JPS63236415A - Voice encoding system - Google Patents

Abstract

PURPOSE:To realize the high quality of voice with a low bit rate, by changing the arrangement of a code bit at every block by comparing a scale value with a threshold value, and changing bit arrangement at every scale value. CONSTITUTION:A voice waveform SS is quantized at an A/D converter 2, and is stored in a buffer 3, and a difference is taken through a maximum value limitation circuit 5, and data is sent to a scale value setting part 7. The scale value setting part 7 finds the scale value, and a bit arrangement control part compares it with the threshold value, and a quasi-moment compression part 11 takes out a prescribed number of bits of the data from the buffer 6, and rounds off low-order bits. An optimization processing part 12 performs correction so as to set an error with an original signal out of transmission bits at the minimum level. After every completion of a processing for one block, expansion and decoding are performed, and the last value is returned to a register 9, and the difference of the forefront of the next block is calculated. Scale data and code data are inputted to a multiplexer 13 at every block, then, transmission data is formed.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a voice encoding system, and is applicable to, for example, voice mail, voice memo, and guidance response devices for various devices.

For example, when digitally processing an audio signal, such as transmitting an audio signal using a high-speed digital line or storing and synthesizing audio signals for a voice response device, this audio signal is converted into a digital signal by some method. need to be converted to .

Basically, audio signals have a frequency band of 0.3 to 3.4KH.
This is an analog signal of
n) encoding).

Then, to convert this digital signal back to the original audio signal,
The signal may be converted into an analog signal using a digital/analog converter with a sampling frequency of 8 Hz and a resolution of 8 bits, and then passed through a low-pass filter to shape the waveform. At this time, the resolution of the analog/digital converter and digital/analog converter (i.e., the bit [
) is higher, the quality of the reproduced audio is higher.

By the way, such a PCM encoded audio signal has 1
The bit rate per second (data rate: hereinafter referred to as bit rate) is 64 Kbps, and transmitting audio signals with such a high bit rate requires a very high-speed transmission path, and it is necessary to store such audio signals. This requires a huge amount of memory. Therefore, various proposals have been made to reduce the bit rate of audio signals.

One of them is a differential PCM encoding method that forms a difference between chronologically adjacent PCM codes. This differential PCM encoding method utilizes redundancy based on the correlation of audio waveforms, and since it is included in a limited range between adjacent samples, it is possible to reduce the number of bits per sample. can. CCIT is one of the adaptive differential PGM encoding methods that further advances this differential PCM encoding method.
Adaptive differential PC based on Y (International Telegraph and Telephone Advisory Committee) recommendations
The M method achieves a bit rate of 32 Kbps.

In addition to this, APC-AB (Adaptive Predic
tion Co-ding with Adaptiv
e Bit A11 location) method, or LSP (Line Spectrum) method using speech analysis and synthesis method.
umPair) method, etc.

However, such adaptive differential PCM method, APC
-The AB method and the LSP method have disadvantages in that the encoding and decoding processes are very complicated, and the equipment for implementing them is very expensive.

On the other hand, one of the high-quality PCM audio transmission systems for broadcasting satellites is the quasi-instantaneous companding system. This quasi-instantaneous companding method is
Divide M-encoded audio data into a predetermined number of blocks in time series, identify scale data representing the most significant digit corresponding to the maximum value of the signal absolute value in each block, and include the most significant digit. This formats data of a predetermined number of bits into code data, and the encoding process is relatively simple, and the number of bits of one sample can be easily reduced. However, such quasi-instantaneous companding method is not efficient enough.

Therefore, a method to improve the efficiency of this quasi-instantaneous companding method is to ``combine the differential PCM method and quasi-instantaneous companding,'' but in general, simply applying quasi-instantaneous companding to the differential PCM method does not allow compression. Missing bits in time cause transmission errors, and the errors accumulate in the integrator on the receiving side, making reception impossible.

Next, this point will be explained.

Let us now consider encoding an audio signal as shown in FIG. 4(a) using an encoding method in which quasi-instantaneous companding is applied to the differential PCM method. First, for differential PCM encoding, this audio signal is sampled at a sampling frequency of, for example, 8 KHz to form a difference value between samples. Here, the difference value between adjacent samples is expressed as signed 8-bit data, that is, 8-bit data in two's complement representation. The conditions for quasi-instantaneous companding are that one block consists of eight samples, and the transmission data is three bits per sample. Also, the scale data is 3 bits.

Here, assuming that the difference values for the eight samples #1 to x8 are obtained as shown in Figure 5 <a>, the one with the maximum absolute value of the difference value within the block is sample #1. Yes, therefore, the scale position PO at this time
8 becomes bit 4, which is the most significant digit of the bit pattern of this sample #l, and the value of scale position @pos is (100
),become.

As a result, the transmission bits (transmission data; that is, code data) of each sample are one digit higher than this scale position PoS, and are 3 bits of data from bit 5 (sign bit) representing the code value, that is, bit 5. ,4.
The data will be 3.

Therefore, in this block, first scale position P
Transmission data (code data) configured by consecutively arranging O8 and the transmission bits of samples #1 to #8 in sequence is as shown in FIG. 3(b).

When decoding such code data, first, one block of code data is decomposed into three bits each.

