JPH06259096A - Audio encoding device - Google Patents

Abstract

PURPOSE:To provide an audio encoding device capable of securing relatively good tone quality even when the compression rate is large in a device compressing digital audio data. CONSTITUTION:This device is provided with a means 50 judging whether an input audio waveform to an audio encoding part is regarded as the same between adjacent analysis frames or not and an encoding auxiliary information storage part 52 minutely storing the auxiliary information of pitch information, a pitch gain, a linear predictive coefficient and an average amplitude of a pseudo residual signal, etc., and is constituted so that, when the waveform is regarded as the same, the signal showing the fact and the signal quantizing only the residual signal, are sent to an encoding part 45, and no auxiliary information is quantized and sent to the encoding part. Then, the device is constituted so that the quantization number of bit to be allocated to the auxiliary information is added to the quantization number of bit of the residual signal. At this time, the auxiliary information stored in the encoding auxiliary information storage part 52 is used for the auxiliary information used in the adaptive prediction parts 34, 37 and the quantization number of bit adaptation part 51 of the audio encoding part.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a voice coder which utilizes an analog line or a digital line, predictively codes a voice signal input through the line, and compresses a voice information amount.

[0002]

2. Description of the Related Art Conventionally, as a high-efficiency coding method for speech,
For example, there is an APC-AB system, etc., but with the development of various voice services in recent years, it has been applied to a voice mail device that stores and transmits voice data as encoded information.
It is in the practical stage. A case will be described below in which a conventional voice encoding device connected to an analog telephone line is performing high-efficiency voice encoding by the APC-AB system.

FIG. 5 is a block diagram of a conventional speech coding unit. Reference numeral 1 is a line interface unit connected to an analog line for controlling the line and separating input / output, and 2 is an A for converting an analog signal into a digital signal. / D converter, 3 is a band division filter that divides an input signal into a plurality of frequency bands, 4 is a linear prediction coefficient extraction unit that extracts a linear prediction coefficient that constitutes a short time prediction filter, and 5 is the short time prediction A predictor based on proximity correlation that performs linear prediction of a speech signal using a filter, 6 is a pitch detection unit that detects the pitch of the speech signal waveform, and 7 is a magnitude of the correlation of the speech signal with respect to the pitch period detected by the pitch detection unit 6. A pitch gain derivation unit for deriving a pitch gain indicating a pitch, and a pitch gain derived by the pitch gain derivation unit 7 for forming a long-term prediction filter Predictor based on pitch correlation for performing linear prediction of the signal, 9 an adaptive prediction unit consisting predictor 8 which based on predictor 5 and pitch correlation based on proximity correlation, 1
0 is a residual signal quantizer that quantizes a residual signal that is the difference between the output signal of the band division filter 3 and the prediction signal output from the adaptive prediction unit 9, and 11 is a pseudo residual corresponding to the residual signal. An inverse filter unit that derives the difference signal, 12 is a partial interval setting unit that equally divides the pitch period detected by the pitch detecting unit 6 into a plurality of intervals, and 13 is a pseudo residual signal derived by the inverse filter unit 11. An average amplitude calculation unit that calculates an average amplitude for each of the partial intervals, and a quantization bit that determines the number of quantization bits of the residual signal for each of the partial intervals according to the average amplitude calculated by the average amplitude calculation unit 13. A number adaptation unit, 15 is an auxiliary information quantizer that quantizes auxiliary information such as pitch information, pitch gain, linear prediction coefficient, and average amplitude, and 16 is a residual signal quantized by the residual signal quantizer 10. And assistance An encoding unit that encodes the auxiliary information quantized by the information quantizer 15 and outputs the encoded data as code data, a multiplexing unit 17 that multiplexes the code data output from the encoding unit 16 in each band, and 18 Multiplexer 1
This is a voice buffer for storing the digital signal output from 7, and the above constitutes the voice encoding unit.

