CN110176241A - Coding method and equipment and signal decoding method and equipment - Google Patents
Coding method and equipment and signal decoding method and equipment Download PDFInfo
- Publication number
- CN110176241A CN110176241A CN201910495957.0A CN201910495957A CN110176241A CN 110176241 A CN110176241 A CN 110176241A CN 201910495957 A CN201910495957 A CN 201910495957A CN 110176241 A CN110176241 A CN 110176241A
- Authority
- CN
- China
- Prior art keywords
- coding
- spectrum
- frequency band
- spectral component
- frequency
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 83
- 238000001228 spectrum Methods 0.000 claims abstract description 122
- 230000003595 spectral effect Effects 0.000 claims abstract description 108
- 238000013139 quantization Methods 0.000 claims description 66
- 230000008859 change Effects 0.000 claims description 6
- 238000010586 diagram Methods 0.000 description 48
- 230000005284 excitation Effects 0.000 description 34
- 230000005236 sound signal Effects 0.000 description 23
- 230000001052 transient effect Effects 0.000 description 20
- 230000008569 process Effects 0.000 description 17
- 238000012545 processing Methods 0.000 description 14
- 230000006870 function Effects 0.000 description 13
- 230000015572 biosynthetic process Effects 0.000 description 12
- 238000011084 recovery Methods 0.000 description 12
- 238000003786 synthesis reaction Methods 0.000 description 12
- 230000009466 transformation Effects 0.000 description 12
- 238000004891 communication Methods 0.000 description 10
- 238000011002 quantification Methods 0.000 description 10
- 238000001914 filtration Methods 0.000 description 6
- 238000005070 sampling Methods 0.000 description 6
- 238000013461 design Methods 0.000 description 5
- 238000012805 post-processing Methods 0.000 description 5
- 238000005086 pumping Methods 0.000 description 5
- 238000007493 shaping process Methods 0.000 description 5
- 230000007704 transition Effects 0.000 description 5
- 230000000712 assembly Effects 0.000 description 4
- 238000000429 assembly Methods 0.000 description 4
- 238000004364 calculation method Methods 0.000 description 4
- 238000000605 extraction Methods 0.000 description 4
- 238000006243 chemical reaction Methods 0.000 description 3
- 238000010276 construction Methods 0.000 description 3
- 238000001514 detection method Methods 0.000 description 3
- 230000006872 improvement Effects 0.000 description 3
- 238000013507 mapping Methods 0.000 description 3
- 230000000873 masking effect Effects 0.000 description 3
- 238000000611 regression analysis Methods 0.000 description 3
- 230000004044 response Effects 0.000 description 3
- 230000011664 signaling Effects 0.000 description 3
- 238000012360 testing method Methods 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 2
- 210000001367 artery Anatomy 0.000 description 2
- JEIPFZHSYJVQDO-UHFFFAOYSA-N ferric oxide Chemical compound O=[Fe]O[Fe]=O JEIPFZHSYJVQDO-UHFFFAOYSA-N 0.000 description 2
- 230000011218 segmentation Effects 0.000 description 2
- 230000003068 static effect Effects 0.000 description 2
- 210000003462 vein Anatomy 0.000 description 2
- 241000208340 Araliaceae Species 0.000 description 1
- 235000005035 Panax pseudoginseng ssp. pseudoginseng Nutrition 0.000 description 1
- 235000003140 Panax quinquefolius Nutrition 0.000 description 1
- 241000209140 Triticum Species 0.000 description 1
- 235000021307 Triticum Nutrition 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 235000013399 edible fruits Nutrition 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 235000008434 ginseng Nutrition 0.000 description 1
- 230000000670 limiting effect Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000036961 partial effect Effects 0.000 description 1
- 230000002441 reversible effect Effects 0.000 description 1
- 238000010183 spectrum analysis Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000001131 transforming effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
- G10L19/00—Speech or audio signals analysis-synthesis techniques for redundancy reduction, e.g. in vocoders; Coding or decoding of speech or audio signals, using source filter models or psychoacoustic analysis
- G10L19/02—Speech or audio signals analysis-synthesis techniques for redundancy reduction, e.g. in vocoders; Coding or decoding of speech or audio signals, using source filter models or psychoacoustic analysis using spectral analysis, e.g. transform vocoders or subband vocoders
- G10L19/0212—Speech or audio signals analysis-synthesis techniques for redundancy reduction, e.g. in vocoders; Coding or decoding of speech or audio signals, using source filter models or psychoacoustic analysis using spectral analysis, e.g. transform vocoders or subband vocoders using orthogonal transformation
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
- G10L19/00—Speech or audio signals analysis-synthesis techniques for redundancy reduction, e.g. in vocoders; Coding or decoding of speech or audio signals, using source filter models or psychoacoustic analysis
- G10L19/02—Speech or audio signals analysis-synthesis techniques for redundancy reduction, e.g. in vocoders; Coding or decoding of speech or audio signals, using source filter models or psychoacoustic analysis using spectral analysis, e.g. transform vocoders or subband vocoders
- G10L19/032—Quantisation or dequantisation of spectral components
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
- G10L19/00—Speech or audio signals analysis-synthesis techniques for redundancy reduction, e.g. in vocoders; Coding or decoding of speech or audio signals, using source filter models or psychoacoustic analysis
- G10L19/002—Dynamic bit allocation
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
- G10L19/00—Speech or audio signals analysis-synthesis techniques for redundancy reduction, e.g. in vocoders; Coding or decoding of speech or audio signals, using source filter models or psychoacoustic analysis
- G10L19/04—Speech or audio signals analysis-synthesis techniques for redundancy reduction, e.g. in vocoders; Coding or decoding of speech or audio signals, using source filter models or psychoacoustic analysis using predictive techniques
- G10L19/16—Vocoder architecture
- G10L19/18—Vocoders using multiple modes
- G10L19/24—Variable rate codecs, e.g. for generating different qualities using a scalable representation such as hierarchical encoding or layered encoding
Abstract
The present invention relates to a kind of coding method and equipment and signal decoding method and equipment.The spectrum coding method, which can comprise the following steps that based on the bit distribution information of each frequency band, selects type of coding；Zero-code is executed for zero-frequency band；It is encoded for information of each non-zero frequency band to the effective frequency component of selection.The spectrum coding method makes it possible to code and decode the spectral coefficient for being adapted to various bit rates and each sub-band size.In addition, can be encoded using TCQ method to frequency spectrum using Bit-Rate Control Algorithm module with fixed bit rate in the codec for supporting multiple rates.The coding efficiency of the codec can be made to maximize by being encoded with accurate target bit rate to high-performance TCQ.
Description
It is February 17, entitled " Signal coding in 2015 that the application, which is the applying date submitted to State Intellectual Property Office,
Method and apparatus and signal decoding method and equipment ", application No. is the divisional applications of 201580020096.0 application.
Technical field
One or more exemplary embodiments are related to audio or voice signal coding and decoding, more particularly, are related to
One kind is for carrying out coding or decoded method and apparatus to spectral coefficient in a frequency domain.
Background technique
The quantizer of various schemes has been suggested to carry out efficient coding to spectral coefficient in a frequency domain.For example, there are nets
Trellis coding quantifies (TCQ), uniform scalar quantization (USQ), factorial pulse code (FPC), algebra VQ (AVQ), taper VQ (PVQ)
Deng, and can be achieved for the optimised lossless encoder of each quantizer.
Summary of the invention
Technical problem
One or more exemplary embodiments include a kind of in a frequency domain to being adapted to various bit rates or various
Sub-band size carries out coding or decoded method and apparatus to spectral coefficient.
One or more exemplary embodiments include that a kind of record thereon has for executing Signal coding or coding/decoding method
Computer-readable program computer readable recording medium.
One or more exemplary embodiments include a kind of using Signal coding or the multimedia device of decoding device.
Solution
According to one or more exemplary embodiments, a kind of spectrum coding method includes: at least based on each frequency band
Bit distribution information selects coding method；Zero-code is executed to zero-frequency band；To about the weight for being directed to each non-zero frequency band selection
The information of frequency component is wanted to be encoded.
According to one or more exemplary embodiments, a kind of frequency spectrum coding/decoding method includes: at least based on each frequency band
Bit distribution information selects coding/decoding method；Null solution code is executed to zero-frequency band；To about the weight obtained for each non-zero frequency band
The information of frequency component is wanted to be decoded.
Beneficial effects of the present invention
The executable coding and decoding to the spectral coefficient for being adapted to various bit rates and each sub-band size.In addition,
The Bit-Rate Control Algorithm module that designed in codec can be supported by using multi tate, by TCQ with fixed bit rate to frequency spectrum
It is encoded.In this case, coding, codec are executed with accurate target bit rate by the high-performance via TCQ
Coding efficiency can be maximized.
Detailed description of the invention
Fig. 1 a and Fig. 1 b are the block diagram of audio coding apparatus and audio decoding apparatus accoding to exemplary embodiment respectively.
Fig. 2 a and Fig. 2 b are the frame of audio coding apparatus and audio decoding apparatus according to another exemplary embodiment respectively
Figure.
Fig. 3 a and Fig. 3 b are the frame of audio coding apparatus and audio decoding apparatus according to another exemplary embodiment respectively
Figure.
Fig. 4 a and Fig. 4 b are the frame of audio coding apparatus and audio decoding apparatus according to another exemplary embodiment respectively
Figure.
Fig. 5 is the block diagram of frequency domain audio encoding device accoding to exemplary embodiment.
Fig. 6 is the block diagram of frequency domain audio decoding device accoding to exemplary embodiment.
Fig. 7 is the block diagram of spectrum coding equipment accoding to exemplary embodiment.
Fig. 8 shows sub-band segmentation.
Fig. 9 is the block diagram of spectrum quantification equipment accoding to exemplary embodiment.
Figure 10 is the block diagram of spectrum coding equipment accoding to exemplary embodiment.
Figure 11 is the block diagram of ISC encoding device accoding to exemplary embodiment.
Figure 12 is the block diagram of ISC information encoding apparatus accoding to exemplary embodiment.
Figure 13 is the block diagram of spectrum coding equipment according to another exemplary embodiment.
Figure 14 is the block diagram of spectrum coding equipment according to another exemplary embodiment.
Figure 15 shows the design of the collection and coded treatment of ISC accoding to exemplary embodiment.
Figure 16 shows the design of the collection and coded treatment of ISC according to another exemplary embodiment.
Figure 17 shows TCQ accoding to exemplary embodiment.
Figure 18 is the block diagram of frequency domain audio decoding device accoding to exemplary embodiment.
Figure 19 is the block diagram of frequency spectrum decoding device accoding to exemplary embodiment.
Figure 20 is the block diagram of frequency spectrum inverse quantization equipment accoding to exemplary embodiment.
Figure 21 is the block diagram of frequency spectrum decoding device accoding to exemplary embodiment.
Figure 22 is the block diagram of ISC decoding device accoding to exemplary embodiment.
Figure 23 is the block diagram of ISC information decoding apparatus accoding to exemplary embodiment.
Figure 24 is the block diagram of frequency spectrum decoding device according to another exemplary embodiment.
Figure 25 is the block diagram of frequency spectrum decoding device according to another exemplary embodiment.
Figure 26 is the block diagram of ISC information encoding apparatus according to another exemplary embodiment.
Figure 27 is the block diagram according to the ISC information decoding apparatus of another configuration embodiment shown.
Figure 28 is the block diagram of the multimedia device of the configuration embodiment according to.
Figure 29 is the block diagram according to the multimedia device of another configuration embodiment shown.
Figure 30 is the block diagram according to the multimedia device of another configuration embodiment shown.
Figure 31 is the flow chart for the method for the configuration embodiment according to encoded to frequency spectrum refinement structure.
Figure 32 is the operation for showing the method for the configuration embodiment according to being decoded to frequency spectrum refinement structure
Flow chart.
