Voice coding method

Description: TECHNICAL FIELD [Detailed description of the invention]

【0001】

[Technical field to which the invention belongs]

The present invention relates to a voice coding method for compressing a multi-channel voice signal.

【0002】

[Conventional technology]

As a method of compressing an audio signal, the present inventor applied in a previous application (Japanese Patent Application No. 9-289159) to a past signal in the time region by using a plurality of predictors having different characteristics with respect to a 1-channel original digital audio signal. Predictive coding that calculates multiple linear predicted values of the current signal from, calculates the predicted residual for each predictor from the original digital audio signal and these multiple linear predicted values, and selects the minimum predicted residual value. Proposing a method.

【0003】

With the above method, a certain compression effect can be obtained when the original digital audio signal has a sampling frequency of 96 kHz and the number of quantization bits is about 20 bits, but recent DVD-Audio discs have twice the sampling frequency ( = 192kHz) is used, and the number of quantization bits tends to be 24 bits. In addition, the sampling frequency and the number of quantization bits in the multi-channel may differ for each channel.

【0004】

[Problems to be Solved by the Invention]

By the way, since the compression rate of a compression method such as the predictive coding method is variable (VBR: variable bit rate), when a multi-channel audio signal is predictively coded, the amount of data for each channel changes significantly over time. To do. Further, when such data is transmitted, it is transmitted as a data stream instead of parallel for each channel. Therefore, it is necessary to enable playback (presentation) of such a variable-length data stream on the playback side (decoding side) in synchronization with each channel.

【0005】

Therefore, an object of the present invention is to provide an audio coding method capable of improving the decoding efficiency on the reproduction side when encoding a multi-channel audio signal with a variable compression rate.

【0006】

[Means for solving problems]

The present invention comprises the means described below in order to achieve the above object. That is, [0007]

A multi-channel audio signal with a certain sampling frequency and the number of quantization bits is obtained as the first sample value in response to the input audio signal for each channel as it is or for each channel correlated with each other, and a plurality of different characteristics are obtained. The linear prediction method predicts the linear predicted values of the past and present signals in the time region, and selects the linear predictive method that minimizes the predicted residuals obtained from the predicted linear predicted values and the voice signal. A step of predictive coding, a subpacket storing predictive coding data including a linear predictive method, a predictive residual, and a predetermined head sample value for each channel selected by the step, and an original analog on the reproduction side. A voice coding method comprising a step of formatting into a data structure having a synchronization information unit including a sampling frequency and the number of quantization bits used when being restored to a voice signal.

【0008】

BEST MODE FOR CARRYING OUT THE INVENTION

Hereinafter, embodiments of the present invention will be described with reference to the drawings. FIG. 1 is a block diagram showing a first embodiment of a voice coding device to which the present invention is applied and a corresponding voice decoding device, FIG. 2 is a block diagram showing a coding unit of FIG. 1 in detail, and FIG. 3 is a diagram. 1. Explanatory diagram showing the bit stream encoded by the coding part of FIG. 2, FIG. 4 is an explanatory diagram showing the format of the DVD pack, FIG. 5 is an explanatory diagram showing the format of the audio pack of DVD, and FIG. 6 is an explanatory diagram showing the format of the audio pack of DVD. An explanatory diagram showing the format of the audio data area of FIG. 5 in detail, FIG. 7 is a block diagram showing the decoding part of FIG. 1 in detail, FIG. 8 is a timing chart showing the write / read timing of the input buffer of FIG. 7, and FIG. 9 is a timing chart. An explanatory diagram showing the amount of compressed data for each access unit, and FIG. 10 is an explanatory diagram showing the access unit and the presentation unit.