Next, the voice decoding unit will be described with reference to FIG. Reference numeral 19 denotes a code sequence separation unit that separates the coded data from the audio buffer 18 for each band, and 20 denotes the code data that has been separated for each band by the code sequence separation unit 19 and the code data of the residual signal and the pitch. The information sequence separation unit 21 separates the information and the code data of the auxiliary information such as the pitch gain, the linear prediction coefficient, and the average amplitude. Reference numeral 21 represents the code data of the auxiliary information separated by the information sequence separation unit 20 as the original auxiliary information. A side information decoder for restoration, 22 is a quantized bit for each partial section of the residual signal, based on the average amplitude information restored by the side information decoder 21, in the same manner as the above speech coding unit. The decoding bit number adaptation unit for calculating the number, 23 is a residual signal decoder for restoring the coded data of the residual signal to the original residual signal, and 24 is the auxiliary information of the pitch gain and the linear prediction coefficient. The made prediction filter, prediction synthesis unit for predicting an audio signal from the output signal of the residual signal decoder 23, the band synthesis filter 25 for fusing the output signals from the prediction composer unit 24 of each band 26
Is a D / A converter that converts a digital signal into an analog signal, and 27 is a line interface unit that is connected to an analog line and controls the line and separates input and output. The above constitutes a speech decoding unit.

The operation of the speech coding apparatus configured as above will be described below. First, in FIG. 5, an analog signal such as voice input to the line interface unit 1 through a telephone line or the like is converted into a digital signal by the A / D converter 2. This digital signal will be divided and processed in short time intervals, that is, every analysis frame (for example, 16 msec). The audio signal (0 to 4 kHz) digitized by the A / D converter 2 is, for example, a QMF (Quadrattu).
A band division filter 3 composed of a re-mirror filter or the like divides into a plurality of bands, for example, three frequency bands of band A (0 to 1 kHz), band B (1 to 2 kHz) and band C (2 to 4 kHz). To be done.

The band A will be described below. The linear prediction coefficient extraction unit 4 extracts the linear prediction coefficient from the original signal of the band A (output signal from the band division filter 3) to form, for example, an even-order fourth-order linear prediction filter. Linear prediction is performed in the predictor 5 based on the proximity correlation using this linear prediction filter. The predictor 5 based on the close correlation performs prediction using sample values in a relatively close range, for example, data of four samples past from the present time. The pitch detector 6 detects the pitch of the original signal in the band A, and the pitch gain deriving unit 7 derives the pitch gain for the detected pitch period. The predictor 8 based on the pitch correlation performs prediction between pitch periods, that is, pitch prediction for predicting the current speech signal from the data of one pitch period past by the prediction filter configured by the derived pitch gain. The signals predicted by the predictor 5 based on the proximity correlation and the predictor 8 based on the pitch correlation are combined and output from the adaptive prediction unit 9. The prediction signal output from the adaptive prediction unit 9 is input to the residual signal quantizer 10 and quantized as a residual signal by taking the difference from the original signal in the band A.

Here, the number of quantization bits of the residual signal quantizer 10 is adapted according to the average amplitude of the pseudo residual signal.
Therefore, it is necessary to calculate the average amplitude of the pseudo residual signal before operating the adaptive prediction unit 9. First, the inverse filter unit 1
In 1, the pseudo residual signal is obtained from the original signal of band A.
The inside of the inverse filter unit 11 is a predictor 5 based on proximity correlation.
And the predictor 8 based on the pitch correlation, and is the same as the adaptive predictor 9. Next, this pseudo residual signal is input to the partial section setting unit 12. The partial section is obtained by equally dividing the pitch cycle Tp into a plurality of sections, for example, four sections, as shown in FIG. 7, and is set so that each section periodically repeats. The first partial section is detected based on the amplitude periodicity of the pseudo residual signal, and the position Td of the partial section is designated by the time length from the beginning of the frame to the beginning of the first partial section. The average amplitude calculation unit 13 calculates the average amplitude of the pseudo residual signal for each partial section, and the quantization bit number adaptation unit 14 allocates the quantization bit number of the residual signal from the average amplitude for each partial section. Calculate. The number of quantization bits R _{ij in the} i band j section is given by the following equation.

R _ij = R _i (R) + F (log ₂ U _ij ) (1) i = A, B, C j = 1, ... 4, where R _i is the average bit rate of the i band and The number of bits R (R = R) allocated to the quantization of the residual signal of the band
Proportional to R _A + R _B + R _C ). R is a constant uniquely determined by the information transmission rate. F is a function proportional to the logarithm of the average amplitude U _ij of the i-band j-section pseudo residual signal. FIG. 8 shows an example of the number of quantization bits for each partial section of the band A determined in this way. The residual signal quantized by the residual signal quantizer 10 according to the number of quantization bits is sent to the encoding unit 16 and encoded, and at the same time, is also fed back to the adaptive prediction unit 9.