Specific embodiment
Since present inventive concept can have a variety of improvement embodiments, preferred embodiment is shown in the accompanying drawings and at this
It is described in the detailed description of inventive concept.However, present inventive concept is not limited in specific embodiment by this, it should manage
Solution, present inventive concept cover all improvement, equivalent form and the alternative forms in the idea and technical scope of present inventive concept.This
Outside, detailed description relevant to known function or construction is excluded so as not to can unnecessarily obscure the theme of present inventive concept.
It will be understood that these elements should not be by these although term first and second is used to describe various elements herein
Term is limited.Term is only used to distinguish a component with other assemblies.
In the following description, technical term is only used for explaining specific illustrative embodiment, without limiting present inventive concept.?
Term used in present inventive concept considers the function of present inventive concept and has been chosen as the general art being widely used at present
Language, but can be modified according to the intention of those of ordinary skill in the art, conventional implementation or the introducing of new technology.In addition, if
Exist under specific circumstances by the optional term of applicant, then in this case, the meaning of the term will be in the present invention
It is described in detail in the corresponding description section of design.Therefore, term should be rather than only on based on the entire content of this specification
The title of each term and be defined.
The term of singular may include plural form, unless being cited in contrast to this."comprising", " comprising " or " tool
Have " meaning indicate attribute, region, fixed number, step, processing, element and/or component, but be not excluded for other attributes, area
Domain, fixed number, step, processing, element and/or component.
Hereinafter, it will be described in detail with reference to the accompanying drawings exemplary embodiment.
Fig. 1 a and Fig. 1 b are the block diagram of audio coding apparatus and audio decoding apparatus accoding to exemplary embodiment respectively.
The audio coding apparatus 110 shown in fig 1 a may include that preprocessor 112, frequency-domain encoder 114 and parameter are compiled
Code device 116.The component can be incorporated at least one module, and can be implemented as at least one processor (not shown).
In fig 1 a, preprocessor 112 can execute filtering, down-sampling etc. to input signal, but not limited to this.Input signal
It may include the mixed signal of voice signal, music signal or voice and music.Hereinafter, for convenient for explanation, input signal quilt
Referred to as audio signal.
Frequency-domain encoder 114 can execute time-frequency conversion, selection and audio signal to the audio signal that preprocessor 112 provides
Number of channels, coding frequency band and the corresponding encoding tool of bit rate, and by using the encoding tool of selection to audio signal
It is encoded.Modified Discrete Cosine Tr ansform (MDCT), modulated lapped transform (mlt) (MLT) or fast Fourier can be used to become for time-frequency conversion
(FFT) is changed, but not limited to this.When given amount of bits is enough, general transition coding scheme can be applied to entire frequency band,
When given amount of bits deficiency, bandwidth extension schemes can be applied to partial-band.When audio signal is stereo channel or more
When sound channel, if given amount of bits is enough, coding is executed for each sound channel, it, can if given amount of bits is insufficient
Using the mixed scheme that contracts.The spectral coefficient of coding is generated by frequency-domain encoder 114.
Parametric encoder 116 can be from the spectral coefficient extracting parameter of the coding provided from frequency-domain encoder 114, and to extraction
Parameter encoded.For example, it can be directed to each sub-band (that is, the unit being grouped to spectral coefficient) extracting parameter, and
And the parameter can have even length or heterogeneous length by reflection critical band.When each sub-band is with non-homogeneous
When length, the sub-band present in low-frequency band can have relatively short length relative to the sub-band present in high frequency band
Degree.Quantity and length including sub-band in a frame can change according to encoding and decoding algorithm, and can influence coding efficiency.Institute
Stating parameter may include such as zoom factor, power, average energy or norm, but not limited to this.The frequency obtained as coding result
Spectral coefficient and parameter form bit stream, and the bit stream can be stored in storage medium, or can be with such as packet
Form is sent by channel.
Audio decoding apparatus 130 shown in Fig. 1 b may include parameter decoder 132, frequency domain decoder 134 and post-processing
Device 136.Frequency domain decoder 134 may include frame error concealment algorithm or bag-losing hide algorithm.The component can be incorporated at least
In one module and at least one processor (not shown) can be implemented as.
In Figure 1b, parameter decoder 132 can go out parameter from the bit stream decoding received, and from decoded parameter with frame
The mistake for having occurred such as wiping or lose is checked whether for unit.Various known methods can be used for error checking, and close
It has been that frame or erasing frame or the information of lost frames are provided to frequency domain decoder 134 in present frame.Hereinafter, for the ease of
It explains, wipes frame or lost frames are referred to as error frame.
When present frame has been frame, frequency domain decoder 134 can by via general transformation decoding process execute decoding come
Generate synthesis spectral coefficient.When present frame is error frame, frequency domain decoder 134 can be hidden via frame error concealment algorithm or packet loss
Algorithm is hidden, by repeating the previously spectral coefficient of good frame (PGF) to error frame or passing through the frequency spectrum system via regression analysis to PGF
Then number, which zooms in and out, repeats the spectral coefficient of PGF to error frame to generate synthesis spectral coefficient.Frequency domain decoder 134 can pass through
Frequency-time domain transformation is executed to synthesis spectral coefficient to generate time-domain signal.
Preprocessor 136 can execute filtering, up-sampling etc. for from the time-domain signal that frequency domain decoder 134 provides to carry out
Sound quality improves, but not limited to this.Preprocessor 136 provides the audio signal of reconstruct as output signal.
Fig. 2 a and Fig. 2 b are the audio coding apparatus and sound with switching construction according to another exemplary embodiment respectively
The block diagram of frequency decoding device.
Audio coding apparatus 210 shown in Fig. 2 a may include preconditioner unit 212, mode determiner 213, frequency domain volume
Code device 214, time-domain encoder 215 and parametric encoder 216.The component can be incorporated at least one module and can be by reality
It is now at least one processor (not shown).
In fig. 2 a, since 212 essence of preprocessor is identical as the preprocessor 112 of Fig. 1 a, it is not repeated that it is retouched
It states.
Mode determiner 213 can determine coding mode by referring to the characteristic of input signal.Mode determiner 213 can root
It determines that the coding mode for being suitable for present frame is speech pattern or music pattern according to the characteristic of input signal, and may further determine that
Coding mode efficient for present frame is Modulation or frequency domain mode.It can short-term characteristic by using frame or multiple frames
Long-time quality perceive the characteristic of input signal, but not limited to this.For example, if input signal is corresponding to voice signal,
Coding mode can be confirmed as speech pattern or Modulation, if input signal be different from voice signal signal (for example,
Music signal or mixed signal) accordingly, then coding mode can be confirmed as music pattern or frequency domain mode.As the spy of input signal
Property to music pattern or corresponding frequency domain mode when, mode determiner 213 can provide preprocessor 212 to frequency-domain encoder 214
Output signal, when the characteristic of input signal is to speech pattern or corresponding Modulation, mode determiner 213 can be to time domain coding
The output signal of the offer preprocessor 212 of device 215.
Since the frequency-domain encoder 114 of frequency-domain encoder 214 and Fig. 1 a is substantially identical, its description is not repeated.
Time-domain encoder 215 can execute Code Excited Linear Prediction (CELP) to the audio signal provided from preprocessor 212
Coding.In detail, algebra CELP can be used for CELP coding, but CELP coding is without being limited thereto.The spectral coefficient of coding by when
Domain encoder 215 generates.
Parametric encoder 216 can be mentioned from the spectral coefficient of the coding provided from frequency-domain encoder 214 or time-domain encoder 215
Parameter is taken, and the parameter of extraction is encoded.Since the parametric encoder 116 of parametric encoder 216 and Fig. 1 a is substantially identical,
Therefore its description is not repeated.The spectral coefficient and parameter obtained as coding result can form bit together with coding mode information
Stream, and the bit stream can be sent or can be stored in storage medium by channel in the form of packet.
Audio decoding apparatus 230 shown in Fig. 2 b may include parameter decoder 232, mode determiner 233, frequency domain decoding
Device 234, time-domain decoder 235 and preprocessor 236.Each of frequency domain decoder 234 and time-domain decoder 235 can be every
It include frame error concealment algorithm or bag-losing hide algorithm in a corresponding domain.The component can be integrated at least one module simultaneously
It can be implemented as at least one processor (not shown).
In figure 2b, parameter decoder 232 can go out parameter from the bit stream decoding sent in the form of packet, and from decoding
Parameter checked whether and go wrong as unit of frame.Various known methods can be used for error checking, and about present frame
It has been that the information of frame or error frame is provided to frequency domain decoder 234 or time-domain decoder 235.
Mode determiner 233 can be checked including coding mode information in the bitstream, and present frame can be supplied to frequency
Domain decoder 234 or time-domain decoder 235.
Frequency domain decoder 234 can the operation when coding mode is music pattern or frequency domain mode, and be in present frame
When frame, spectral decoder 234 can generate synthesis spectral coefficient by executing decoding via general transformation decoding process.Working as
Previous frame is error frame, and when the coding mode of previous frame is music pattern or frequency domain mode, frequency domain decoder 234 can pass through frame
Error concealment algorithm or bag-losing hide algorithm, by error frame repeat previously good frame (PGF) spectral coefficient or pass through through
It is zoomed in and out by spectral coefficient of the regression analysis to PGF and generates synthesis frequency then to repeat the spectral coefficient of PGF to error frame
Spectral coefficient.Frequency domain decoder 234 can generate time-domain signal by executing frequency-time domain transformation to synthesis spectral coefficient.
Time-domain decoder 235 can the operation when coding mode is speech pattern or Modulation, and be normal in present frame
Time-domain signal is generated by executing decoding via general CELP decoding process when frame.It is error frame in present frame, and previously
When the coding mode of frame is speech pattern or Modulation, time-domain decoder 235 can execute frame error concealment algorithm in the time domain
Or bag-losing hide algorithm.
Preprocessor 236 time-domain signal provided from frequency domain decoder 234 or time-domain decoder 235 can be executed filtering,
Up-sampling etc., but not limited to this.Preprocessor 236 provides the audio signal of reconstruct as output signal.
Fig. 3 a and Fig. 3 b are the frame of audio coding apparatus and audio decoding apparatus according to another exemplary embodiment respectively
Figure.
Audio coding apparatus 310 shown in Fig. 3 a may include preprocessor 312, linear prediction (LP) analyzer 313, mould
Formula determiner 314, frequency domain excitation encoder 315, time domain excitation encoder 316 and parametric encoder 317.The component can be collected
At at least one module and at least one processor (not shown) can be implemented as.
In fig. 3 a, since the preprocessor 112 of preprocessor 312 and Fig. 1 a is substantially identical, it is not repeated that it is retouched
It states.
LP analyzer 313 can extract LP coefficient by executing LP analysis to input signal, and can be from the LP coefficient of extraction
Generate pumping signal.The pumping signal can be provided to frequency domain excitation encoder 315 according to coding mode and time domain excitation is compiled
One in code device 316.
Since the mode determiner 213 of mode determiner 314 and Fig. 2 a is substantially identical, its description is not repeated.
Frequency domain motivate encoder 315 can the operation when coding mode is music pattern or frequency domain mode, and due in addition to
Input signal is other than pumping signal, and frequency domain motivates the frequency-domain encoder 114 of encoder 315 and Fig. 1 a substantially identical, therefore not
Repeat its description.
Time domain excitation encoder 316 can the operation when coding mode is speech pattern or Modulation, and due to time domain
It motivates the time-domain encoder 215 of encoder 316 and Fig. 2 a substantially identical, therefore is not repeating its description.