【0009】

Here, as the multi-channel method, for example, the following four methods are known. (1) 4-channel system Like the Dolby Surround system, there are 3 channels of front L, C, and R + 1 channel of rear S, for a total of 4 channels. (2) 5-channel system Like the Dolby AC-3 system without SW channels, there are 3 channels of front L, C, and R + 2 channels of rear SL and SR, for a total of 5 channels. (3) 6-channel system 6-channel system (L, C, R, SW (Lfe), SL, SR like DTS (Digital Theater System) system and Dolby AC-3 system) (4) 8-channel system Like the SDDS (Sony Dynamic Digital Sound) system, there are 6 channels of front L, LC, C, RC, R, and SW + 2 channels of rear SL and SR, for a total of 8 channels. [0010]

The 6-channel (ch) mix & matrix circuit 1'on the coding side shown in FIG. 1 has front left (Lf), center (C), front right (Rf), surround left (Ls), as an example of a multi-channel signal. 6ch PCM data of surround light (Rs) and Lfe (Low Frequency Effect) is classified into 2ch "1" and "2" related to the front group and 4ch "3" to "6" related to other groups by the following equation (1). 2ch "1" and "2" are output to the first coding unit 2'-1, and 4ch "3" to "6" are output to the second coding unit 2'-2. "1" = Lf + Rf "2" = Lf-Rf "3" = C- (Ls + Rs) / 2 "4" = Ls + Rs "5" = Ls-Rs "6" = Lfe-a × C However, 0 a 1 ... (1) [0011]

The first and second coding units 2'-1 and 2'-2 that make up the coding unit 2'are 2ch "1", "2" and 4ch "3" to "2", respectively, as shown in detail in FIG. The PCM data of "6" is predictively coded for each channel, and the predictively coded data is transmitted to the decoding side in a bit stream as shown in FIG. 3 via a recording medium 5 or a communication medium 6 such as a satellite line or a telephone line. On the decoding side, the first and second decoding units 3'-1 and 3'-2 that make up the decoding unit 3'have the 2ch "1" and "2" related to the front group, respectively, as shown in detail in FIG. Decode the predicted encoded data of 4ch "3" to "6" related to other groups into PCM data for each channel.

【0012】

Next, the original 6ch (Lf, C, Rf, Ls, Rs, Lfe) is restored based on the equation (1) by the mix & matrix circuit 4', and the original 6ch and the coefficient mij (i = 1,2, Stereo 2ch data (L, R) is generated by the following equation (2) by j = 1,2 ~ 6). L = m11 Lf + m12 Rf + m13 C + m14 Ls + m15 Rs + m16 Lfe R = m21 Lf + m22 Rf + m23 C + m24 Ls + m25 Rs + m26 Lfe ... (2) [0013]

The coding units 2'-1 and 2'-2 will be described in detail with reference to FIG. The PCM data of each channel "1" to "6" is stored in the 1-frame buffer 10 for each frame. Then, the sample data of each channel "1" to "6" of one frame is applied to the prediction circuits 13D1, 13D2, 15D1 to 15D4, respectively, and the head sample data of each frame of each channel "1" to "6" is applied. (Stored in the restart header described later) is applied to the unpacking circuit 8 and the formatting circuit 19. Further, the sampling frequency (fs) and the number of quantization bits (Qb) when the PCM data is A / D converted are applied to the packing circuit 18 and the formatting circuit 19. The prediction circuits 13D1, 13D2, and 15D1 to 15D4 each use a plurality of predictors (not shown) having different characteristics for the PCM data of each channel "1" to "6" to change the past signal to the current signal in the time region. The multiple linear prediction values of are calculated, and then the prediction residual for each predictor is calculated from the original PCM data and the plurality of linear prediction values. Subsequent buffer / selectors 14D1, 14D2, and 16D1 to 16D4 temporarily store each prediction residual calculated by the prediction circuits 13D1, 13D2, and 15D1 to 15D4, respectively, and select signal / DTS (decoding time stamp). The minimum predicted residual value is selected for each subframe specified by the generator 17.

【0014】

The selection signal / DTS generator 17 applies the bit number flag of the predicted residual to the packing circuit 18 and the formatting circuit 19, and also the predictor selection flag indicating the predictor with the smallest predicted residual and the equation ( A correlation coefficient a in 1) and a DTS indicating the time for the decoding side to extract stream data from the input buffer 22a (FIG. 7) are applied to the formatting circuit 19. The packing circuit 18 packs the predicted residuals for 6 channels selected by the buffer selectors 14D1, 14D2, 16D1 to 16D4 with a specified number of bits based on the bit number flag specified by the selection signal / DTS generator 17. .. The PTS generator 17c also generates a PTS (presentation time stamp) indicating the time for the decoding side to extract PCM data from the output buffer 110 (FIG. 7) and outputs it to the formatting circuit 19.