The scale position PO8 is identified from the value of the first 3 bits. When expanding the following 3-bit code data to 8-bit data, the MSB (most significant bit) of the code data is matched to the digit one higher than the scale position PO8, and each digit higher than the MSB is expanded. Set the value of the sign bit in the digit, and its LSB (least significant bit)
Set O in each lower digit than .

As a result, the data shown in FIG. 3(c) is obtained as decoded data. Comparing this data with the data before encoding, it is found that the decoded data lacks information (DC component) of lower digits than the transmission bits (see FIG. 4(b)). In other words, there are missing bits in the information.

When an audio signal is reproduced based on such decoded data, errors for missing bits accumulate and cause a negative DC shift, as shown by the dashed line in Fig. 4(C), The waveform becomes more downward sloping than the original waveform shown, and as a result, information cannot be transmitted properly.

One way to solve this problem is to use differential companding PCM (DC
-PCM) J (Takaran et al., Journal of the Institute of Electronics and Communication Engineers '84
/10 Vol. J67-B No. 10) has been proposed.

However, this method is effective when compressing differential data of about 15 bits to about 8 bits, and cannot be applied to a low bit rate encoding method that compresses differential data of 8 bits to about 3 bits.

In other words, in the case of such a low bit rate, the scale position may vary significantly between blocks, such as when the amplitude of the audio waveform changes significantly between blocks, so the accumulated error signal is The value may be larger than the valid data to be transmitted. In such a case, the data to be transmitted will be buried in the error signal, making it impossible to achieve proper data transmission.

In order to solve the above-mentioned disadvantages of the prior art, the applicant of the present invention first provided an audio encoding system that can reproduce high-quality audio at a low bit rate and through simple processing.

FIG. 6 is a diagram for explaining an example of a speech encoding device previously proposed by the applicant. This example applies quasi-instantaneous companding to the differential PCM encoding method, and The compressed difference data that remains in blocks during companding is sequentially decoded and compared with the original signal to correct each sample point so that the difference data has fewer errors within the number of compressed bits. In this example, the audio signal is sampled at a sampling frequency of 8 KHz, and the difference value between the samples is expressed as 8-bit data in two's complement representation, and 8 samples constitute one block for quasi-instantaneous companding. The transmission data is 3 bits per sample, and the scale data is 3 bits.

In the figure, an input audio signal SS is band-limited by a low-pass filter 1 and then applied to an analog/digital converter 2 where it is converted into an 8-bit digital signal DS. This analog/digital converter 2 performs linear quantization at a sampling frequency of 8K)lz.

The digital signal O8 is accumulated in a buffer 3 having a storage capacity of 8 samples forming one block.
The digital signal DSd stored in is applied to the plus input terminal of the adder/subtractor 4 for forming differential data.

The 9-bit difference data D output from this adder/subtractor 4
Ds is converted into 8-bit differential data DD by the maximum value limiting circuit 5. The reason why the maximum value of differential data is limited in this way is as follows.

In other words, when the differential data DDs is directly compressed quasi-instantaneously, a correspondingly large scale value is set for a block that suddenly contains large differential data, and therefore the data after quasi-instantaneous compression for other small differential data is Followability deteriorates. As a result, the restored audio signal gives an aurally jerky feel. Therefore, by limiting the maximum value of the difference data to a certain extent in this way, such auditory problems can be solved.

This difference data DD is applied to one input terminal of a buffer 6 having a storage capacity of 8 samples, a scale value setting section 7 for setting a scale value for quasi-instantaneous compression, and an adder 8.

The output of this adder 8 is added to a register 9, and the output of this register 9 is applied to the minus input terminal of the adder/subtractor 4,
It is added to the other input terminal and to a buffer 10 having a storage capacity of 8 samples.

In this way, an integrated value of the difference data DD limited to 8 bits by the maximum value limiting circuit 5 is formed by the adder 8, and this data is used as data of the immediately preceding sample for forming the difference data DDs. ing.

The data stored in the buffer 6 is applied to a quasi-instantaneous compression unit 11 which performs quasi-instantaneous compression encoding for each sample.

The scale value setting unit 7 identifies the one with the largest absolute value among the eight consecutive samples of the difference data DD output from the maximum value limiting circuit 5, determines the most significant digit of the bit pattern, and sets that bit. The position is output as 3-bit scale data DK. This scale data DK is stored in the quasi-instantaneous compression section 11. An optimization processing section 12 for converting compressed difference data DC output from the quasi-instantaneous compression section 11 into optimal data; one input terminal of a multiplexer 13 for shaping one block's worth of data into a predetermined signal format; It is also added to a quasi-instantaneous decompression unit 14 for decompressing the optimized compressed differential data.

Furthermore, the data stored in the buffer 10 is decoded data formed by integrating the difference data DD output from the register 9, and is treated as the original audio signal (hereinafter referred to as the original signal) to be encoded by the optimization processing unit. It has been added to 12.

The quasi-instantaneous compression unit 11 converts the 8-bit difference data DD added from the buffer 6 into scale data D added from the scale value setting unit 7 for each sample.
MS the 1 bit higher digit than the scale position where K is
The 3-bit data designated as B is extracted and outputted to the optimization processing section 12 as compressed difference data DC.