Parametric encoder 317 can be from the frequency for the coding for motivating encoder 315 or time domain excitation encoder 316 to provide from frequency domain
Spectral coefficient extracting parameter, and the parameter of extraction is encoded.Due to the parametric encoder 116 of parametric encoder 317 and Fig. 1 a
It is substantially identical, therefore its description is not repeated.The spectral coefficient and parameter obtained as coding result can be believed together with coding mode
Breath forms bit stream, and bit stream can be sent in the form of packet by channel, or can be stored in storage medium.
Audio decoding apparatus 330 shown in Fig. 3 b may include parameter decoder 332, mode determiner 333, frequency domain excitation
Decoder 334, time domain excitation decoder 335, LP synthesizer 336 and preprocessor 337.Frequency domain motivates decoder 334 and time domain
Motivating each of decoder 335 can include frame error concealment algorithm or bag-losing hide algorithm in each corresponding domain.Institute
Stating component can be integrated at least one module and can be implemented as at least one processor (not shown).
In fig 3b, parameter decoder 332 can go out parameter from the bit stream decoding sent in the form of packet, and can be from decoding
Parameter checked whether and go wrong as unit of frame.Various known methods can be used for error checking, and about present frame
It has been that the information of frame or error frame is provided to frequency domain excitation decoder 334 or time domain excitation decoder 335.
Mode determiner 333 can be checked including coding mode information in the bitstream, and present frame can be supplied to frequency
Motivate decoder 334 or time domain excitation decoder 335 in domain.
Frequency domain motivate decoder 334 can the operation when coding mode is music pattern or frequency domain mode, and in present frame
When being frame, frequency domain excitation decoder 334 can generate synthesis frequency spectrum by executing decoding via general transformation decoding process
Coefficient.It is error frame in present frame, and when the coding mode of previous frame is music pattern or frequency domain mode, frequency domain excitation decoding
Device 334 can be by frame error concealment algorithm or bag-losing hide algorithm, by the frequency spectrum system for repeating previously good frame (PGF) to error frame
It counts or by zooming in and out via spectral coefficient of the regression analysis to PGF then to repeat the spectral coefficient of PGF to error frame
To generate synthesis spectral coefficient.Frequency domain excitation decoder 334 can be made by executing frequency-time domain transformation to synthesis spectral coefficient to generate
For the pumping signal of time-domain signal.
Time domain excitation decoder 335 can the operation when coding mode is speech pattern or Modulation, and be in present frame
The pumping signal as time-domain signal is generated by executing decoding via general CELP decoding process when good frame.It is in present frame
Error frame, and when the coding mode of previous frame is speech pattern or Modulation, time domain excitation decoder 335 can be in the time domain
Execute frame error concealment algorithm or bag-losing hide algorithm.
LP synthesizer 336 can be by believing the excitation for motivating decoder 334 or time domain excitation decoder 335 to provide from frequency domain
Number execute LP synthesis to generate time-domain signal.
Preprocessor 337 can execute filtering, up-sampling etc. to the time-domain signal provided from LP synthesizer 336, but be not limited to
This.Preprocessor 337 provides the audio signal of reconstruct as output signal.
Fig. 4 a and Fig. 4 b are the audio coding apparatus and sound with switching construction according to another exemplary embodiment respectively
The block diagram of frequency decoding device.
Audio coding apparatus 410 shown in Fig. 4 a may include preprocessor 412, mode determiner 413, frequency-domain encoder
414, LP analyzer 415, frequency domain excitation encoder 416, time domain excitation encoder 417 and parametric encoder 418.The component can
It is integrated at least one module and at least one processor (not shown) can be implemented as.Show due to being contemplated that in Fig. 4 a
Audio coding apparatus 410 out can carry out group by the audio coding apparatus 310 of audio coding apparatus 210 and Fig. 3 a to Fig. 2 a
It closes to obtain, therefore the description of the operation of common elements is not again repeated, and pattern determining unit 413 will now be described
Operation.
Mode determiner 413 can determine the coding mould of input signal by referring to the characteristic and bit rate of input signal
Formula.Mode determiner 413, which can be based on present frame according to the characteristic of input signal, to be speech pattern or music pattern and is based on
Coding mode efficient for present frame is Modulation or frequency domain mode, and coding mode is determined as CELP mode or another
Mode.Coding mode can be determined as CELP mode when the characteristic of input signal is corresponding to speech pattern by mode determiner 413,
Coding mode is determined as frequency domain mode when the characteristic of input signal is to music pattern and corresponding high bit rate, and is believed in input
Number characteristic coding mode is determined as audio mode to when music pattern and corresponding low bit rate.Mode determiner 413 can be
Input signal is supplied to frequency-domain encoder 414 when being frequency domain mode by coding mode, when coding mode is audio mode via
Input signal is supplied to frequency domain excitation encoder 416 by LP analyzer 415, and when coding mode is CELP mode via LP points
Input signal is supplied to time domain excitation encoder 417 by parser 415.
Frequency-domain encoder 414 can be compiled with the audio of frequency-domain encoder 114 or Fig. 2 a in the audio coding apparatus 110 of Fig. 1 a
Frequency-domain encoder 214 in decoding apparatus 210 is corresponding, and frequency domain motivates encoder 416 or time domain excitation encoder 417 can be with Fig. 3 a's
Frequency domain excitation encoder 315 or time domain excitation encoder 316 in audio coding apparatus 310 is corresponding.
Audio decoding apparatus 430 shown in Fig. 4 b may include parameter decoder 432, mode determiner 433, frequency domain decoding
Device 434, frequency domain excitation decoder 435, time domain excitation decoder 436, LP synthesizer 437 and preprocessor 438.Frequency domain decoder
434, frequency domain excitation each of decoder 435 and time domain excitation decoder 436 can include that frame is poor in each corresponding domain
Wrong hidden algorithm or bag-losing hide algorithm.The component can be incorporated at least one module and can be implemented as at least one
Processor (not shown).As being contemplated that audio decoding apparatus 430 shown in Fig. 4 b can be by the audio decoder to Fig. 2 b
The audio decoding apparatus 330 of equipment 230 and Fig. 3 b are combined to obtain, thus to the description of the operation of common elements no longer by
It repeats, and the operation of mode determiner 433 will now be described.
Mode determiner 433 can be checked including coding mode information in the bitstream, and present frame can be supplied to frequency
Domain decoder 434, frequency domain excitation decoder 435 or time domain excitation decoder 436.
Frequency domain decoder 434 can be with the audio solution of frequency domain decoder 134 or Fig. 2 b in the audio decoding apparatus 130 of Fig. 1 b
Frequency domain decoder 234 in decoding apparatus 230 is corresponding, and frequency domain motivates decoder 435 or time domain excitation decoder 436 can be with Fig. 3 b's
Frequency domain excitation decoder 334 or time domain excitation decoder 335 in audio decoding apparatus 330 is corresponding.
Fig. 5 is the block diagram of frequency domain audio encoding device accoding to exemplary embodiment.
Frequency domain audio encoding device 510 shown in Fig. 5 may include transient detector 511, converter 512, Modulation recognition
Device 513, energy coding device 514, frequency spectrum normalizer 515, bit distributor 516, spectral encoders 517 and multiplexer 518.Institute
Stating component can be incorporated at least one module and can be implemented as at least one processor (not shown).Frequency domain audio coding
The repertoire of frequency domain audio encoder 214 shown in Figure 2 and the partial function of parametric encoder 216 can be performed in equipment 510.
Other than signal classifier 513, frequency domain audio encoding device 510 can be by the encoder disclosed in ITU-T G.719 standard
Configuration substitution, and converter 512 can be used with 50% overlapping the duration mapping window.In addition, in addition to transition
Other than detector 511 and signal classifier 513, frequency domain audio encoding device 510 can be disclosed in ITU-T G.719 standard
The configuration of encoder substitutes.In several cases, although not shown, but as in ITU-T G.719 standard, noise level is estimated
Meter unit can be further included the rear end of spectral encoders 517 to estimate unallocated in bit allocation process to have bit
The noise level of spectral coefficient, and the noise level estimated is inserted into bit stream.
Referring to Fig. 5, transient detector 511 can be detected by being analyzed input signal shows that transient response continues
Time, and generate in response to testing result the transient state signaling information of every frame.Various known methods, which can be used to detection transient state, to be continued
Time.Accoding to exemplary embodiment, transient detector 511 can first determine that whether present frame is transient frame, then to true
The present frame for being set to transient frame is verified.The transient state signaling information device 518 that can be re-used includes that in the bitstream, and can be provided that
To converter 512.
Converter 512 can determine the window size that will be used to convert according to the testing result to transient durations,
And time-frequency conversion is executed based on determining window size.For example, short window can be applied to have detected that transient durations
Sub-band, and long window can be applied to that the sub-bands of transient durations is not detected.It as another example, can be by short window
Mouth is applied to the frame including transient durations.
Signal classifier 513 can analyze the frequency spectrum provided from converter 512 as unit of frame, with the every frame of determination
It is whether corresponding to harmonic wave frame.Various known methods can be used to determine harmonic wave frame.Accoding to exemplary embodiment, signal classifier
The frequency spectrum provided from converter 512 can be divided into multiple sub-bands by 513, and obtain the peak value energy peace of each sub-band
Equal energy value.Later, signal classifier 513 can get peak value energy estimated rate bigger than the average energy value of every frame or more
Sub-band quantity, and by the quantity of the sub-band of acquisition be greater than or equal to predetermined value frame be determined as harmonic wave frame.It is described pre-
Fixed-ratio and the predetermined value can be determined in advance and testing or emulating.The harmonic wave signaling information device 518 that can be re-used includes
In the bitstream.
Energy coding device 514 can obtain energy as unit of each sub-band, and the energy can be carried out quantization and
Lossless coding.According to embodiment, norm value corresponding with the average frequency spectrum energy as unit of each sub-band is used as institute
Energy is stated, and zoom factor or power can be used as, but energy is without being limited thereto.The norm value of each sub-band can be provided that
To frequency spectrum normalizer 515 and bit distributor 516, and the device 518 that can be re-used includes in the bitstream.
Frequency spectrum normalizer 515 can carry out normalizing to frequency spectrum by using the norm value obtained as unit of each sub-band
Change.
Bit distributor 516 can by using the norm value obtained as unit of each sub-band, with graduation of whole numbers of units or with
Sub-multiple unit distributing bit.In addition, bit distributor 516 can be come by using the norm value obtained as unit of each sub-band
Masking threshold is calculated, and estimates to feel the amount of bits needed (that is, admissible bit number by using the masking threshold
Amount).Bit distributor 516 can limit: for each sub-band, the amount of bits distributed is no more than admissible bit number
Amount.Bit distributor 516 can from the sub-band with larger norm value order-assigned bit, and according to the sense of each sub-band
Know importance, the norm value of each sub-band is weighted to adjust distributed amount of bits, so that more bit numbers
Amount is assigned to the sub-band of perceptual important.The norm value of bit distributor 516 being quantized is supplied to from energy coding device 514
It can be distributed after being pre-adjusted for bit, to consider psychologic acoustics weighted sum masking effect according to ITU-TG.719 standard
Fruit.
Spectral encoders 517 can be come by using the amount of bits of each sub-band distributed to normalized frequency spectrum into
Row quantization, and quantized result is reversibly encoded.For example, TCQ, USQ, FPC, AVQ and PVQ or their combination and being directed to
The optimised lossless encoder of each quantizer can be used for spectrum coding.In addition, grid coding may be alternatively used for spectrum coding,
But spectrum coding is without being limited thereto.In addition, can also be used various according to the environment for having corresponding codec or user demand is realized
Spectrum coding method.Information about the frequency spectrum encoded by spectral encoders 517 can be re-used device 518 including in the bitstream.
Fig. 6 is the block diagram of frequency domain audio encoding device accoding to exemplary embodiment.
Frequency domain audio encoding device 600 shown in Fig. 6 may include preprocessor 610, frequency-domain encoder 630, time domain volume
Code device 650 and multiplexer 670.Frequency-domain encoder 630 may include transient detector 631, converter 633 and spectral encoders 635.