【0015】

The subsequent formatting circuit 19 formats the user data as shown in FIGS. 3 to 6. The user data (subpackets) shown in FIG. 3 are variable rate bit stream (substream) BS0 containing predicted encoded data of 2ch "1" and "2" for the forward group, and 4ch "3" ~ for other groups. It is composed of a variable rate bit stream (substream) BS1 containing the predicted encoded data of "6" and a bitstream header (restart header) provided before the substreams BS0 and BS1. Also, one frame of substreams BS0 and BS1 Frame header and The first sample data of one frame of each channel "1" to "6" and Predictor selection flag for each subframe of each channel "1" to "6" and -The bit number flag for each subframe of each channel "1" to "6" and Predicted residual data string (variable bit number) for each channel "1" to "6", -The coefficient a of ch "6" is multiplexed. According to such predictive coding, when the original signal is, for example, sampling frequency (fs) = 96 kHz, quantization bit number (Qb) = 24 bits, and 6 channels, a compression rate of 71% can be realized.

【0016】

When the variable rate bitstream data predicted and encoded by the coding units 2'-1 and 2'-2 shown in FIG. 2 is recorded on a DVD audio disc as an example of a recording medium, the audio shown in FIG. 4 ( A) Packed in packs. This pack contains 2034 bytes of user data (A packet, V packet), 4 bytes of pack start information, 6 bytes of SCR (System Clock Reference) information, and 3 bytes of Mux rate (3 bytes). It is composed by adding a total of 14 bytes of pack header of rate) information and 1 byte of stuffing (1 pack = total of 2048 bytes). In this case, the time of the A pack in the same title can be managed by making the SCR information which is the time stamp continuous in the same title as "1" in the first pack.

【0017】

As shown in detail in Fig. 5, the A packet of compressed PCM consists of a packet header of 9 to 22 bytes, a private header of compressed PCM, and 1 to 2015 bytes of audio data (compressed PCM) in the format shown in Fig. 3. ing. Then, DTS and PTS are set in the packet header of FIG. 5 (specifically, PTS is set in the 10th to 14th bytes of the packet header, and DTS is set in the 15th to 19th bytes). The private header of compressed PCM is 1-byte substream ID and 2-byte UPC / EAN-ISRC (Universal Product Code / European Article Number-International Standard Recording Code) number, UPC / EAN-ISRC data, and 1 byte private header length and -A 2-byte first access unit pointer and 4 bytes of audio data information (ADI) and It is composed of 0 to 7 bytes of stuffing bytes.

【0018】

Then, the front access unit search pointer for searching the access unit one second later and the rear access unit search pointer for searching the access unit one second before are both set in the ADI in 1 byte. Specifically, the front access unit search pointer is set in the first byte of ADI, and the rear access unit search pointer is set in the eighth byte. In this way, ADI can store up to 2015 bytes of audio data because compressed PCM reduces it to 4 bytes.

【0019】

The audio data area of the compressed PCM (PPCM) audio packet shown in FIG. 5 is composed of a plurality of PPCM access units as shown in FIG. 6, and the PPCM access unit is composed of PPCM sync information and subpackets. The first subpacket in the PPCM access unit consists of a directory, substream "BS0", CRC (1 or 2 bytes), substream "BS1", CRC and extra information, and substream "BS0". , "BS1" are composed only of PPCM blocks. The second and subsequent sub-packets in the PPCM access unit are also composed of a directory, sub-stream "BS0", CRC, sub-stream "BS1", CRC and extra information, and sub-streams "BS0" and "BS1". Consists of a restart header and a PPCM block. And the extra information has at least a size adjustment function. That is, when the incoming data is a fixed rate (CBR), the number of samplings per packet is set to 40, 80, or 160 by the sampling frequency fs as described above, and therefore the determined number of samplings. Depending on the case, the data length per packet and the size of the sub-packet may not match, and in order to match it with the size of the sub-packet, for example, 0, 0 ... etc. are added to adjust the size. Further, it is also possible to use text data or the like as the data for this size adjustment.