The optimization processing unit 12 sequentially decodes one block of compressed difference data DC added from the quasi-instantaneous compression unit 11 and compares it with the original signal added from the buffer 10.
The compressed difference data DC is corrected for each sample point so that it becomes difference data with few errors within the number of compressed bits, and this is added to the other input end of the multiplexer 13 and the quasi-instantaneous decompression unit 14 as optimized difference data DCo. - As shown in FIG. 7, the multiplexer 13 is constructed by arranging the scale data DK output from the scale value setting section 7 at the beginning, followed by sequentially arranging the optimized difference data DCo of each sample. The signal is formed as one block of encoded data DL and output to the next stage device (for example, a data transmission device or a data storage device, etc.).

The quasi-instantaneous decompression unit 14 uses the optimal 3-bit data output from the optimization processing unit 12 so that the MSB matches the digit one higher than the digit indicated by the scale data DK added from the scale value setting unit 7. difference data DC.

At the same time, the code data value of the optimized difference data DCo is placed in the higher digits, and 0 is placed in the lower digits.
is arranged to form 8-bit decoded data DE, and the decoded data DE is added to the integrating section 15.

The integrating section 15 integrates the decoded data DE output from the quasi-instantaneous decompression section 14, forms restored data SD when the encoded data DL is actually decoded and restored, and stores this restored data SD in a register. It is output to 9. register 9
Now, this restored data SD is fetched just before finishing the processing of one block and starting the processing of the next block.

As a result, the accumulation of errors due to missing bits, which is characteristic of quasi-instantaneous companding, can be eliminated when forming the first sample data of the next block, and as a result, more accurate encoded data DL can be formed. be able to.

In this way, in this embodiment, the optimization processing unit 12 corrects the compressed difference data DC so that it follows the original audio signal, and the restoration data SD formed by the integration unit 15 corrects the cumulative error within the block. Since it is reflected in the next block, low bit rate audio encoding processing using quasi-instantaneous companding can be realized with higher accuracy.

Next, the optimization process of compressed difference data executed by the optimization processing unit 12 will be explained.

Now, let us consider encoding a speech waveform as shown in FIG. 8(a). Sample # with reference to sample #0
1. #2. When the difference in #3 is formed using 8 bits each, the absolute value of the difference in this case is the largest in sample #1. Therefore, the 3-bit compressed difference data at this time is formed based on sample #1 as the 4th standard, and the scale value is the most significant digit position of the bit pattern.

Now, when the compressed difference data is made up of 3 bits, the data that can be expressed has a quantization width of 1 bit lower than the scale value, so each sample has a quantization width of $1. #2. The value #3 is replaced with data that can be expressed with this quantization width. For example, the compressed difference data of sample #1 is a value P12 (= (010), which is lower than the actual value P11 among the data that can be expressed. However, in this case, L
SB is 1 bit lower than the scale position, the same applies hereafter)
become.

By the way, among the data that can be expressed with this quantization width, P
The data corresponding to the value P13 (=(011)z), which is one value larger than I3, is closer to the actual pH value of sample #1. Therefore, if this value P13 is used as the compressed difference data of sample #1, the error in the audio signal (decoded value) when decoded can be reduced.

That is, the error in the decoded value at this time can be suppressed to 172, which is the quantization width of the compressed difference data, at most.

Similarly, sample #2. Considering #3, its decoded value is the value of the signal before encoding (value P21 in sample #2).
．． For sample #3, the compressed difference data closest to the value P31) may be selected.

That is, in this case, for sample #2, the value P2
The value P compared to the decoded value based on the value P22 smaller than 1
Since the decoded value based on the value P23 larger than 21 is closer to the value P21, the value P which is the decoded value of sample #l
13 and the value P23 (=(oto)i) is set as compressed difference data. Also, for sample #3, the value P
31 matches the possible decoded value of the compressed difference data, so the value P23 and the value P31, which are the decoded values of sample #2,
Set the difference (=(010),) with the compressed difference data.

In this way, compressed difference data with improved followability to the original audio signal can be created. An example of the optimization difference bit routine that is the process for this purpose is shown in Figure 9 (
Shown in a) and ('b).

First, compressed difference data (DC) is input from the quasi-instantaneous compression unit 11 (step 101), and the value of the compressed difference data d is
Determine whether the number of compression bits (in this case 3 bits) is greater than the maximum positive value MAX (= (011)) or smaller than the maximum negative value MIN (= (100) 2). Judgment 102, 103), Judgment 102
When the result is YES, the value M is added to the compressed difference data d.
Process 104) to substitute AX, and the result of judgment 103 is YE.
When the value becomes S, the value MIN is assigned to the compressed difference data d (processing 105).

Next, from the compressed difference data d, LSB(=(001)2
Processing 106, 107) to form a value dm that is smaller than the compressed difference data d and a value dp that is larger by the LSB than the compressed difference data d.
If a becomes smaller than the value MIN, determine whether to substitute the value WIN to the value d+u 108. Process 109), value ctp
becomes larger than the value MAX, the value dp is set to the value MAX.
Substitute X (judgment 110° process 111).

After forming the values dp and da+ in this way, 8-bit values dd, ddp, and ddm are calculated when the values d, dp, and da are expanded quasi-instantaneously based on the scale data DK (processing 112). And these values dd, ddp, ddm
and the decoded value dao of the data one sample before, respectively, to obtain local decoded values da and dap corresponding to the respective values d, dp, and dm.

A dam is formed (process 113). Note that local decoding processing is realized in processes 112 and 113, and for this purpose, the decoded value dao of the immediately preceding sample is stored.