The component can be incorporated at least one module and can be implemented as at least one processor (not shown).
Referring to Fig. 6, preprocessor 610 can execute filtering, down-sampling etc. to input signal, but not limited to this.Preprocessor
610 can determine coding mode according to characteristics of signals.Preprocessor 610 can determine the coding for being suitable for present frame according to characteristics of signals
Mode is speech pattern or music pattern, and may further determine that coding mode efficient for present frame is Modulation or frequency
Domain model.Perceptual signal characteristic can be come by using the short-term characteristic of frame or the long-time quality of multiple frames, but not limited to this.Example
Such as, if input signal is corresponding to voice signal, coding mode can be confirmed as speech pattern or Modulation, if input
Signal is corresponding to signal (that is, music signal or the mixed signal) of voice signal is different from, then coding mode can be confirmed as sound
Happy mode or frequency domain mode.When characteristics of signals is to music pattern or corresponding frequency domain mode, preprocessor 610 can be to Frequency Domain Coding
Device 630 provides input signal, and when characteristics of signals is to speech pattern or corresponding Modulation, preprocessor 610 can be compiled to time domain
Code device 660 provides input signal.
Frequency-domain encoder 630 can be handled the audio signal provided from preprocessor 610 based on transition coding scheme.
In detail, transient detector 631 can detect transient component from audio signal, and determine whether present frame is corresponding to transition frame.
Converter 633 can determine the length or shape of mapping window based on frame type (that is, the transient state information provided from transient detector 631)
Shape, and audio signal can be transformed into frequency domain based on determining mapping window.As the example of transformation tool, discrete cosine is improved
Transformation (MDCT), modulated lapped transform (mlt) (MLT) or Fast Fourier Transform (FFT) (FFT) can be used.In general, can be by short variation
Window is applied to the frame including transient component.Spectral encoders 635 can execute coding to the audible spectrum for transforming to frequency domain.It will ginseng
Spectral encoders 635 are more fully described according to Fig. 7 and Fig. 9.
Time-domain encoder 650 can execute Code Excited Linear Prediction (CELP) to the audio signal provided from preprocessor 610
Coding.In detail, algebra CELP can be used for CELP coding, but CELP coding is without being limited thereto.
Multiplexer 670 can be by spectrum component or signal component and as in frequency-domain encoder 630 or time-domain encoder 650
The result encoded and the variable index generated is multiplexed, to generate bit stream.The bit stream can be stored in storage
In medium, or it can be sent in the form of packet by channel.
Fig. 7 is the block diagram of frequency spectrum decoding device accoding to exemplary embodiment.Spectrum coding equipment shown in Fig. 7 can be with
The spectral encoders 635 of Fig. 6 are corresponding, can be included in another Frequency Domain Coding equipment, or can be implemented separately.
Spectrum coding equipment shown in Fig. 7 may include energy estimator 710, Energy Quantization and coding unit 720, bit
Distributor 730, frequency spectrum normalizer 740, spectrum quantification and coding unit 750 and noise filling device 760.
Referring to Fig. 7, energy estimator 710 original spectral coefficients can be divided into multiple sub-bands and estimated energy (for example,
The norm value of each sub-band).Each sub-band can have uniform length in frame.When each sub-band is with heterogeneous
When length, the quantity including the spectral coefficient in sub-band can increase from low-frequency band to high frequency band.
Energy Quantization and coding unit 720 can be quantified and be encoded to the norm value of the estimation of each sub-band.It can lead to
Cross various tools (such as vector quantization (VQ), scalar quantization (SQ), Trellis coding quantization (TCQ), lattice vector quantization (LVQ)
Deng) norm value quantified.In addition Energy Quantization and coding unit 720 can execute lossless coding to further increase
Code efficiency.
Bit distributor 730 based on the norm value of the quantization of each sub-band, can consider the permissible bit of frame to distribute
Bit needed for coding.
Frequency spectrum normalizer 740 can be normalized frequency spectrum based on the norm value obtained for each sub-band.
Spectrum quantification and coding unit 750 can carry out normalized frequency spectrum based on the bit of the distribution of each sub-band
Quantization and coding.
Noise can be added to due to the permissible bit in spectrum quantification and coding unit 750 by noise filling device 760
Constrain and be quantified as zero component.
Fig. 8 shows sub-band segmentation.
Referring to Fig. 8, when input signal is using the sample frequency of 48KHz and has the frame sign of 20ms, for every frame
The quantity of processed sample is become 960.That is, when the MDCT by using the overlapping with 50% is to input signal
When being converted, 960 spectral coefficients are obtained.The ratio of overlapping can be provided with differently configured according to encoding scheme.In a frequency domain, directly
Frequency band to 24KHz theoretically can be processed, and is contemplated that the range of audibility to indicate the frequency band until 20KHz.It is arrived 0
In the low-frequency band of 3.2KHz, sub-band includes 8 spectral coefficients.In the frequency band of 3.2KHz to 6.4KHz, sub-band includes 16
A spectral coefficient.In the frequency band of 6.4KHz to 13.6KHz, sub-band includes 24 spectral coefficients.In 13.6KHz to 20KHz
Frequency band in, sub-band include 32 spectral coefficients.For the predetermined frequency band being arranged in encoding device, it can be performed and be based on norm
The coding of value, and the high frequency band for being higher than the predetermined frequency band can be applied based on kinds of schemes (such as, bandwidth expansion)
Coding.
Fig. 9 is the block diagram for showing the configuration of spectrum quantification equipment accoding to exemplary embodiment.
Equipment shown in Fig. 9 may include quantizer selecting unit 910, USQ 930 and TCQ 950.
In Fig. 9, quantizer selecting unit 910 can according to by the characteristic of the signal being quantized (that is, input signal) each
Most efficient quantizer is selected among kind quantizer.As the characteristic of input signal, the bit of each frequency band can be used to distribute letter
Breath, frequency band dimension information etc..According to selection as a result, can by by the signal being quantized be supplied to USQ 930 and TCQ 950 it
One, so that corresponding quantization is performed.
Figure 10 is the block diagram of the configuration of spectrum coding equipment accoding to exemplary embodiment.Equipment shown in Figure 10 can be with
The spectrum quantification and coding unit 750 of Fig. 7 is corresponding, can be included in another Frequency Domain Coding equipment, or can be by independently real
It is existing.
Equipment shown in Figure 10 may include coding method selecting unit 1010, Zero-code unit 1020, unit for scaling
1030, ISC coding unit 1040, quantized components recovery unit 1050 and inverse unit for scaling 1060.Here, quantized components restore single
Member 1050 and inverse unit for scaling 1060 can be provided.
In Figure 10, coding method selecting unit 1010 can select coding method by considering input signal characteristics.It is defeated
Entering characteristics of signals may include the bit to each bandwidth assignment.It, can be by normalizing based on the encoding scheme for being directed to each frequency band selection
The frequency spectrum of change is supplied to Zero-code unit 1020 or unit for scaling 1030.According to embodiment, each sample of frequency band is distributed to
Average number of bits amount is greater than or equal to predetermined value (for example, 0.75) can be to respective tones by determining that frequency band is extremely important
Band uses USQ, and TCQ can be used for every other frequency band.Here, it is contemplated that frequency band length or frequency band size are described to determine
Average number of bits amount.The label of a bit can be used the coding method of selection is arranged.
All encoding samples are zero (0) by the frequency band that Zero-code unit 1020 can be zero for distributed bit.
Unit for scaling 1030 can adjust bit rate by zooming in and out based on the bit for distributing to frequency band to frequency spectrum.At this
In the case of kind, normalized frequency spectrum can be used.Unit for scaling 1030 can be by considering to be assigned to including in each of frequency band
The average number of bits amount of sample (that is, spectral coefficient) executes scaling.For example, average number of bits amount is bigger, bigger scaling can
It is performed.
According to embodiment, unit for scaling 1030 can determine suitable scaling according to the bit distribution for each frequency band
Value.
In detail, firstly, frequency band length and bit distribution information can be used to estimate the number of pulses of present band.This
In, pulse can indicate unit pulse.Before being estimated, the actually required bit (b) of present band can be by equation 1 come based on
It calculates.
Wherein, n indicates frequency band length, and m indicates number of pulses, and i indicates the non-zero with important spectral coefficient (ISC)
The quantity of position.
It can be based on the quantity for for example obtaining non-zero position according to the probability of equation 2.
In addition, amount of bits needed for non-zero position can be estimated by equation 3.
bnzp=log2(pNZP(i)) (3)
Finally, can be by having with the value b for the immediate value of bit for distributing to each frequency band come the quantity of strobe pulse.
Next, the estimation of the quantity to the pulse obtained for each frequency band and the absolute value of input signal can be passed through
To determine original zoom factor.Input signal can be zoomed in and out according to original zoom factor.If for the original by scaling
The sum of the number of pulses of beginning signal (that is, signal of quantization) is not identical as the number of pulses of estimation, and the scaling of update may be used
The factor is handled to execute pulse redistribution.It is handled according to pulse redistribution, if for the umber of pulse of present band selection
Amount is less than the number of pulses for the estimation obtained of each frequency band, then by reducing zoom factor, number of pulses increases, no
Then, if the number of pulses for present band selection is greater than the number of pulses for the estimation obtained of each frequency band, lead to
Increase zoom factor is crossed, number of pulses reduces.In this case, the smallest position of distortion of original signal is made by selection,
Zoom factor can be increased or reduce predetermined value.
Since the distortion function of TSQ needs relative size rather than accurate distance, the distortion function of TSQ can be such as
In the sum of the squared-distance being obtained in each frequency band between quantized value and non-quantized value shown in formula 4.
Wherein, piIndicate actual value, qiIndicate quantized value.
Euclidean distance can be used to determine optimal quantization value in the distortion function of USQ.In this case, including zoom factor
Improvement equation to can be used for making computation complexity minimum, and 5 calculated distortion function of equation can be passed through.
If the number of pulses and desirable value for each frequency band mismatch, need to increase while keeping minimum metric
Add deduct the pulse of few predetermined quantity.This can be reached by increasing or deleting single pulse and then repeated until number of pulses
It is performed until desirable value according to recursive fashion.
In order to increase or delete a pulse, need to obtain n distortion value to select optimal distortion value.For example, such as equation 6
Shown in, distortion value j can be corresponding to pulse is added to j-th of position in frequency band.
In order to avoid equation 6 is performed n times, deviation can be used as shown in equation 7.
In equation 7,It can only be calculated only once.In addition, n indicates frequency band length (that is, in frequency band
Number of coefficients), p indicates original signal (that is, input signal of quantizer), and q indicates quantized signal, g indicate scaling because
Son.Finally, the position j that distortion d is minimized can be selected, q is thus updatedj。
For control bit rate, the spectral coefficient that can be scaled by using process simultaneously selects suitable ISC to execute coding.
In detail, the bit distribution for each frequency band can be used to select the spectrum component for quantization.In this case, may be used
Based on selecting spectrum component according to the various combinations of the distribution of spectrum component and variance.Next, actual non-zero can be calculated
Position.It can be by being analyzed the nonzero digit to obtain non-zero position, and select in this way to amount of zoom and redistribution operation
It sets and is referred to alternatively as ISC.In short, optimal to obtain by the amplitude for analyzing the signal for having gone through scaling and redistributing processing
Zoom factor and non-zero location information corresponding with ISC.Here, the quantity of non-zero location information instruction non-zero position and position
It sets.If the quantity of pulse is not under control by scaling and redistribution processing, the arteries and veins to selection can be handled by TCQ
Capable quantization is rushed in, and quantized result can be used to adjust spare bits.The processing is shown as follows.