【0020】

PPCM sync information (hereinafter, also referred to as synchronization information) includes the following information. -Number of samples per packet: 40, 80 or 160 is selected depending on the sampling frequency fs. -"0" when the data rate is VBR (identifier indicating that the data in the subpacket is compressed data of VBR), "1" when the data rate is CBR (data in the subpacket is a fixed rate) Identifier indicating that) Sampling frequency fs and number of quantization bits Qb Channel allocation information [0021] [0021]

Next, the decoding units 3'-1 and 3'-2 will be described with reference to FIG. 7. The variable rate bitstream data BS0 and BS1 in the above format are separated by the deformatting circuit 21. Then, the first sample data of one frame of each channel "1" to "6" and the predictor selection flag are applied to the prediction circuits 24D1, 24D2, 23D1 to 23D4, respectively, and the number of bits of each channel "1" to "6". The flag is applied to the unpacking circuit 22. Further, the SCR, the DTS, and the predicted residual data string are applied to the input buffer 22a, and the PTS is applied to the output buffer 110. In addition, an identifier indicating whether the data rate is VBR or CBR is applied to each predictor 24D1, 24D2, 23D1, 23D2, 23D3, 23D4, in which the input / output data processing program corresponding to the identifier is determined and processed. It will be. In the case of VBR, it is necessary to switch the processing program and load the input data each time, which requires time for processing, but in the case of CBR, since the rate is fixed, there is no need to switch the processing program and the processing can be performed. It will be faster. Further, the sampling frequency fs and the quantization bit number Qb are applied to the D / A converter 102. Here, the plurality of predictors (not shown) in the prediction circuits 24D1, 24D2, 23D1 to 23D4 have the same characteristics as the plurality of predictors in the prediction circuits 13D1, 13D2, 15D1 to 15D4 on the coding side, respectively. , Those with the same characteristics are selected by the predictor selection flag.

【0022】

The deforming circuit 21 first separates the audio packet from the audio pack, then separates the stream data (predicted residual data string) from the audio packet and extracts the bitstreams BS0 and BS1. Further, the SCR is taken out, and as shown in FIG. 8, it is taken into the input buffer 22a for each access unit and accumulated according to the timing by the SCR. Here, the amount of data in one access unit is (1 / 96kHz) seconds when fs = 96kHz, for example, but it has a variable length as shown in detail in FIGS. 9 and 10A. Then, the stream data stored in the input buffer 22a is read out by the FIFO based on the DTS and applied to the unpacking circuit 22.

【0023】

The unpacking circuit 22 separates the predicted residual data strings of each channel "1" to "6" based on each bit number flag and outputs them to the prediction circuits 24D1, 24D2, and 23D1 to 23D4, respectively. In the prediction circuits 24D1, 24D2, 23D1 to 23D4, respectively, according to the current prediction residual data of each channel "1" to "6" from the unpacking circuit 22, and the predictor selection flag among the multiple internal predictors. The previous predicted value predicted by each selected one is added to calculate the current predicted value, and then the PCM data of each sample is calculated based on the first sample data of one frame and stored in the output buffer 110. Will be done. The PCM data stored in the output buffer 110 is read out based on the PTS and output. Therefore, the variable-length access unit shown in FIG. 10 (a) is extended to have a constant length shown in FIG. 10 (b). The presentation unit is output.

【0024】

Further, the PCM data is converted into an analog signal by the D / A converter 102 based on the sampling frequency fs and the number of quantization bits Qb in the PPCM sync information. At the same time, the CBR identifier is detected in the PPCM sync information, the position of the extra data in the directory is detected, and further, for example, 0,0 ... data and extra data for size adjustment such as text data are detected. Then, if it is text data, extra data is supplied from this unpacking circuit 22 to a text data decoding circuit (not shown), where it is decoded and taken out as text data and output through the output buffer 110. Will be. On the other hand, if the extra data is 0,0 ... data, no processing is performed. If the text data decoder circuit is not prepared, this process is passed. Further, when the search playback is instructed via the operation unit 101, the control unit 100 controls the front access unit search pointer (1 second ahead) and the rear access unit search pointer (1 second ahead) shown in FIG. Play the access unit based on (before). The search pointer may be 2 seconds ahead or 2 seconds ahead instead of 1 second ahead or 1 second ago.