Next, the value dai of the original signal corresponding to the sample is read from the buffer 10, and the absolute values Da, Dp, and D11 of the differences between the value dai of the original signal and each local decoded value da, dap, and dam are calculated. Then, it is checked whether the value dai of the original signal is larger than the local decoded value da (decision 115).

When the result of this judgment 115 is YES, it is checked whether the absolute value Da is larger than the absolute value op (judgment 1
16) When the result of this judgment 116 is YES, the value dP is substituted for the value d. 117) The process returns to process 106, and when the result of judgment 116 is NO, the value d is assigned as the optimized difference data DCo. Output (process 118)
.

Moreover, when the result of judgment 115 is No, the absolute value D
Check whether a is larger than the absolute value Dp (Judgment 11
9), when the result of this judgment 119 is YES,
Assign the value dll to the value d (process 120) and return to process 106.

When the result of judgment 119 is NO, processing 118 is executed.

That is, as shown on the left side of FIG. 8(b).

When the value dai of the original signal is closer to the decoded value dap than the decoded value da, the compressed difference data d is set to a value d larger by the LSB.
p and outputs the value as optimized difference data DCo. Conversely, when the value dai of the original signal is closer to the decoded value dam than the decoded value da, the compressed difference data d 3 LSB is updated to a smaller value dm, and the value is output as the optimized difference data DCo. Note that if the optimum value cannot be obtained in one process, this process is repeatedly executed.

In this way, the operation of adding and subtracting the LSB of the compressed bits to the compressed difference data obtained by quasi-instantaneous companding is repeated for each sample point, and the difference data is processed to reduce the error between the decoded value and the original signal. is being corrected.

Note that the output data of the buffer 3 before being input to the maximum value limiting circuit 5 can also be used as the data stored in the buffer 10 as the original signal.

Therefore, let us consider, for example, encoding the audio signal shown in Fig. 10(a) (same as Fig. 4(a)) by quasi-instantaneous compression using such an optimized differential bit routine6. , the difference values of the eight samples #1 to #8 of this audio signal are obtained as shown in FIG. 11(a), and therefore, the scale position PO8 at this time is the most significant digit of the bit pattern of sample 11. Bit 4 becomes bit 4, and the value of scale position PO8 becomes (100). As a result, the compressed difference data DC formed by the quasi-instantaneous compression section 11
are shown in FIG. 11(b) for each of samples #1 to #8.

This compressed difference data DC is corrected for each of the samples 11 to IJ8 into optimized difference data DCo as shown in FIG. 11(c) by the above-mentioned optimization difference bit routine.

As a result, in the audio decoding device described later, the value of each sample 1/L-18 of this optimized difference data DCo is expanded into 8-bit difference data as shown in FIG. 11(d). Based on the difference data, the audio signal is decoded as shown by the solid line in FIG. 10(e). Here, a reproduced audio signal when the above-mentioned compressed difference data DC is used as encoded data as it is is shown by a dashed line in FIG. 10(c).

As can be seen from these comparisons, the optimized difference data DC
The audio signal reproduced based on the method o clings to the unencoded audio signal shown by the broken line in FIG. 10(c), and the followability to the unencoded audio signal is greatly improved. .

FIG. 12 shows an example of an audio decoding device, which decodes the encoded data DL encoded by the aforementioned audio encoding device and outputs an audio signal. be.

In the figure, encoded data OL output from a preceding stage device (shown in the figure) such as a data receiving device or a data storage device is applied to a demultiplexer 21, and the first three bits of each block are converted to a scale value. The other code data (compressed difference data) is identified as SC and added to the scale value input terminal of the quasi-instantaneous decompression unit 22.
is applied to the code data input terminal of the quasi-instantaneous decompression section 22.

The quasi-instantaneous decompression unit 22 divides the added code data into 3 bits each, and adds the 3 bits to the bit position corresponding to the input scale data sc in the 8-bit data.
Arrange the bit data, place the contents of the sign bit in the upper digits of the code data, and place O in the lower digits to make 8.
Decompress it to bit data (see Figure 11(d)), and
The bit data is output to the integrating section 23.

The integrating section 23 integrates the sequentially input 8-bit data to form a signal value for each sample of the audio signal, and outputs this to the digital/analog converter 24.

The digital/analog converter 24 converts the received signal value into 8
It is converted into a corresponding analog signal (level signal) at a conversion frequency of KHz, and outputted to the low-pass filter 25. This analog signal is waveform-shaped by a low-pass filter 25 and then output as a reproduced audio signal to a subsequent device (for example, an audio output device).

The configuration of the audio decoding device for decoding encoded data in this way is extremely simple, and therefore, it is possible to realize this audio decoding device using, for example, a general-purpose 8-bit microprocessor. It is also possible to keep costs to a minimum.

By the way, in quasi-instantaneous companding, the scale position is set based on the bit pattern that maximizes the absolute value of the signal value within one block, so for example, if a sample suddenly takes a large value in that block, If there is a scale value corresponding to that large value, the scale value corresponding to that large value is set, so that the followability of the encoded data with respect to other small values becomes poor, which may cause an inconvenience that is not pleasing to the auditory sense.

Next, an example will be described in which such inconvenience can be resolved.