Quantity for non-zero position is not identical as the number of pulses of the estimation for each frequency band and is greater than predetermined value
(for example, 1) and quantizer select the condition of information instruction TCQ, can be quantified by practical TCQ to adjust spare bits.In detail
It says, in situation corresponding with the condition, TCQ quantification treatment is first carried out to adjust spare bits.If quantified by TCQ
The exact amount of the pulse of the present band of acquisition is less than the number of pulses estimated previously obtained for each frequency band, then leads to
It crosses and TCQ is quantified predetermined zoom factor is multiplied with the value (for example, 1.1) greater than 1 to increase zoom factor, it is otherwise, logical
It crosses and practical TCQ is quantified predetermined zoom factor is multiplied with the value (for example, 0.9) less than 1 to reduce zoom factor.Work as needle
The pulse of the TCQ present band for quantifying to obtain is passed through with by repeating the processing to the number of pulses for the estimation that each frequency band obtains
When quantity is identical, spare bits are updated by calculating the bit used in practical TCQ quantification treatment.It obtains through this process
Non-zero position can be corresponding to ISC.
ISC coding unit 1040 can the information to the quantity of the ISC about final choice and the letter about non-zero position
Breath is encoded.It in this process, can be using lossless coding to improve code efficiency.ISC coding unit 1040 can be used selected
The quantizer of the non-zero frequency band for distributed bit non-zero selected executes coding.In detail, ISC coding unit 1040
The ISC of each frequency band can be selected for normalized frequency spectrum, and based on quantity, position, amplitude and symbol to the ISC about selection
Information encoded.In such a case, it is possible to be compiled in a manner of being different from quantity, position and symbol to ISC amplitude
Code.For example, one of USQ and TCQ can be used to carry out quantization and by arithmetic coding in ISC amplitude, and the quantity of ISC, position and symbol
It number can be by arithmetic coding.If it is determined that special frequency band includes important information, USQ may be used, TCQ otherwise can be used.According to reality
Example is applied, one of TCQ and USQ can be selected based on characteristics of signals.Here, characteristics of signals may include distributing to the bit of each frequency band
Or frequency band length.If distribute to the average number of bits amount including each sample in frequency band more than or equal to threshold value (for example,
0.75) it, then can determine that frequency band includes very important information, and USQ therefore can be used.Even if long with short frequency band
In the case where the low-frequency band of degree, USQ can also be according to circumstances used.According to another embodiment, the first joint can be used according to bandwidth
One of scheme and the second scheme for combining.For example, the first scheme for combining can be used for NB and WB, for SWB and FB, can be used
Second scheme for combining, wherein, can other than for the original bit of each frequency band distribution information in the first scheme for combining
Quantizer is selected by the secondary bit distribution in addition using the spare bits to the frequency band from previous coding to be handled,
In the second scheme for combining, for the frequency band for using USQ is determined, TCQ can be used to least significant bit (LSB).In the first joint
In scheme, it can be distributed to select to handle two frequency bands by the spare bits to the frequency band from previous coding
Secondary bit distribution.In the second scheme for combining, USQ can be used to remaining bit.
Quantized components recovery unit 1050 can by the position ISC, amplitude and symbolic information are added to the component of quantization come
Restore the component of actual quantization.Here, can zero setting position spectral coefficient (spectral coefficient for being encoded as zero) distribution zero.
Inverse unit for scaling 1060 can be exported and normalized input by carrying out reverse scaling to the quantized components of recovery
The quantization spectral coefficient of the rank same levels of frequency spectrum.Identical scaling can be used in unit for scaling 1030 and inverse unit for scaling 1060
The factor.
Figure 11 is the block diagram for showing the configuration of ISC encoding device accoding to exemplary embodiment.
Equipment shown in Figure 11 may include ISC selecting unit 1110 and ISC information coding unit 1130.The equipment of Figure 11
Can be corresponding to the ISC coding unit 1040 of Figure 10, or autonomous device can be implemented as.
In Figure 11, ISC selecting unit 1110 can select ISC to adjust ratio based on preassigned from by the frequency spectrum of scaling
Special rate.ISC selecting unit 1110 can be by obtaining actual non-zero position from the spectrum analysis zoom degree by scaling.This
In, ISC can be corresponding to the actual non-zero spectral coefficient before scaling.ISC selecting unit 1110 can be based on for each bandwidth assignment
Bit, spectral coefficient (that is, non-zero position) that will be encoded is selected by considering distribution and the variance of spectral coefficient.For
ISC selects that TCQ can be used.
ISC information coding unit 1130 based on the ISC of selection to ISC information (that is, the quantity information of ISC, location information,
Amplitude information and symbol) it is encoded.
Figure 12 is the block diagram for showing the configuration of ISC information encoding apparatus accoding to exemplary embodiment.
Equipment shown in Figure 12 may include location information coding unit 1210, amplitude information coding unit 1230 and symbol
Coding unit 1250.
In Figure 12, location information coding unit 1210 can be to the ISC's selected by ISC selecting unit (1110 of Figure 11)
Location information (that is, location information of non-zero frequency spectral coefficient) is encoded.Location information may include quantity and the position of the ISC of selection
It sets.Arithmetic coding can be used for encoding location information.New buffer can be configured by collecting the ISC of selection.It is right
It is collected in ISC, zero-frequency band and non-selected frequency spectrum can be excluded.
Amplitude information coding unit 1230 can encode the amplitude information of the ISC newly configured.In this case, may be used
Quantization is executed by selection one of TCQ and USQ, and then can in addition execute arithmetic coding.In order to improve the effect of arithmetic coding
The quantity of non-zero location information and ISC can be used in rate.
Symbolic information coding unit 1250 can encode the symbolic information of the ISC of selection.Arithmetic coding can be used for
Symbolic information is encoded.
Figure 13 is the block diagram for showing the configuration of spectrum coding equipment according to another exemplary embodiment.Shown in Figure 13
Equipment can be corresponding to the spectrum quantification of Fig. 7 and coding unit 750, can perhaps be included in another Frequency Domain Coding equipment or
It can be implemented separately.
Equipment shown in Figure 13 may include unit for scaling 1330, ISC coding unit 1340, quantized components recovery unit
1350 and inverse unit for scaling 1360.Compared with Figure 10, in addition to Zero-code unit 1020 and coding method selecting unit 1010 are saved
Slightly and ISC coding unit 1340 is using except TCQ, and the operation of each component is identical.
Figure 14 is the block diagram for showing the configuration of spectrum coding equipment according to another exemplary embodiment.Shown in Figure 14
Equipment can be corresponding to the spectrum quantification of Fig. 7 and coding unit 750, can perhaps be included in another Frequency Domain Coding equipment or
It can be implemented separately.
Equipment shown in Figure 14 may include coding method selecting unit 1410, unit for scaling 1430, ISC coding unit
1440, quantized components recovery unit 1450 and inverse unit for scaling 1460.Compared with Figure 10, in addition to Zero-code unit 1020 is omitted
Except, the operation of each component is identical.
Figure 15 shows the design of the collection and coded treatment of ISC accoding to exemplary embodiment.Firstly, zero-frequency band (that is, will be by
It is quantified as zero frequency band) it is omitted.Next, the ISC that can be selected by using the spectrum component present in the non-zero frequency band is come
Configure new buffer.TCQ and corresponding lossless coding can be executed to the ISC newly configured as unit of frequency band.
Figure 16 shows the design of the collection and coded treatment of ISC according to another exemplary embodiment.Firstly, zero-frequency band (that is,
To be quantified as zero frequency band) it is omitted.Next, can be selected by using the spectrum component present in the non-zero frequency band
ISC configures new buffer.USC or TCQ and corresponding lossless volume can be executed to the ISC newly configured as unit of frequency band
Code.
Figure 17 shows TCQ accoding to exemplary embodiment, and there are two the other four coset grids of state eight-shaped of zero level with tool
Structure is corresponding.The detailed description of corresponding TCQ is disclosed in the Pat. Registration that number is US 7605725.
Figure 18 is the block diagram for showing the configuration of frequency domain audio decoding device accoding to exemplary embodiment.
Frequency domain audio decoding device 1800 shown in Figure 18 may include frame error detection unit 1810, frequency domain decoding unit
1830, time domain decoding unit 1850 and post-processing unit 1870.Frequency domain decoding unit 1830 may include frequency spectrum decoding unit 1831,
Memory updating unit 1833, inverse transformation block 1835 and overlap-add (OLA) unit 1837.Each component can be incorporated in
It can be realized in a few module and by least one processor (not shown).
Referring to Fig.1 8, frame error detection unit 1810 from the bit stream received can detect whether that frame error has occurred.
Frequency domain decoding unit 1830 can be operated when coding mode is music pattern or frequency domain mode, frame occurring
FEC algorithm or PLC algorithm are enabled when mistake, and time domain is generated by general transformation decoding process when there is not frame error
Signal.In detail, frequency spectrum decoding unit 1831 can execute frequency spectrum decoding by using decoded parameter to synthesize spectral coefficient.
9 and 20 frequency spectrum decoding unit 1831 will be more fully described referring to Fig.1.
The synthesis spectral coefficient of the renewable present frame as normal frame of memory updating unit 1833 is directed to subsequent frame more
Newly believed using the information of decoded gain of parameter, the quantity of continuous error frame so far, the characteristics of signals of every frame, frame type
Breath etc..Here, characteristics of signals may include transient response and static characteristic, and frame type may include transition frame, static frames or harmonic wave frame.
Inverse transformation block 1835 can generate time-domain signal by executing time-frequency inverse transformation to the spectral coefficient of synthesis.
OLA unit 1837 can execute OLA processing by using the time-domain signal of previous frame, generate for the final of present frame
Time-domain signal as OLA handle as a result, and final time-domain signal is supplied to post-processing unit 1870.
Time domain decoding unit 1850 can the operation when coding mode is speech pattern or Modulation, when there is frame error
FEC or PLC algorithm is enabled, and time-domain signal is generated by general CELP decoding process when there is not frame error.
Post-processing unit 1870 can be to the time-domain signal provided from frequency domain decoding unit 1830 or time domain decoding unit 1850
Filtering or up-sampling are executed, but not limited to this.Post-processing unit 1870 can provide the audio signal of recovery as output signal.
Figure 19 is the block diagram for showing the configuration of frequency spectrum decoding device accoding to exemplary embodiment.Equipment shown in Figure 19
Can be corresponding to the frequency spectrum decoding unit 1831 of Figure 18, or can be included in another frequency domain decoding device, or by independently real
It is existing.
Frequency spectrum decoding device 1900 shown in Figure 19 may include energy decoding and inverse quantization unit 1910, bit distributor
1930, frequency spectrum decoding and inverse quantization unit 1950, noise filling device 1970 and spectral shaping unit 1990.Here, noise filling
Device 1970 can be located at the rear end of spectral shaping unit 1990.Each component can be incorporated at least one module and can be by least
One processor (not shown) is realized.
Referring to Fig.1 9, energy decoding and inverse quantization unit 1910 (can such as perform in the encoding process nothing to energy
The parameter for damaging coding, for example, norm value) losslessly encoding is carried out, and inverse quantization is carried out to decoded norm value.It can be used and compiling
Inverse quantization is executed for the corresponding scheme of quantization scheme of norm value in code processing.
Bit distributor 1930 can norm value or inverse quantization based on quantization norm value, distribute needed for each sub-band
The bit of quantity.In this case, for the amount of bits of each subband allocation can and the ratio that distributes in the encoding process
Special quantity is identical.
Frequency spectrum decoding and inverse quantization unit 1950 can be by using the amount of bits for each subband allocation to coding
Spectral coefficient carry out losslessly encoding and execute inverse quantization processing to decoded spectral coefficient generating normalized spectral coefficient.
Noise filling device 1970 can be in the portion for needing noise filling of each sub-band among normalized spectral coefficient
Noise is filled in point.
Spectral shaping unit 1990 can carry out shaping to normalized spectral coefficient by using the norm value of inverse quantization.It can
Final decoded spectral coefficient is obtained by frequency spectrum shaping processing.