【0025】

When the variable rate bitstream data predicted and encoded by the coding units 2'-1 and 2'-2 shown in FIG. 2 is transmitted via the network, the coding side is for transmission as shown in FIG. Is packetized (step S41), then a packet header is added (step S42), and then this packet is sent out on the network (step S43).

【0026】

On the decryption side, as shown in FIG. 12 (A), the header is removed (step S51), then the data is restored (step S52), and then this data is stored in the memory and waits for decryption (step S53). Then, when decoding is performed, as shown in FIG. 12 (B), deformatting is performed (step S61), then the input / output control of the input buffer 22a is performed (step S62), and then unpacking is performed (step S61). S63). At this time, if there is a search playback instruction, the search pointer is decoded. The predictor is then selected based on the flag and decoded (step S64), then the I / O control of the output buffer 110 (step S65), then the original multichannel is restored (step S66), and then this. Output (step S67), and so on.

【0027】

In the above embodiment, 2ch "1" and "2" related to the front group are used. "1" = Lf + Rf "2" = Lf-Rf Instead, the multi-channel was downmixed by Eq. (2) to generate stereo 2ch data (L, R), and then Eq. (1)' "1" = L + R "2" = LR "3" to "5" are the same "6" = Lfe-C ... (1)' It may be converted by the above and predictively coded (second embodiment). In this case, the mix & matrix circuit 4'on the decoding side can generate the channel L by adding the channels "1" and "2", and can generate the channel R by subtracting the channels "1" and "2".

【0028】

Further, as a third embodiment, as shown in FIG. 13, instead of 2ch 1 and 2, the multi-channel is downmixed by the equation (2) to generate stereo 2ch data (L, R). , The stereo 2ch (L, R) and 4ch "3" to "6" may be predictively encoded. In the second and third embodiments, the front left (Lf) and the front right (Rf) are not transmitted to the decoding side, so the decoding side generates them by the equations (1) and (2).

【0029】

Next, a fourth embodiment will be described with reference to FIGS. 14, 15, and 16. In the above embodiment, one group of correlated signals "1" to "6" is predictively coded, but in this fourth embodiment, a plurality of groups of correlated signals are generated. It is configured to be predictively coded and select the group of predictively coded data with the highest compression ratio. Further, in this embodiment, the coding within the one group is not converted into 2ch related to the front group and 4ch related to the other group as in the case of each of the above-mentioned examples. FIG. 14 is shown as a diagram corresponding to FIG. 1 described above in a configuration in which the coding process summarized in the above is performed. Further, FIG. 15 shows a detailed block of the coding unit, but in the case of this embodiment, n correlation circuits 1-1 to 1-n are provided on the mix & matrix circuit 1'side. ing. These n correlation circuits 1-1 to 1-n can use, for example, 6ch (Lf, C, Rf, Ls, Rs, Lfe) PCM data with n types of 6ch signals "1" to "6" having different correlations. Convert to.

【0030】

For example, the first correlation circuit 1-1 is converted as follows. "1" = Lf "2" = C- (Ls + Rs) / 2 "3" = Rf-Lf "4" = Ls-a x Lfe "5" = Rs-b x Rf "6" = Lfe Further, the nth correlation circuit 1-n is converted as follows. "1" = Lf + Rf "2" = C-Lf "3" = Rf-Lf "4" = Ls-Lf "5" = Rs-Lf "6" = Lfe-C [0031]

In addition, a prediction circuit 15 and a buffer selector 16 are provided for each of the correlation circuits 1-1 to 1-n, and the group with the highest compression ratio is correlated selection based on the data amount of the minimum value of the prediction residual for each group. Selected by signal generator 17b. At this time, the formatting circuit 19 adds the selection flag (correlation circuit selection flag, correlation coefficient a, b of the correlation circuit) to multiplex.