Now, in order to eliminate the above-mentioned inconvenience, it is sufficient to form encoded data with the least error in that block, and for this purpose, it is sufficient to set an appropriate scale position (value).

Therefore, when performing quasi-instantaneous compression based on the initially set scale position, when performing quasi-instantaneous compression based on the scale position when one digit higher than that scale position is set as the scale position, and when performing quasi-instantaneous compression based on the scale position, When the scale position is set to one digit lower than the scale position of , the error between the decoded value and the value of the audio signal before encoding is calculated for each sample when quasi-instantaneous compression is performed based on the scale position. Based on these calculations, an evaluation value at each scale position is formed, and based on this evaluation value, the best one is selected from among those at each scale position. Note that as this evaluation value, the sum of the absolute values of the differences between the original audio signal and the decoded value in each sample, the sum of the squares of the differences, or the square root (error power) can be used.

In other words, 1. For example, when there is an audio signal as shown in FIG. 13(C), each sample #1 to #8 of this audio signal
Assume that the scale value is determined based on sample #1 having the largest difference value, and the state is shown in the figure.

On the other hand, when the scale value is decreased by one, the quantization width of the sample value is decreased by one step, resulting in the state shown in (b) of the same figure, and when the scale value is increased by one, the quantization width of the sample value is decreased by one step. The width increases by one step, resulting in the state shown in FIG. 3(c).

Furthermore, when the optimization processing using the above-mentioned optimization difference bit routine is executed in each state, the sign bit in each sample becomes as shown in the following table.

Here, the scale value in the case of Fig. 13(a) is expressed as SC
-□ is the scale value in the case of (b) in the same figure, which is smaller by 1 than SC0, and SC□ is the scale value in the case of (b) in the same figure, which is larger by 1 than SC0.
The scale values for case c) are shown respectively.

When the scale value is changed in this way, the sum of the differences between the original audio signal and the decoded value, or the error power, naturally changes depending on the change in the audio signal in that block. The minimum value has the best followability for the audio signal in the block.

For example, according to experiments, statistically speaking, the scale value SCo
The error power of the scale value 5C-1 is the lowest in about 60% of the total number of blocks, and the lowest error power of the scale value 5C-1 is 3 out of the total blocks.
It is about 0%, and the error power of the scale value SC is the smallest at about 10% of the total number of blocks.

By selecting the scale value to be used for each block in this manner, the sound quality (audible) of the reproduced audio signal is improved.

FIG. 14 is a diagram showing an example of a speech encoding device that takes into account the above points. In the figure, the same parts as in FIG. will be added and its explanation will be omitted. This device also operates on a block-by-block basis with quasi-instantaneous compression.

In the figure, the scale value setting unit 7a determines the scale value in a block consisting of eight consecutive samples of the difference data DD output from the maximum value limiting circuit 5, and
Forming values one larger and one smaller than the scale value, scale data DK, , DKl, respectively.
Output as DK-1. These scale data DK
,,OK,,0K-1 are the scale values SC0,
They correspond to SC□ and 5C-1, respectively.

The scale data DKll is applied to one input terminal of the quasi-instantaneous compression unit 11°, the optimization processing unit 12I, the quasi-instantaneous decompression unit 14°, and the selector 31, and the scale data DKll is
Quasi-instantaneous compression section 11 Optimization processing section 12 Quasi-instantaneous decompression section 1
41 and the low input terminal of selector 31.

The scale data DK-1 is applied to the quasi-instantaneous compression section 11-1, the optimization processing section 12-1, the quasi-instantaneous expansion section 14-1, and other input terminals of the selector 31.

The quasi-instantaneous compression unit 11° forms compressed difference data DC from the output data of the buffer 6 based on the scale data DK, and the optimization processing unit 12 applies the above-described optimized difference bit routine to this compressed difference data DC. to form optimized difference data DCo, and this optimized difference data DC
o0 is applied to the quasi-instantaneous expansion section 14° and one input terminal of the selector 32. The quasi-instantaneous decompression unit 14° decompresses the input optimization difference data DCo into 8-bit data DE0 using the scale data DK as a reference, and this data DE is integrated and decoded by the integration unit 15°.
The decoded value SD is applied to one input terminal of the comparator 33 and one input terminal of the selector 34.

In addition, the quasi-instantaneous compression unit 111 converts the compressed difference data DC□ from the output data of the buffer 6 based on the scale data DK.
The optimization processing unit 12□ forms this compressed difference data D
The above-described optimized difference bit routine is applied to the C process to form optimized difference data DCo, and this optimized difference data DCo is applied to one input terminal of the quasi-instantaneous decompression unit 141 and the selector 32. The quasi-instantaneous expansion unit 14□ expands the input optimization difference data DCo into 8-bit data DE1 using the scale data DK1 as a reference, and this data DE1 is integrated and decoded by the integration unit 151, and its decoded value is SD1 is applied to one input terminal of the comparator 33 and one input terminal of the selector 34.

Similarly, the quasi-instantaneous compression unit 11-1 performs scale data DK-
0 from the output data of the buffer 6, and the optimization processing unit 12-0 applies the above-mentioned optimized difference bit routine to this compressed difference data DC-1 to create the optimized difference data. This optimized difference data DCo-0 is applied to the quasi-instantaneous decompression unit 14-1 and one input terminal of the selector 32.