On the other hand, auxiliary information such as pitch information, pitch gain, linear prediction coefficient, and average amplitude of pseudo residual signal is also quantized by the auxiliary information quantizer 15 and then encoded by the encoder 16. Become. The number of quantization bits at this time is uniquely determined by the information transmission rate.
The code data output from the coding unit 16 of each band is multiplexed by the multiplexing unit 17 and stored in the voice buffer 18 as code data.

The above-described signal processing process is the same for the band B and the band C, but the long-term prediction filter and the sub-interval setting in the predictor 8 based on the pitch correlation use those configured in the band A. The pitch detection unit 6, the pitch gain derivation unit 7, and the partial section setting unit 12 do not operate. The above operation is the description of the voice encoding unit.

Next, referring to FIG. 6, the audio buffer 18
An example of reconstructing the audio signal from is shown. First voice buffer 1
The code data from No. 8 is separated into the code data series for each band by the code series separation unit 19. Further, this coded data is separated into coded data of the residual signal and coded data of auxiliary information such as pitch information, pitch gain, linear prediction coefficient, and average amplitude in the information series separation unit 20 of each band.

The signal processing of band A will be described below. The coded data of the auxiliary information is the auxiliary information decoder 21.
And is restored to the original auxiliary information such as pitch information, pitch gain, linear prediction coefficient and average amplitude. Thereafter, the decoding bit number adaptation unit 22 calculates the decoding bit number R _ij of the residual signal from the restored average amplitude U _ij based on the equation (1), and based on this, calculates the residual signal The coded data is restored to the original residual signal in the residual signal decoder 23. The prediction synthesizing unit 24 is composed of auxiliary information such as the decoded pitch information, pitch gain, and linear prediction coefficient, and is composed of a predictor 28 based on proximity correlation and a predictor 29 based on pitch correlation as shown in FIG. Which is the adaptive prediction unit 9 of the speech coding unit.
Is the same as the predictor 5 based on the proximity correlation and the predictor 8 based on the patch correlation. This prediction synthesis unit 24
Then, the decoded residual signal is input and the audio signal for each band is reproduced. The above-described signal processing process is exactly the same for band B and band C.

The band synthesizing filter 25 fuses the audio signals reproduced for each band, converts the digital signal into an analog signal by the D / A converter 26, and outputs it as a reproduced signal to the line interface unit 27. .

[0014]

In the conventional speech coding apparatus, when performing the speech coding processing as described above, the auxiliary information per analysis frame and the number of quantization bits assigned to the residual signal are used for information transmission. It is uniquely determined in proportion to the speed. Therefore, if the information transmission rate is low,
The number of quantization bits assigned to the auxiliary information and the residual signal also decreases. When the number of quantization bits assigned to the quantization of the auxiliary information decreases, the prediction accuracy of the adaptive prediction unit 9 decreases and the quantization error of the residual signal increases. In addition, since the number of quantization bits assigned to the residual signal quantizer 10 is small, there is a problem that the quantization error is further increased and the sound quality is significantly deteriorated.

When the information transmission rate is low, the number of quantization bits assigned to the quantization of the auxiliary information is increased,
It is also known that even if the prediction accuracy in the adaptive prediction unit 9 is improved, there is a limit to the sound quality improvement.

Therefore, it is also known that providing the means for increasing the number of quantization bits assigned to the residual signal quantizer 10 to reduce the quantization error leads to the improvement of sound quality.

The present invention solves the above problem, and an object of the present invention is to provide a speech coding apparatus capable of ensuring a relatively good sound quality even when the information transmission rate is low.

[0018]

In order to achieve the above object, the speech coding apparatus of the present invention is such that the speech waveform continuously changes between analysis frames, and therefore the auxiliary information representing the characteristics thereof is constantly maintained. Focusing on the change, in the configuration of the speech coding unit of the conventional example, a means for judging whether or not the input speech waveforms to the speech coding apparatus are considered to be the same between adjacent analysis frames, and a speech coding unit. And a decoding auxiliary information storage unit and a decoding auxiliary information storage unit for temporarily storing auxiliary information in the speech decoding unit, and when they are regarded as the same, only a signal indicating that and a residual signal It is configured so that it is quantized and transmitted to the encoding unit, and the auxiliary information is not quantized and is not transmitted to the encoding unit. Then, the number of quantization bits to be assigned to the auxiliary information is added to the number of quantization bits to be assigned to the residual signal quantizer, and the number of quantization bits is increased. At this time, the auxiliary information stored in the encoded auxiliary information storage unit is used as the auxiliary information used in the adaptive prediction unit, the quantization bit number adaptation unit, and the like. Also,
In the speech decoding unit, the auxiliary information stored in the decoding auxiliary information storage unit is used as the auxiliary information used in the prediction synthesis unit and the decoding bit number adaptation unit.

