CN1890712A - Audio signal coding - Google Patents
Audio signal coding Download PDFInfo
- Publication number
- CN1890712A CN1890712A CNA200480035931XA CN200480035931A CN1890712A CN 1890712 A CN1890712 A CN 1890712A CN A200480035931X A CNA200480035931X A CN A200480035931XA CN 200480035931 A CN200480035931 A CN 200480035931A CN 1890712 A CN1890712 A CN 1890712A
- Authority
- CN
- China
- Prior art keywords
- data
- frequency
- window
- granularity
- signal
- 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.)
- Pending
Links
- 230000005236 sound signal Effects 0.000 title description 6
- 230000006870 function Effects 0.000 claims abstract description 61
- 238000005070 sampling Methods 0.000 claims description 76
- 230000008707 rearrangement Effects 0.000 claims description 16
- 230000009466 transformation Effects 0.000 claims description 13
- 238000000034 method Methods 0.000 claims description 12
- 238000006243 chemical reaction Methods 0.000 claims description 9
- 238000000844 transformation Methods 0.000 claims description 4
- 230000003247 decreasing effect Effects 0.000 claims 1
- 230000015572 biosynthetic process Effects 0.000 abstract 1
- 239000008187 granular material Substances 0.000 abstract 1
- 238000003786 synthesis reaction Methods 0.000 abstract 1
- 235000019580 granularity Nutrition 0.000 description 46
- 239000002131 composite material Substances 0.000 description 10
- 238000010586 diagram Methods 0.000 description 6
- 230000000694 effects Effects 0.000 description 4
- 238000001228 spectrum Methods 0.000 description 4
- 230000008569 process Effects 0.000 description 3
- 238000011002 quantification Methods 0.000 description 3
- 238000007792 addition Methods 0.000 description 2
- 238000013139 quantization Methods 0.000 description 2
- 230000001105 regulatory effect Effects 0.000 description 2
- 230000004044 response Effects 0.000 description 2
- OVOUKWFJRHALDD-UHFFFAOYSA-N 2-[2-(2-acetyloxyethoxy)ethoxy]ethyl acetate Chemical compound CC(=O)OCCOCCOCCOC(C)=O OVOUKWFJRHALDD-UHFFFAOYSA-N 0.000 description 1
- 241000255777 Lepidoptera Species 0.000 description 1
- 230000002411 adverse Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 150000001875 compounds Chemical class 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 230000002349 favourable effect Effects 0.000 description 1
- 230000008520 organization Effects 0.000 description 1
- 230000008447 perception Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 230000003595 spectral effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000001052 transient effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; 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
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; 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
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/14—Fourier, Walsh or analogous domain transformations, e.g. Laplace, Hilbert, Karhunen-Loeve, transforms
- G06F17/147—Discrete orthonormal transforms, e.g. discrete cosine transform, discrete sine transform, and variations therefrom, e.g. modified discrete cosine transform, integer transforms approximating the discrete cosine transform
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; 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
- G10L19/022—Blocking, i.e. grouping of samples in time; Choice of analysis windows; Overlap factoring
-
- H—ELECTRICITY
- H03—ELECTRONIC CIRCUITRY
- H03M—CODING; DECODING; CODE CONVERSION IN GENERAL
- H03M7/00—Conversion of a code where information is represented by a given sequence or number of digits to a code where the same, similar or subset of information is represented by a different sequence or number of digits
- H03M7/30—Compression; Expansion; Suppression of unnecessary data, e.g. redundancy reduction
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; 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
- G10L19/0204—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 subband decomposition
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Signal Processing (AREA)
- Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- Human Computer Interaction (AREA)
- Acoustics & Sound (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Mathematical Physics (AREA)
- Spectroscopy & Molecular Physics (AREA)
- Mathematical Analysis (AREA)
- Computational Mathematics (AREA)
- Mathematical Optimization (AREA)
- Pure & Applied Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Discrete Mathematics (AREA)
- Algebra (AREA)
- Databases & Information Systems (AREA)
- Software Systems (AREA)
- General Engineering & Computer Science (AREA)
- Compression, Expansion, Code Conversion, And Decoders (AREA)
Abstract
One aspect of the invention provides a decoder for MPEG-1 layer III data signals. In the preferred embodiment, the decoder performs a single inverse MDCT on all 576 frequency lines of a respective granule for type 0, 1 and 3 MP3 window functions, and performs three inverse MDCTs on three sets of 192 frequency lines for type 2 window functions. It is found that the use of ''long'' inverse MDCTs provides an adequate approximation of a hybrid filterbank which comprises a plurality of ''short'' inverse MDCTs and a synthesis filterbank. As a result, an output signal may be constructed without the need for a filterbank. Another aspect of the invention provides an encoder for generating MPEG-1 layer III type data signals in which ''long'' MDCTs are used to replace the hybrid filterbank. As a result, MPEG-1 layer III type data signals may be generated without the need for a filterbank.
Description
Technical field
The present invention relates to the Code And Decode of data-signal.The present invention especially but non-exclusively relate to the equipment that is used for Code And Decode MPEG-1 layer III data-signal.
Background technology
MPEG-1 layer III (being commonly referred to MP3) is widely used audio codec.At ISO/IEC JTC1/SC29/WG11 MPEG, " Information Technology-Coding of Moving Pictures and Associated Audio for DigitalStorage Media at up to about 1.5Mbit/s " (part 3: audio frequency of IS11172-3, the industry standard of MP3 has been described MPEG-1,1992).This standard can from International Organization for Standardization (
Www.i so.ch) the middle acquisition, thereby be hereby incorporated by.
Fig. 1 represents the simplified block diagram of typical conventional MPEG-1 layer III scrambler 10, only shows that those help to understand element of the present invention in the scrambler 10.Scrambler 10 is configured to receive the PCM input signal, comprises the audio sample of 1152 of a string or frames.Input signal is provided for (heterogeneous) analysis filterbank (filterbank) 12, and it is filtered into 32 evenly spaced, as to overlap frequency bands with input signal, and to produce the sub-band signal component of 32 down-samplings, each comprises the sampling of 36 sub-frequency bands.