Figure 20 is the block diagram for showing the configuration of frequency spectrum inverse quantization equipment accoding to exemplary embodiment.
Equipment shown in Figure 20 may include inverse DCT selecting unit 2010, USQ 2030 and TCQ2050.
In Figure 20, inverse DCT selecting unit 2010 can be according to the spy of input signal (that is, by by signal of inverse quantization)
Property selects most efficient inverse DCT among various inverse DCTs.Bit distribution information, bit size for each frequency band
Information etc. can be used as the characteristic of input signal.According to selection as a result, USQ 2030 can will will be supplied to by the signal of inverse quantization
One of with TCQ 2050, so that corresponding inverse quantization is performed.
Figure 21 is the block diagram for showing the configuration of frequency spectrum decoding device accoding to exemplary embodiment.Equipment shown in Figure 21
It can be decoded to the frequency spectrum of Figure 19 and inverse quantization unit 1950 is corresponding, or can be included in another frequency domain decoding device, or by
Independently realize.
Equipment shown in Figure 21 may include coding/decoding method selecting unit 2110, zero decoding unit 2130, ISC decoding unit
2150, quantized components recovery unit 2170 and inverse unit for scaling 2190.Here, quantized components recovery unit 2170 and inverse scaling are single
Member 2190 can be provided selectively.
In Figure 21, coding/decoding method selecting unit 2110 can select decoding side based on the bit for each bandwidth assignment
Method.Normalized frequency spectrum can be supplied to zero decoding unit 2130 or ISC is solved based on the coding/decoding method for being directed to each frequency band selection
Code unit 2150.
All samples are decoded as zero by the frequency band that zero decoding unit 2130 can be zero for the bit of distribution.
ISC decoding unit 2150 can carry out the frequency band that the bit of distribution is not zero by using the inverse DCT of selection
Decoding.ISC decoding unit 2150 can obtain the information about important frequencies component, and base for each frequency band of the frequency spectrum of coding
The information about important frequencies component obtained for each frequency band is decoded in quantity, position, amplitude and symbol.It can be with
Important frequency component amplitudes are decoded in a manner of being different from quantity, position and symbol.For example, important frequencies component amplitude
It can be arithmetic decoded, and can be used one of USQ and TCQ by inverse quantization, and the quantity of important frequencies component, position and symbol can quilts
Arithmetic decoding.The identical result in ISC coding unit 1040 that can be used and be shown in FIG. 10 executes the choosing of inverse DCT
It selects.The frequency band that ISC decoding unit 2150 can be not zero by using the bit that a pair of TCQ and USQ is distributed carries out inverse quantization.
Quantized components recovery unit 2170 can restore actual amount based on the position of the ISC of recovery, amplitude and symbolic information
The component of change.It here, can be by zero dispensing zero position (that is, non-quantized part as the spectral coefficient for being decoded as zero).
Inverse unit for scaling (not shown) can be further included to carry out inverse scaling to the quantized components of recovery, to export
The spectral coefficient of quantization identical with normalized frequency spectrum rank.
Figure 22 is the block diagram for showing the configuration of ISC decoding device accoding to exemplary embodiment.
Equipment shown in Figure 22 may include number of pulses estimation unit 2210 and ISC information decoding unit 2230.Figure 22
Shown in equipment can be corresponding to the ISC decoding unit of Figure 21, or autonomous device can be implemented as.
In Figure 22, number of pulses estimation unit 2210 can be worked as by using frequency band size and bit distribution information to determine
The estimated value of number of pulses needed for preceding frequency band.That is, due to the bit distribution information of present frame and the bit of encoder
It is identical to distribute information, therefore decoding can be executed by using identical bit distribution information to obtain estimating for identical number of pulses
Evaluation.
ISC information decoding unit 2230 can the number of pulses based on estimation to ISC information (that is, the quantity information of ISC, position
Confidence breath, amplitude information and symbol) it is decoded.
Figure 23 is the block diagram for showing the configuration of ISC information decoding apparatus accoding to exemplary embodiment.
Equipment shown in Figure 23 may include location information decoding unit 2310, amplitude information decoding unit 2330 and symbol
Decoding unit 2350.
In Figure 23, location information decoding unit 2310 can be by including in the bitstream relevant to location information
It indexes to restore quantity and the position of ISC.Arithmetic decoding can be used to be decoded to location information.Amplitude information decoding unit
2330 can index related with amplitude information in the bitstream carries out arithmetic decoding to including, and by selection TCQ and USQ it
One to carry out inverse quantization to decoded index.In order to improve the efficiency of arithmetic decoding, the number of non-zero location information and ISC can be used
Amount.Symbol decoding unit 2350 can be by restoring to being decoded including index related with symbolic information in the bitstream
The symbol of ISC.Arithmetic decoding can be used to be decoded to symbolic information.According to embodiment, umber of pulse needed for non-zero frequency band
Amount can be estimated and be used to be decoded location information, amplitude information or symbolic information.
Figure 24 is the block diagram for showing the configuration of frequency spectrum decoding device according to another exemplary embodiment.Shown in Figure 24
Equipment can be decoded to the frequency spectrum of Figure 19 and inverse quantization unit 1950 is corresponding, or can be included in another frequency domain decoding device,
Or it can be implemented separately.
Equipment shown in Figure 24 may include that ISC decoding unit 2150, quantized components recovery unit 2170 and inverse scaling are single
Member 2490.Compared with Figure 21, in addition to coding/decoding method selecting unit 2110 and zero decoding unit 2130 are omitted, ISC decoding unit
2150 using other than TCQ, and the operation of each element is identical.
Figure 25 is the block diagram for showing the configuration of frequency spectrum decoding device according to another exemplary embodiment.Shown in Figure 25
Equipment can be decoded to the frequency spectrum of Figure 19 and inverse quantization unit 1950 is corresponding, or can be included in another frequency domain decoding device,
Or it can be implemented separately.
Equipment shown in Figure 25 may include that coding/decoding method selecting unit 2510, ISC decoding unit 2550, quantized components are extensive
Multiple unit 2570 and inverse unit for scaling 2590.Compared with Figure 21, other than zero decoding unit 2130 is omitted, each element
It operates identical.
Figure 26 is the block diagram for showing the configuration of ISC information encoding apparatus according to another exemplary embodiment.
The equipment of Figure 26 may include probability calculation unit 2610 and lossless encoding unit 2630.
In Figure 26, probability calculation unit 2610 can be by using the quantity of ISC, the quantity of pulse and TCQ information, root
The probability value for amplitude coding is calculated according to equation 8 and equation 9.
Wherein,It indicates the quantity of remaining ISC after coding among the ISC sent for each frequency band,
It indicates the quantity of remaining pulse after coding among the pulse sent, M for each frequency bandsIt indicates in trellis state
Existing amplitude set under S.In addition, j indicates the amplitude of the pulse of present encoding.
Lossless encoding unit 2630 can by using the probability value of acquisition to TCQ amplitude information (that is, amplitude and path are believed
Breath) it is reversibly encoded.The quantity of the pulse of each amplitude passes throughValue andValue encodes.Here,Value indicates previous
The probability of the final pulse of amplitude,Indicate probability corresponding with other pulses in addition to the final pulse.Finally, logical
The index for crossing the probability value coding obtained is exported.
Figure 27 is the block diagram of the configuration of ISC information decoding apparatus according to another exemplary embodiment.
The equipment of Figure 27 may include probability calculation unit 2710 and lossless decoding unit 2730.
In Figure 27, probability calculation unit 2710 can be by using ISC information (quantity i and position) TCQ information, pulse
Quantity m and frequency band size m, which is calculated, is used for the decoded probability value of amplitude.For this purpose, can by using the number of pulses that had previously obtained and
Frequency band size obtains required bit information b.It thereafter, can quantity by using bit information b, ISC of acquisition, the position ISC
And TCQ information, it is calculated based on equation 8 and 9 and is used for the decoded probability value of amplitude.
Lossless decoding unit 2730 can be by using the probabilistic information and hair obtained in a manner of identical with encoding device
The index information sent carries out losslessly encoding to TCQ amplitude information (that is, amplitude and routing information).For this purpose, using probability value first
The arithmetic coding model for being directed to quantity information is obtained, and by using acquisition for carrying out arithmetic decoding to TCQ amplitude information
Model TCQ amplitude information is decoded.Particularly, the quantity of the pulse of each amplitude passes throughValue andValue is compiled
Code.Here,Value indicates the probability of the final pulse of previous amplitude.In addition,It indicates and in addition to the final pulse
The corresponding probability of other pulses.Finally, by the decoded TCQ amplitude information of probability value of acquisition (that is, amplitude information and road
Diameter information) it is exported.
Figure 28 is the block diagram of the multimedia device including coding module accoding to exemplary embodiment.
Referring to Figure 28, multimedia device 2800 may include communication unit 2810 and coding module 2830.In addition, multimedia fills
It sets 2800 and may also include storage unit 2850 and obtained to store as the result that the use according to audio bitstream is encoded
Audio bitstream.In addition, multimedia device 2800 may also include microphone 2870.That is, storage unit 2850 and wheat
Gram wind 2870 can selectively be included.Multimedia device 2800 may also include arbitrary decoder module (not shown), for example, with
Decoder module in the decoder module for executing general decoding function or accoding to exemplary embodiment.Decoder module 2830 can by with packet
At least one processor (not shown) that the other assemblies (not shown) in multimedia device 2800 becomes one is included to come in fact
It is existing.
Communication unit 2810 can receive at least one of the audio signal provided from outside or bit stream of coding, or
In the transmittable audio signal of reconstruction or the bit stream of coding obtained as the result encoded in coding module 2830
At least one.
Communication unit 2810 is configured as through wireless network (such as wireless Internet, wireless intranet, radiotelephony network
Network, WLAN (LAN), Wi-Fi, Wi-Fi direct (WFD), the third generation (3G), forth generation (4G), bluetooth, infrared data association
Meeting (IrDA), radio frequency identification (RFID), ultra wide band (UWB), Zigbee or near-field communication (NFC)) or cable network it is (such as wired
Telephone network or wired internet), transmit data to external multimedia apparatus or server and from external multimedia apparatus
Or server receives data.
Accoding to exemplary embodiment, coding module 1830 can be based on quantity, position, amplitude and symbol, for normalized
Frequency spectrum is selected ISC as unit of frequency band and encoded to the information of the important spectral component of the selection of each frequency band.It can be according to
The schemes different from the scheme encoded to quantity, position and symbol encode the amplitude of important spectral component.For example,
Quantization and arithmetic coding can be carried out come the amplitude to important spectral component by using one selected from USQ and TCQ, and can led to
Arithmetic coding is crossed to encode the quantity of important spectral component, position and symbol.Accoding to exemplary embodiment, coding module
2830 can execute scaling to normalized frequency spectrum based on the bit distribution for each frequency band, and select from by the frequency spectrum of scaling
ISC。
Storage unit 2850 can store the bit stream of the coding generated by coding module 2830.In addition, storage unit 2850
Various programs needed for operation multimedia device 2800 can be stored.
Microphone 2870 can provide the audio signal from user or outside to coding module 2830.
Figure 29 is the block diagram of the multimedia device including decoder module accoding to exemplary embodiment.
Referring to Figure 29, multimedia device 2900 may include communication unit 2910 and decoder module 2930.In addition, according to conduct
Decoded result and the use of the audio signal of reconstruction obtained, multimedia device 2900 may also include for storing the sound rebuild
The storage unit 2950 of frequency signal.In addition, multimedia device 2900 may also include loudspeaker 2970.That is, storage unit
2950 and loudspeaker 2970 can also selectively be included.Multimedia device 2900 may also include coding module (not shown), example
Such as, for executing the coding module or coding module accoding to exemplary embodiment of general encoding function.Decoder module 2930 can
By with include that at least one processor for becoming one of the other assemblies (not shown) in multimedia device 2900 (does not show
Lai Shixian out).