[0019]

According to the present invention, the number of quantized bits to be assigned to the auxiliary information is added to the number R of bits to be assigned to the quantization of the residual signal in the entire band, and then the calculation of the equation (1) is performed by the above configuration. By doing so, a larger number of quantization bits than before is assigned to the residual signal quantizer, the residual signal can be quantized precisely, the quantization error becomes smaller, and the voice waveform can be faithfully reproduced. Can be played. Further, it may be considered that the auxiliary information constantly changes between adjacent analysis frames as described above. Therefore, as long as it is determined that the same voice waveform is used, there is little error even if the same auxiliary information is used. From the above, the error can be suppressed as a whole.

[0020]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram of a voice encoding unit according to an embodiment of the present invention. Reference numeral 30 is a line interface unit connected to an analog line for controlling the line and separating input / output, and 31 is for converting an analog signal into a digital signal. A / D converter, 32 is a band division filter that divides an input signal into a plurality of frequency bands, 33 is a linear prediction coefficient extraction unit that extracts a linear prediction coefficient that constitutes a short time prediction filter, and 34 is the short Predictor based on proximity correlation for linear prediction of speech signal using temporal prediction filter, 35
Is a pitch detector for detecting the pitch of the voice signal waveform, 36
Is a pitch gain deriving unit for deriving a pitch gain indicating the magnitude of the correlation of the voice signal with respect to the pitch period detected by the pitch detecting unit, and 37 is a long-term prediction filter using the pitch gain derived by the pitch gain deriving unit 36. A predictor based on pitch correlation for linear prediction of a speech signal, an adaptive predictor 38 including a predictor 34 based on proximity correlation and a predictor 37 based on pitch correlation, and 39 is an output signal of the band division filter 32. And a residual signal quantizer that quantizes a residual signal that is a difference between the prediction signal output from the adaptive prediction unit 38, and an inverse filter unit 40 that derives a pseudo residual signal corresponding to the residual signal, Reference numeral 41 is a partial section setting section for equally dividing the pitch cycle detected by the pitch detection section 35 into a plurality of sections, and 42 is a pseudo residual derived from the inverse filter section 40. The average amplitude calculating section for calculating an average amplitude of the signal for each said subinterval, 45 residual signal quantizer 39
An encoding unit that encodes the residual signal quantized in the above and the auxiliary information quantized in the auxiliary information quantizer 44 in the below-described quantization bit number processing unit 43, and outputs the encoded data as coded data. In the above, a multiplexing unit for multiplexing the code data output from the encoding unit 45 and a voice buffer 47 for storing the digital signal output from the multiplexing unit 46 are the same as the conventional example. .

The quantization bit number processing unit 43 is shown in FIG. In FIG. 2, reference numeral 50 is a speech waveform discrimination unit that determines whether the speech waveform input to the speech encoding unit can be regarded as the same as the speech waveform of the immediately preceding analysis frame, and 51 is the average calculated by the average amplitude calculation unit 42. A quantization bit number adaptation unit that determines the quantization bit number of the residual signal for each of the partial intervals according to the amplitude, 44 is pitch information, pitch gain,
Auxiliary information quantizer for quantizing auxiliary information such as linear prediction coefficient and average amplitude, 52 is the auxiliary information quantizer 44
It is a coding auxiliary information storage unit for temporarily storing the auxiliary information quantized in. The band B and the band C do not have the voice waveform discriminating unit 50. The above constitutes the speech encoding unit.

Next, FIG. 3 is a block diagram of a voice decoding unit in one embodiment of the present invention, and 60 is a voice buffer 47.
A code sequence separation unit for separating the coded data from each of the bands for each band, 61 denotes the code data separated for each band by the code sequence separation unit 60, the code data of the residual signal, the pitch information, the pitch gain, the An information sequence separation unit that separates the linear prediction coefficient and the code data of the auxiliary information such as the average amplitude, 6
Reference numeral 2 denotes a side information decoder that restores the coded data of the side information separated by the information sequence separation unit 61 to the side information that is the original, and 63 is based on the average amplitude information that is restored by the side information decoder 62. , A decoding bit number adaptation unit that calculates the number of quantization bits for each partial section of the residual signal in the same manner as the speech coding unit described above, and 64 represents the residual signal based on the code data of the residual signal. A residual signal decoder for recovering
Is a prediction synthesis unit that predicts a speech signal from the output signal of the residual signal decoder 64 by a prediction filter configured by pitch information and auxiliary information of a linear prediction coefficient, and 66 is a prediction synthesis unit 65 for each band. A band synthesis filter that fuses output signals, 67 is a D / A converter that converts a digital signal into an analog signal, and 68 is a line interface unit that is connected to an analog line and controls the line and separates input and output. This is exactly the same as the conventional example. Reference numeral 69 denotes a decoding auxiliary information storage unit that temporarily stores the auxiliary information separated by the information sequence separation unit 61, and the above constitutes a speech decoding unit.

The operation of the speech coding apparatus configured as above will be described below. First, an analog signal such as voice input to the line interface unit 30 through a telephone line or the like is converted into a digital signal by the A / D converter 31. This digital signal will then be used for a short time interval or analysis frame (eg 16 mse).
It will be divided and processed for each c). A / D
Audio signal digitized by converter 31 (0-4
kHz) is, for example, QMF (Quadrature M).
A plurality of bands, for example, band A (0
˜1 kHz), band B (1-2 kHz) and band C
It is divided into three frequency bands (2 to 4 kHz).

The band A will be described below. First, the pitch detector 35 detects the pitch of the original signal in the band A. This pitch information is sent to the quantized bit number processing unit 43, and the voice waveform discrimination unit 50 in the quantized bit number processing unit 43 sets this value within ± 5% of the pitch detected in the previous analysis frame. Determine if it is included. ± 5
If it is included in the range of%, it is determined that the voice waveform of the current analysis frame is the same as the voice waveform of the previous analysis frame,
Otherwise, the speech waveform of the current analysis frame is not the same as the speech waveform of the immediately preceding analysis frame.

The case where the speech waveform of the current analysis frame is considered not to be the same as the speech waveform of the previous analysis frame ((a) of FIG. 2) will be described. First, a signal indicating that the voice waveform determination unit 50 has determined that the waveforms are not the same is sent to the band B and the band C. In this state, the linear prediction coefficient is extracted by the linear prediction coefficient extraction unit 33 from the original signal of the band A (output signal from the band division filter 32) to form a linear prediction filter. Linear prediction is performed in the predictor 34 based on the close correlation using this linear prediction filter. The pitch gain deriving unit 36 derives a pitch gain based on the pitch information detected by the pitch detecting unit 35. The predictor 37 based on the pitch correlation performs prediction between pitch periods, that is, pitch prediction that predicts the current speech signal from the data of one pitch period past by the prediction filter configured by the derived pitch gain. The signals predicted by the predictor 34 based on the proximity correlation and the predictor 37 based on the pitch correlation are combined and output from the adaptive prediction unit 38. The difference between the prediction signal output from the adaptive prediction unit 38 and the original signal in band A is calculated, and the difference is input to the residual signal quantizer 39 and quantized as a residual signal. The above is the same operation as the conventional example.

The number of quantization bits of the residual signal quantizer 39 is adapted according to the average amplitude of the pseudo residual signal,
This derivation method is also the same as the conventional example, and the inverse filter unit 4
0, the average amplitude of the pseudo residual signal for each partial section is calculated by the average amplitude calculating section 42 through the partial section setting section 41. Thereafter, in the quantization bit number adaptation unit 51, the operation of assigning the quantization bit number for each partial section from the average amplitude is performed by the equation (1). After that, in the same way as the conventional example,
According to the quantization bit number determined by the equation (1) in the quantization bit number adaptation unit 51, the residual signal quantizer 39
The residual signal quantized in (1) is sent to the coding unit 45 and coded, and at the same time, fed back to the adaptive prediction unit 38. On the other hand, auxiliary information such as pitch information, pitch gain, linear prediction coefficient, and average amplitude of pseudo residual signal is also quantized by the auxiliary information quantizer 44 in the same manner as in the conventional example, and then encoded by the encoding unit 45. Be converted. At the same time, the quantized auxiliary information is stored in the encoded auxiliary information storage unit 5
Stored in 2. That is, the auxiliary information in the current analysis frame is always newly rewritten and stored in the encoded auxiliary information storage unit 52. The number of quantization bits at this time is uniquely determined by the information transmission rate. The code data output from the coding unit 45 of each band is multiplexed by the multiplexing unit 46 and stored in the voice buffer 47 as code data.

The above-mentioned signal processing process is the same for the band B and the band C, but the predictor 37 based on the pitch correlation is used.
For setting the long-term prediction filter and partial interval in
The pitch detection unit 3
5, pitch gain derivation unit 36 and partial section setting unit 41
Is not performed. Further, since it is determined in the band A whether the input voice waveform is considered to be the same as the voice waveform of the immediately preceding analysis frame, the operation of the voice waveform discriminating unit 50 is not performed in the band B and the band C. . The above operation is the description of the voice encoding unit.

Next, the case where the audio signal is restored from the audio buffer 47 will be described with reference to FIG. First, the code data from the voice buffer 47 is converted into the code sequence separation unit 60.
Is separated into code data series for each band. Further, the coded data is classified into coded data of the residual signal and coded data of auxiliary information such as pitch information, pitch gain, linear prediction coefficient, and average amplitude in the information series separation unit 61 of each band. At this time, the coded data of the auxiliary information is always stored in the decoding auxiliary information storage unit 69.
That is, the coded data of the auxiliary information in the current analysis frame is always newly rewritten and stored in the decoding auxiliary information storage unit 69. The subsequent operation is exactly the same as the conventional example.

Next, the case where the speech waveform discriminating unit 50 in the quantizing bit number processing unit 43 determines that the speech waveform is the same as the speech waveform of the immediately preceding analysis frame ((b) of FIG. 2) will be described. First, the signal indicating that the quantization bit number processing unit 50 has determined that the waveforms are the same is transmitted to the band B and the band C. In this state, the linear prediction filter used in the predictor 34 based on the proximity correlation is configured by the information of the linear prediction coefficient stored in the coding auxiliary information storage unit 52 and performs linear prediction. The prediction filter used in the predictor 37 based on the pitch correlation is composed of the pitch gain information stored in the coding auxiliary information storage unit 52, and the prediction between pitch cycles, that is, the data of one pitch cycle past, Pitch prediction for predicting a voice signal is performed. The signals predicted by the predictor 34 based on the proximity correlation and the predictor 37 based on the pitch correlation are combined and output from the adaptive prediction unit 38. The difference between the prediction signal output from the adaptive prediction unit 38 and the original signal in band A is calculated, and the difference is input to the residual signal quantizer 39 and quantized. At this time, the auxiliary information is not quantized, and only the signal indicating that they have the same voice waveform is encoded.
Sent to.

The number of quantization bits of the residual signal quantizer 39 is adapted according to the average amplitude of the pseudo residual signal,
This is also based on the average amplitude information stored in the encoding auxiliary information storage unit 52, and in the quantization bit number adaptation unit 51, the operation of assigning the quantization bit number to each partial section is performed by the equation (2). . That is, the number of quantization bits R _ij of the residual signal is given by the following equation.

R _ij = R _j (R + Rs) + F (log ₂ U _ij ) (2) Here, R _i is the average bit rate of the i band, and when the residual signal shown in the conventional example is quantized. It is proportional to the sum of the total number of quantized bits R to be allocated and the number of bits Rs to be allocated to the auxiliary information in the above embodiment. F is a function proportional to the logarithm of the average amplitude U _ij of the pseudo residual signal in the i band j section and is similar to the conventional example. FIG. 4 shows an example of the number of quantization bits for each partial section of the band A determined in this way. As compared with FIG. 8 of the conventional example, the number of quantization bits in each section increases, so that the quantization error can be significantly suppressed.

Thereafter, in the same manner as in the conventional example, the residual signal quantized by the residual signal quantizer 39 according to the quantization bit number determined by the equation (2) in the quantization bit number adaptation unit 51 is At the same time as being sent to the encoding unit 45 and being encoded, it is also fed back to the adaptive prediction unit 38.

As described above, the information to be encoded is the signal indicating the same voice waveform and the residual signal, and this processing is performed as long as the same voice waveform continues. This process is the same for the band B and the band C.

Next, the case where the audio signal is restored from the audio buffer 47 will be described with reference to FIG. First, the code data from the voice buffer 47 is converted into the code sequence separation unit 60.
Is separated into code data series for each band. Further, this coded data is separated into coded data of the residual signal and coded data of a signal indicating that it has the same speech waveform as the immediately preceding analysis frame in the information sequence separation unit 61 of each band. When a signal indicating that it has the same speech waveform as the immediately preceding analysis frame is detected, it is stored in the decoding auxiliary information storage unit 69 as auxiliary information such as pitch information, pitch gain, linear prediction coefficient, and average amplitude. The code data is read, and the auxiliary information is restored accordingly. That is, while it is determined that the audio waveform is the same as the previous analysis frame, the same auxiliary information is always restored. The subsequent operation is exactly the same as the conventional example.

[0035]

As described above, according to the present invention, the speech waveform control unit of the quantization bit number processing unit detects whether the input speech waveform is the same speech waveform as the immediately preceding analysis frame, and detects the same speech waveform. In this case, by using the number of bits that should be assigned to the auxiliary information in the residual signal quantizer, it is possible to greatly suppress the quantization error of the residual signal, and significantly improve the sound quality deterioration when the information transmission rate is low. However, the signal-to-quantization noise ratio (S / N ratio) can be improved.

[Brief description of drawings]

FIG. 1 is a block diagram showing a speech coding unit of a speech coding apparatus according to an embodiment of the present invention.

FIG. 2 is a block diagram showing a configuration in a quantization bit number processing unit of a speech encoding unit in the same speech encoding device.

FIG. 3 is a block diagram showing a speech decoding unit of a speech coding apparatus according to an embodiment of the present invention.

FIG. 4 is a diagram for explaining the number of quantization bits for each partial section by the speech encoding unit in the same speech encoding apparatus.

FIG. 5 is a block diagram showing a speech encoding unit of a conventional speech encoding device.

FIG. 6 is a block diagram showing a speech decoding unit of a conventional speech encoding device.

FIG. 7 is a diagram illustrating a partial interval setting method by a speech encoding unit of a conventional speech encoding device.

FIG. 8 is a diagram for explaining the number of quantization bits for each partial section by the speech encoding unit of the conventional speech encoding device.

FIG. 9 is a specific block diagram of a prediction synthesis unit in a speech decoding unit of a conventional speech encoding device.

[Explanation of symbols]

 30 line interface section 31 A / D converter 32 band division filter 33 linear prediction coefficient extraction section 34 predictor based on proximity correlation 35 pitch detection section 36 pitch gain derivation section 37 predictor based on pitch correlation 38 adaptive prediction section 39 residual error Signal quantizer 40 Inverse filter unit 41 Partial interval setting unit 42 Average amplitude calculation unit 43 Quantization bit number processing unit 44 Auxiliary information quantizer 45 Encoding unit 46 Multiplexing unit 47 Speech buffer 50 Speech waveform discriminating unit 51 Quantization bit Number adaptation unit 52 coding auxiliary information storage unit 60 code sequence separation unit 61 information sequence separation unit 62 auxiliary information decoder 63 decoding bit number adaptation unit 64 residual signal decoder 65 prediction synthesis unit 66 band synthesis filter 67 D / A converter 68 Line interface unit 69 Decoding auxiliary information storage unit

─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁵ Identification code Office reference number FI Technical display location H03M 7/30 8522-5J H04B 14/06 D 4101-5K

Claims (1)

[Claims] 1. A prediction means for adaptively predicting an input speech signal processed for each analysis frame by a digital filter, and a residual signal which is a difference between the speech signal predicted by the prediction means and the input speech signal. Means for generating a pseudo residual signal, means for setting a partial interval by equally dividing the pitch period of the input audio signal, and means for detecting an average amplitude of the pseudo residual signal for each partial interval. And means for quantizing the residual signal, means for adaptively allocating the number of quantized bits at the time of quantizing the residual signal in correspondence with the average amplitude for each of the partial intervals, and the predicting means. Information and means for quantizing the information used in the means for adaptively allocating the quantization bit number of the residual signal, and the input audio signal is regarded as the same as the input audio signal of the immediately preceding analysis frame. If it is determined that the information is the same as the information used in the prediction means and the information used in the means for adaptively allocating the quantization bit number of the residual signal, And a means for effectively utilizing the number of quantization bits to be assigned to the quantization by means for adaptively assigning the number of quantization bits of the residual signal.