With regard to each sub-band signal component, windowization (just) MDCT (improved discrete cosine transform) is carried out by MDCT unit 14.Four window types are used for regulating variable time slice.(standard) steady component for signal can adopt so-called normal window, and for the astable part of signal, can adopt a series of so-called short windows.For switching to short window from normal window and when short window switches to normal window, preventing to be interrupted, defined the window of two instantaneous types, promptly so-called beginning, stop window.For normal, begin or stop window, go up execution MDCT in 36 inputs (i.e. 36 sub-frequency bands sampling), the MDCT coefficient of 18 outputs of generation is commonly referred to frequency line (frequency line).For short window, (that is, in three group of 12 sub-frequency bands sampling) carries out MDCT in three groups of 12 inputs, and produces three groups 6 output MDCT coefficients, or frequency line.One group 576 MDCT coefficient is called a granularity.With regard to the common MP3 frame that comprises 1152 input samples,, produce two granularities as the result of encoding process overlapping characteristic.For per 576 input samples, add up to produce 18 * 32=576 MDCT coefficient or frequency line.
Normal, begin or stop under the situation of window, the MDCT frequency line is provided for anti-aliased butterfly (anti-aliasing butterflie) 16, to reduce the influence of the aliasing that the wave filter that overlaps by downsampling filter group 12 parts causes.At last, quantize and suitable quantification and the coding of coding unit 18 execution frequency lines, so that produce the output signal of specified bit stream format.Quantize and be coded in Bit Allocation in Discrete unit 20 controls of carrying out bit distribution algorithm to carry out down, regulated and control by psychoacoustic model.
Fig. 2 represents the simplified block diagram of conventional MPEG-1 layer III demoder 30, only shows those and helps to understand element of the present invention.Demoder 30 is arranged to receive the input signal of specified bit stream format.Decoding and going quantifying unit 32 to carry out the decoding of bit streams and go quantizes, to produce frequency line or MDCT coefficient.For every group of 576 MDCT frequency lines that produced by scrambler 10, reproduce 576 frequency lines separately.
Frequency line be provided for the rearrangement unit 34, its under the situation of short type window, the frequency line of in each granularity, resequencing.Normal, begin or stop under the situation of window, frequency line is provided aliasing butterfly 36, anti-operation is carried out in its anti-aliased operation that anti-aliased butterfly 16 is carried out.
IMDCT (improved inverse discrete cosine transform) is carried out in IMDCT unit 38 on frequency line, to produce 32 sub-frequency bands component of signals, each component comprises the sampling of 36 sub-frequency bands.Corresponding to for normal, as to begin or the stop window MDCT frequency line, IMDCT unit 38 is thought and has been imported 18 frequency lines, and produces the sampling of 36 sub-frequency bands for those.Corresponding to for the frequency line of short window MDCT, IMDCT unit 38 is thought and has been imported 3 groups of 6 frequency lines, and produces 3 group of 12 sub-frequency bands sampling for those.
The overlapping and add operation mutually of windowization (windowed) operation and standard are performed in the sub-band sampling by windowization and overlap-add unit 40.The information of the window of relevant which kind of type of use is carried in the relevant supplementary of bit stream (side information).At last, the sub-band sampling is provided for (heterogeneous) composite filter group 42, comprise the output signal that PCM samples with generation, and composite filter group 42 also comprises 32 times of up-samplings.
Bank of filters 12,42 comprises the prototype low-pass filter, its by cosine modulation to form high frequency band.The tandem compound of sub-band filter group and MDCT unit is called hybrid filter-bank, because its part is formed, partly is made up of conversion by bank of filters.In scrambler 10, analysis filterbank 12 and MDCT unit 14 comprise the hybrid analysis bank of filters jointly, and in demoder 30, IMDCT unit 38 and composite filter group 42 comprise jointly mixes the composite filter group.In view of the calculating that utilizes hybrid filter-bank to bring, then the complicacy of implementing, for MP3, this is the shortcoming of generally acknowledging.Therefore be desirable to provide MP3 scrambler and/or the demoder that needs less calculating.
Summary of the invention
Therefore, a first aspect of the present invention provides a kind of demoder that is used for data-signal, described data-signal by data-signal is provided to the sub-band filter group and by sub-band signal in each generation on carry out separately that the frequency direct transform is encoded, described demoder comprises: be used to decode and remove the data-signal that quantizes the to receive device with the granularity that produces a plurality of frequency lines; Be used on each granularity, carrying out one or more frequency inverse transformations to produce the device of a plurality of data samplings; And be used for described data sampling is applied the window function of one or more types to produce the device of a plurality of window data samplings; Wherein, with regard at least the first type window function, described frequency inverter is provided on all frequency lines of granularity separately and carries out the single frequency inverse transformation, and wherein said demoder also comprises the device that is used for constructing from described window data sampling output signal.
A second aspect of the present invention provides a kind of method of decoding data signal, described data-signal is by providing data-signal and being encoded by the frequency direct transform of carrying out on the sub-band signal of each generation separately to the sub-band filter group, described method comprises: decode and the data-signal that goes to quantize to receive to produce the granularity of a plurality of frequency lines; On each granularity, carry out one or more frequency inverse transformations to produce a plurality of data samplings; The window function that described data sampling is applied one or more types is to produce a plurality of window data samplings; And from described window data sampling, construct output signal; Wherein, with regard at least the first type window function, on all frequency lines in the granularity separately, carry out the single frequency inverse transformation.
Each of first and second aspects of the present invention allows to generate output signal and need not bank of filters.In a preferred embodiment, encoded data signals comprises the data-signal of MPEG-1 layer III, and frequency just comprise improved discrete cosine transform (MDCT) and improved inverse discrete cosine transform (IMDCT) respectively with inverse transformation.
In a preferred embodiment, the frequency direct transform is opposite comprises improved discrete cosine transform (MDCT), and encoded data signals comprises the data-signal of MPEG-1 layer III.