【0032】

Then, FIG. 16 shows a data area corresponding to FIG. 6 described above, and in this embodiment, the substream BS1 is not used, and only the substream BS0 is configured.

【0033】

Further, on the decoding side shown in FIG. 17, n correlation circuits 4-1 to 4-n (or coefficients a and b can be changed with respect to the correlation circuits 1-1 to 1-n on the coding side. One abbreviated correlation circuit) is provided. When the prediction circuits of n groups shown in FIG. 15 have the same configuration, it is not necessary to provide the prediction circuits for n groups as shown in FIG. 17, and the prediction circuits for one group may be used. Then, one of the correlation circuits 4-1 to 4-n is selected based on the selection flag transmitted from the coding device, or the coefficients a and b are set and the original 6ch (Lf, C, Rf, Ls, Rs, Lfe) is restored, and multi-channel is downmixed by Eq. (2) to generate stereo 2ch data (L, R).

【0034】

Further, in the first embodiment described above, one type of correlation signals "1" to "6" is predictively coded, but the group of the signals "1" to "6" is used. A group of original signals (Lf, C, Rf, Ls, Rs, Lfe) may be predictively coded and the group with the higher compression ratio may be selected.

【0035】

[Effect of the invention]

As described above, according to the present invention, a subpacket containing compressed data and a data structure having a synchronization information unit including its sampling frequency and the number of quantization bits are formatted. Therefore, a multi-channel audio signal is used. The decoding efficiency on the reproduction side can be improved when the data is encoded with a variable compression rate.

[Simple explanation of drawings]

[Figure 1]

It is a block diagram which shows the 1st Embodiment of the voice coding apparatus to which this invention is applied, and the voice decoding apparatus corresponding thereto.

[Figure 2]

It is a block diagram which shows the coding part of FIG. 1 in detail.

[Fig. 3]

It is explanatory drawing which shows the bit stream encoded by the coding part of FIG. 1 and FIG.

[Fig. 4]

It is explanatory drawing which shows the format of a DVD pack.

[Fig. 5]

It is explanatory drawing which shows the format of the audio pack of a DVD.

[Fig. 6]

It is explanatory drawing which shows the format of the audio data area of FIG. 5 in detail.

[Fig. 7]

It is a block diagram which shows the decoding part of FIG. 1 in detail.

[Fig. 8]

It is a timing chart which shows the write / read timing of the input buffer of FIG.

[Fig. 9]

It is explanatory drawing which shows the compressed data amount for each access unit.

[Fig. 10]

It is explanatory drawing which shows the access unit and the presentation unit.

[Fig. 11]

It is a flowchart which shows the voice transmission method.

[Fig. 12]

It is a flowchart which shows the voice transmission method.

[Fig. 13]

It is a block diagram which shows the 3rd Embodiment of the voice coding apparatus to which this invention is applied, and the voice decoding apparatus corresponding thereto.

[Fig. 14]

It is a block diagram which shows the 4th Embodiment of the voice coding apparatus to which this invention is applied, and the voice decoding apparatus corresponding thereto.

[Fig. 15]

It is a block diagram which shows the 4th Embodiment of the voice coding apparatus to which this invention is applied.

[Fig. 16]

It is explanatory drawing of another Example corresponding to FIG.

[Fig. 17]

It is a block diagram which shows the audio decoding apparatus of 4th Embodiment.

[Explanation of symbols]

1'6ch mix & matrix circuit 13D1,13D2,15D1 ~ 15D4 Prediction circuit (Compresses with buffer / selector 14D1,14D2,16D1 ~ 16D4) 14D1,14D2,16D1 ~ 16D4 Buffer / selector 17 Selection signal / DTS generator (timing generator) 17c PTS generator (timing generator) 19 Formatting circuit (formatting means) 21 Deformatted circuit (separation means) 22 Unpacking circuit 22a input buffer 24D1,24D2,23D1 ~ 23D4 Prediction circuit (extension means) 100 Control unit 102 D / A converter 110 output buffer