The quasi-instantaneous decompression unit 14-1 receives the input optimization difference data D.
Co-, is expanded to 8-bit data DE-, using scale data DK-2 as a reference, and this data DE-
1 is integrated and decoded by the integrating section 15-8, and the decoded value SD-, is applied to one input terminal of the comparing section 33 and one input terminal of the selector 34.

Further, the difference data DD output from the maximum value limiting circuit 5
is accumulated by one block by the buffer 10a,
The output of this buffer 10a is sequentially integrated by an integrating section 15a to form an uncompressed audio signal (that is, an original signal), and the corresponding data SDa is sent to the optimization processing sections 12°, 12. ．． 12-1 and the comparison section 33.

In this way, the comparator 33 has data SDa corresponding to the original signal, scale value SC, (scale data DK, )
Optimized difference data DCo corresponding to. Decoded value SD when decoding , scale value SC (scale data DK
, ), and the decoded value SD1 when the optimized differential data DCo1 corresponding to
The decoded value SD-□ obtained when Co-□ is decoded is added to each sample.

The comparison unit 33 compares the data SDa and the decoded values SD, ,SD,
, 5D-1, their decoded values SD, ,SD
, , SD- from the data SDa for each sample, calculate the error power in one block for each decoded value SD, , 501, 5D-1, and take the minimum value among them. Discern things.

Then, when the error power for the decoded value S00 becomes the minimum, the selector 31 selects the scale data DK.
, and outputs it to the multiplexer 13, and the selector 32 selects the optimized difference data DCo, and outputs it to the multiplexer 13. Further, the code value SD is selected by the selector 34 and is used as data to be taken into the register 9.

Further, when the error power for the decoded value SD1 is the minimum, the selector 31 selects the scale data DK and outputs it to the multiplexer 13, and the selector 32 selects the optimized difference data DCo1 and outputs it to the multiplexer. Output to 13. Further, the code value S01 is selected by the selector 34, and is used as data to be taken into the register 9.

Similarly, when the error power for the decoded value 5D-1 is the minimum, the selector 31 selects -1 for the scale data D and outputs it to the multiplexer 13, and the selector 32 selects the optimized difference data DCo.
-1 is selected and output to the multiplexer 13.

Further, the code value 5D-4 is selected by the selector 34 and is used as data to be taken into the register 9.

Therefore, from the multiplexer 13, encoded data OL with the smallest error power in that block is output.
is output.

FIG. 14 shows an example of a comparison routine executed by the comparison section 33. First, data SDa and decoded values SDa, SD1
゜5D-2 is input for each sample (processing 20
1) Based on the input data, calculate the following equation to obtain the decoded value SD0. Error power PMSo, PMS0. corresponding to SD□, 5D-1, respectively. Each PMS-1 is calculated (process 202).

Here, k=o, 1. -1 and j represent sample numbers within the block, 5Daj represents data at each sample within the block, and 5Dkj represents the code value SDk at each sample within the block.

And which error power RMS, , RMS, ,
Distinguishing whether RMS-1 is the smallest 203, 20
4,205), minimum error power RMSo, RMS,
, Scale data DK corresponding to RMS-1, ,DK
, ,DK-, and the optimized difference data DCo, .

Select DCol and DCo-0 (process 206
, 207, 208).

In addition, in this comparison routine, each decoded value S
In addition to the above-mentioned error power, the evaluation value based on the error between D, , SDl, 5D-1 and the data SDa may be, for example, the data SDa and the decoded value SD0 . The sum of the differences between SDi, SD-, and the like can be used.

The above-mentioned audio compression encoding method applies quasi-instantaneous companding to differential PCM audio encoding and corrects the transmitted bits to compress a natural audio signal. The allocation of transmission bit rates is always constant, and the only way to increase compression efficiency is to uniformly lower the transmission bits or lower the sampling frequency, and it is impossible to increase the compression rate as is.

■--Lid This invention was made in view of the above-mentioned circumstances.
In particular, the purpose of this invention is to provide an audio encoding system that encodes and decodes natural audio using low bit rate and simple processing.

In order to achieve the above object, the present invention divides an input audio signal into blocks each having a predetermined number of samples, converts the values at the sample points into feature quantities for each block, and
In the audio encoding method, the maximum value is found, the highest value is identified as scale data, data of a predetermined number of bits is formatted into code data based on that digit, and each block is transmitted based on the scale data. Changing the bit allocation of transmission bits per PCM
A difference value between adjacent samples of the encoded audio data and the previous sample is formed, and the difference value is divided into a predetermined number of blocks in time series, and the maximum absolute value of the difference value in each block is calculated. The scale data representing the most significant digit corresponding to the value is identified, and the predetermined scale data including the digit indicated by the scale data is The number of bits is formatted into code data to form a plurality of code data sequences, and the error between the decoded value corresponding to the individual code data constituting each code data sequence and the audio data corresponding to the code data is Each of the plurality of code data sequences is corrected by correcting the code data so as to have the smallest error power, and among the corrected code data sequences, the code data sequence in which the error power within the block is the smallest is used as the code of the block. In an audio encoding method that uses the decoded value of the last sample of the code data series selected in the block as a reference when calculating the difference value of the first sample of the subsequent block, the scale value is used as the threshold. Based on the embodiments of the present invention, the method is characterized in that the code bit allocation for each block is changed by comparison, or the bit allocation for code data for each block is changed for each scale value. explain.

FIG. 1 is a block diagram for explaining one embodiment of the present invention. This embodiment improves the conventional technique shown in FIG. Since the bit allocation control section 16 that controls the allocation of compression bits based on the value is added, the parts in the figure that have the same effect as the circuit shown in FIG. The same reference numerals will be used to omit the detailed explanation.

In normal quasi-instantaneous companding, scale data and several samples of encoded data are transmitted as one block, but in that case, the number of bits of encoded data is set in advance and is the same for all blocks. It is. Therefore, in the present invention, compression efficiency is further improved by changing the number of bits of code data depending on the value of scale data.

To explain based on FIG. 1, the audio waveform SS is LP-F
1, the band is compressed to 1/2 or less of the sampling frequency of the A/D converter, and then quantized by the A/D converter 2. This A/D converted data DS is stored in a buffer 3.

One block of data is passed from the buffer 3 to the maximum value limiting circuit 5, a difference is taken, and the data is sent to the scale value setting section 7. The scale value setting section 7 takes the absolute value of each difference data, and calculates the maximum effective digit, that is, the scale value, from the maximum value. Based on the scale value, the bit allocation control unit 16 compares it with a predetermined threshold (TS), and if the scale value is larger than the threshold, m bits,
If it is smaller, a control signal is sent to the quasi-instantaneous compression unit 11 to set it to n bits.

The quasi-instantaneous compression unit 11 extracts a predetermined number of bits from the upper part of the data from the buffer 6, which is determined by a threshold value, and discards the lower bits. In the optimization processing section, transmission bits (m
Correction is performed so that the error from the original signal is minimized within the bits or n bits).

- When block processing is completed, the block is expanded by the logic of the quasi-instantaneous expansion unit 14 and locally decoded by the integration unit 15, so that the last value is returned to the register 9 and the difference at the beginning of the next block is calculated. Scale data and code data are input to a multiplexer for each block to form transmission data.

Below, we will specifically consider the case of compressing 12-bit data to 4-bit data, and form one block with 8 samples.
If the scale value is larger than the threshold value, 4 bits will be allocated, and if it is smaller, 3 bits will be allocated.

In this case, 4 bits are required as scale data, so if the scale data is larger than the threshold, 4+4X
8 = 36 bits, or in the case of a small number, 4 + 3 x 8 = 28 bits, which will be transmitted per block. Therefore, 6
(=(36-28)) bits/block can be saved.

In this case, the bit allocation is changed by comparing the scale value with the threshold value, but a predetermined bit allocation may be performed for each scale.

Moreover, although differential data is used as code data within a block in the above, this encoded PCM code data may be used as is.

FIG. 2 is a diagram for explaining an example of a decoding method for decoding the encoded data DL shown in FIG.
is a demultiplexer, 42 is a bit control logic, 4
3 is an expansion unit, 44 is a register, 45 is a digital-to-analog (D/A) converter, and 46 is a low-pass filter (L
At PF), the transmission signal DL is input to the demultiplexer 41, and the leading scale data is first taken out.

The scale data is sent to the bit control logic 42, the number of bits of the code data that follows is determined, and the number of bits is sent to the demultiplexer 41 to cut out the code data that follows the scale data. The extracted code is expanded based on the scale value and sequentially added to previous data to perform decoding.

FIG. 3 is a block diagram for explaining another embodiment of the present invention, and this embodiment is an improvement on the prior art shown in FIG. 14.

In the figure, the parts that have the same effect as in Figure 14 have the first
The same reference numerals as in the case of FIG. 4 are given, and detailed explanation thereof will be omitted. As shown in the figure, this embodiment is provided with a bit allocation control logic section 35, which has a reference scale value determined by the scale value setting section 7a, and this value is determined by the bit allocation control logic section 35. This determines the assignment of code bits. In this case, the bit allocation control unit 35 determines the allocated bits by using a threshold value based on the value of the scale data or for each scale value.

Explaining based on FIG. 3, the audio waveform SS is band-compressed to 1/2 or less of the sampling frequency of the A/D converter 2 by LPFI, and then quantized by the A/D converter. This A
The /D converted data DS is stored in the buffer 3, one block of data from the buffer 3 is passed through the maximum value limiting circuit 5, a difference is taken, and the data is sent to the scale value setting section 7a. The scale value setting section 7a takes the absolute value of each difference data and calculates the effective maximum digit, that is, the scale value from the maximum value. Based on the scale value, the bit allocation control unit 3
In step 5, a comparison is made with a predetermined threshold (TS), and if the scale value is larger than the predetermined threshold, a control signal is issued to the quasi-instantaneous compression unit to allocate m bits, and if it is smaller, to allocate n bits; , and the scale value determined by the scale value setting section 7a.

Three scale values, ie, a value obtained by subtracting the scale value by 1 and a value obtained by adding 1 to the scale value, are set, and each code data is shaped into the previously determined number of allocated bits using each scale value.

In the optimization processing section, the code data is corrected so that the error from the original signal is minimized within the allocated number of bits.The code data corresponding to the three scale values is decoded by the integrator section, and the code data is decoded by the comparison section. Select the scale value and its code data with the smallest error from the original signal. The scale value and code data are input to multiplexer 13 and transmitted according to a certain new formula. In this case, bit allocation at sample points within each block based on the scale value may be performed by setting a predetermined number of bits for each scale value without using a threshold.

As is clear from the above description, according to the present invention, it is possible to realize a high quality audio encoding system at a low bit rate.

[Brief explanation of drawings]

FIG. 1 is a block diagram for explaining an example of encoding processing according to the present invention, FIG. 2 is a block diagram for explaining an example of decoding processing, and FIG. 3 is a block diagram for explaining an example of encoding processing according to the present invention. FIG. 4 is a block diagram for explaining another embodiment of the encoding process according to the present invention, and FIG. 5 is an explanatory diagram of the prior art, and FIGS.
is a signal arrangement diagram for explaining the encoding/decoding situation of the prior art, FIGS. 6 to 15 are diagrams for explaining the prior art of the present invention, and FIG. 6 is a diagram of a speech encoding device. Figure 7 is a signal arrangement diagram illustrating the signal format of encoded data, Figure 8 (a) is a waveform diagram to explain the optimization process, and Figure 8 (b) is the optimization process. Signal arrangement diagrams for explaining the action of the differential bit routine, Fig. 9(a), (
b) is a flowchart showing an example of the optimization difference bit routine; FIG. 10 is an explanatory diagram of the effect of the optimization process;
Figures 1 (a) to (d) are signal allocation diagrams showing the state of optimization, Figure 12 is a block diagram showing an example of an audio decoding device, and Figures 13 (a) to (c) are FIG. 14 is a block diagram showing a speech coding device according to another embodiment, and FIG. 15 is a flowchart showing an example of a comparison routine. 1.25...Low pass filter, 2...Analog/
Digital converter, 3.6.10, 1011, 10.
．． 10-1. IQaJOb. 10c... Buffer, 4... Adder/subtractor, 5... Maximum value limiting circuit, 7.7a... Scale value setting section, 8...
...Adder. 9... register, 11,11. ．． 111,11-□
...Quasi-instantaneous compression section. 12.12°, 12□, 12-1... optimization processing section,
13...Multiplexer, 14, 14°, 141
．． 14-, , 22... Quasi-instantaneous extension part. 15, 15°, 151.15-□, 15a, 23...
- Integration section, 16...Bit allocation control section, 31, 32
, 34...Selector. 33... Comparison unit, 35... Bit allocation control logic. Patent applicant Ricoh Co., Ltd. Figure 5 Figure 77 LX slo $+ $2 Mi(b
) Figure 9 Figure 9 (b) Figure 10 Figure 11 Figure 12 Figure 13 (a) (b) Figure 13 (C) Figure 15 Procedure Kokeguchi Orthography (Method) 1. Indication of the case 1985 Patent Application No. 70855 2, Name of the invention Audio encoding method 3, Person making the amendment Relationship to the case Patent applicant Ota Ri Nakamagome Address 1-3-6 Nakamagome, Ota-ku, Tokyo Name (Name) (674) Ricoh Co., Ltd. Representative Hama 1)
Hiro 4, Agent address: 1-2-7 Furocho, Naka-ku, Yokohama 231
Seat Train Yokohama 807

Claims (6)

[Claims] (1) Divide the input audio signal into blocks of a predetermined number of samples, convert the values at the sample points for each block into feature quantities, find the maximum value, and use the top digit as scale data. In the audio encoding method that identifies the digit, formats the data with a predetermined number of bits into coded data based on the digit, and transmits it, the assignment of the number of transmission bits for each block is changed based on the scale data. Characteristic voice encoding method. (2) When allocating transmission bits for each block based on scale data, a predetermined threshold TS is set, and a large bit allocation is applied to blocks with scale data larger than this threshold TS, and a large bit allocation is applied to blocks with scale data larger than this threshold TS. 2. The audio encoding method according to claim 1, wherein a small bit is allocated to each block. (3) When allocating transmission bits to each block based on scale data, the bit allocation to each block is performed after setting the bit allocation corresponding to each value of the scale value in advance. Claim No. (
The audio encoding method described in section 1). (4) Using the difference value of the PCM encoded audio data as the feature quantity for each block, the error between the data obtained by decoding the encoded data shaped from the scale data and the original audio is reduced. According to claim (1), when the first value of the following block is calculated, the decoded value of the last sample of the block is used as a reference. Audio encoding method. (5) The audio encoding method according to claim (2), characterized in that PCM encoded audio data is used as the feature amount. (6) Form a difference value between adjacent samples of the PCM encoded audio data and the previous sample, divide the difference value into a predetermined number of blocks in time series, and calculate the difference value in each block. Identify scale data representing the most significant digit corresponding to the maximum absolute value of A predetermined number of bits, including digits, are formatted into code data to form a plurality of code data sequences, and the decoded values corresponding to individual code data constituting each code data sequence and the above-mentioned audio corresponding to the code data are generated. Correcting each of the plurality of code data series by correcting the code data so that the error with the data is minimized,
Among these corrected code data sequences, the code data sequence with the smallest error power within the block is selected as the code data sequence of the block, and when calculating the difference value of the first sample of the subsequent block, the code data sequence with the smallest error power within the block is selected. In a speech encoding method that uses the decoded value of the last sample of the selected coded data sequence as a reference, by comparing the scale value with a threshold value, the allocation of code bits for each block can be changed or the allocation of code bits for each block can be changed based on the scale value. A speech encoding method characterized by changing bit allocation of encoded data.