A third aspect of the present invention provides a kind of scrambler that is used to comprise the input signal of a plurality of data samplings, and described scrambler comprises: be used for described data sampling is applied the window function of one or more types to produce the device of a plurality of window data samplings; Be used on the window data sampling carrying out the device of one or more improved discrete cosine transforms (MDCT) with the granularity that produces a plurality of frequency lines; And be used to encode and quantize the device of each granularity with the data-signal that produces MPEG-1 layer III type; Wherein, with regard at least the first type window function, described MDCT device is provided in about the single MDCT of execution on all window data samplings of the data-signal of the described reception of its generation granularity separately.
A fourth aspect of the present invention provides a kind of the input signal that comprises a plurality of data samplings is carried out Methods for Coding, and described method comprises: the window function that described data sampling is applied one or more types is to produce a plurality of window data samplings; On the window data sampling, carry out one or more improved discrete cosine transforms (MDCT) to produce the granularity of a plurality of frequency lines; Coding also quantizes each granularity to produce the data-signal of MPEG-1 layer III type; Wherein, with regard at least the first type window function, on all window data samplings of the data-signal of the described reception that produces granularity separately about it, carry out single MDCT.
Third and fourth aspect of the present invention allows to generate the data-signal of MPEG-1 layer III type, and need not to use bank of filters.
A fifth aspect of the present invention provides a kind of system or codec that is used for the Code And Decode data-signal, and described system comprises the scrambler of third aspect present invention and the demoder of first aspect present invention.
In the dependent claims, narrated the preferred feature of each aspect of the present invention.
To those skilled in the art, following to the description of specific embodiment and with reference to accompanying drawing by reading, favourable aspect of the present invention will become more obvious.
Description of drawings
Now by example and embodiment with reference to the accompanying drawings to describe the present invention, wherein:
Fig. 1 is the block diagram of conventional MPEG-1 layer III scrambler;
Fig. 2 is the block diagram of conventional MPEG-1 layer III demoder;
Fig. 3 is the diagrammatic representation from the MDCT coefficient of the MPEG-1 layer III hybrid analysis bank of filters of delta pulse;
Fig. 4 is that Fig. 3 is in the diagrammatic representation of (multiply by-1) MDCT coefficient afterwards that the odd number sub-band is negated;
Order after the MDCT coefficient that Fig. 5 shows short window is resequenced in conventional MPEG-1 layer III demoder;
Fig. 6 is the block diagram of the MPEG-1 layer III signal of demoder, and described demoder has embodied one aspect of the present invention;
Order after the MDCT coefficient that Fig. 7 shows short window is resequenced in the decoding device of Fig. 6; And
Fig. 8 is the scrambler that is used to produce MPEG-1 layer III type signal that embodies third aspect present invention.
Embodiment
In conventional MPEG-1 layer III (MP3) system, typical Frame comprises two granularities, and each granularity is made up of 576 frequency lines or MDCT coefficient.As mentioned above, according to the coding of conventional MP3, these 576 frequency lines comprise 18 frequency lines of one group separately of each sub-band in 32 sub-frequency bands.When using the window of short type, 18 frequency lines of every group are made up of 6 every group 3 class frequency lines.In the scrambler 10 of Fig. 1, carry out conversion by hybrid filter-bank 12,14.Depend on required window type, one or more MDCT at each sub-band are carried out in MDCT unit 14.The MDCT that carries out by MDCT unit 14 can think and comprise " short " MDCT, wherein, each only (relatively little) of frame data separately part go up and carry out each MDCT.For window normal, that begin or stop type, on 36 input samples of sub-band, carry out single MDCT to produce 18 frequency lines.For the window of short type, carry out three MDCT conversion, each is carrying out to produce one group of 6 frequency line separately on one group of 12 input sample at sub-band separately all.Therefore, in the demoder 30 of conventional MP3, the contrary MDCT that carries out by IMDCT unit 38 can think and comprise " short " contrary MDCT because each contrary MDCT only produce about Frame through decoding and go execution on the part separately of the frequency line that quantizes.For window normal, that begin or stop type, on 18 frequency lines of sub-band, carry out single contrary MDCT to produce 36 time-domain samplings.For the window of short type, carry out three contrary MDCT conversion, each is all carried out on one group of 6 frequency line separately of sub-band to produce one group of 12 time-domain sampling separately.
On the contrary, in the embodiment of one aspect of the invention, the method of decoding MP3 data is provided, wherein, on frequency line through decoding and going to quantize or MDCT coefficient, carry out one or more " length " contrary MDCT, and about frequency line or the MDCT coefficient of whole data granularity generation through decoding and going to quantize.Granularity for 576 frequency lines or MDCT coefficient, normal when needs, when beginning or stopping the window of type, on all 576 frequency lines, carry out single " length " contrary MDCT to produce 1152 time-domain samplings, and for the window of short type, on one group of 192 frequency line separately, carry out three " length " contrary MDCT, to produce one group of 384 time-domain sampling separately.No matter but be which kind of situation, on all frequency lines of as a whole granularity, carry out one or more contrary MDCT, rather than on the separately frequency line relevant, carry out with sub-band separately.It is found that some pre-service by frequency line and by suitable windowization and overlap-add operation, the output of " long " contrary MDCT can be used to provide on the perception of PCM output signal of expectation approximate value quite accurately.Can be during encoding process principle like the application class, thereby get the needs that offset scrambler median filter group.It below is more detailed description.
In forming process of the present invention, draw following observation :-desirable bank of filters comprises the rectangular passband that does not overlap.If used bank of filters is desirable among the MP3, then can very accurately be similar to hybrid filter-bank by above-mentioned single " length " MDCT.The combination of bank of filters and anti-aliased butterfly has provided relative good being similar to of ideal filter group.Therefore, hybrid filter-bank can be substituted by single " length " MDCT in conjunction with anti-aliased butterfly.By these observation, infer total Code And Decode processing, more specifically, the operation of hybrid filter-bank separately can be similar to by the cosine modulation conversion.Especially suppose that total Code And Decode processing can be similar to by MDCT.
If this supposition is correct, if promptly hybrid filter-bank can be approximated to be MDCT, then the response to the delta pulse should comprise cosine waveform.The analytical expression that is used for (just) MDCT is as follows:
Wherein, n is the time index (time index) that is used for conventional MP3 scrambler, the index of its indication sub-band sampling; N is the length or the size of conversion; K is a frequency index; X[n] be time-domain signal, it comprises the sub-band time-domain signal of being made up of the sub-band sampling in the MP3 of routine scrambler; And c[k] be the frequency spectrum of frequency domain MDCT.
(irrelevant with windowization) can be followingly described in the delta pulse:
x(n)=1 n=n′
x(n)=0 n≠n′ [2]
With [2] substitution [1], obtain:
Fig. 3 result of caption hybrid analysis bank of filters after the anti-aliased butterfly of delta pulse.As seen frequency spectrum shown in Figure 3 is made up of the waveform of cosine type, and the waveform of cosine type have corresponding to odd number, promptly replace (alternate) or every the waveform of the sub-band of negated (multiply by-1).The output of this characteristic and hybrid filter-bank is shared, and the output of known hybrid filter-bank comprises the alternately sub-band component of negating.In fact, for composite filter group 42 in the demoder 30 every one sub-band, negated (promptly multiply by-1) every one input value, with compensation by the 12 caused frequency inversions of the analysis filterbank in the scrambler 10.As a result, the phase differential between the adjacent sub-bands becomes and is similar to 180 degree, promptly multiply by-1.More detailed description is the row article as follows: " the Aliasing reduction insub-bands of cascaded filter banks with decimation " of B.Edler, ElectronicsLetters, on June 4th, 1992,28, the 12 phases of volume).
Fig. 4 with caption the hybrid filter-bank after the anti-aliased butterfly of delta pulse multiply by-1 at the sub-band component of negating and compensate this frequency spectrum after negating.After compensation, c[k] comprise the cosine function of distortion a little.The distortion that can see in Fig. 4 results from the aliasing that causes owing to the down-sampling in the analysis filterbank, and it partly compensates by anti-aliased butterfly and by the fact that analysis filterbank does not have an ideal linearity phase propetry.Therefore, by some pre-service of MDCT coefficient, the operation of hybrid filter-bank can be similar to MDCT.As following described in more detail, in a preferred embodiment, one or more " long " MDCT is used for replacing the operation of the mixing composite filter group 38,42 of demoder 30.Equally, one or more " long " MDCT can be used to replace the operation of the hybrid analysis bank of filters 12,14 of scrambler 10.
In Fig. 6, show the decode data-signal of MPEG-1 layer III and the equipment of similar encoded data signals of being used to that embodies one aspect of the invention, be typically expressed as 60.Only show those for understanding element essential to the invention.Decoding device or demoder 60 comprise decoding and go quantifying unit 62 that it is configured to receive the data-signal or the similar encoded data signals of MPEG-1 layer bit stream form.Decoding and go the bit stream of 62 pairs of receptions of quantifying unit to carry out suitable decoding (being generally the Huffman decoding of MP3 defined) and re-quantization is to recover a plurality of frequency lines or MDCT coefficient.When this bit stream comprises MP3 consistance (conformant) data, decode and remove MP3 decoding and the re-quantization that quantifying unit 62 can operative norm.Usually for a frame that comprises 1152 input audio samples, two granularities of 576 frequency lines are recovered (because the overlap-add of carrying out operation by unit 62 in windowization, 576 input samples provide 576 MDCT coefficients effectively, so system is by threshold sampling).
Demoder 60 comprises rearrangement unit 64, is used for resequencing as required by decoding and removing the frequency line of quantifying unit 62 generations.The rearrangement that the rearrangement reversing is normally carried out by scrambler.It below is more detailed description.Rearrangement unit 62 can from frame is associated separately supplementary determine the rearrangement that needs what type.
Contrary MDCT unit IMDCT 68 is provided, is used on the frequency line of rearrangement, carrying out one or more contrary MDCT.As mentioned above, IMDCT unit 68 is configured to each enterprising line operate of whole granularity at frequency line, carry out single contrary MDCT when stopping the window of type (at needs normal, short or) on all frequency lines in this granularity, or in this granularity, carry out a plurality of contrary MDCT (when needs short type window) on the child group of the respective numbers of all frequency lines.The MP3 bit stream that comprises 576 frequency lines for a granularity, for normal, begin or stop window, single contrary MDCT is carried out in IMDCT unit 68 on whole granularity, produce 1152 time-domain samplings, and in every group form by 192 frequency lines three son groups, carry out three contrary MDCT on separately one, produce three sequence or groups separately of 384 time-domain samplings.Component of signal that the output of IMDCT unit 68 comprises a plurality of (in the current example being 1152) has recovered or sampling, it can be used to construct the output signal of PCM.
In order to construct the output signal of PCM, on the signal sampling that produces by IMDCT unit 68, carry out windowization and overlap-add operation.Therefore, demoder 60 also comprises windowization and overlap-add unit 70, is described in more detail below its operation.
It is to be noted that 42 pairs of the composite filter groups of the conventional MP3 decoding device 30 alternately component of signal or the sub-band sound channels of sub-band are negated, with the frequency inversion of the analysis filterbank 12 of compensated encoder 10.Therefore, in the embodiment of the demoder 60 that is intended to decoding standard MP3 consistance data, demoder 60 comprises the unit 66 of negating, and is used for the component of signal or the sound channel that replace sub-band are negated, and promptly relevant MDCT coefficient multiply by-1.The unit 66 of negating has been shown among Fig. 6, between rearrangement unit 64 and IMDCT unit 68, but replacedly, can be positioned at other places, for example decoded and going between quantifying unit 62 and the rearrangement unit 64.
It should be appreciated that analysis filterbank 12 has the sub-band of overlapping.This effect is reduced by anti-aliased butterfly 16 usually, and scrambler 10 generally includes anti-aliased butterfly 16.
In order to understand rearrangement unit 64 and windowization and overlap-add unit 70 better, the existing MP3 windowization of describing routine in more detail.Four kinds of different window types (with subsidiary length) in MP3, have been stipulated, i.e. ' normal (normally) ', ' start (beginning) ', ' short (weak point) ' and ' stop (stopping) '.The sequence of the particular type of window or different windows type is selected to that a part of characteristic that is fit to be about to apply the data of (one or more) window.For example, the short type window is applied in usually to the data division corresponding to transient state in the sound signal.Which kind of window type is the supplementary that is associated with the Frame that provides indicate will be used with granularity.Required window type is all influential to the length of MDCT (in like manner, contrary MDCT) or size and windowization/overlap-add operation.
For MP3, window function z (n) can be described below:
Window (type 0) for normal type:
Window (Class1) for the beginning type:
For the window (type 2) of short type, three short windows are encoded simultaneously:
For the window that stops type (type 3):
Each window function in equation [4], [5], [6] and [7] is commonly referred to be the single window function, surpasses one window even they can comprise applying.This should be 36 (i.e. 36 windows), find out the function [4], [5] and [7] of index n from 0 to 35 thus from length of window.For function [6], the pattern length of 12 windows of three weak points is 36, and for p from 0 to 2, n runs to 11 from 0 thus.Therefore, the size of the total length respective frequency sub-bands component of signal of each window type (sampling of 36 sub-frequency bands).
For the window of type 2, promptly lack the sequence of window, in scrambler 10,576 MDCT coefficients of each granularity (32 sub-frequency bands multiply by 3 windows and multiply by 6 MDCT coefficients) are sorted, so that allow more efficient coding.Therefore in demoder, rearrangement accordingly takes place, so that make the rearrangement of carrying out by scrambler reverse.In the MP3 decoding device 30 of routine, the MDCT coefficient or the frequency line of granularity, according to frequency line, then according to the window index, then according to sub-band, resequenced according to the granularity that increases progressively.This is illustrated that in Fig. 5 it shows the structure of the part of granularity 50, and in granularity 50, each frequency line or MDCT coefficient can be given and from 0 to 575 the index of frequency line separately.At the highest or the most coarse granularity place, frequency line is sorted according to the sub-band index of indicating them to belong to which sub-band and from 0 to 31.In each sub-band, frequency line is sorted according to the window index which window of indication will be used to frequency line and from 0 to 2.In each window, frequency line is provided for the order of MDCT according to the indication frequency line and from 0 to 5 frequency line subindex is sorted.Therefore, first frequency line in the granularity 50 (promptly, frequency line with minimum frequency line index (=0)), be that subindex is 0, the window index be 0 and sub-band index be 0 frequency line, the subindex of second frequency line (frequency line coefficient=1) is 1, the window index be 0 and sub-band index be 0, or the like, the last frequency line in granularity 50, it has the highest frequency line index 575, has subindex and be 5, window index 2 and a sub-band index 31.
In demoder 60, rearrangement unit 64 is configured to according to being different from the resequence frequency line of granularity of the above-mentioned mode that is used for conventional demoder.For " weak point " (type 2) window, rearrangement unit 64 is according to frequency line, then according to sub-band and last window, according to the granularity rearrangement frequency line that increases progressively.This is illustrated in Fig. 7, as can be seen from Figure 7 granularity 50 ' in, frequency line according to the window index, then according to sub-band index, then according to the frequency band subindex, according to the ordering of the highest rank.
The PCM output signal that constructs in conjunction with IMDCT unit 68 by windowization and overlap-add unit 70 is now described.Suppose in the example below that initial p CM signal comprises a plurality of frames of 1152 audio samples, each frame is transformed into two granularities of 576 frequency lines (or MDCT coefficient).Therefore, IMDCT unit 68 is at the enterprising line operate of the granularity of 576 MDCT coefficients, comprises the signal of 1152 samplings with generation, and these 1152 samplings are provided for windowization and overlap-add unit 70 then.
The MDCT coefficient X of l group or l granularity
l(k) expression, wherein k=0...575.The output signal that produces after decoding l group (l is since 1) MDCT coefficient by windowization and overlap-add unit 70 is described to (use overlap-add):
y
l(n+576·l)=y
l-1(n+576·l)+x
l(n)
y
l(n+576(l-1))=y
l-1(n+576(l-1))+x
l(n) [8]
Wherein, index n=0...1151, y
L-1(n) be in the output signal of decoding l-1 after organizing, and x
l(n) be at MDCT coefficient X by IMDCT unit 68
l(k) enterprising line operate and the signal that produces.Output signal y
0(n) be both initialized to zero for all n.
Signal x
l(n) window type of appointment is depended in generation.When the window type of l granularity was 0,1 or 3, IMDCT unit 68 was at X
l(k) carry out contrary MDCT on 576 input coefficients that provide, to produce the momentary signal x of 1152 points
Tmp(n), as described in the equation [9]:
N=0...N-1 wherein, N=1152.
When the window type of l group was 2 (i.e. " weak point " windows), three contrary MDCT were carried out in IMDCT unit 68 on one group of 192 input coefficient separately, and each coefficient is by X
l(k) provide, to produce the momentary signal x that each is made up of 384 points
Tmp, 0(n), x
Tmp, 1(n) and x
Tmp, 2(n), each momentary signal is as described in the equation [10]:
P=0...2 wherein, n=0...N-1, N=384.
It is offered the momentary signal x of windowization and overlap-add unit 70 effectively
Tmp(n), x
Tmp, p(n).
When the window type of l group is 0, windowization and overlap-add unit 70 following signal calculated x
l(n):
Wherein, the divisor 1152 in [11] is corresponding to the transform length N of IMDCT.
When the window type of l group is 1, windowization and overlap-add unit 70 following signal calculated x
l(n):
x
l(n)=x
tmp(n) n=576...767.
x
l(n)=0 n=960...1151
[12]
Wherein, the divisor 1152 in [12] is corresponding to the transform length N of IMDCT, and divisor 384 is corresponding to N/3, and 576 corresponding to N/2.
When the window type of l group was 2, windowization and overlap-add unit 70 came signal calculated x by at first calculating three momentary signals
l(n):
Wherein, divisor 384 is corresponding to the transform length N of IMDCT.
Signal x
l(n) be constructed as follows then:
x
l(n)=0 n=0...191
x
l(n)=x
l,tmp,0(n-192) n=192...383
x
l(n)=x
l,tmp,0(n-192)+x
l,tmp,1(n-384) n=384...575
x
l(n)=x
l,tmp,1(n-384)+x
l,tmp,2(n-576) n=576...767
x
l(n)=x
l,tmp,2(n-576) n=768...959
x
l(n)=0 n=960...1151
[14]
When the window type of l group is 3, windowization and overlap-add unit 70 following signal calculated x
l(n):
x
l(n)=0 n=0...191
x
l(n)=x
tmp(n) n=384...575
[15]
Wherein, divisor 1152 is corresponding to the length N of IMDCT, and divisor 384 is corresponding to N/3.As seen, equation [11], [12], [13] and [15] are general types:
x
l(n)=z(n)x
tmp(n) [16]
Wherein, x
l(n) be the signal of windowization, x
Tmp(n) be not by the signal of windowization, z (n) is a window function.It is to be noted that the window function z (n) of equation [11], [12], [13] and [15] is similar to the window function z (n) that describes respectively usually in equation [4], [5], [6] and [7].Yet, longer and divisor separately is correspondingly bigger according to window function z (n) length of window separately in separately transform length N equation [11], [12], [13] and [15].Can think, the window function z (n) of equation [11], [12], [13] and [15] is included in up-sampling (up-sampled) version of the window function z (n) that describes respectively in equation [4], [5], [6] and [7], the degree of up-sampling depends on transform length/length of window separately, N.It should be appreciated that each all comprises single window function the window function of equation [11], [12], [13] and [15], even its application may comprise the application of an above window.
In addition, 70 of windowization and overlap-add unit carry out an application of specified window type, i.e. sampling to whole granularity applies a window function.This is opposite with conventional demoder 30, and in conventional demoder 30, window function all applies at each sub-band.
Discovery comprises high-quality sound signal by the PCM output signal that windowization and overlap-add unit produce, although it is with the MP3 standard is not quite identical or bit correctly (bit-true) not.Especially, some phase distortions and aliasing occur, compare, cause relatively little distortion spectrum and time domain distortion with MP3 consistance signal.Yet, find that these distortions or artifact do not have important adverse effect to people's sensing audio signal.In fact, in demoder 60, the contrary MDCT that mixing composite filter group is had " length " phase distortion of some spectral aliasings substitutes.By eliminating the demand to many " weak point " MDCT and composite filter group, the computational complexity of demoder 60 is greatly reduced.For example, each output sampling needs about 22.11 multiplication and 26.73 sub-additions through the general conventional MP3 decoding device of optimizing.Corresponding demoder 60 each output sampling through optimizing only need 8 multiplication and 20.5 sub-additions.As a result, demoder 60 provides higher decoding efficiency, the latter cause still less power consumption or lower DSP demand.The complicacy of demoder 60 further reduces, and does not wherein need aliasing butterfly (because their existence in demoder 30 help composite filter group 42 reconstruct PCM output signals).
As mentioned above, principle of the present invention can be used for the scrambler of MP3 type equally.Therefore, another aspect of the present invention provides a kind of coding audio signal that is used for the signal of generation MPEG-1 layer III type or the equipment of bit stream.It is to be noted that this bit stream is not the MP3 bit stream of standard, although it is consistent with MP3---the decoded signal that obtains is being different from the MP3 standard aspect phase response and the aliasing.In essence, " length " MDCT of phase distortion is used for replacing the analysis hybrid filter-bank 12,14 of conventional scrambler 10.Fig. 8 shows the simplified block diagram that embodies the present invention's scrambler 80 in this respect.Scrambler 80 comprises window unit 82, and it carries out windowization on the PCM input sample that receives.The window function class is similar to those functions of describing in the equation [4], [5], [6] and [7], although length of window is according to required MDCT transform size and difference.For window normal, that begin or stop type, " length " MDCT is carried out in MDCT unit 84 on all 1152 input samples of the frame that receives, to produce 576 frequency lines.For short window, MDCT unit 84 is carried out three " long " MDCT on one group separately in three groups of input samples of 384 every group, to produce one group of 192 frequency line separately.Scrambler 80 can comprise that conventional MP3 quantizes and coding unit 86 and Bit Allocation in Discrete unit 88.Can between MDCT unit 82 and quantification and coding unit 86, provide the unit 85 of negating, be used for to replace, promptly negate every one sub-band.
Should be appreciated that the effect of the unit 66 of negating in the demoder 60 is to compensate alternately the intrinsic of sub-band and negates, and alternately intrinsic the negating of sub-band taken place in conventional MP3 scrambler.Correspondingly, negate in the scrambler 80 effect of unit 85 is to create alternately the negating of sub-band of taking place usually in conventional scrambler 10.Yet, alternately sub-band negate optional, so in alternative embodiment, the unit 66,85 of negating can omit.
Should be appreciated that the MPEG-1 layer III data that demoder 60 not only can the decoding standard unanimity, and the MPEG-1 layer III categorical data that can decode nonstandard, for example nonstandard data that produce by scrambler 80.
The present invention is not limited to the signal or the MDCT of MPEG-1 layer III data.For example, the demoder that embodies first aspect present invention can be provided in and carry out work on the encoded data signal that is produced by following scrambler (comprising non-MPEG-1 layer III scrambler): this scrambler provides the data-signal (particularly but nonessential be sound signal) of un-encoded to the sub-band filter group, and impels subsequently on the sub-band signal of each generation, promptly carries out the correspondent frequency direct transform on hybrid filter-bank.As long as go accordingly to quantize and decoding in the demoder execution, then the quantification after whom needn't be consistent with MP3 with coding.Similarly, the frequency direct transform needn't comprise MDCT, as long as this demoder adopts compatible frequency inverse transformation.It is to be noted in this, term " granularity " mainly is the term of MP3, but the technician will readily appreciate that, in the context of the embodiment of non-MP3, term " granularity " can be interpreted as any equivalence group (term " frame " is equivalent to " granularity " usually) of frequency line or coefficient as used herein.
Preferably but not necessarily, sub-band filter group and frequency transformation are sampled critically, and window function overlapping 50% (therefore, conversion demonstrates the time domain aliasing and eliminates (TDAC) attribute), and more preferably, have actual value.Also preferably but not necessarily, carry out aliasing and reduce,, at scrambler the sub-band signal of conversion is carried out aliasing and reduce for example by anti-aliased butterfly.
Aforementioned description relates to monaural signal, but the present invention can easily be used for the coding and the decoding of stereo or multichannel by handling each sound channel separately in the above described manner.
Embody encoder of the present invention, can realize the program code that for example uses a computer, hardware or their combination with any mode easily.
The present invention is not limited to embodiment described here, can make amendment without departing from the scope of the invention or change.
Claims (25)
1. demoder that is used for data-signal, described data-signal by data-signal is provided to the sub-band filter group and by sub-band signal in each generation on the frequency direct transform carried out separately be encoded, described demoder comprises: be used to decode and remove the data-signal that quantizes the to receive device with the granularity that produces a plurality of frequency lines; Be used on each granularity, carrying out one or more frequency inverse transformations to produce the device of a plurality of data samplings; And be used for described data sampling is applied the window function of one or more types to produce the device of a plurality of window data samplings; Wherein, with regard at least the first type window function, described frequency inverter is provided on all frequency lines of granularity separately and carries out the single frequency inverse transformation, and wherein said demoder also comprises the device that is used for constructing from described window data sampling output signal.
2. demoder as claimed in claim 1, wherein said sub-band filter group comprises the sub-band filter group of threshold sampling.
3. demoder as claimed in claim 1 or 2, the coding of wherein said data-signal also are included on the described sub-band signal to be carried out aliasing and reduces.
4. as the described demoder of arbitrary aforementioned claim, wherein said frequency direct transform comprises the conversion of threshold sampling.
5. demoder as claimed in claim 4, wherein said window function overlaps 50% about data sampling.
6. demoder as claimed in claim 5, wherein said output signal constructing apparatus applies one or more overlap-add operations to produce described output signal to described window data-signal.
7. as the described demoder of arbitrary aforementioned claim, wherein said frequency direct transform comprises improved discrete cosine transform (MDCT), and described frequency inverse transformation comprises improved inverse discrete cosine transform (IMDCT).
8. as the described demoder of arbitrary aforementioned claim, wherein with regard to described at least first type window function, described window function bringing device is configured to all data samplings that produce about granularity are separately applied the single window function.
9. as the described demoder of arbitrary aforementioned claim, wherein said at least the first type window function comprises the version that the length adjustment of MPEG-1 layer III type 0, Class1 and type 3 window functions is crossed.
10. as the described demoder of arbitrary aforementioned claim, wherein with regard at least the second type window function, described frequency inverter is provided in the frequency inverse transformation of carrying out on each class frequency line of granularity separately, and all frequency lines of described granularity belong to a group or other groups in described many groups.
11. demoder as claimed in claim 10, wherein with regard to described at least second type window function, described window function bringing device is configured to applying the single window function on all data samplings that produce about each class frequency line.
12. as claim 10 or 11 described demoders, wherein said at least the second type window function comprises the version that the length adjustment of MPEG-1 layer III type 2 window functions is crossed, and the frequency line of described granularity belongs to one group or other groups in three groups.
13. as the described demoder of arbitrary aforementioned claim, wherein in the granularity each frequency line and a plurality of frequency sub-bands separately one be associated, described demoder also comprises the device that is used for the frequency line in the rearrangement granularity when being about to apply described at least the second type window function, described reordering apparatus be configured to according to frequency line belong to any group, relevant with which frequency sub-bands, then according to the order of frequency line, according to the granularity of the successively decreasing frequency line of resequencing.
14. as the described demoder of arbitrary aforementioned claim, also comprise being used for device that the frequency line relevant with the alternating frequency sub-band negated.
15. as the described demoder of the claim 14 that is subordinated to claim 13, the wherein said device of negating is set between described reordering apparatus and the described frequency inverter.
16. decoding data signal method, described data-signal by data-signal is provided to the sub-band filter group and by sub-band signal in each generation on the frequency direct transform carried out separately be encoded, described method comprises: decode and the data-signal that goes to quantize to receive to produce the granularity of a plurality of frequency lines; On each granularity, carry out one or more frequency inverse transformations to produce a plurality of data samplings; The window function that described data sampling is applied one or more types is to produce a plurality of window data samplings; And from described window data sampling, construct output signal; Wherein with regard at least the first type window function, on all frequency lines in the granularity separately, carry out the single frequency inverse transformation.
17. a scrambler that is used to comprise the input signal of a plurality of data samplings, described scrambler comprises: be used for described data sampling is applied the window function of one or more types to produce the device of a plurality of window data samplings; Be used on the window data sampling carrying out the device of one or more improved discrete cosine transforms (MDCT) with the granularity that produces a plurality of frequency lines; And be used to encode and quantize the device of each granularity with the data-signal that produces MPEG-1 layer III type; Wherein with regard at least the first type window function, described MDCT device is provided in about the single MDCT of execution on all window data samplings of the data-signal of the described reception of its generation granularity separately.
18. scrambler as claimed in claim 17, wherein with regard to described at least first type window function, described window function bringing device is provided in to produce on all window data samplings of the data-signal that receives separately of granularity separately about it and applies the single window function.
19. as claim 17 or 18 described scramblers, wherein said at least the first type window function comprises the version that the length adjustment of MPEG-1 layer III type 0, Class1 and type 3 window functions is crossed.
20. as the described scrambler of any one claim in the claim 17~19, wherein with regard at least the second type window function, described MDCT device is provided in about it and produces the MDCT that carries out on the window data sampling group separately of granularity separately separately, belongs to one group or other groups in described many groups about its all window data samplings that produce granularity separately.
21. scrambler as claimed in claim 20, wherein with regard at least the second type window function, described window function bringing device is configured to all window data samplings of organizing are separately applied the single window function.
22. as claim 20 or 21 described scramblers, wherein said at least the second type window function comprises the version that the length adjustment of MPEG-1 layer III type 2 window functions is crossed, and belongs to one group or other groups in three groups about its described window data sampling that produces granularity separately.
23. as the described scrambler of any one claim in the claim 17~22, wherein in the granularity in each frequency line and a plurality of frequency sub-bands separately one be associated, described scrambler also comprises and is used for device that the frequency line relevant with the alternating frequency sub-band negated.
24. one kind is carried out Methods for Coding to the input signal that comprises a plurality of data samplings, described method comprises: the window function that described data sampling is applied one or more types is to produce a plurality of window data samplings; The window data sampling is carried out one or more improved discrete cosine transforms (MDCT) to produce the granularity of a plurality of frequency lines; Coding also quantizes each granularity to produce the data-signal of MPEG-1 layer III type; Wherein with regard at least the first type window function, on all window data samplings of the data-signal of the described reception that produces granularity separately about it, carry out single MDCT.
25. a system that is used for the Code And Decode data-signal, described system comprise scrambler as claimed in claim 17 and demoder as claimed in claim 1.
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
EP03104535 | 2003-12-04 | ||
EP03104535.4 | 2003-12-04 |
Publications (1)
Publication Number | Publication Date |
---|---|
CN1890712A true CN1890712A (en) | 2007-01-03 |
Family
ID=34639327
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CNA200480035931XA Pending CN1890712A (en) | 2003-12-04 | 2004-11-30 | Audio signal coding |
Country Status (5)
Country | Link |
---|---|
EP (1) | EP1692686A1 (en) |
JP (1) | JP2007515672A (en) |
KR (1) | KR20060131767A (en) |
CN (1) | CN1890712A (en) |
WO (1) | WO2005055203A1 (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102243872A (en) * | 2010-05-10 | 2011-11-16 | 炬力集成电路设计有限公司 | Method and system for encoding and decoding digital audio signals |
Families Citing this family (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2013186344A2 (en) | 2012-06-14 | 2013-12-19 | Dolby International Ab | Smooth configuration switching for multichannel audio rendering based on a variable number of received channels |
EP3067886A1 (en) * | 2015-03-09 | 2016-09-14 | Fraunhofer-Gesellschaft zur Förderung der angewandten Forschung e.V. | Audio encoder for encoding a multichannel signal and audio decoder for decoding an encoded audio signal |
JP7385531B2 (en) * | 2020-06-17 | 2023-11-22 | Toa株式会社 | Acoustic communication system, acoustic transmitting device, acoustic receiving device, program and acoustic signal transmitting method |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US7369989B2 (en) * | 2001-06-08 | 2008-05-06 | Stmicroelectronics Asia Pacific Pte, Ltd. | Unified filter bank for audio coding |
-
2004
- 2004-11-30 JP JP2006542091A patent/JP2007515672A/en active Pending
- 2004-11-30 EP EP04799284A patent/EP1692686A1/en not_active Withdrawn
- 2004-11-30 CN CNA200480035931XA patent/CN1890712A/en active Pending
- 2004-11-30 KR KR1020067010745A patent/KR20060131767A/en not_active Application Discontinuation
- 2004-11-30 WO PCT/IB2004/052602 patent/WO2005055203A1/en not_active Application Discontinuation
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102243872A (en) * | 2010-05-10 | 2011-11-16 | 炬力集成电路设计有限公司 | Method and system for encoding and decoding digital audio signals |
Also Published As
Publication number | Publication date |
---|---|
JP2007515672A (en) | 2007-06-14 |
WO2005055203A1 (en) | 2005-06-16 |
KR20060131767A (en) | 2006-12-20 |
EP1692686A1 (en) | 2006-08-23 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN1258172C (en) | Device and method for encoding a time-discrete audio signal and method for decoding coded audio data | |
US7343287B2 (en) | Method and apparatus for scalable encoding and method and apparatus for scalable decoding | |
CN1065381C (en) | Digital audio signal coding and/or decoding method | |
JP5400143B2 (en) | Factoring the overlapping transform into two block transforms | |
CN1914669A (en) | Audio signal decoding using complex-valued data | |
EP2625688B1 (en) | Apparatus and method for processing an audio signal and for providing a higher temporal granularity for a combined unified speech and audio codec (usac) | |
US20050114126A1 (en) | Apparatus and method for coding a time-discrete audio signal and apparatus and method for decoding coded audio data | |
CN1310210C (en) | Audio coding system using characteristics of a decoded signal to adapt synthesized spectral components | |
EP2028648A2 (en) | Multi-channel audio encoding and decoding | |
CN101053019A (en) | Encoding and decoding of audio signals using complex-valued filter banks | |
CN1875402A (en) | Audio signal encoding or decoding | |
EP2372706B1 (en) | Method and apparatus for encoding excitation patterns from which the masking levels for an audio signal encoding are determined | |
CN1036886C (en) | Method of and system for coding analogue signals | |
CN1878001A (en) | Apparatus and method of encoding audio data and apparatus and method of decoding encoded audio data | |
CN1926609A (en) | Adaptive hybrid transform for signal analysis and synthesis | |
CN1897467A (en) | Signal encoding apparatus and method, signal decoding apparatus and method, programs and recording mediums | |
CA2717584A1 (en) | Method and apparatus for processing an audio signal | |
CN1806239A (en) | Device and method for conversion into a transformed representation or for inversely converting the transformed representation. | |
CN1677490A (en) | Intensified audio-frequency coding-decoding device and method | |
CN1787383A (en) | Methods and apparatuses for transforming, adaptively encoding, inversely transforming and adaptively decoding an audio signal | |
CN1132146C (en) | Method and apparatus for synthesizing speech | |
JP2004531151A (en) | Method and apparatus for processing time discrete audio sample values | |
CN1677491A (en) | Intensified audio-frequency coding-decoding device and method | |
JPWO2009125588A1 (en) | Encoding apparatus and encoding method | |
CN1890712A (en) | Audio signal coding |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
WD01 | Invention patent application deemed withdrawn after publication |
Open date: 20070103 |