Communication unit 2910 can receive at least one of the audio signal provided from outside or bit stream of coding, or
The audio signal of the reconstruction obtained as the result being decoded in decoder module 2930 can be transmitted or as the result of coding
And at least one of the audio bitstream obtained.Communication unit 2910 can be implemented as 2810 essence of communication unit with Figure 28
It is similar.
Accoding to exemplary embodiment, decoder module 2930 can receive the bit stream provided by communication unit 2910, for
The frequency spectrum of coding obtains the information of important spectral component as unit of frequency band, and based on quantity, position, amplitude and symbol to acquisition
The information of important spectral component be decoded.It can be according to the side different from the scheme being decoded to quantity, position and symbol
Case is decoded the amplitude of important spectral component.For example, can be come by using one selected from USQ and TCQ to important frequency
The amplitude of spectral component carries out arithmetic decoding and inverse quantization, and can quantity, position and semiology analysis arithmetic to important spectral component
Decoding.
Storage unit 2950 can store the reconstructed audio signals generated by decoder module 2930.In addition, storage unit 2950
Various programs needed for operation multimedia device 2900 can be stored.
The audio signal of the reconstruction generated by decoder module 2930 can be output to outside by loudspeaker 2970.
Figure 30 is the block diagram of the multimedia device including coding module and decoder module accoding to exemplary embodiment.
Referring to Figure 30, multimedia device 3000 may include communication unit 3010, coding module 3020 and decoder module 3030.
In addition, multimedia device 3000 may also include the audio bitstream that storage unit 2850 is obtained to store the result as coding
Or as the audio signal of reconstruction for using the result being decoded according to audio bitstream and obtaining.In addition, multimedia device
3000 may also include microphone 3050 and/or loudspeaker 3060.Coding module 3020 and decoder module 3030 can by be included in
At least one processor (not shown) that other assemblies (not shown) in multimedia device 3000 becomes one is realized.
The group of multimedia device 2800 shown in the component and Figure 28 of the multimedia device 3000 as shown in Figure 30
The component of multimedia device 2900 shown in part or Figure 29 is corresponding, therefore omits the detailed description.
Each of multimedia device 2800,2900 and 3000 may include voice shown in Figure 28, Figure 29 and Figure 30
Communicate special-purpose terminal (such as phone or mobile phone), broadcast or music dedicated unit (such as TV or MP3 player) or voice
The hybrid terminal device of special-purpose terminal and broadcast or music dedicated unit is communicated, but not limited to this.In addition, multimedia device
2800, each of 2900 and 3000 can be used as the energy converter arranged between client computer, server or client computer and server.
When multimedia device 2800,2900 and 3000 is such as mobile phone, although not shown, but multimedia device
2800,2900 and 3000 may also include user input unit (such as keyboard), for showing through user interface or mobile phone
The processor of the display unit of the information of processing and the function for controlling mobile phone.In addition, mobile phone may also include
Has the function of the camera unit of image pickup function and for executing at least one component needed for mobile phone.
When multimedia device 2800,2900 and 3000 is such as TV, although not shown, but multimedia device 2800,
2900 and 3000 may also include user input unit (such as keyboard), the display unit for showing received broadcast message and
For controlling the functional processor of institute of TV.In addition, TV may also include at least one component of the function for executing TV.
Figure 31 is the process for showing the operation of the method encoded to frequency spectrum refinement structure accoding to exemplary embodiment
Figure.
Coding method may be selected in operation 3110 referring to Figure 31.For this purpose, can be used about each frequency band information and
Bit distribution information.Here, coding method may include quantization scheme.
In operation 3130, determine whether present band is frequency band (that is, zero-frequency band) that bit is assigned as zero, and if work as
Preceding frequency band is zero-frequency band, then the method proceeds to operation 3250, otherwise, if present band is non-zero-frequency band, the side
Method proceeds to operation 3270.
All samples in operation 3150, zero-frequency band can be encoded as zero.
In operation 3170, the frequency band as non-zero frequency band can be encoded based on the quantization scheme of selection.It, can according to embodiment
The quantity of the pulse of each frequency band is estimated by using frequency band length and bit distribution information, the quantity of non-zero position is determined and estimates
The required amount of bits of non-zero position is counted to determine the final amt of pulse.Next, can be based on the number of the pulse of each frequency band
The absolute value of amount and input signal determines original zoom factor, and can be based on original zoom factor, pass through scaling and arteries and veins
Redistribution processing is rushed to update zoom factor.Spectral coefficient is zoomed in and out using the zoom factor of final updated, and can
Suitable ISC is selected using the spectral coefficient by scaling, can be selected based on the bit distribution information for each frequency band
The spectrum component that will be quantized.Next, can by USC and TCQ scheme for combining come the amplitude of the ISC to collection carry out quantization and
Arithmetic coding.Here, in order to improve the efficiency of arithmetic coding, the quantity of non-zero position and the quantity of ISC can be used.USC and TCQ
Scheme for combining may include the first scheme for combining and the second scheme for combining according to bandwidth.First scheme for combining makes it possible to by making
With selecting quantizer to the secondary bit allocation process of the spare bits from previous band, and it can be used for NB and WB,
Two scheme for combining be for being determined as being used for LSB using the frequency band of USQ, TCQ and USQ is used for the schemes of other bits, and
It can be used for SWB and FB.The symbolic information of the ISC of selection can be with the equal probabilities for plus sign and minus symbol and by arithmetic
Coding.
After operation 3170, the operation for restoring the operation of the component of quantization and carrying out inverse scaling to frequency band can be further
Including.In order to restore the component of actual quantization, position, symbol and amplitude information can be added to the component of quantization.Zero can be divided
It is fitted on zero position.It can be used with the same zoom factor for scaling and extract inverse zoom factor, and the actual quantization restored
Component can be by inverse scaling.Can have by the signal of inverse scaling identical with the rank of normalized frequency spectrum (that is, input signal)
Rank.
The operation of each component of above-mentioned encoding device can according to circumstances be further added to the operation of Figure 31.
Figure 32 is the process for showing the operation of the method being decoded to frequency spectrum refinement structure accoding to exemplary embodiment
Figure.Position, quantity, symbol can be based on according to the operation of Figure 32 in order to which the refinement structure to normalized frequency spectrum carries out inverse quantization
It is decoded with information of the amplitude to ISC and the ISC about selection for each frequency band.Here, arithmetic decoding can be passed through
And USQ and TCQ scheme for combining is decoded amplitude information, and by arithmetic decoding to position, quantity and symbolic information
It is decoded.
In detail, coding/decoding method may be selected in operation 3210 referring to Figure 32.For this purpose, can be used about each frequency band
Information and bit distribution information.Here, coding/decoding method may include inverse quantization scheme.Can by be applied to above-mentioned encoding device
The identical processing of quantization scheme selection is to select inverse quantization scheme.
In operation 3230, determine whether present band is frequency band (that is, zero-frequency band) that bit is assigned as zero, and if work as
Preceding frequency band is zero-frequency band, then the method proceeds to operation 3250, otherwise, if present band is non-zero-frequency band, the side
Method proceeds to operation 3270.
All samples in operation 3250, zero-frequency band can be decoded as zero.
In operation 3270, the frequency band as non-zero frequency band can be decoded based on the inverse quantization scheme of selection.According to embodiment,
It can be by using the quantity of frequency band length and bit distribution information estimation or the pulse for determining each frequency band.This can by with application
It is executed in the identical processing of the scaling of above-mentioned encoding device.Next, the location information of ISC can be restored (that is, the quantity of ISC
And position).This can be similar with above-mentioned encoding device handled, and identical probability value can be used for suitably solving
Code.Next, can be decoded by arithmetic decoding come the amplitude of the ISC to collection and by USC and TCQ scheme for combining come anti-
Quantization.Here, the quantity of non-zero position and the quantity of ISC can be used for arithmetic decoding.USC and TCQ scheme for combining may include root
According to the first scheme for combining and the second scheme for combining of bandwidth.First scheme for combining makes it possible to by addition using to from previous
The secondary bit allocation process of the spare bits of frequency band selects quantizer, and can be used for NB and WB, and the second scheme for combining is
For being determined as being used for LSB using the frequency band of USQ, TCQ and USQ is used for the scheme of other bits, and can be used for SWB and
FB.The symbolic information of the ISC of selection can be arithmetic decoded with the equal probabilities for plus sign and minus symbol.
After operation 3270, the operation for restoring the operation of the component of quantization and carrying out inverse scaling to frequency band can be further
Including.In order to restore the component of actual quantization, position, symbol and amplitude information can be added to the component of quantization.Without general
The frequency band of the data sent can use zero padding.Next, can estimate the quantity of the pulse in non-zero frequency band, and can be based on estimating
The quantity of the pulse of meter is decoded the location information of quantity and position including ISC.Can by losslessly encoding and USC and
TCQ scheme for combining is decoded amplitude information.For non-zero amplitude-value, it can finally restore symbol and quantized components.For extensive
The norm information sent can be used to execute inverse scaling for multiple actual quantization component.
The operation of each component of above-mentioned decoding device can according to circumstances be further added to the operation of Figure 32.
The above exemplary embodiments can be written as computer executable program, and can be implemented in by using nonvolatile
Property computer readable recording medium execute described program general purpose digital computer in.In addition, the number that can be used in embodiment
It can be recorded in non-transitory computer readable recording medium in various ways according to structure, program instruction or data file.
Non-transitory computer readable recording medium is any data storage dress for storing and can reading data by computer system thereafter
It sets.The example of non-transitory computer readable recording medium includes magnetic medium (such as hard disk, floppy disk and tape), optical recording Jie
It matter (such as CD ROM and DVD), magnet-optical medium (such as CD) and is specially configured to store and execute the hardware of program instruction
Device (such as ROM, RAM and flash memory).In addition, non-transitory computer readable recording medium can be for sending designated program
The transmission medium of the signal of instruction, data structure etc..The example of program instruction can not only include the machine language by compiler-creating
It says code, may also include the higher-level language code that can use the execution such as interpreter by computer.
Although being particularly shown and having described exemplary embodiment, those skilled in the art should be managed
Solution, in the case where not departing from the spirit and scope of present inventive concept defined by claim, it can be carried out form and
Various changes in details.It should be understood that exemplary embodiment described herein should be considered only as descriptive meaning, rather than
For the purpose of limitation.The description of features or aspect in each exemplary embodiment should be generally viewed as can be applied to other
Other similar features or aspect in exemplary embodiment.
Claims (18)
1. a kind of spectrum coding method, comprising:
Based on the average number of bits amount for the spectrum component being assigned in frequency band come in uniform scalar quantization USQ and grid coding amount
Change the coding method that selection among TCQ is directed to the frequency band；
The spectrum component in the frequency band is zoomed in and out based on the bit for being assigned to the frequency band；
The important spectral component in the frequency band is selected based on the spectrum component by scaling in the frequency band；And
The information about the important spectral component in the frequency band is encoded by using the coding method of selection.
2. spectrum coding method as described in claim 1, wherein the step of selecting the important spectral component in frequency band packet
It includes: selecting the important spectral component by analyzing amount of zoom from the spectrum component by scaling.
3. spectrum coding method as described in claim 1, wherein the information about the important spectral component includes: described
Quantity, position, amplitude and the symbol of the important spectral component in frequency band.
4. spectrum coding method as claimed in claim 3, wherein the amplitude of the important spectral component according to it is described important
The quantity of spectrum component, the position encoding scheme different with the encoding scheme of symbol are encoded.
5. spectrum coding method as claimed in claim 3, wherein encoded to the information about the important spectral component
The step of include: by using USQ and TCQ among selected coding method come to the amplitude of the important spectral component into
Row coding.
