frame; And

- coherence data indicating the coherence situation between the first channel and the second channel in the inactive frame.

Brief description of drawings

Fig. 1 shows an example in an encoder, in particular for classifying a frame as active or inactive.

Fig. 2 shows an example of an encoder and decoder.

Fig. 3a-3f show examples of multi-channel signal generators that can be used in a decoder.

Fig. 4 shows an example of an encoder and decoder.

Fig. 5 shows an example of a noise parameter quantization stage.

Fig. 6 shows an example of a noise parameter dequantization stage.

Carrying out the invention

This document describes, among other things, new technology for, for example, DTX and CNG for discretely encoded stereo signals. Instead of down-mixing the stereo signal, the noise parameters for both channels are extracted, jointly encoded, and transmitted. In a decoder (or more generally, a multi-channel oscillator), three independent comfort noise signals can be mixed based on a single wideband inter-channel coherence value that is transmitted, for example, along two sets of noise parameters. Some aspects of the examples may cover, in some examples, at least one of the following aspects:

- CNG in the decoder by mixing, for example, three independent noise signals. After decoding the stereo SID and reconstructing the noise parameters for the left and right channels, two noise signals can be generated, for example, as a mixture of correlated and decorrelated noise. To achieve this, one common noise source for both channels (serving as a source of correlated noise) and two separate noise sources (providing decorrelated noise) can be mixed together. The mixing process can be controlled by the inter-channel coherence value carried in the stereo SID. After mixing, the two mixed noise signals are spectrally shaped using the reconstructed noise parameters for the left and right channels, respectively.

- A combined encoding of noise parameters can be extracted from two channels of a stereo signal. To keep the stereo SID bit rate low, the noise parameters may be further compressed before being encoded into the stereo SID. This can be achieved, for example, by converting the left/right channel representation of the noise parameters into an average/side representation and encoding the side noise parameters with fewer bits than the average noise parameters.

- SID for two-channel DTX (stereo SID). This SID may contain noise parameters for both channels of the stereo signal along with a single wideband inter-channel coherence value and a flag indicating equal noise parameters for both channels.

It is shown that the following examples can be implemented in devices, systems, methods, controllers, and persistent storage units storing instructions that, when executed by a processor, cause the processor to perform the disclosed technologies (eg, methods such as flows of operations).

In particular, at least one of the following blocks may be controlled by a controller.

Examples

Before explaining aspects of the present examples in detail, a brief overview of some of the most important ones is provided.

1) Fig. 3a-3f show examples of multi-channel signal generators (eg, generated by at least one first signal or channel and one second audio signal or channel) that generate a multi-channel audio signal (eg, in a decoder). A multi-channel audio signal (initially in the form of multiple, decorrelated channels) may be affected (eg, scaled) by the amplitude element(s). The magnitude of the impact may be based on coherence data between the first and second audio signals estimated at the encoder. The first and second audio signals may be mixed with a common mixing signal (which may also be decorrelated and affected, eg, scaled, by the coherence data). The amount of impact for the mixing signal may be such that the first and second audio signals are scaled by a high weighting factor (for example, 1 or less, but, for example, close to 1), while the mixing signal is scaled by a low weighting factor (for example, 0 or more, but, for example, close to 0), and vice versa. The amount of influence for the mixing signal may be such that high coherence measured at the encoder causes the first and second audio signals to be scaled by a low weighting factor (eg, 0 or more, but, for example, close to 0), and high coherence measured at encoder causes the first and second audio signals to be scaled by a high weighting factor (eg, 1 or less, but eg, close to 1). Technologies according to FIG. 3a-3f can be used to implement a comfort noise generator (CNG).

2) Fig. 1, 2 and 4 show examples of encoders. The encoder can classify an audio frame as active or inactive. If the audio frame is inactive, then only some parametric noise data is encoded into the bit stream (eg, to provide a parametric noise shape that provides a parametric representation of the noise shape without having to provide the noise signal itself), and coherence data between the two channels may also be provided.

3) Fig. 2 and 4 show examples of decoders. The decoder can generate an audio signal (comfort noise), for example, by the following:

a. using one of the technologies shown in FIG. 3a-3f (1) above) (in particular taking into account the coherence value provided by the encoder and its application as a weighting factor in the amplitude element(s)); And

b. generating a shaped audio signal (comfort noise) using parametric noise data encoded in a bit stream.

In particular, it is not necessary for the encoder to provide the full audio signal for the inactive frame, but only a coherence value and a parametric representation of the noise shape, thereby reducing the number of bits that must be encoded in the bitstream.

Signal generator (eg decoder side), CNG

Fig. 3a-3f show examples of a CNG, or more generally, a multi-channel signal generator 200 for generating a multi-channel signal 204 having a first channel 201 and a second channel 203. (In the present description, the generated audio signals 221 and 223 are considered noise, but other types of signals are also possible , which do not represent noise). Refer first to FIG. 3f, which is general, while FIG. 3a-3e show specific examples.

The first audio source 211 may be a first noise source and may be referred to herein as generating a first audio signal 221, which may be a first noise signal. The mixing noise source 212 may generate the mixing noise signal 222 . The second audio source 213 may generate a second audio signal 223, which may be a second noise signal. The multi-channel signal generator 200 may mix a first audio signal 221 (first noise signal) with a mixing noise signal 222 and a second audio signal 223 (second noise signal) with a mixing noise signal 222. (In addition or alternatively, the first audio signal 221 may be mixed with a version 221a of the mixing noise signal 222, and the second audio signal 223 may be mixed with the version 221b of the mixing noise signal 222, wherein the versions 221a and 221b may differ, for example, by 20% from each other ; each of versions 221a and 221b may, for example, be an upscaled and/or downscaled version of the overall signal 222). Accordingly, the first channel 201 of the multi-channel signal 204 may be obtained from the first audio signal 221 (the first noise signal) and the mixing noise signal 222. Likewise, the second channel 203 of the multi-channel signal 204 may be obtained from the second audio signal 223 mixed with the noise mix signal 222. It should also be noted that the signals here may be in the frequency domain, and k denotes a particular index or coefficient (associated with a particular frequency bin).

As can be seen from Fig. 3a-3f, the first audio signal 221, the mixing noise signal 222, and the second audio signal 223 may be decorrelated with each other. This can be achieved, for example, by decorrelating the same signal (eg in a decorrelator) and/or by independently shaping noise (examples are given below).

The mixer 208 may be implemented to mix the first audio signal 221 and the second audio signal 223 with the noise mix signal 222 . The mixing may be of a signal summing type (eg, in adder stages 206-1 and 206-3) after the first audio signal 221, mixing noise signal 222, and second audio signal 223 have been weighted by scaling (eg, in amplitude elements 208-1, 208 -2, 208-3). Mixing is of the “sum after weighing” type. Fig. 3a-3f show the actual signal processing that is applied to generate the noise signals N _l [k] and N _r [k], where the sum element (+) denotes the summation over samples of the two signals (k is the index of the frequency bin).

Amplitude elements 208-1, 208-2, and 208-3 (or weighting elements or scaling elements) may be obtained, for example, by scaling the first audio signal 221, the mixing noise signal 222, and the second audio signal 223 by suitable coefficients and may output a weighted version 221' the first audio signal 221, a weighted version 222' of the mixing noise signal 222, and a weighted version 223' of the second audio signal 223. Suitable coefficients may be sqrt(coh) and sqrt(1-coh) and may be obtained, for example, from coherence information encoded in the transmission in the overhead of a particular descriptor frame (see also below) (sqrt refers here to the square root operation). The coherence "coh" is explained in detail below and, for example, may be what is indicated by "c" or "c _ind " or "c _q " below, for example, encoded in the 232 bit stream coherence information 404 (see below, in combination with Fig. 2 and 4). In particular, the noise mixing signal 222 may be scaled, for example, by a weighting factor that is the square root of the coherence value, while the first audio signal 221 and the second audio signal 222 may be scaled by a weighting factor that is the square root of the value. complementary to one of the coherence coh. Regardless, the noise mix signal 222 may be considered a common mode signal, a portion of which is mixed into a weighted version 221' of the first audio signal 221 and a weighted version 223' of the second audio signal 223 so as to obtain a first channel 201 of the multi-channel signal 204 and a second channel 203 of the multi-channel signal 204, respectively. In some cases, the first noise source 211 or the second noise source 213 may be configured to generate the first noise signal 221 or the second noise signal 223 such that the first noise signal 221 and/or the second noise signal 223 are decorrelated with respect to the mixing noise signal 222 ( see below with reference to Figures 3b-3e).

At least one or each of the first audio source 211, the second audio source 213, and the mixing noise source 212 may be a Gaussian noise source.

In the example of FIG. 3a, a first audio source 211 (here indicated by 211a) may comprise or couple to a first noise generator, and a second audio source 213 (213a) may comprise or couple to a second noise generator. The mixing noise source 212 (212a) may comprise or be coupled to a third noise generator. The first noise generator 211 (211a), the second noise generator 213 (213a), and the third noise generator 212 (212a) may generate mutually decorrelated noise signals.

In examples, at least one of the first audio source 211 (211a), the second audio source 213 (213a), and the mixing noise source 212 (212a) may operate using a pre-stored noise table, which may thereby provide a random sequence.

In some examples, at least one of the first audio source 211, the second audio source 213, and the mixing noise source 212 may generate a complex spectrum for a frame using a first noise value for the real part and a second noise value for the imaginary part. Optionally, the at least one noise generator may generate a complex spectral noise value (eg, a coefficient) for frequency bin k using, for one of the real part and the imaginary part, a first random value with index k, and using, for the other from the real part and the imaginary part, the second random value with index (k+M). The first noise value and the second noise value may be included in a noise array, for example, extracted from a random number sequence generator or noise table, or from a noise process, ranging from a start index to an end index, wherein the start index is less than M, and wherein the end index equal to or less than 2xM (which is twice the size of M). M and k can be integers (with k being the index of the particular frequency element of the bit resolution in the frequency domain representation of the signal).

Each audio source 211, 212, 213 may include at least one audio source generator (noise generator) that generates noise, for example, in terms of N ₁ [k], N ₂ [k], N ₃ [k].

The multi-channel signal generator 200 of FIG. 3a-3f can be used, for example, for a decoder 200a, 200b (200''). In particular, the multichannel signal generator 200 can be considered as part of the comfort noise generator (CNG) 220 in FIG. 4. The decoder 200 may be used generally to decode signals that are encoded by the encoder, either by generating signals to be generated by energy information obtained from the bit stream so as to generate an audio signal that corresponds to the original input audio signal input to encoder In some examples, a classification is provided between speech frames (or generally non-empty audio signals) and silence insert descriptor frames. As explained above and below, silent insert descriptor (SID) frames (so-called “inactive frames 308”, which may be encoded, for example, as SID frames 241 and/or 243) are generally provided under bit rate information and therefore are provided less frequently than normal speech frames (so-called “active frames 306”, see also below). In addition, the information that is present in silent insert descriptor frames (SID, inactive frames 308) is generally limited (and may substantially correspond to energy information regarding the signal).

However, it should be understood that it is possible to augment the content of SID frames with multi-channel noise 204 generated by the multi-channel signal generator. As such, audio sources 211, 212, 213 may process signals (eg, noise) that may be independent and decorrelated to each other. Despite this, the first audio signal 221, the mixing noise signal 222, and the second audio signal 223 can be scaled by the coherence information provided by the encoder and inserted into the bit stream. As can be seen from Fig. 3a-3f, the coherence value may be equal to the mixing noise signal 222, introduces a common mode signal into both the first audio signal 221 and the second audio signal 223, thereby allowing the first channel 201 and the second channel 203 of the multi-channel signal 204 to be obtained. coherence generally ranges from 0 to 1:

- A coherence of 0 means that the original first audio channel (eg, L, 301) and the second audio channel (eg, R, 303) are completely decorrelated with each other, and the amplitude element 208-2 of the noise signal 222 of the mix should be scaled by 0 noise mixing signal 222, which results in the first audio signal 221 and the second audio signal 223 not having to be mixed with the overall mode signal at all (by mixing with a signal that is constantly equal to 0), and the output channels 201, 203 having to be substantially equal to the first noise signal 221 and a second noise signal 223 of the multi-channel signal 204.

- Coherence equal to 1 means that the original first audio channel (for example, L, 301) and the second audio channel (for example, R, 303) must be the same, and the amplitude elements 208-1 and 208-3 must scale by 0 the input signals, and the first and second channels are then equal to the mixing noise signal 222 (which is scaled by 1 in amplitude element 208-2).

- Coherences intermediate between 0 and 1 should result in intermediate mixes between the two situations described above.

Some aspects and options of the 206 and/or CNG 220 mixer are explained below.

The first audio source (211) may be a first noise source and the first audio signal (221) may be a first noise signal, or the second audio source (213) may be a second noise source and the second audio signal (223) may be a second noise signal. The first noise source (211) or the second noise source (213) may be configured to generate a first noise signal (221) or a second noise signal (223) such that the first noise signal (221) or the second noise signal (223) is decorrelated relative to the mixing noise signal (222).

The mixer (206) may be configured to generate the first channel (201) and the second channel (203) such that the magnitude of the mixing noise signal (222) in the first channel (201) is equal to the magnitude of the mixing noise signal (222) in the second channel ( 203) or is within a range of 80-120 percent relative to the magnitude of the noise signal (222) of the mix in the second channel (203) (for example, portions 221a and 221b differ within the range of 80-120 percent from each other and from the original noise signal 222 mixes).

In some cases:

- the amount of action performed by the first amplitude element (208-1) and the amount of action performed by the second amplitude element (208-3) are equal to each other (for example, when there is no distinction between parts 221a and 221b), or

- the amount of action performed by the second amplitude element (208-3) differs by less than 20 percent relative to the amount performed by the first amplitude element (208-1) (for example, when the difference between parts 221a and 221b is less than 20%).

The mixer (206) and/or CNG 220 may include a control input for receiving a control parameter (404, c). The mixer (206) may therefore be configured to control the amount of the mixing noise signal (222) in the first channel (201) and the second channel (203) in response to the control parameter (404, c).

In fig. 3a-3f show that the mixing noise signal 222 is subject to a factor sqrt(coh), and the first and second audio signals 221, 223 are subject to a factor sqrt(1-coh).

As explained above, FIG. 3a shows a CNG 220a in which the first source 211a (211), the second source 213a (213), and the mixing noise source 212a (212) comprise different oscillators. This is not strictly required and several variations are possible.

To summarize:

1. First variant CNG 220b (Fig. 3b):

a. the first audio source 211b (211) may include a first noise generator for generating a first audio signal (221) as a first noise signal,

b. the second audio source 213b (213) may include a decorrelator for decorrelating the first noise signal (221) to generate a second audio signal (213) as the second noise signal (eg, the second audio signal is obtained from the first audio signal after decorrelation), and

c. The mixing noise source 212b (212) may comprise a second noise generator (which is initially decorrelated with respect to the first noise generator);

2. Second variant CNG 220c (Fig. 3c):

a. the first audio source 211c (211) may include a first noise generator for generating a first audio signal (221) as a first noise signal,

b. the second audio source 213c (213) may include a second noise generator for generating a second audio signal (223) as a second noise signal (eg, the second noise generator is initially decorrelated with respect to the first noise generator), and

c. The mixing noise source 212c (212) may comprise a decorrelator for decorrelating the first noise signal (221) or the second noise signal (223) to generate a mixing noise signal (222);

3. Third variant CNG 220d (Fig. 3d and 3e):

a. one of the first audio source 211d or 211e (211), the second audio source 213d or 213e (213) and the mixing noise source 212d or 212e (212) may include a noise generator for generating a noise signal,

b. the other of the first audio source 211d or 211e (211), the second audio source 213d or 213e (213), and the mixing noise source 212d or 212e (212) may comprise a first decorrelator for decorrelating the noise signal, and

c. yet another of the first audio source 211d or 211e (211), the second audio source 213d or 213e (213), and the mixing noise source 212d or 212e (212) may comprise a second decorrelator for decorrelating the noise signal,

d. the first decorrelator and the second decorrelator may be different from each other, such that the output signals of the first decorrelator and the second decorrelator are decorrelated from each other;

4. Fourth variant CNG 220 (Fig. 3a):

a. the first audio source 211a (211) contains a first noise generator,

b. second audio source 213a (213) includes a second noise generator,

c. mixing noise source 212a (212) includes a third noise generator,

d. the first noise generator, the second noise generator, and the third noise generator may be generated mutually decorrelated noise signals (eg, the tree generators are initially decorrelated from each other).

5. Fifth option:

a. one of the first audio source (211), the second audio source (213), and the mixing noise source (212) may comprise a pseudo-random number sequence generator for generating a pseudo-random number sequence in response to the seed,

b. at least two of the first audio source (211), the second audio source (213), and the mixing noise source (212) may initialize the pseudo-random number sequence generator using different seeds.

6. Sixth option:

a. at least one of the first audio source (211), the second audio source (213) and the mixing noise source (212) may be operated using a previously stored noise table,

b. optionally, at least one of the first audio source (211), the second audio source (213) and the mixing noise source (212) may generate a complex spectrum for the frame using the first noise value for the real part and the second noise value for the imaginary part,

c. optionally, the at least one noise generator may generate a complex spectral noise value for frequency bin k using, for one of the real part and the imaginary part, a first random value with index k, and using, for the other, a real part and an imaginary part part, a second random value with index (k+M) (the first noise value and the second noise value are included in a noise array, for example, extracted from a random number sequence generator or from a noise table, or from a noise process, in the range from the initial index to the final index, where the starting index is less than M, and where the ending index is equal to or less than 2xM, M and k are integers),

As can be seen from Fig. 4, decoder 200'' (200a, 200b) may include, in addition to the CNG 220 of FIG. 3, also an input interface 210 for receiving encoded audio data in a sequence of frames containing an active frame and an inactive frame after the active frame; and an audio decoder for decoding encoded audio data for the active frame to generate a decoded multi-channel signal for the active frame, wherein the first audio source 211, the second audio source 213, the mixing noise source 212, and the mixer 206 are active in the inactive frame to generate a multi-channel signal for the inactive frame.

In particular, active frames are frames that are classified by the encoder as having speech (or any other form of non-noise audio), and inactive frames are frames that are classified as having silence or only noise.

Any of the example CNG 220 (220a-220e) can be controlled by a suitable controller.

Encoder

The encoder is explained below. The encoder can encode active frames and inactive frames. For inactive frames, the encoder may encode parametric noise data (eg, noise shape and/or coherence value) without encoding the entire audio signal. It should be noted that the encoding of inactive audio frames may be reduced relative to active audio frames, so as to reduce the amount of information that must be encoded in the bit stream. Also, parametric noise data (eg, noise shape) for inactive frames may have less information per frequency band and/or may have fewer bins than data encoded in active frames. The noise parameter data may be determined in the left/right region or in another region (eg, in the middle/side region), for example, by providing a first linear combination between the noise parameter data of the first and second channels and a second linear combination between the noise parameter data of the first and second channels. channels (in some cases, it is also possible to provide gain information that is not associated with the first and second line combinations, but is determined in the left/right region). The first and second linear combinations are generally linearly independent of each other.

The encoder may include an activity detector that classifies whether a frame is active or inactive.

Fig. 1, 2 and 4 show examples of encoders 300a and 300b (which are also indicated at 300 if it is not necessary to distinguish between encoder 300a and encoder 300b). Each audio encoder 300 may generate an encoded multi-channel audio signal 232 for a sequence of frames of the input signal 304. The input signal 304 is here considered divided between the first channel 301 (also referred to as the left channel or "l", where "l" is a letter, the uppercase version of which represents "L" is the first letter of the word "left" in English) and the second channel 303 (or "r" where "r" is the letter whose uppercase version is "R" and is the first letter of the word "right" in English language).

The encoded multi-channel audio signal 232 may be specified in a sequence of frames, which, for example, may be in the time domain (for example, each sample "n" may represent a specific point in time, and samples of one frame may form a sequence, for example, the sampling sequence of the input audio signal or the sequence after filtering the input audio signal).

Encoder 300 (300a, 300b) may include an activity detector 380, which is not shown in FIG. 2 and 4 (although in some examples it is implemented in it), but is shown in FIG. 1. Fig. 1 shows that each frame of input signal 304 can be classified as either “active frame 306” or “inactive frame 308”. The inactive frame 308 is such that the signal is considered silent (and, for example, there is only silence or noise), while the active frame 306 may have some detection of a silent audio signal (for example, speech, music, etc.).

In the encoded multi-audio signal 232 encoded (e.g., in a bitstream) by the encoder 300, information regarding whether the frame is an active frame 306 or a silent frame 308 may be signaled, for example, in so-called "side information 402 (p_frame)" comfort noise generation”, also called “auxiliary information”.

Fig. 1 shows a preprocessing stage 360 that can determine (eg, classify) whether a frame is an active frame 306 or a silent frame 308. It should be noted here that the channels 301 and 303 of the input signal 304 are indicated by capital letters such as L (301, left channel) and R (303, right channel) to indicate that they are in the frequency domain. As can be seen in FIG. 1, a spectral analysis stage 370 may be applied (a first spectral analysis 370-1 to the first channel 301, L; and a second stage 370-3 to the second channel 303, R). Spectral analysis stage 370 may be performed on each frame of input signal 304 and may be based, for example, on harmonicity measurements. Specifically, in some examples, spectral analysis is performed by stage 370 on the first channel 301, and may be performed separately from the spectral analysis performed on the second channel 303 of the same frame.

In some cases, spectral analysis stage 370 may include calculating energy-dependent parameters such as average energy over a range of predefined frequency bands and overall average energy.

An activity detection stage 380 may be employed (which may be considered voice activity detection in the case of a voice search). The first activity detection stage 380-1 may be applied to the first channel 301 (and, in particular, to the measurements performed on the first channel), and the second activity detection stage 380-3 may be applied to the second channel 303 (and, in particular, to the measurements , performed for the second channel). In examples, activity detection stage 380 may estimate the energy of background noise in input signal 304 and use this estimate to calculate a signal-to-noise ratio that is compared to a signal-to-noise ratio threshold to determine whether a frame is classified as active or inactive (i.e., e. the calculated signal-to-noise ratio is above the signal-to-noise ratio threshold, which implies that the frame is classified as active; and the calculated signal-to-noise ratio is below the signal-to-noise ratio threshold, which implies that the frame is classified as inactive). In examples, stage 380 may compare the harmonicity obtained by spectral analysis stages 370-1 and 370-3, respectively, with one or two harmonicity thresholds (e.g., a first threshold for the first channel 301 and a second threshold for the second channel 303). In both cases, it may be possible to classify not only each frame, but also each channel of each frame as an active channel or an inactive channel.

Decision 381 may be executed, and based on it, a decision may be made (as identified by switch 381'') as to whether discrete stereo processing 306a or discontinuous transfer stereo (stereo-DTX) processing 306b should be performed. Particularly in the case of the active frame (and discrete stereo processing 306a), encoding may be performed according to any strategy or processing standard or process and is therefore not analyzed in more detail here. Most of the following explanation applies to the stereo DTX 306b.

Specifically, in the examples, a frame is classified (in step 381) as an inactive frame only if both channels 301 and 303 are classified as inactive in steps 380-1 and 380-3, respectively. Therefore, problems in the activity detection solution as explained above are eliminated. In particular, it is not necessary to signal the active/inactive classification for each channel for each frame (thereby reducing signaling) and synchronization between channels is internally obtained. In addition, if the decoder is as explained herein, it is possible to use the coherence between the first and second channels 301 and 303 and generate some noise signals that are correlated/decorrelated according to the coherence obtained for the signal 304. The elements of the encoder 300 are explained in detail below. 300a, 300b), which are used to encode the inactive frame. As explained, any other technology may be used to encode active frames 308 and is therefore not explained here.

In general, the encoder 300a, 300b (300) may include a noise parameter calculation module 3040 for computing noise parameter data 401, 403 for the first and second channels 301, 303. A noise parameter calculation module 3040 may calculate noise parameter data 401, 403 (e.g., indices and/or gains) for the first channel 301 and the second channel 303. The noise parameter calculation module 3040 may thus provide encoded audio data 232 in a sequence of frames, which may include active frames 306 and inactive frames 308 (which may be after active frames 306). Specifically, in the case of inactive frames 308, encoded audio data 232 may be encoded as one or two silent insertion descriptor (SID) frames 241, 243. In some examples (eg, FIG. 2), only one separate SID frame is provided, in some others, two SID frames are provided (eg, in FIG. 4).

Inactive frame 308 may include, but is not limited to, at least one of the following:

- auxiliary information for the formation of comfortable noise (for example, 402, p_frame);

- comfort noise parameter data 401 for the first channel 301 or a first linear combination of the comfort noise parameter data for the first channel 301 and the comfort noise parameter data for the second channel (v _{l, ind} , v _{m, ind} , p_noise, gain g _{l, q} );

- comfort noise parameter data 403 for the second channel 303 or a second linear combination of the comfort noise parameter data for the first channel 301 and the comfort noise parameter data (v _{r, ind} , v _{s, ind} , p_noise, gain g _{r, q} ) for the second channel;

- information (c, 404) coherence (coherence data).

In some examples, the first silent insert descriptor frame 241 may include the first two items of the above list, and the second silent insert descriptor frame 243 may include the last two features in particular data fields. However, different protocols may provide different data fields or different bitstream organization. However, in some cases (eg, FIG. 2), there may be only one separate inactive frame for noise parameters for both channels.

It is shown that the coherence information (e.g., the “silence insertion descriptor” portion) may include one single value (e.g., encoded in a small number of bits, e.g., four bits) that indicates coherence information (e.g., correlation data) for example, the coherence between the first channel 301 and the second channel 303 of the same inactive frame 308. On the other hand, the comfort noise parameter data 401, 403 may indicate, for each channel 301, 303, the signal energy for the inactive frame 308 (for example, it may actually provide an envelope) or in any case may provide noise shape information. The envelope or noise shape information may take the form of several coefficients for frequency elements of resolution and gain for each channel. The noise shape information may be obtained in stage 312 (see below) using the original input channels (301, 303), and thereafter mid/side encoding is performed on the noise shape parameter vectors. It is shown that in the decoder it may be possible to generate certain noise channels (eg, 201, 203, as indicated in FIG. 3), which can be affected by the coherence information 404. The noise channels 201, 203 generated by the CNG 220 (220a-220) can therefore be modified by the signal modification module 250 controlled by control noise data (comfort noise parameter data 401, 403, 2312) that indicate the signal energies for the first audio channel L _out and second audio channel R _out .

Audio encoder 300 (300a, 300b) may include a coherence calculator 320 that may obtain coherence information (404) to be encoded in the bit stream (eg, in signal 232, frame 241 or 243). The coherence information (c, 404) may indicate a coherence situation between the first channel 301 (eg, the left channel) and the second channel 303 (eg, the right channel) in the inactive frame 308. Examples of the foregoing will be explained.

Encoder 300 (300a, 300b) may include an output interface 310 configured to generate a multi-channel audio signal 232 (bit stream) with encoded audio data for an active frame 306 and, for an inactive frame 308, first parameter data 401 (p_noise, left) ( parametric comfort noise data) second parametric data (p_noise, right 403) noise and c (404) coherence data. The first parameter data 401 may be the parameter data of a first channel (eg, a left channel) or a first linear combination of a first and a second channel (eg, a middle channel). The second parameter data 403 may be parameter data of a second channel (eg, a right channel) or a second linear combination of the first and second channel (eg, a side channel) different from the first linear combination.

In the bit stream 232, ancillary information 402 may also be provided including an indicator as to whether the current frame is an active frame 306 or an inactive frame 308, for example, to inform the decoder as to the decoding technologies to be used.

In particular, in FIG. 4 shows a noise parameter calculation module 3040 (noise parameter calculation stage) including a first noise parameter calculation module stage 304-1 in which comfort noise parameter data 401 for the first channel 301 can be calculated, and a second parameter calculation module stage 304-3. noise, in which a second comfort noise parameter 403 for the second channel 303 can be calculated. FIG. 2 shows an example in which noise parameters are processed and quantized in a joint manner. The internals (e.g., converting noise shape vectors to M/S representation) are shown in FIG. 5. Essentially, one can have a first channel noise shape M and a second channel noise shape S, which can be encoded as mid indices and side indices, while the gain for the left channel noise shape 301 and the gain for the right channel noise shape 303 may also be encoded.

The coherence calculation unit 320 may calculate coherence data c (404) (coherence information) that indicates the coherence situation between the first channel L and the second channel R. In this case, the coherence calculation unit 320 may operate in the frequency domain.

As can be seen, the coherence calculating module 320 may include a channel coherence calculating stage 320'' in which a coherence value c (404) is obtained. After this, a uniform quantization stage 320'' can be used. Therefore, a quantized version c _ind of the coherence value c can be obtained.

Below are some explanations regarding how coherence should be obtained and how it should be quantized.

Coherence calculation module 320 may, in some examples:

- calculate the real intermediate value and the imaginary intermediate value from the complex spectral values for the first channel and the second channel (303) in the inactive frame;

- calculate a first energy value for the first channel and a second energy value for the second channel (303) in the inactive frame; And

- calculate coherence data (404, c) using a real intermediate value, an imaginary intermediate value, a first energy value and a second energy value, and/or

- smoothing at least one of the real intermediate value, the imaginary intermediate value, the first energy value and the second energy value and calculating coherence data using the at least one smoothed value.

The coherence calculator 320 may square the smoothed real intermediate value and square the smoothed imaginary intermediate value and sum the squared values to obtain a first component number. Coherence calculation module 320 may multiply the smoothed first and second energy values to obtain a second component number, and combine the first and second component numbers to obtain a resultant number for the coherence value on which the coherence data is based. The coherence calculation module 320 may calculate the square root of the resultant number to obtain a coherence value on which the coherence data is based. Below are examples of formulas.

The following explains how the shape of the noise waveform (or other signal energy) is obtained, which must be prepared by rendering in the decoder. Essentially, the shape (or other energy-related information) of the noise of the original input signal 302 is encoded, which in the decoder is to be applied to and shape the noise 252 so as to render noise 252 (the output audio signal) that resembles the original noise. signal 304.

It should first be noted that signal 304 as such is not encoded into the 232 bit stream by the encoder. However, noise information (eg, energy information, envelope information) can be encoded into a 232 bit stream to then generate a noise signal that has a noise shape encoded by an encoder.

The noise shape obtaining block 312 may be applied to the encoder input signal 304. The "noise shape acquisition" block 312 may compute a low-resolution parametric representation 1312 of the spectral envelope of the noise in the input signal 304. This may be accomplished, for example, by calculating the energy values in the frequency domain representation bands of the input signal 304. The energy values may be converted to logarithmic representation (if necessary) and can be compressed into a smaller number (N) of parameters, which are subsequently used in the decoder to generate comfortable noise. These low-resolution noise representations are referred to herein as “noise shapes 1312.” Therefore, what comes after the "get noise shape" block 312 should be understood not as representing the input signal 304, but as representing its noise shape (parametric representations of the spectral envelopes of the noise in the corresponding channels). This is important because the encoder can only transmit this lower resolution representation of the noise spectral envelope in the SID frame. Thus, in FIG. 2, the entire "noise parameter calculation module" portion (3040) can be understood to operate only on these vectors of noise-related parameters (eg, identified as _vl , _vr ,vm _,ind and vs _,ind ) and not on the representations signals for signal 304.

Fig. 5 shows an example of a “noise parameter calculation module” (combined noise shape quantization) portion 3040. L/RM/S converter stage 314 may be used to obtain an average channel representation v _m of the noise shape 1312 (the first linear combination of the L and R channel noise shapes) and a side channel representation v _r of the noise shape 1312 (the second linear combination of the noise shapes for the channel noise shapes L and R). The method for obtaining them is shown below. Accordingly, the noise waveform 304 may result in being separated into two channels v _m and v _r .

Then, in normalization stage 316, at least one of the middle channel representation v _m of noise shape 1312 and the side channel representation v _r of noise shape 1312 may be normalized to obtain a normalized version v _{m, n of} the middle channel representation v _m of noise shape 1312 and/or a normalized version of the v _{r, n} side channel representation of the v _r shape 1312 of the noise.

A quantization stage 318 (e.g., vector quantization, VQ) may then be applied to a normalized version of the signal 1304, e.g., in the form of a quantized version v _{m, ind} of the normalized channel average representation v _{m, n} of the noise form 1312 and a quantized version v _{s, ind} of the normalized side channel representation v _{s, n} of noise form 1312. Vector quantization can be used (for example, through a multi-stage vector quantizer). Therefore, the indices v _{m, ind} [k] (k is the index of a particular frequency bin) may describe the average representation of the noise shape, and the indices v _{s, ind} [k] may describe the side representation of the noise shape. The indices v _{m, ind} [k] and v _{s, ind} [k] can therefore be encoded in the 232 bit stream as a first linear combination of the comfort noise parameter data for the first channel and the comfort noise parameter data for the second channel and the second linear combination of data comfort noise parameters for the first channel and comfort noise parameters data for the second channel.

At dequantization stage 322, dequantization may be performed on a quantized version v _{m, ind} of the normalized mid-channel representation v _{m, n} of the noise shape 1312 and a quantized version v _{s, ind} of the normalized side channel representation v _{s, n} of the noise shape 1312.

The M/SL/R converter 324 can be applied to dequantized versions of the dequantized mid and side v _{m, q} and v _{s, q} representations of the noise shape 1312 to produce the v' _l and v' _r versions of the noise shape 1312 in the original (left and right) channels .

The gains g _l and g _r can then be calculated in stage 326. In particular, the gains are valid for all noise shape samples (v' _l and v' _r ) of the same channel of the same inactive frame 306. The gains g _l and g _r can be obtained by taking into account a population (or nearly a population) of frequency bins in the v' representations _l and v' _r noise shapes.

Gain g _l can be obtained by comparison:

- values of the frequency elements of the noise waveform resolution of the first channel 301 in the L/R region (before the L/R-M/S converter 314);

- with the values of the frequency elements of the resolution of the noise form 1312, after re-conversion in the L/R domain, of the first channel 301 (after the M/S-L/R converter 324).

Likewise, the gain g _r can be obtained by comparison:

- values of the noise shape coefficients of the second channel 303 in the L/R region (before the L/R-M/S converter 314);

- with the values of the noise form coefficients 1312, after repeated conversion in the L/R region, of the second channel 303 (after the M/S-L/R converter 324).

Below is an example of how to get boosts. However, the gain, in the linear domain, for example, can be proportional to the geometric mean of the fractions, with each fraction being a fraction between the noise shape factors of a particular channel in the L/R domain (before the L/R-M/S converter 314) and the coefficients the same channel after re-conversion into the L/R region after the 324 M/S-L/R converter. In the logarithmic domain, for each channel, the gain can be obtained as proportional to the algebraic average between the differences between the coefficients of the FD version of the noise shape in the L/R domain (before the 314 L/R-M/S converter) and the coefficients of the noise shape after reconversion in the L/ domain R after converter 324 M/S-L/R. In general, in the logarithmic or scalar domain, gain may provide a relationship between the left or right channel version of the noise waveform before the L/R-M/S transformation and the quantization with the left or right channel version of the noise waveform after dequantization and the inverse M/S-L/R transformation.

A quantization stage 328 may be applied to the gain g _l to obtain a quantized version thereof, indicated by g _{l, q} , to the gain g _r to obtain a quantized version thereof, indicated by g _{r, q} , which may be obtained from the unquantized gain g _r . The gains g _{l, q} and g _{r, q} may be encoded into a 232 bit stream (eg, as comfort noise parameter data 401 and/or 403), which must be read by a decoder.

In some examples, it is also possible to compare the energy of a vector of sidechannel noise shapes (e.g., before normalization, e.g., between stages 314 and 316) with a given energy threshold α (which may be a positive real value) (which in this case is 0.1. but may also be another value, for example a value between 0.05 and 0.15). At comparison block 435, it is possible to determine whether the side view v_sthe noise shape of the inactive frame 308 is enough energy. If the energy of the lateral representation v_snoise shape is less than the α energy threshold, then the binary results ("no-side flag"), as auxiliary information 402, are signaled in a 232 bit stream. Here the no-side flag is assumed to be 1 if the side view energy is v_snoise shapes are less than the α energy threshold, and the no-side=0 flag if the side view energy v_snoise shapes are greater than the α energy threshold. In some cases, the flag may be 1 or 0 according to a particular application in case the energy is exactly equal to the energy threshold. Block 436 negates the binary value of the no-side flag 436 (if block 436's input is 1, then output 436'' is 0; if block 436's input is 0, then output 436'' is 1). Block 436 is shown as outputting 436'' the opposite value of the flag. Accordingly, if the energy of the lateral representation v_snoise shape is greater than the energy threshold, then the value of 436'' can be equal to 1, and if the energy of the side view v_snoise shape is less than a given threshold value, then the value of 436'' is equal to 0. It should be noted that the dequantized value of v_{s, q} can be multiplied by the binary value 436''. This represents simply one possible way of obtaining that if the side view energy v_sform of noise is less than a given threshold value α of energy, then the resolution elements of the dequantized side representation v_{s, q} noise forms are artificially reset to zero (pin 437'' of block 437 should be equal to 0). On the other hand, if the energy of the lateral representation v_sthe noise shape is large enough (>α), then the output 437'' of block 437 (multiplier) can be exactly equal to v_{s, q}. Accordingly, if the energy of the lateral representation v_snoise shapes less than a given threshold α energy, side view v_snoise form (and in particular its dequantized version v_{s, q}) is not taken into account when obtaining left/right representations of the noise shape. (It is shown that in addition or alternatively, the decoder may also have a similar mechanism that nulls the side representation coefficients of the noise shape). It should be noted that the no-side flag may also be encoded in the 232 bit stream as part of the side information 402.

It should be noted that the energy of the side representation of the noise shape is shown as being measured (via block 435) before the noise shape is normalized (at block 316), and the energy is not normalized before it is compared with a threshold value. In principle, it could also be measured by block 435 after normalizing the noise shape (for example, block 435 could be input by v _{s, n} instead of v _s ).

With reference to the threshold value α used to compare the energy of the side representation of the noise shape, a value of 0.1, in some examples, can be arbitrarily selected. In examples, the threshold value α may be selected after experimentation and fine tuning (eg, through calibration). In some examples, in principle, any number that works for the number format (floating point or fixed point) or the precision of a particular implementation can be used. Therefore, the threshold value α may be an implementation-specific parameter that may be entered after calibration.

It should be noted that the output interface (310) may be configured to:

- generating an encoded multi-channel audio signal (232) having encoded audio data for the active frame (306) using a first set of coefficients for a first number of frequency bins; And

- generating first parametric noise data, second parametric noise data, or a first linear combination of the first parametric noise data and second parametric noise data and a second linear combination of the first parametric noise data and second parametric noise data using a second set of coefficients describing a second number of frequency bins,

- in this case, the first number of frequency bins is greater than the second number of frequency bins.

In fact, reduced resolution can be used for inactive frames, thereby further reducing the number of bits used to encode the bit stream. The same applies to the decoder.

Any of the encoder examples may be controlled by a suitable controller.

Decoder

The decoders are explained below according to examples. The decoder may include, for example, a comfort noise generator 220 (220a-220e) explained above, such as shown in FIG. 3a-3f. Comfort noise 204 (multi-channel audio signal) may be generated at signal modification module 250 to produce output signal 252. Of interest here is the demonstration of operations for generating noise in inactive frames 308 rather than operations for active frames 206.

Fig. 4 shows a first example of decoder 200'', here indicated by 200'' (200b). It should be noted that the decoder 200' includes a comfort noise generator 220, which may include a generator 220 (220a-220e) according to any of FIG. 3a-3f. Following the generator 220 (220a-220e), a signal modification module 250 (not shown, but shown in FIG. 4) may be present to generate a shaped multi-channel noise 204 according to the energy parameters encoded in the comfort noise parameter data (401, 403). Via decoder input interface 210, decoder 200'' may obtain from the 232 bit stream comfort noise parameter data (401, 403), which may include comfort noise parameter data describing signal energy (eg, for the first channel and the second channel, or for the first linear combination and a second linear combination of the first and second channels, the first and second linear combinations being linearly independent of each other). Through decoder input interface 210, decoder 200'' can receive coherence data 404 that indicates coherence between different channels of FIG. 4 shows that in the 232 bit stream for encoding inactive frames, two different silent descriptor frames 241 and 243 are provided, respectively, but it is possible to use more than two descriptor frames or just one separate descriptor frame. The output of the decoder 200b is a multi-channel output.

Referring to FIG. 2, the decoder 200'' (here indicated at 200a) is explained below, which is an example of a decoder 200 that can be used to generate an output signal 252, for example, in the form of noise.

First, for example, decoder 200a (200'') may include an input interface 210 for receiving encoded audio data 232 (bit stream) in a sequence of frames 306, 308 encoded by encoder 300a or 300b. Decoder 200a (200''), for example, may be or, more generally, be part of a multi-channel signal generator 200, which may be or include a comfort noise generator 220 (220a-220e) of any one of FIGS. 3a-3f.

First, fig. 2 shows a stereo comfort noise generator (CNG) 220 (220a-220e). In particular, the comfort noise generator 220 (220a-220e) may be similar to the comfort noise generator of FIG. 3a-3f or one of its variants. Here, coherence information 404 (e.g., c or more specifically c _q, also indicated by "coh" or c _ind ) obtained from encoder 300a or 300b may be used to generate a multi-channel signal 204 (in channels 201, 203) that explained above. The multi-channel signal 204 generated by the CNG 220 (220a-220e) may in fact be further modified, for example, by taking into account comfort noise parameter data 401 and 403, such as noise shape information for the first (left) channel and the second (right) channel of the multi-channel signal that should be generated. In particular, it is shown that it is possible to obtain the middle indices v _{m, ind} (401) and side indices v _{s, ind} (403) generated by the encoder 300a (and, in particular, by the noise parameter calculation module 3040) in stage 316 and /or 318, and gains g _{l, q} and g _{r, q} obtained in stage 326 and/or 328.

As shown in FIG. 2, supporting information 402 may allow it to be determined whether the current frame is an active frame 306 or an inactive frame 308. The elements of FIG. 2 refer to processing of inactive frames 308, and it is understood that any technology can be used to generate output in active frames 306, which is therefore not the purpose of this document.

As shown in FIG. 2, several examples of comfort noise data are obtained from a 232 bit stream. The comfort noise data may include, as explained above, coherence information 404 (data), parameters 401 and 403 (v _{m, ind} and v _{s, ind} ) indicating noise shape, and/or gains (g _{l, q} and g _{r, q} ).

Stage 212-C may dequantize the quantized version c _ind of the coherence information 404 to obtain the dequantized coherence information c _q .

Stage 2120 (combined noise shape dequantization) may allow other comfort noise data obtained from the 232 bit stream to be dequantized. Refer to FIG. 6. The dequantization stage 212 is formed by other dequantization stages, indicated herein by numerals 212-M, 212-S, 212-R, 212-L. Stage 212-M may dequantize channel average noise waveform parameters 401 and 403 to obtain dequantized noise waveform parameters v _{m, q} and v _{s, q} . Stage 212-S may provide a dequantized version of the v _{s, q} parameters 403 (v _{s, ind} ) of the side channel noise waveform. In some examples, a no-side flag may be used to reset the output of stage 212-S in the event that block 435 in encoder 300a detects the energy of the noise waveform vector v _s as less than a predetermined threshold value α. In the event that the energy is less than a given threshold α and the no-side flag signals this, the dequantized version v _{s, q} of the noise waveform vector v _s can be set to zero (which is conceptually shown as multiplying by flag 536' obtained from block 536. which has the same function as encoder block 436, even though block 536 actually reads the no-side flag encoded in the auxiliary information of bitstream 232 without performing a comparison with the threshold α at all). Therefore, if the side channel energy at the encoder is determined to be less than a predetermined threshold value α, the dequantized version v _{s, q} of the noise shape vector v _s is artificially nulled and the value at output 537'' of scaler block 537 is zero. Otherwise, if the energy is greater than a predetermined threshold, then the output 537'' is equal to the quantized version v _{s, q} of the side channel indices 403 (v _{s, ind} ) of the side channel noise waveform. In other words, the values of the noise waveform vector v _{s, ind} are neglected if the side channel energy is below a predetermined energy threshold α.

The M/SL/R stage 516 performs the M/SL/R transformation so as to obtain the L/R v' _l , v' _r version of the parametric data (noise waveform). An amplifier stage 518 (formed by stages 518-L and 518-L) may then be used such that in stage 518-L, channel v' _l is scaled by gain g _{l, d} , whereas in stage 518-R, channel v' _r is scaled by the gain of g _{r, q} . Therefore, the power channels v _{l, q} and v _{r, q} can be obtained as the output of the amplifier stage 518. Stage blocks 518-L and 518-R are shown with "+" since the transfer of values is assumed to be in the logarithmic domain, and the scaling of the values therefore it is indicated in the summation. However, the amplifier stage 518 indicates that the reconstructed noise waveform vectors v _{l, q} and v _{r, q} are scaled. The reconstructed noise shape vectors v _{l, q} and v _{r, q} are here collectively indicated by 2312 and are a reconstructed version of the noise shape 1312 originally obtained by the “get noise shape” block 312 in the encoder. In general terms, each gain is constant across all indices (coefficients) of the same channel of the same inactive frame.

It should be noted that the indices v _{m, ind} , v _{s, ind} and gains g _{l, q} , g _{r, q} are noise shape factors and provide information regarding frame energy. They essentially mean the parametric data associated with the input signal 304 that is used to generate the signal 252, but they do not represent the signal 304 or the signal 252 that is to be generated. In other words, the noise channels v _{r, q} and v _{l, q} describe an envelope that should be applied to the multi-channel signal 204 generated by the CNG 220.

Returning to FIG. 2, the reconstructed vectors v _{l, q} and v _{r, q} of the noise shapes (2312) are used in the signal modification module 250 to obtain the modified signal 252 through noise shaping 204. In particular, the first channel 201 of the generated noise 204 may be formed by the channel v _{l, q} in stage 250-L and channel 203 of the generated noise 204 in stage 250-R to produce multi-channel audio output 252 (L _out and R _out ).

In the examples, the comfort noise signal 204 itself is not generated in the logarithmic domain: only noise shapes can use the logarithmic representation. Conversion from logarithmic domain to linear domain can be performed (although not shown).

Also, conversion from frequency domain to time domain can be performed (although not shown).

Decoder 200'' (200a, 200b) may also include a time-spectral converter (eg, signal modification module 250) for converting the resulting first channel 201 and the resulting second channel 203, spectrally adjusted and coherently adjusted, into corresponding time domain representations that must be combined or concatenated with the time domain representations of the corresponding channels of the decoded multi-channel signal for the active frame. This conversion of the generated comfort noise into a time domain signal occurs after the signal modification block 250 in FIG. 2. "Combination or concatenation" with a part essentially means that before or after an inactive frame that uses one of these CNG technologies, active frames may also be provided (another processing path in Fig. 1), and to form a continuous output without any interruptions or audible clicks, etc., the frames should be concatenated correctly.

In some examples:

- the encoded audio signal (232) for the active frame (306) has a first set of coefficients describing a first number of frequency bins; And

- the encoded audio signal (232) for the inactive frame (308) has a second set of coefficients describing a second number of frequency bins.

The first number of frequency bins may be greater than the second number of frequency bins.

Any of the example decoder may be controlled by a suitable controller.

Processing stages: first version

The noise parameters encoded in two SID frames for two channels are calculated as specified in EVS [6], for example, according to LP-CNG or FD-CNG or both. The generation of noise energy in the decoder is also the same with EVS, for example, according to LP-CNG or FD-CNG or both of them.

In the encoder, the coherence of the two channels is further calculated, uniformly quantized using four bits, and sent in a 232-bit stream. At the decoder, operation in the CNG mode can then be controlled by the transmitted coherence value 404. Three sources N ₁ , N ₂ , N ₃ (211a, 212a, 213a; 211b, 212b, 213b; 211c, 212c, 213c; 211d, 212d, 213d; 211e, 212e, 213e) of Gaussian noise can be used as shown in FIG. . 3a-3f. When channel coherence is high, mostly correlated noise may be added to both channels 221' and 223', while more decorrelated noise is added if coherence 404 is low.

For all inactive frames 306, parameters for generating comfort noise (noise parameters) may be continuously estimated in the encoder (eg, 300, 300a, 300b). This may be accomplished, for example, by applying a frequency domain noise estimation algorithm (e.g., [8]), e.g., as described in [6] separately for both input channels (e.g., 301, 303) to compute two sets of noise parameters (e.g. , 401, 403), which are also explained as parametric noise data. In addition, the coherence (c, 404) of the two channels can be calculated (for example, in the coherence calculation module 320) as follows: Given the M-point DFT spectra of the two input channels (L, R can be 301, 303), four intermediate values can be calculated, for example:

,

and the energies of two channels:

Here, it can be M=256, denotes the real part of a complex number, denotes the imaginary part of a complex number, and denotes complex conjugation. These intermediate values can then be smoothed, for example using the corresponding values from the previous frame:

This pass may be part of a "channel coherence calculation" block 320'' in the encoder. It is a temporary smoothing of internal parameters to prevent large sudden jumps in parameters between frames. In other words, a low pass filter is applied to the parameters here.

Instead of constants 0.95 and 0.05, other constants in the range 0.95±0.03 and 0.05±0.03 can be used.

Alternatively, you can specify:

,

Where , and β+γ=1, for example, β=0.95 and γ=0.05.

The coherence (c, 404) (which can be between 0 and 1) can then be calculated (for example, in the calculation module (320) of the coherence as follows:

,

and uniformly quantized (for example, in quantizer 320'') using, for example, four bits as follows:

Encoding of the estimated noise parameters 1312, 2312 for both channels can be done separately, for example, as described in [6]. The two SID frames 241, 243 can then be encoded and sent to the decoder. The first SID frame 241 may contain estimated L channel noise parameters 401 and (eg, four) bits of ancillary information 402, eg, as described in [6]. In the second SID frame 243, the R channel noise parameters 403 may be sent along with the four-bit quantized coherence value c, 404 (different numbers of bits may be selected in other examples).

At a decoder (eg, 200'', 200a, 200b), both the SID frame noise parameters (401, 403) and the first frame side information 402 may be decoded, for example, as described in [6]. The coherence value 404 in the second frame may be dequantized in stage 212-C as follows:

(in Fig. 2, is replaced by c _q ).

To generate comfort noise (e.g., in the generator 220 or in any of the generators 220a-220e, which may include the generator of any of FIGS. 3a-3e), according to the example, three Gaussian noise sources 211, 212, 213 can be used, as shown in Fig. 3. The noise sources 211, 212, 213 may be adaptively summed together (eg, in adder stages 206-1 and 206-3), eg, based on a coherence value (c, 404). The DFT spectra of the left and right channel noise signals N _l [k], N _r [k] can be calculated as follows:

,

Where (which is the index of a particular frequency bin, whereas each channel has M frequency bins), and j ² = -1 (ie j is the imaginary unit), and "x" is a normal multiplication. Here, “frequency element” means the number of complex values in the spectra N _l and N _r , respectively. M is the length of the FFT or DFT transform that is used, so that the length of the spectra is M. It should be noted that the noise inserted into the real part and the noise inserted into the imaginary part may be different. Thus, for a spectrum length M, 2xM values (one real and one imaginary) generated from each noise source are required. Or, in other words: N _l and N _r are complex-valued vectors of length M, while N ₁ , N ₂ and N ₃ are real-valued vectors of length 2xM.

Subsequently, the noise signal 204 in the two channels is spectrally generated (eg, in stages 250-L, 250-R in FIG. 2) using the corresponding noise parameters (2312) decoded from the corresponding SID frame and then converted back to the time domain (eg , as described in [6]) to generate comfortable frequency domain noise.

Any of the example processing may be performed by a suitable controller.

Processing stages: second version

Aspects of the processing steps, as explained above, may be integrated with at least one of the following aspects. Here it is mainly necessary to refer to FIG. 2 and 5, but also refer to FIGS. 4.

A block diagram of the overall encoder infrastructure is illustrated in FIG. 1. For each frame in the encoder, the current signal can be classified as active or inactive by separately performing VAD for each channel, as described in [6]. The VAD solution can then synchronize between the two channels. In the examples, a frame is classified as inactive frame 308 only if both channels are classified as inactive. Otherwise, it is classified as active and both channels are jointly encoded in an MDCT based system using M/S for each frequency band as described in [10]. When switching from an active frame to an inactive frame, signals may enter the SID encoding path, as shown in FIG. 3.

Parameters (eg, 1312, 401, 403, q _{l, q} , g _{r, q} ) for generating comfort noise (eg, noise parameters) can be continuously estimated in the encoder (eg, 300, 300a, 300b) for active and inactive frames ( 306, 308). This may be accomplished, for example, by applying a frequency domain noise estimation process, such as the process explained in [8] and/or as described in [6], for example, separately for both input channels 301, 303 to calculate two sets of parameters noise, including forms (Mi or 401 and/or I _s or 403) of spectral noise, for example, in the logarithmic domain for each channel.

In addition, the coherence (404, c) of the two channels can be calculated (for example, in the coherence calculation module 320) as follows: Given the M-point DFT spectra of the two input channels , the four intermediate values can be calculated as:

,

and the energies of two channels:

Here, it can be M=256 (other values for M can be used), denotes the real part of a complex number, denotes the imaginary part of a complex number, and denotes complex conjugation. These intermediate values are then smoothed based on 10 ms subframes. If denotes the corresponding value from the previous subframe, the smoothed values can be calculated as follows:

Instead of constants 0.95 and 0.05, other constants in the range 0.95±0.03 and 0.05±0.03 can be used.

Alternatively, you can specify:

,

Where , β+xγ=1, for example β=0.95 and γ=0.95 (β> γ, for example β> 3xγ or β> 6xγ).

Coherence can then be calculated (for example in 320'') as follows:

,

and uniformly quantized (e.g. at 320'') using four bits (but different numbers of bits are possible) as follows:

,

Where denotes rounding down to the nearest whole number (minimum level function)

Encoding of the estimated noise shapes of both channels can be performed jointly. From the left (v _l ) and right (v _r ) channel noise shapes, different channels can be obtained (for example, through a linear combination), for example, the middle channel (v _m ) noise shape and the side channel (v _s ) noise shape can be calculated , (for example, at block 314) as follows:

,

where N denotes the length of vectors of noise shapes (eg, for each inactive frame 308), for example, in the frequency domain. N denotes the length of a noise shape vector, for example, estimated as specified in EVS [6], which can be between 17 and 24. Noise shape vectors can be thought of as a more compact representation of the spectral envelope of the noise in the input frame, or, more abstractly, as as a parametric spectral description of the noise signal using N parameters. N is not related to the length of the FFT or DFT transform.

These noise waveforms can then be normalized (eg, at stage 316) and/or quantized. For example, they can be vector quantized (for example, in stage 318), for example, using multistage vector quantizers (MSVQ) (an example is described in [6, page 442].

The MSVQ used in stage 318 to quantize the shape of v _m (to obtain v _{m, ind} 401) may have 6 stages (but another number of stages is possible) and/or use 37 bits (but another number of bits are possible), for example, when implemented for mono channels in [6], while the MSVQ used, in stage 318, to quantize the shape of v _s (to obtain v _{s, ind} 403) can be reduced to 4 stages (or, in any case, to the number of stages less than the number of stages used in stage 318) and/or may use a total of 25 bits (or, in any case, a number of bits less than the number of bits used in stage 318 to encode the form v _m ).

MSVQ codebook indices may be carried in the bit stream (eg, in data 232 and, more specifically, in comfort noise parameter data 401, 403). The indices are then dequantized, resulting in dequantized forms v _{m, q} and v _{m, q} of noise.

In the case of background noise, which is a single noise source in the center of the stereo image, the estimated noise shapes v _m , v _s of both channels are expected to be almost equal or even equal. The resulting S-channel noise waveform should then contain only zeros. However, the vector quantizer (stage 322) used to quantize the current implementation of v _s may be such that it cannot model a vector with all zeros, and after dequantization, the dequantized shape (v _{s, q} ) of the noise v _s results may no longer be all zeros. This may lead to perceptual problems with the representation of such centered background noise. To work around this shortcoming of VQ 322, the value of no_side (no_side flag) may be calculated (and may also be signaled in the bit stream) depending on the energy of the unquantized shape vector v _s (eg, the energy of the noise shape vector v _s after stage 314 and/ or before step 316). The no_side flag could be as follows:

The threshold α energy value may be, just by way of example, 0.1 or another value in the range [0.05, 0.15]. However, the threshold value α can be arbitrary, and in implementation may depend on the number format used (eg, fixed point or floating point) and/or on the signal normalizations that may be used. In the examples, a positive real value may be used depending on how sharp the "silence" S-channel definition used is. Therefore, the interval can be (0, 1). The value no_side can be used to indicate whether the noise waveform v _s should be used to reconstruct the channel noise waveforms v _l and vr (eg, in a decoder). If no_side is 1, the dequantized shape of v _s is set to zero (eg, by scaling the channel v _{s, q} to the value 436'' in FIG. 2, which is the logical value NOT(no_side)); no_side is carried (signalled) in a 232 bit stream, for example, as side information 402. Then, an inverse M/S transform (for example, stage 324) can be applied to the dequantized vectors v _{m, q} and v _{s, q of} noise shapes (the latter may be replaced by 0 if the energy is low, therefore indicated by 437'' in Fig. 2) to obtain the intermediate vectors v' _l and v' _r as follows:

Using these intermediate vectors v' _l and v' _r and the unquantized vectors v _l and vr of the noise shapes, two gain values are calculated as follows:

.

The two gain values can then be linearly quantized (for example, at stage 328) as follows:

,

(other quantizations are possible).

The quantized gains may be encoded in the SID bitstream (e.g., as part of the comfort noise parameter data 401 or 403, and more specifically, g _{l, q} may be a portion of the first noise parametric data, and g _{r, q} may be a portion of the second parametric noise data noise data), for example, using seven bits for the gain value g _{l, q} and/or seven bits for the gain value g _{r, q} (different values are also possible for each gain value).

At a decoder (e.g., 200'', 200a, 200b), the quantized noise shape vectors (e.g., a portion of the comfort noise parameter data 401 or 403 and, more specifically, the first noise parameter data and the second noise parameter data) may be dequantized, e.g. stages 212 (specifically, in any of the partial stages 212-M, 212-S).

The gain values may be dequantized, for example, in stage 212 (specifically, in any of the partial stages 212-L, 212-R) as follows:

(the value of 45 depends on the quantization and may differ for different quantizations). (In Fig. 2, g _{l, d} and g _{r, d} are used instead of g _{l, deq} and g _{r, deq} ).

The coherence value 404 may be dequantized (eg, in stage 212-C) as follows:

If the no_side flag (in supporting information 402) is 1, the dequantized form v _{s, q} v _s is set to zero (value 537'') before calculating the intermediate vectors v' _l and v' _r (eg, in step 516). The corresponding gain value is then summed with all elements of the corresponding intermediate vector to form the dequantized noise forms v _{l, q} and v _{r, q} , collectively indicated at 522, as follows:

(Summation is due to the fact that we are in the logarithmic domain, and corresponds to multiplication by a coefficient in the linear domain).

To generate comfort noise, three Gaussian noise sources N ₁ , N ₂ , N ₃ (for example, 211a, 212a, 213a in Fig. 3a, 211b, 212b, 212c in Fig. 3b, etc.) can be used as shown in any of FIG. 3a-3f (or any of the other technologies may be used). When channel coherence is high, mostly correlated noise is added to both channels, whereas more decorrelated noise is added if coherence is low.

Using three noise sources, the DFT spectra of the left and right channel noise signals N _l (201) and N _r (203) can be calculated as follows:

,

Where , and j ² =-1. Here, M denotes the length of the DFT block. To generate independent noise in the real and imaginary parts of the complex spectrum, 2xM values (two for each frequency bin) per frame must be generated by each noise source. Therefore, N ₁ , N ₂ and N ₃ (respectively at 211, 212, 213 in Fig. 3f) can be considered as real-valued noise vectors having a length of 2xM, while N _r and N _k (respectively at 201, 203 ) are complex-valued vectors with length M.

Subsequently, the noise signals in the two channels can be spectrally conditioned (eg, in signal modification module 252) using the appropriate noise shape (v _{l, q} or v _{r, q} ) decoded from the 232 bit stream, and then converted back from the logarithmic domain into the scalar domain and from the frequency domain into the time domain, for example as described in [6], to generate a stereo comfort noise signal.

Any of the example processing may be performed by a suitable controller.

Some benefits

The present invention may provide technology for generating stereo comfort noise, particularly suitable for discrete stereo encoding schemes. By combining encoding and transmitting noise shape parameters for both channels, stereo CNG can be used without the need for mono downmixing.

Together with two separate sets of noise parameters, mixing one common and two separate noise sources controlled by a single coherence value provides the ability to reliably reconstruct a stereo image of background noise without the need to convey highly detailed stereo parameters typically found only in parametric audio encoders. Since only this one parameter is used, SID encoding is simple without the need for complex compression techniques, while keeping the SID frame size small.

Some important aspects:

In some examples, at least one of the following aspects is obtained:

1. Generate comfort noise for a stereo signal by mixing three Gaussian noise sources, one for each channel, and a third common noise source to create correlated background noise.

2. Control the mixing of noise sources with the coherence value that is transmitted with the SID frame.

3. Transfer of separate noise shape parameters for both stereo channels by means of combined noise shape encoding using the M/S method. Reducing the bit rate of SID frames by S-shape encoding with fewer bits than M.

Other technologies

It is also possible to implement a method for generating a multi-channel signal having a first channel and a second channel containing:

- generating a first audio signal using the first audio source;

- generating a second audio signal using a second audio source;

- generation of a mixing noise signal using a noise source during mixing; And

- mixing the mixing noise signal and the first audio signal to obtain a first channel, and mixing the mixing noise signal and the second audio signal to obtain a second channel.

An audio encoding method may also be implemented to generate an encoded multi-channel audio signal for a sequence of frames comprising an active frame and an inactive frame, the method comprising:

- analyzing the multi-channel signal to determine the frame of the frame sequence as representing an inactive frame;

- calculating first parametric noise data for a first channel of the multi-channel signal and calculating second parametric noise data for a second channel of the multi-channel signal;

- calculating coherence data indicating a coherence situation between the first channel and the second channel in the inactive frame; And

- generating an encoded multi-channel audio signal having encoded audio data for the active frame and, for the inactive frame, first noise parametric data, second noise parametric data, and coherence data.

The invention may also be implemented in a persistent storage module storing instructions which, when executed by a computer (or processor or controller), cause the computer (or processor or controller) to carry out the above method.

The invention may also be implemented in a multi-channel audio signal organized into a sequence of frames, wherein the sequence of frames comprises an active frame and an inactive frame, wherein the encoded multi-channel audio signal comprises:

- encoded audio data for the active frame;

- first parametric noise data for the first channel in the inactive frame;

- second parametric noise data for the second channel in the inactive frame; And

- coherence data indicating the coherence situation between the first channel and the second channel in the inactive frame. The multi-channel audio signal may be obtained using one of the technologies disclosed above and/or below.

Advantages of the Embodiments

Inserting a common noise source for two channels to simulate correlated noise to generate the final comfort noise plays an important role in simulating stereo recording of background noise.

Embodiments of the invention may also be considered a procedure for generating comfort noise for a stereo signal by mixing three Gaussian noise sources, one for each channel, and a third common noise source to create correlated background noise, or, additionally or separately, controlling the mixing of the noise sources with a coherence value. which is transmitted with the SID frame, either, additionally or separately, as follows: In a stereo system, separate background noise generation results in completely decorrelated noise that sounds unpleasant and significantly different from the actual background noise, causing abrupt audible transitions when switching to/from the background active mode into DTX mode backgrounds. In an embodiment, on the encoder side, in addition to the noise parameters, the coherence of the two channels is calculated, uniformly quantized, and summed with the SID frame. At the decoder, CNG mode operation is then controlled by the transmitted coherence value. Three sources N_1, N_2, N_3 of Gaussian noise are used; when channel coherence is high, mostly correlated noise is added to both channels, whereas more decorrelated noise is added if coherence is low.

It should be noted here that all alternatives or aspects explained above and all aspects specified by independent claims in the following claims can be used separately, i.e. without alternatives or purposes other than the intended alternative, purpose or independent claim. However, in other embodiments, two or more of the alternatives or aspects or independent claims may be combined with each other, and, in other embodiments, all aspects or alternatives and all independent claims may be combined with each other.

The encoded signal according to the invention may be stored on a digital storage medium or a non-transitory storage medium, or may be transmitted over a transmission medium such as a wireless transmission medium or a wired transmission medium, for example, the Internet.

Although certain aspects are described in the context of an apparatus, it will be appreciated that these aspects also represent a description of the corresponding method, wherein the block or apparatus corresponds to a method step or a feature of a method step. Likewise, aspects described in the context of a method step also provide a description of the corresponding block or element, or feature of the corresponding device.

Depending on certain implementation requirements, embodiments of the invention may be implemented in hardware or software. An implementation may be performed using a digital storage medium, such as a floppy disk, DVD, CD, ROM, PROM, EPROM, EEPROM, or flash memory, having stored electronically readable control signals that interface (or enable interaction) with a programmable computer system such that the appropriate method is carried out.

Some embodiments of the invention include a storage medium having electronically readable control signals that allow interaction with a programmable computer system in a manner that implements one of the methods described herein.

In general, embodiments of the present invention may be implemented as a computer program product with program code, wherein the program code is configured to implement one of the methods where the computer program product is executed on a computer. The program code, for example, may be stored on a computer-readable medium.

Other embodiments comprise a computer program for performing one of the methods described herein stored on a computer-readable medium or a non-transitory storage medium.

In other words, an embodiment of the method of the invention is thus a computer program having program code for carrying out one of the methods described herein when the computer program is running on a computer.

Therefore, a further embodiment of the methods of the invention is a storage medium (digital storage medium or computer readable medium) containing a recorded computer program for performing one of the methods described herein.

Therefore, a further embodiment of the method of the invention is a data stream or signal sequence representing a computer program for implementing one of the methods described herein. The data stream or signal sequence, for example, may be configured to be transmitted over a data connection, such as the Internet.

An additional embodiment comprises processing means, such as a computer or programmable logic device, configured to implement one of the methods described herein.

A further embodiment comprises a computer having a computer program installed for performing one of the methods described herein.

In some embodiments, a programmable logic device (eg, a field programmable gate array) may be used to perform some or all of the functionality of the methods described herein. In some embodiments, a field programmable gate array may interface with a microprocessor to implement one of the methods described herein. In general, the methods are preferably implemented by any hardware device.

The above-described embodiments are merely illustrative of the principles of the present invention. It should be understood that modifications and changes to the arrangements and details described herein will be apparent to those skilled in the art. Accordingly, they are intended to be limited only by the scope of the claims below and not by the specific details provided by the description and explanation of the embodiments herein.

Bibliography

[1] ITU-T G.729 Annex B "A silence compression scheme for G.729 optimized for terminals conforming to ITU-T Recommendation V.70. International Telecommunication Union (ITU)", series G, 2007.

[2] ITU-T G.729.1 Annex C "DTX/CNG scheme: International Telecommunication Union (ITU)", series G, 2008.

[3] ITU-T G.718 "Frame error robust narrow-band and wideband embedded variable bit-rate coding of speech and audio from 8-32 kbit/s, International Telecommunication Union (ITU)", series G, 2008.

[4] "Mandatory Speech Codec speech processing functions; Adaptive Multi-Rate (AMR) speech codec; Transcoding functions", 3GPP Technical Specification TS 26.090, 2014.

[5] "Adaptive Multi-Rate - Wideband (AMR-WB) speech codec; Transcoding functions", 3GPP, 2014.

[6] 3GPP TS 26.445 "Codec for Enhanced Voice Services (EVS); Detailed algorithmic description".

[7] Z. Wang et al. "Linear prediction based comfort noise generation in the EVS codec", in IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Brisbane, QLD, 2015.

[8] A. Lombard, S. Wilde, E. Ravelli, S. Döhla, G. Fuchs and M. Dietz "Frequency-domain Comfort Noise Generation for Discontinuous Transmission in EVS", in IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Brisbane, QLD, 2015.

[9] A. Lombard, M. Dietz, S. Wilde, E. Ravelli, P. Setiawan and M. Multrus "Generation of the comfort noise with high spectro-temporal resolution in discontinuous transmission of audio signals." US Patent No. 9,583,114B2, June 19, 2015.

[10] E. NORVELL and F. JANSSON "SUPPORT FOR GENERATION OF COMFORT NOISE. AND GENERATION OF COMFORT NOISE", publication WO 2019/193149 A1, April 5, 2019.

Claims (173)

1. Multichannel signal generator (200) for generating a multichannel signal (204) having a first channel (201) and a second channel (203), containing: - a first audio source (211) for generating a first audio signal (221); - a second audio source (213) for generating a second audio signal (223); - a mixing noise source (212) to generate a mixing noise signal (222); And - a mixer (206) for mixing the mixing noise signal (222) and the first audio signal (221) to obtain the first channel (201), and for mixing the mixing noise signal (222) and the second audio signal (222) to obtain the second channel (203), - in this case the mixer (206) contains: - a first amplitude element (208-1) for influencing the amplitude of the first audio signal (221); - a first adder (206-1) for summing the output signal (221) of the first amplitude element and at least part of the mixing noise signal (222); - a second amplitude element (208-3) for influencing the amplitude of the second audio signal (223); - a second adder (206-3) for summing the output (223) of the second amplitude element (208-3) and at least part of the mixing noise signal (222), - in this case, the magnitude of the impact performed through the first amplitude element (208-1) and the magnitude of the impact performed through the second amplitude element (208-3) are equal to each other, or the magnitude of the impact performed through the second amplitude element (208-3) , differs by less than 20 percent relative to the value performed by the first amplitude element (208-1), - wherein the mixer (206) contains a third amplitude element (208-2) for influencing the amplitude of the mixing noise signal (222), - in this case, the magnitude of the impact performed by means of the third amplitude element (208-2) depends on the magnitude of the impact performed by means of the first amplitude element (208-1) or the second amplitude element (208-3) in such a way that the magnitude of the impact performed by means the third amplitude element (208-2) becomes larger when the amount of action performed by the first amplitude element or the amount of action performed by the second amplitude element (208-3) becomes smaller. 2. The channel signal generator of claim 1, wherein the first audio source (211) is a first noise source, and the first audio signal (221) is a first noise signal, and/or the second audio source (213) is a second noise source, and the second audio signal (223) is a second noise signal, - wherein the first noise source (211) and/or the second noise source (213) is configured to generate the first noise signal (221) and/or the second noise signal (223) in such a way that the first noise signal (221) and/or the second noise signal (223) is decorrelated with respect to the mixing noise signal (222). 3. The multi-channel signal generator according to claim 1 or 2, in which the mixer (206) is configured to generate the first channel (201) and the second channel (203) in such a way that the magnitude of the mixing noise signal (222) in the first channel (201) equal to the magnitude of the mixing noise signal (222) in the second channel (203) or within a range of 80-120 percent relative to the magnitude of the mixing noise signal (222) in the second channel (203). 4. The multi-channel signal generator according to one of the preceding paragraphs, in which the mixer (206) contains a control input for receiving a control parameter (404, s), and the mixer (206) is configured to control the magnitude of the noise signal (222) of the mix in the first channel (201) and the second channel (203) in response to the control parameter (404, s). 5. The multi-channel signal generator as claimed in one of the preceding claims, wherein each of the first audio source (211), the second audio source (213), and the mixing noise source (212) is a Gaussian noise source. 6. Multichannel signal generator according to one of the previous paragraphs, - wherein the first audio source (211) includes a first noise generator for generating a first audio signal (221) as a first noise signal, wherein the second audio source (213) includes a decorrelator for decorrelating the first noise signal (221) to generate a second audio signal (213) as as a second noise signal, and wherein the mixing noise source (212) comprises a second noise generator. 7. Multichannel signal generator according to one of claims. 1-5, - wherein the first audio source (211) includes a first noise generator (211) for generating a first audio signal (221) as a first noise signal, wherein the second audio source (213) contains a second noise generator (213) for generating a second audio signal (223) as as a second noise signal, and wherein the mixing noise source (212) comprises a decorrelator for decorrelating the first noise signal (221) or the second noise signal (223) to generate a mixing noise signal (222). 8. Multichannel signal generator according to one of claims. 1-5, - wherein one of the first audio source (211), the second audio source (213) and the mixing noise source (212) contains a noise generator for generating a noise signal, and the other of the first audio source (211), the second audio source (213) and the source The mixing noise source (212) comprises a first decorrelator for decorrelating the noise signal, and wherein another of the first audio source (211), the second audio source (213), and the mixing noise source (212) includes a second decorrelator for decorrelating the noise signal, wherein the first the decorrelator and the second decorrelator are different from each other such that the output signals of the first decorrelator and the second decorrelator are decorrelated from each other. 9. Multichannel signal generator according to one of claims. 1-5, wherein the first audio source (211) includes a first noise generator, wherein the second audio source (213) contains a second noise generator, and wherein the mixing noise source (212) includes a third noise generator, wherein the first noise generator, the second the noise generator and the third noise generator are configured to generate mutually decorrelated noise signals. 10. Multichannel signal generator according to one of the previous paragraphs, - wherein one of the first audio source (211), the second audio source (213) and the mixing noise source (212) comprises a pseudo-random number sequence generator configured to generate a pseudo-random number sequence in response to the seed, and at least two of the first audio source (211), the second audio source (213) and the mixing noise source (212) are configured to initialize a pseudo-random number sequence generator using different seeds. 11. Multichannel signal generator according to one of claims. 1-6, - wherein at least one of the first audio source (211), the second audio source (213) and the mixing noise source (212) is configured to operate using a previously stored noise table. 12. Multichannel signal generator according to one of claims. 1-6, - wherein at least one of the first audio source (211), the second audio source (213) and the mixing noise source (212) is configured to generate a complex spectrum for the frame using the first noise value for the real part and the second noise value for the imaginary part . 13. Multichannel signal generator according to one of claims. 11 and 12, in which the at least one noise generator is configured to generate a complex spectral noise value for frequency bin k using, for one of the real part and the imaginary part, a first random value with index k, and using, for the other from the real part and the imaginary part, a second random value with index (k+M), wherein the first noise value and the second noise value are included in a noise array, for example, extracted from a random number sequence generator or from a noise table, or from a noise process, in the range from the start index to the end index, wherein the start index is less than M, and wherein the end index is equal to or less than 2M, wherein M and k are integers. 14. Multichannel signal generator according to one of the previous paragraphs, - in which the magnitude of the influence performed by means of the third amplitude element (208-2) is the square root of a given value (c _q ), and the magnitude of the action performed by means of the first amplitude element (208-1), and the magnitude of the influence performed by means of the second amplitude element (208-3), represents the square root of the difference between 1 and the specified value (c _q ). 15. Multichannel signal generator according to one of the previous paragraphs, additionally containing: - an input interface (210) for receiving encoded audio data (232) in a sequence of frames (306, 308) containing an active frame (306) and an inactive frame (308) after the active frame (306); And - an audio decoder (200'', 200a, 200b) for decoding encoded audio data for the active frame (306) to generate a decoded multi-channel signal for the active frame, wherein the first audio source (211), the second audio source (213), the mixing noise source (212) and the mixer (206) are active in the inactive frame (308) to generate a multi-channel signal (204) for the inactive frame. 16. Multichannel signal generator according to claim 15, in which: - encoded audio data (232) for the active frame (306) has a first set of coefficients describing a first number of frequency bins; And - the encoded audio data (232) for the inactive frame (308) has a second set of coefficients describing a second number of frequency bins, - in this case, the first number of frequency bins is greater than the second number of frequency bins. 17. Multichannel signal generator (200) for generating a multichannel signal (204) having a first channel (201) and a second channel (203), containing: - a first audio source (211) for generating a first audio signal (221); - a second audio source (213) for generating a second audio signal (223); - a mixing noise source (212) to generate a mixing noise signal (222); - a mixer (206) for mixing the mixing noise signal (222) and the first audio signal (221) to obtain the first channel (201), and for mixing the mixing noise signal (222) and the second audio signal (222) to obtain the second channel (203), - an input interface (210) for receiving encoded audio data (232) in a sequence of frames (306, 308) containing an active frame (306) and an inactive frame (308) after the active frame (306); And - an audio decoder (200'', 200a, 200b) for decoding encoded audio data for the active frame (306) to generate a decoded multi-channel signal for the active frame, - wherein the first audio source (211), the second audio source (213), the mixing noise source (212) and the mixer (206) are active in the inactive frame (308) to generate a multi-channel signal (204) for the inactive frame, - wherein the encoded audio data (232) for the inactive frame (308) contains silence insertion descriptor data (p_noise, c) containing comfort noise data (c, p_noise) indicating the signal energy (1312) for each channel of the two channels (301, 303 ) or for each of a first linear combination of the first and second channels and a second linear combination of the first and second channels for the inactive frame and indicating coherence (404, c) between the first channel (301) and the second channel (303) in the inactive frame, and - wherein the mixer (206, 220) is configured to mix (206-1, 206-3) the mixing noise signal (222) and the first audio signal (221) or the second audio signal (223) based on the comfort noise data indicating coherence (404 , c), and - wherein the generator (200, 220, 220a-220e) of multi-channel signals additionally contains a signal modification module (250) for modifying the first channel (201) and the second channel (203), or the first audio signal (221) or the second audio signal (223), or noise signal (222) mixing, - wherein the signal modification module (250) is configured to be controlled by comfort noise data (p_noise) indicating signal energies for the first audio channel (301) and the second audio channel (303) or indicating signal energies for the first linear combination of the first and second channels and the second linear combination of the first and second channels. 18. Multichannel signal generator according to any one of claims. 15-17, in which the audio data (232) for the inactive frame contains: - a first frame (241) of a silence insert descriptor for the first channel (201) and a second frame (243) of a silence insert descriptor for the second channel (203), wherein the first frame (241) of the silence insert descriptor contains: - data (p_noise) of comfort noise parameters for the first channel (201) and/or for the first linear combination of the first and second channels, and - auxiliary information (p_frame) for the formation of comfortable noise for the first channel and the second channel (203), and - wherein the second frame (243) of the silence insertion descriptor contains: - data (p_noise) of comfort noise parameters for the second channel (203) and/or for the second linear combination of the first and second channels, and - coherence information (404, c) indicating coherence between the first channel (201) and the second channel (203) in the inactive frame, and - wherein the multi-channel signal generator contains a controller for controlling the formation of a multi-channel signal (204) in an inactive frame using auxiliary information (p_frame) for the formation of comfortable noise for the first frame (241) of the silence insert descriptor for determining the mode of generating comfortable noise for the first channel (201) and the second channel (203) and/or for the first linear combination of the first and second channels and the second linear combination of the first and second channels, using the coherence information (404, c) in the second frame (243) of the silence insert descriptor to specify the coherence (404, c) between the first channel (201) and the second channel (203) in the inactive frame, and using the comfort noise parameter data (p_noise) from the first frame (241) of the silence insert descriptor, and using the comfort noise parameter data (p_noise) from the second a silent insert descriptor frame (243) for specifying the power situation (v _{l, q} ) of the first channel (301) and the power situation (v _{r, q} ) of the second channel (303). 19. Multichannel signal generator according to any one of claims. 15-18, in which the audio data (232) for the inactive frame contains: - at least one silent insert descriptor frame (241) for a first linear combination of the first and second channels and a second linear combination of the first and second channels, - wherein at least one frame (241) of the silence insertion descriptor contains: - data (p_noise) of comfortable noise parameters for the first linear combination of the first and second channels, and - auxiliary information (p_frame) for the formation of comfortable noise for the second linear combination of the first and second channels, - wherein the multi-channel signal generator contains a controller for controlling the generation of a multi-channel signal (204) in an inactive frame using auxiliary information (p_frame) for generating comfortable noise for the first linear combination of the first and second channels and the second linear combination of the first and second channels, using information ( 404, c) coherence in the second frame (243) of the insertion of silence descriptor to specify the coherence (404, c) between the first channel (201) and the second channel (203) in the inactive frame, and using data (p_noise) comfort noise parameters at least from at least one silence insertion descriptor frame (241), and using the comfort noise parameter data (p_noise) from at least one silence insertion descriptor frame (243) to specify the energy situation (v _l , _q ) of the first channel (301) and the energy situation (v _{r, q} ) of the second channel (303). 20. Multichannel signal generator according to any one of claims. 17-19, further comprising a time-spectral converter for converting the resulting first channel and the resulting second channel, spectrally variable and coherently variable, into corresponding time domain representations to be combined or concatenated with the time domain representations of the corresponding channels of the decoded multi-channel signal for the active frame. 21. Multichannel signal generator according to any one of claims. 15-20, in which the audio data for the inactive frame contains: - frame (241, 243) of the silence insertion descriptor, wherein the frame (241, 243) of the silence insertion descriptor contains data (p_noise) of the comfortable noise parameters for the first and second channels (201, 203) and auxiliary information (p_frame) for the formation of comfortable noise for the first channel (203) and the second channel (203) and/or for the first linear combination of the first and second channels and the second linear combination of the first and second channels and coherence information (404, c) indicating the coherence between the first channel (201) and the second channel (203) in an inactive frame, and - wherein the generator (200) of multi-channel signals contains a controller for controlling the formation of a multi-channel signal (202) in an inactive frame using auxiliary information (p_frame) for the formation of comfortable noise for the frame (241, 243) of the silence insertion descriptor for determining the mode of generating comfortable noise for the first channel (201) and the second channel (203), using the coherence information (404, s) in the silence insert descriptor frame (241) to specify the coherence (404, s) between the first channel (201) and the second channel (203) in the inactive frame, and using the data (p_noise) comfort noise parameters from the frame (241, 243) of the silence insertion descriptor to specify the energy situation (v _{l, q} ) of the first channel (301) and the energy situation (v _{r, q} ) of the second channel (303 ). 22. Multichannel signal generator according to any one of claims. 15-21, - wherein the encoded audio data (232) for the inactive frame contains silence insert descriptor data (p_noise, c), containing comfort noise data (c, p_noise), indicating the signal energy for each channel in the middle/side view, and data (404, c ) coherences indicating coherence between the first channel and the second channel in a left/right representation, wherein the multi-channel signal generator is configured to convert the middle/side representation of the energy of the signals into a left/right representation of the energy of the signals in the first channel (301) and the second channel (303 ), - wherein the mixer (206, 220) is configured to mix (206-1, 206-3) the mixing noise signal (222) into the first audio signal (221) and the second audio signal (223) based on the coherence data (404, c) for obtaining a first channel (201) and a second channel (203), and - wherein the multi-channel signal generator additionally contains a signal modification module (250), configured to modify the first and second channels (201, 203) by generating the first and second channels (201, 203) based on the energy of the signals in the left/right region. 23. The multi-channel signal generator of claim 22, configured to, if the audio data contains signaling signaling indicating that the energy in the side channel is less than a predetermined threshold value, reset (337) the coefficients (v _{s, q} ) of the side channel . 24. The multi-channel signal generator according to claim 22 or 23, wherein the audio data for the inactive frame contains: - at least one frame (241, 243) of a silence insertion descriptor, wherein at least one frame (241, 243) of a silence insertion descriptor contains data (p_noise, v _{m, ind} , q _{l, q} , q _{r, q} , v _{s, ind} ) parameters of comfortable noise for the middle and side channels (v _{m, q} , v _{s, q} ) and auxiliary information (p_frame) for the formation of comfortable noise for the middle and side channels (v _{m, q} , v _{s, q} ) and coherence information (404, c) indicating coherence between the first channel (201) and the second channel (203) in the inactive frame, and - wherein the generator (200) of multi-channel signals contains a controller for controlling the formation of a multi-channel signal (202) in an inactive frame using auxiliary information (p_frame) for the formation of comfortable noise for the frame (241, 243) of the silence insertion descriptor for determining the mode of generating comfortable noise for the first channel (201) and the second channel (203), using the coherence information (404, s) in the silence insert descriptor frame (241) to specify the coherence (404, s) between the first channel (201) and the second channel (203) in the inactive frame, and using the data (p_noise) comfort noise parameters or their processed version from the frame (241, 243) of the silence insertion descriptor to specify the energy situation (v _{l, q} ) of the first channel (301) and the energy situation (v _{r, q} ) second channel (303). 25. Multichannel signal generator according to any one of claims. 15-24, further configured to scale the energy coefficients (1312, v' _l , v' _r ) of the signals for the first and second channel by means of gain information (g _{l, q} , q _{r, q} ) encoded by the data (401, 403) comfortable noise parameters for the first and second channels. 26. The multichannel signal generator according to any of the previous paragraphs, configured to converting the generated multi-channel signal (252) from a frequency domain version to a time domain version. 27. The channel signal generator as claimed in any one of the preceding claims, wherein the first audio source (211) is a first noise source and the first audio signal (221) is a first noise signal, or the second audio source (213) is a second noise source and the second the audio signal (223) is a second noise signal, - wherein the first noise source or the second noise source is configured to generate the first noise signal (201) or the second noise signal (203) such that the first noise signal (201) or the second noise signal (203) is at least partially correlated, And - wherein the mixing noise source (212) is configured to generate a mixing noise signal (222) with a first part (221a) of mixing noise and a second part (221b) of mixing noise, wherein the second part (221b) of mixing noise is at least is at least partially decorrelated with respect to the first part of the mixing noise (221b); And - wherein the mixer (206) is configured to mix the first noise part (221a) when mixing the mixing noise signal (222) and the first audio signal (221) to obtain the first channel (201), and mix the second noise part (221b) when mixing the noise a mixing signal (222) and a second audio signal (223) to obtain a second channel (203). 28. Multichannel signal generator (200) for generating a multichannel signal (204) having a first channel (201) and a second channel (203), containing: - a first audio source (211) for generating a first audio signal (221); - a second audio source (213) for generating a second audio signal (223); - a mixing noise source (212) to generate a mixing noise signal (222); And - a mixer (206) for mixing the mixing noise signal (222) and the first audio signal (221) to obtain the first channel (201), and for mixing the mixing noise signal (222) and the second audio signal (222) to obtain the second channel (203), - wherein the first audio source (211) is a first noise source, and the first audio signal (221) is a first noise signal, or the second audio source (213) is a second noise source, and the second audio signal (223) is a second noise signal, - wherein the first noise source or the second noise source is configured to generate the first noise signal (201) or the second noise signal (203) such that the first noise signal (201) or the second noise signal (203) is at least partially correlated, And - wherein the mixing noise source (212) is configured to generate a mixing noise signal (222) with a first part (221a) of mixing noise and a second part (221b) of mixing noise, wherein the second part (221b) of mixing noise is at least is at least partially decorrelated with respect to the first part of the mixing noise (221b); And - wherein the mixer (206) is configured to mix the first noise part (221a) when mixing the mixing noise signal (222) and the first audio signal (221) to obtain the first channel (201), and mix the second noise part (221b) when mixing the noise a mixing signal (222) and a second audio signal (223) to obtain a second channel (203). 29. A method for generating a multi-channel signal having a first channel and a second channel (203), comprising the steps of: - generating a first audio signal (221) using the first audio source (211); - generating a second audio signal (223) using a second audio source (213); - generating a mixing noise signal (222) using a mixing noise source (212); And - mixing (206) the mixing noise signal (222) and the first audio signal (221) to obtain the first channel (201), and mixing the mixing noise signal (222) and the second audio signal (223) to obtain the second channel (202), wherein the method contains stages in which: - using a first amplitude element (208-1) influencing the amplitude of the first audio signal (221); - using a first adder (206-1) summing the output signal (221) of the first amplitude element and at least part of the mixing noise signal (222); - using a second amplitude element (208-3) influencing the amplitude of the second audio signal (223); - using a second adder (206-3), summing the output (223) of the second amplitude element (208-3) and at least part of the mixing noise signal (222), - in this case, the magnitude of the impact performed through the first amplitude element (208-1) and the magnitude of the impact performed through the second amplitude element (208-3) are equal to each other, or the magnitude of the impact performed through the second amplitude element (208-3) , differs by less than 20 percent relative to the value performed by the first amplitude element (208-1), - wherein the mixing (206) uses a third amplitude element (208-2) affecting the amplitude of the mixing noise signal (222), - in this case, the magnitude of the impact performed by means of the third amplitude element (208-2) depends on the magnitude of the impact performed by means of the first amplitude element (208-1) or the second amplitude element (208-3) in such a way that the magnitude of the impact performed by means the third amplitude element (208-2) becomes larger when the amount of action performed by the first amplitude element or the amount of action performed by the second amplitude element (208-3) becomes smaller. 30. An audio encoder (300, 300a, 300b) for generating an encoded multi-channel audio signal (232) for a sequence of frames comprising an active frame (306) and an inactive frame (308), the audio encoder comprising: - an activity detector (380) for analyzing the multi-channel signal (304) to determine (381) a frame of the sequence of frames as representing an inactive frame (308); - a noise parameter calculation module (3040) for calculating the first noise parametric data (p_noise, v _{m, ind} ) for the first channel (301, 201) of the multi-channel signal (304) and for calculating the second noise parametric data (p_noise, v _{s, ind} ) for the second channel (303) of the multi-channel signal (320); - a coherence calculation module (320) for calculating coherence data (404, c) indicating the coherence situation between the first channel (301, 201) and the second channel (303, 203) in the inactive frame (308); And - output interface (310) for generating an encoded multi-channel audio signal (232) having encoded audio data for the active frame (306) and for the inactive frame (308), first noise parametric data (p_noise, v _{m, ind} ), second parametric data (p_noise , v _{s, ind} ) noise and/or a first linear combination of the first parametric noise data and the second parametric noise data and a second linear combination of the first parametric noise data and the second parametric noise data and the coherence data (c, 404), wherein modulus (3040) calculating the noise parameters is configured to convert at least some of the first noise parameter data and the second noise parameter data from a left/right view to a middle/side view with a middle channel and a side channel. 31. The audio encoder of claim 30, wherein the noise parameter calculation module (3040) is configured to reconvert the middle/side representation (M, S) of at least some of the first noise parametric data and the second noise parametric data into a left/right representation , - wherein the noise parameter calculation module (3040) is configured to calculate from the re-converted left/right representation the first gain information (g _l ) for the first channel (301) and the second gain information ( _gr ) for the second channel (303) and provide , first gain information (g _l ) for the first channel (301) included in the first noise parameter data, and second gain information ( _gr ) included in the second noise parameter data. 32. Audio encoder (300) according to claim 31, in which the module (3040) for calculating noise parameters is configured to calculate: - first information (g _l ) gain through comparison: - a version (v' _l ) of the first parametric noise data for the first channel (301) re-converted from the middle/side view to the left/right view; - with a version (v _l ) of the first parametric noise data for the first channel (301) before conversion from the middle/side view to the left/right view; and/or - second gain information (g _r ) by comparison: - a version (v' _r ) of the second parametric noise data for the second channel (301) re-converted from the middle/side view to the left/right view; - with a version (v _r ) of the second parametric noise data for the second channel (301) before conversion from the middle/side view to the left/right view. 33. An audio encoder (300, 300a, 300b) for generating an encoded multi-channel audio signal (232) for a sequence of frames comprising an active frame (306) and an inactive frame (308), the audio encoder comprising: - an activity detector (380) for analyzing the multi-channel signal (304) to determine (381) a frame of the sequence of frames as representing an inactive frame (308); - a noise parameter calculation module (3040) for calculating the first noise parametric data (p_noise, v _{m, ind} ) for the first channel (301, 201) of the multi-channel signal (304) and for calculating the second noise parametric data (p_noise, v _{s, ind} ) for the second channel (303) of the multi-channel signal (320); - a coherence calculation module (320) for calculating coherence data (404, c) indicating the coherence situation between the first channel (301, 201) and the second channel (303, 203) in the inactive frame (308); And - output interface (310) for generating an encoded multi-channel audio signal (232) having encoded audio data for the active frame (306) and for the inactive frame (308), first noise parametric data (p_noise, v _{m, ind} ), second parametric data (p_noise , v _{s, ind} ) noise and/or a first linear combination of the first parametric noise data and the second parametric noise data and a second linear combination of the first parametric noise data and the second parametric noise data, and the coherence data (c, 404), wherein modulus (320 ) coherence calculations are made with the ability to: - calculating the real intermediate value and the imaginary intermediate value from the complex spectral values for the first channel and the second channel (303) in the inactive frame; - calculating a first energy value for the first channel (301) and a second energy value for the second channel (303) in the inactive frame; And - calculating coherence data (404, c) using a real intermediate value, an imaginary intermediate value, a first energy value and a second energy value, or - smoothing at least one of the real intermediate value, the imaginary intermediate value, the first energy value and the second energy value and calculating coherence data using the at least one smoothed value, - wherein the coherence calculating module (320) is configured to square the smoothed real intermediate value and square the smoothed imaginary intermediate value and sum the squared values to obtain the first component number, - wherein the coherence calculation module (320) is configured to multiply the smoothed first and second energy values to obtain a second component number and combine the first and second component numbers to obtain a resultant number for the coherence value on which the coherence data is based. 34. An audio encoder (300, 300a, 300b) for generating an encoded multi-channel audio signal (232) for a sequence of frames comprising an active frame (306) and an inactive frame (308), the audio encoder comprising: - an activity detector (380) for analyzing the multi-channel signal (304) to determine (381) a frame of the sequence of frames as representing an inactive frame (308); - a noise parameter calculation module (3040) for calculating the first noise parametric data (p_noise, v _{m, ind} ) for the first channel (301, 201) of the multi-channel signal (304) and for calculating the second noise parametric data (p_noise, v _{s, ind} ) for the second channel (303) of the multi-channel signal (320); - a coherence calculation module (320) for calculating coherence data (404, c) indicating the coherence situation between the first channel (301, 201) and the second channel (303, 203) in the inactive frame (308); And - output interface (310) for generating an encoded multi-channel audio signal (232) having encoded audio data for the active frame (306) and for the inactive frame (308), first noise parametric data (p_noise, v _{m, ind} ), second parametric data (p_noise , v _{s, ind} ) noise and/or a first linear combination of the first parametric noise data and the second parametric noise data and a second linear combination of the first parametric noise data and the second parametric noise data and coherence data (c, 404), - wherein the noise parameter calculation module (3040) is configured to compare the energy of the second linear combination between the first noise parameter data and the second noise parameter data with a given energy threshold value(s) and: - in the event that the energy of the second linear combination between the first noise parametric data and the second noise parametric data is greater than a predetermined energy threshold(s), the coefficients of the side channel noise shape vector are set to zero (437); And - in case the energy of the second linear combination between the first noise parametric data and the second noise parametric data is less than a predetermined energy threshold (ex), the coefficients of the side channel noise shape vector are stored. 35. Audio encoder according to any one of paragraphs. 30-34, in which the coherence calculation module (320) is configured to calculate (320'') the coherence value (404, c) and quantize (320'') the coherence value (320'') to obtain a quantized value (c _ind ) coherence, wherein the output interface (310) is configured to use the quantized coherence value (c _ind ) as coherence data in the encoded multi-channel signal. 36. Audio encoder according to any one of paragraphs. 30-35, - wherein the coherence calculation module (320) is configured to calculate the real intermediate value as a sum over the real parts of the products of complex spectral values for the corresponding frequency bins of the first channel and the second channel (303) in the inactive frame, or - calculating an imaginary intermediate value as a sum over the imaginary parts of the products of complex spectral values for the corresponding frequency bins of the first channel and the second channel (303) in the inactive frame. 37. The audio encoder of claim 33, wherein the coherence calculation module is configured to calculate the square root of the resultant number to obtain a coherence value on which the coherence data is based. 38. Audio encoder according to one of paragraphs. 30-37, - wherein the coherence calculation module (320) is configured to quantize the coherence value (404, c) using a uniform quantizer (320'') to obtain the quantized coherence value (c _ind ) as n bits as coherence data. 39. The audio encoder of claim 38, wherein the uniform quantizer (320'') is configured to calculate n bits such that the value for n is equal to the value of the bits occupied by the comfort noise generating auxiliary information (p_frame) for the first frame (241) silent insert handle. 40. Audio encoder according to one of paragraphs. 30-39, in which the output interface (310) is configured to generate a first silence insert descriptor frame (241) for the first channel (301, L) and a second silence insert descriptor frame (243) for the second channel (303, R), with In this case, the first frame (241) of the silence insert descriptor contains data (p_noise) of the comfort noise parameters for the first channel (301, L) and auxiliary information (p_frame) for the formation of comfortable noise for the first channel (301, L) and the second channel (303, R) , and wherein the second frame (243) of the silence insert descriptor contains comfort noise parameter data (p_noise) for the second channel (303) and coherence information (404, c) indicating coherence between the first channel and the second channel (303) in the inactive frame. 41. Audio encoder according to one of paragraphs. 30-39, - in which the output interface (310) is configured to generate a silence insertion descriptor frame (241, 243), wherein the silence insertion descriptor frame contains data (p_noise) of comfort noise parameters for the first and second channels (301, 303) and auxiliary information ( p_frame) comfort noise generation for the first channel (301, L) and the second channel (303, R) and coherence information (404, c) indicating coherence between the first channel (301, L) and the second channel (303, R) in the inactive frame. 42. Audio encoder according to one of paragraphs. 30-39, - in which the output interface (310) is configured to generate a first silence insert descriptor frame (241) for the first channel (301, L) and the second channel and a second silence insert descriptor frame (243) for the first channel and the second channel (303, R ), wherein the first frame (241) of the silence insert descriptor contains data (p_noise) of the comfort noise parameters for the first channel and the second channel and auxiliary information (p_frame) for the formation of comfortable noise for the first channel (301, L) and the second channel (303, R ), and wherein the second frame (243) of the silence insert descriptor contains comfort noise parameter data (p_noise) for the first channel and the second channel (303) and coherence information (404, c) indicating the coherence between the first channel and the second channel (303) in an inactive frame. 43. Audio encoder (300) according to one of claims. 30-42, in which the activity detector (380) is configured to: for at least one frame of the sequence of frames: - analyzing (370-1) the first channel (301, L) of the multi-channel signal (304) to classify the first channel (301, L) as active or inactive, and - analyzing (370-2) the second channel (303, R) of the multi-channel signal (304) to classify the second channel (303, R) as active or inactive, and - defining (381) a frame as inactive if both the first channel (301, L) and the second channel (303, R) are classified as inactive, and otherwise as active. 44. Audio encoder (300) according to one of claims. 30-43, in which the noise parameter calculation module (3040) is configured to calculate first gain information (g _l ) for the first channel (301) and second gain information ( _gs ) for the second channel (g _l ) and provide noise parameter data as first gain information (g _l ) for the first channel (301) and second gain information (g _s ). 45. Audio encoder according to one of paragraphs. 30-44, configured to encode a second linear pattern between the first noise parameter data and the second noise parameter data with fewer bits than the number of bits through which the first linear pattern between the first noise parameter data and the second noise parameter data is encoded. 46. Audio encoder according to one of paragraphs. 30-45, - in which the output interface (310) is configured to: - generating an encoded multi-channel audio signal (232) having encoded audio data for the active frame (306) using a first set of coefficients for a first number of frequency bins; And - generating first parametric noise data, second parametric noise data, or a first linear combination of the first parametric noise data and second parametric noise data and a second linear combination of the first parametric noise data and second parametric noise data using a second set of coefficients describing a second number of frequency bins, - in this case, the first number of frequency bins is greater than the second number of frequency bins. 47. An audio encoding method for generating an encoded multi-channel audio signal for a sequence of frames containing an active frame and an inactive frame, the method comprising the steps of: - analyzing the multi-channel signal to determine a frame of the frame sequence as representing an inactive frame; - calculate the first parametric noise data for the first channel of the multi-channel signal and/or for the first linear combination of the first and second channels of the multi-channel signal and calculate the second parametric noise data for the second channel (303) of the multi-channel signal and/or for the second linear combination of the first and second channels of the multi-channel signal signal; - calculating coherence data indicating the coherence situation between the first channel and the second channel (303) in the inactive frame; And - generating an encoded multi-channel audio signal having encoded audio data for the active frame and, for the inactive frame, first noise parametric data, second noise parametric data and coherence data, - wherein the noise parameter calculation module (3040) is configured to convert at least some of the first noise parameter data and the second noise parameter data from a left/right view to a middle/side view with a middle channel and a side channel. 48. A permanent storage module that stores instructions that, when executed on a computer or processor, implement the method of claim 29. 49. A permanent storage module storing instructions that, when executed on a computer or processor, implements the method of claim 47.