6. spectrum coding method as claimed in claim 3, wherein encoded to the information about the important spectral component
The step of include: to be encoded by using arithmetic coding come quantity, position and the symbol to the important spectral component.
7. spectrum coding method as described in claim 1, wherein encoded to the information about the important spectral component
The step of include: according to bandwidth, using a scheme in the first combined coding scheme and the second combined coding scheme to about
The information of the important spectral component is encoded.
8. spectrum coding method as described in claim 1, wherein encoded to the information about the important spectral component
The step of include:
Come by using TCQ to the important frequency according to bandwidth when the coding method for the frequency band of selection is USQ
The least significant bit LSB of the amplitude of spectral component is encoded, and uses USQ to other ratios of the amplitude of the important spectral component
Spy encodes.
9. spectrum coding method as claimed in claim 8, wherein the bandwidth is SWB or FB.
10. a kind of spectrum coding equipment, comprising:
At least one processor, is configured as:
Based on the average number of bits amount for the spectrum component being assigned in frequency band come in uniform scalar quantization USQ and grid coding amount
Change the coding method that selection among TCQ is directed to the frequency band；
The spectrum component in the frequency band is zoomed in and out based on the bit for being assigned to the frequency band；
The important spectral component in the frequency band is selected based on the spectrum component by scaling in the frequency band；And
The information about the important spectral component in the frequency band is encoded by using the coding method of selection.
11. spectrum coding equipment as claimed in claim 10, wherein at least one described processor is configured as: by from
The spectrum component by scaling analyzes amount of zoom to select the important spectral component.
12. spectrum coding equipment as claimed in claim 10, wherein the information about the important spectral component includes: institute
State quantity, position, amplitude and the symbol of the important spectral component in frequency band.
13. spectrum coding equipment as claimed in claim 12, wherein the amplitude of the important spectral component according to it is described heavy
The encoding scheme for wanting the quantity of spectrum component, position different with the encoding scheme of symbol is encoded.
14. spectrum coding equipment as claimed in claim 12, wherein at least one described processor is configured as: by making
The amplitude of the important spectral component is encoded with the selected coding method among USQ and TCQ.
15. spectrum coding equipment as claimed in claim 12, wherein at least one described processor is configured as: by making
It is encoded with arithmetic coding come quantity, position and the symbol to the important spectral component.
16. spectrum coding equipment as claimed in claim 10, wherein at least one described processor is configured as: according to band
Width, using a scheme in the first combined coding scheme and the second combined coding scheme to about the important spectral component
Information is encoded.
17. spectrum coding equipment as claimed in claim 10, wherein at least one described processor is configured as:
Come by using TCQ to the important frequency according to bandwidth when the coding method for the frequency band of selection is USQ
The least significant bit LSB of the amplitude of spectral component is encoded, and uses USQ to other ratios of the amplitude of the important spectral component
Spy encodes.
18. spectrum coding equipment as claimed in claim 17, wherein the bandwidth is SWB or FB.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910495957.0A CN110176241B (en) | 2014-02-17 | 2015-02-17 | Signal encoding method and apparatus, and signal decoding method and apparatus |
Applications Claiming Priority (5)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| US201461940798P | 2014-02-17 | 2014-02-17 | |
| US61/940,798 | 2014-02-17 | ||
| CN201910495957.0A CN110176241B (en) | 2014-02-17 | 2015-02-17 | Signal encoding method and apparatus, and signal decoding method and apparatus |
| PCT/KR2015/001668 WO2015122752A1 (en) | 2014-02-17 | 2015-02-17 | Signal encoding method and apparatus, and signal decoding method and apparatus |
| CN201580020096.0A CN106233112B (en) | 2014-02-17 | 2015-02-17 | Coding method and equipment and signal decoding method and equipment |
Related Parent Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201580020096.0A Division CN106233112B (en) | 2014-02-17 | 2015-02-17 | Coding method and equipment and signal decoding method and equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110176241A true CN110176241A (en) | 2019-08-27 |
| CN110176241B CN110176241B (en) | 2023-10-31 |
Family
ID=57257234
Family Applications (2)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201580020096.0A Active CN106233112B (en) | 2014-02-17 | 2015-02-17 | Coding method and equipment and signal decoding method and equipment |
| CN201910495957.0A Active CN110176241B (en) | 2014-02-17 | 2015-02-17 | Signal encoding method and apparatus, and signal decoding method and apparatus |
Family Applications Before (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201580020096.0A Active CN106233112B (en) | 2014-02-17 | 2015-02-17 | Coding method and equipment and signal decoding method and equipment |
Country Status (4)
| Country | Link |
|---|---|
| EP (1) | EP3109611A4 (en) |
| JP (1) | JP6633547B2 (en) |
| KR (3) | KR20240008413A (en) |
| CN (2) | CN106233112B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2021114847A1 (en) * | 2019-12-10 | 2021-06-17 | 腾讯科技（深圳）有限公司 | Internet calling method and apparatus, computer device, and storage medium |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110867190B (en) * | 2013-09-16 | 2023-10-13 | 三星电子株式会社 | Signal encoding method and device and signal decoding method and device |
| CN106233112B (en) * | 2014-02-17 | 2019-06-28 | 三星电子株式会社 | Coding method and equipment and signal decoding method and equipment |
| JP6763849B2 (en) | 2014-07-28 | 2020-09-30 | サムスン エレクトロニクス カンパニー リミテッド | Spectral coding method |
| WO2019091576A1 (en) * | 2017-11-10 | 2019-05-16 | Fraunhofer-Gesellschaft zur Förderung der angewandten Forschung e.V. | Audio encoders, audio decoders, methods and computer programs adapting an encoding and decoding of least significant bits |
| JP7173134B2 (en) | 2018-04-13 | 2022-11-16 | 日本電信電話株式会社 | Encoding device, decoding device, encoding method, decoding method, program, and recording medium |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP0628195A1 (en) * | 1992-01-17 | 1994-12-14 | The Massachusetts Institute Of Technology | Method and apparatus for encoding, decoding and compression of audio-type data |
| CN101836251A (en) * | 2007-10-22 | 2010-09-15 | 高通股份有限公司 | Scalable speech and audio encoding using combinatorial encoding of MDCT spectrum |
| WO2013048171A2 (en) * | 2011-09-28 | 2013-04-04 | 엘지전자 주식회사 | Voice signal encoding method, voice signal decoding method, and apparatus using same |
| WO2013062392A1 (en) * | 2011-10-27 | 2013-05-02 | 엘지전자 주식회사 | Method for encoding voice signal, method for decoding voice signal, and apparatus using same |
| US20130290003A1 (en) * | 2012-03-21 | 2013-10-31 | Samsung Electronics Co., Ltd. | Method and apparatus for encoding and decoding high frequency for bandwidth extension |
Family Cites Families (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP3685823B2 (en) * | 1993-09-28 | 2005-08-24 | ソニー株式会社 | Signal encoding method and apparatus, and signal decoding method and apparatus |
| US6717990B1 (en) * | 2000-01-05 | 2004-04-06 | General Dynamics Decision Systems, Inc. | Communication system and method for multi-rate, channel-optimized trellis-coded quantization |
| US6847684B1 (en) * | 2000-06-01 | 2005-01-25 | Hewlett-Packard Development Company, L.P. | Zero-block encoding |
| KR100871999B1 (en) * | 2001-05-08 | 2008-12-05 | 코닌클리케 필립스 일렉트로닉스 엔.브이. | Audio coding |
| RU2387024C2 (en) * | 2004-11-05 | 2010-04-20 | Панасоник Корпорэйшн | Coder, decoder, coding method and decoding method |
| KR100851970B1 (en) * | 2005-07-15 | 2008-08-12 | 삼성전자주식회사 | Method and apparatus for extracting ISCImportant Spectral Component of audio signal, and method and appartus for encoding/decoding audio signal with low bitrate using it |
| JP4640020B2 (en) * | 2005-07-29 | 2011-03-02 | ソニー株式会社 | Speech coding apparatus and method, and speech decoding apparatus and method |
| ATE500588T1 (en) * | 2008-01-04 | 2011-03-15 | Dolby Sweden Ab | AUDIO ENCODERS AND DECODERS |
| KR102078865B1 (en) * | 2011-06-30 | 2020-02-19 | 삼성전자주식회사 | Apparatus and method for generating a bandwidth extended signal |
| CN106233112B (en) * | 2014-02-17 | 2019-06-28 | 三星电子株式会社 | Coding method and equipment and signal decoding method and equipment |
| JP6763849B2 (en) * | 2014-07-28 | 2020-09-30 | サムスン エレクトロニクス カンパニー リミテッド | Spectral coding method |
-
2015
- 2015-02-17 CN CN201580020096.0A patent/CN106233112B/en active Active
- 2015-02-17 KR KR1020247000605A patent/KR20240008413A/en active Application Filing
- 2015-02-17 CN CN201910495957.0A patent/CN110176241B/en active Active
- 2015-02-17 EP EP15749031.9A patent/EP3109611A4/en not_active Ceased
- 2015-02-17 JP JP2016569544A patent/JP6633547B2/en active Active
- 2015-02-17 KR KR1020227012038A patent/KR102625143B1/en active IP Right Grant
- 2015-02-17 KR KR1020167022489A patent/KR102386738B1/en active IP Right Grant
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP0628195A1 (en) * | 1992-01-17 | 1994-12-14 | The Massachusetts Institute Of Technology | Method and apparatus for encoding, decoding and compression of audio-type data |
| CN101836251A (en) * | 2007-10-22 | 2010-09-15 | 高通股份有限公司 | Scalable speech and audio encoding using combinatorial encoding of MDCT spectrum |
| WO2013048171A2 (en) * | 2011-09-28 | 2013-04-04 | 엘지전자 주식회사 | Voice signal encoding method, voice signal decoding method, and apparatus using same |
| WO2013062392A1 (en) * | 2011-10-27 | 2013-05-02 | 엘지전자 주식회사 | Method for encoding voice signal, method for decoding voice signal, and apparatus using same |
| US20130290003A1 (en) * | 2012-03-21 | 2013-10-31 | Samsung Electronics Co., Ltd. | Method and apparatus for encoding and decoding high frequency for bandwidth extension |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2021114847A1 (en) * | 2019-12-10 | 2021-06-17 | 腾讯科技（深圳）有限公司 | Internet calling method and apparatus, computer device, and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| EP3109611A1 (en) | 2016-12-28 |
| CN106233112A (en) | 2016-12-14 |
| KR20220051028A (en) | 2022-04-25 |
| KR102386738B1 (en) | 2022-04-14 |
| KR20160122160A (en) | 2016-10-21 |
| JP6633547B2 (en) | 2020-01-22 |
| KR20240008413A (en) | 2024-01-18 |
| EP3109611A4 (en) | 2017-08-30 |
| CN110176241B (en) | 2023-10-31 |
| KR102625143B1 (en) | 2024-01-15 |
| JP2017506771A (en) | 2017-03-09 |
| CN106233112B (en) | 2019-06-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| KR102151749B1 (en) | Frame error concealment method and apparatus, and audio decoding method and apparatus | |
| CN106233112B (en) | Coding method and equipment and signal decoding method and equipment | |
| TWI585748B (en) | Frame error concealment method and audio decoding method | |
| CN104011793B (en) | Hiding frames error method and apparatus and audio-frequency decoding method and equipment | |
| US10194151B2 (en) | Signal encoding method and apparatus and signal decoding method and apparatus | |
| US11616954B2 (en) | Signal encoding method and apparatus and signal decoding method and apparatus | |
| US10811019B2 (en) | Signal encoding method and device and signal decoding method and device | |
| KR102386737B1 (en) | Signal encoding method and apparatus and signal decoding method and apparatus | |
| US10902860B2 (en) | Signal encoding method and apparatus, and signal decoding method and apparatus |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |