JPS61150000A - Voice encoding system and apparatus - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

DETAILED DESCRIPTION OF THE INVENTION (Field of Industrial Application) The present invention relates to an encoding method and apparatus for encoding an audio signal with high quality at a low bit rate.

(Prior art) Audio signals are transmitted at a low transmission bit rate (for example, 4.8 kbp).
Vocoder (vO
cODER) is known. Regarding the principle of this method, for example, see M.R. Schrader (M.
, R,5 CROEDER)6 ``VOCODER8: ANALYSIS AND 5
YNTHESIS 0FSPEECH The moon and the paper (
PROC, IEEE, p, p, 720-734,
MAY, 1966; Reference January, etc. In addition, LPC VOCODER is known as a vocoder that uses linear predictive analysis, and its contents can be found, for example, in ``Linear Prediction Vocoder'' by J. D. MARKEL et al. Based Abonza Autocorrelation Method (ALINEARPR
EDICTION VOCODERBASED UPO
NTHE AUTOCORRELATION ME
A paper entitled THOD) J (IEEE 'I'RAN
S, A, S, S, P, , p, p, 12
4-134° APPIL, 1974; It is explained in detail in February, etc. The present invention is a VOCODE
It is an improved sound source part of R, and LPC VOCODE
Since it is closely related to R, hereafter LPCVOCODER
An outline will be explained focusing on the configuration of the synthesis section.

FIG. 4 is a block diagram showing the combining unit (receiving unit) of the LPC VOCODER (7) described in Document 2. The synthesis section includes a sound source generation section 500 and a synthesis filter circuit 510. The sound source generating section 500 includes an impulse generator 501, a noise generator 502, a voiced/unvoiced switching circuit 503, and a gain circuit 504. With VOCODER, the audio signal is divided into voiced and unvoiced signals every short period of time (20m 5ec).
If it is divided into species and is voiced, an impulse vocalizer 501
A pulse train having a time interval of pitch period Pd is generated from . On the other hand, if there is no voice, white noise is generated from the noise generator 502. Voiced/unvoiced control is performed by switching circuit 5.
It will be held in 03. A gain G is applied to the signal generated in this manner in a gain circuit 504, and the signal is outputted to a synthesis filter 510 as a sound source signal d(n). The synthesis filter 510 synthesizes and outputs the sound x(n) using the sound source signal -d(n) and the filter parameter Ki. Here, pitch period Pd, voiced/unvoiced switching signal (V/UV), gain G, filter parameter Ki
is calculated at predetermined time intervals on the analysis side (sending side) and transmitted to the receiving side.

(Problems to be Solved by the Invention) In the LPC VOCODER described above, the transmission information is the pitch period, voiced/unvoiced signal, gain, and filter parameters, and since the audio signal can be synthesized from these information, the transmission bit rate can be adjusted. (e.g. 4.8k
bps). However, with this conventional method, it is difficult to synthesize high-quality speech. Since this method divides speech into two extremes, voiced and unvoiced, it has the disadvantage that if an error in voiced/unvoiced discrimination occurs, it will cause a significant deterioration in quality.

In addition, when the sound source signal is voiced, the sound source is represented by one impulse per pitch, and since it does not include phase information, individuality is considerably lost, and the synthesized sound is a so-called F mechanical sound. It was J. In addition, the sound source could not be represented well at the transition between voiceless and voiced, resulting in deterioration. Furthermore, if the pitch period is determined by a deviation, there is a drawback that a large quality deterioration is caused. SUMMARY OF THE INVENTION An object of the present invention is to provide a high-efficiency speech encoding method and apparatus that can synthesize high-quality speech even at a low transmission bit rate with a relatively small amount of calculation.

(Means for Solving the Problems) The audio encoding method of the present invention inputs a discrete audio signal on the transmitting side, and extracts a spectrum parameter representing a short-time spectral envelope and a pitch parameter representing a pitch from the audio signal. is extracted and encoded, and based on the audio signal, the pitch parameter, and the spectrum parameter, a representative sound source pulse train for representing the audio signal is determined according to the pitch parameter. A predetermined number of parameters are determined and encoded for each interval, and the code representing the pitch parameter and the code representing the spectrum parameter are combined and output, and the reception (
In R1-, the combined code is input, and the code representing the pitch parameter, the code representing the spectrum parameter, and the code representing the representative sound source pulse train are separated and decoded, and the decoded pitch parameter and the code representing the representative sound source pulse train are separated and decoded. Based on the decoded sound source pulse train, the sound source pulse train is subjected to processing to give a temporally smooth change to restore the driving sound source signal, and the restored driving sound source signal and the decoded spectral parameter are also combined. The method is characterized in that the audio signal is synthesized with the audio signals.

Further, the encoding device of the present invention includes a parameter calculation circuit that extracts and encodes a spectral parameter representing a short-time spectral envelope and a pitch parameter representing a pitch from an input speech A word, and the audio signal and the pitch parameter. and the spectral parameters, a predetermined number of representative sound source pulse sequences for representing the audio signal are obtained for a representative section of the time sections obtained according to the pitch parameter. The present invention is characterized by comprising a drive signal calculation circuit for encoding, and a multiplexer circuit for outputting a combination of the output code of the parameter calculation circuit and the output code of the drive signal calculation circuit.

Furthermore, the decoding device of the present invention receives as input a code sequence in which a code representing a pitch parameter, a code representing a spectral parameter, and a code representing a representative sound source pulse train are combined, and the code representing the pitch parameter and the spectral parameter are combined. a demultiplexer circuit that separates and decodes a code representing the sound source pulse train and a code representing the sound source pulse train; A driving sound source signal restoration circuit that restores the driving sound source signal by performing a process of changing the driving sound source signal, and a synthesis filter circuit that synthesizes and outputs an audio signal based on the driving sound source signal and the decoded spectral parameter. It is characterized by

(Function) The present invention always generates a sound source signal by combining a plurality of pulses as shown in FIG. : (multi-pulse), and is characterized by not switching between pulses and noise between voiced and unvoiced parts. In FIG. 5, the horizontal axis represents discrete time and the vertical axis represents amplitude. Speech is synthesized from this multipulse and a synthesis filter. For example, analysis-by-synthesis (
ANALYSIS-by-3YNTHESIS; A
-b-8) is known, and its details are given by Benis Atal (B, S, ATA).
“Anni-21 Model of L.P.” by Mr. L) et al.
Sea Excitation 7 O Producing Natural Sounding Speech at Low Bit Rates (A NEW MODELOFLPCEXCI
TATIONFORPRODUCINGNATU-RA
LSOUNDINGSPEECHATLOWBITRA
TES) J (PROC, 1, C, A, S
, S, P, , p, p, 614-617゜1
982; Document 3), etc. In addition, a method based on correlation calculation is known as a method for obtaining high-speed calculations, and the details can be found in Japanese Patent Application No. 57-231606 (Reference 4).
), etc., so the explanation will be omitted here.

Furthermore, the present invention utilizes the periodicity of the audio signal, and on the transmitting side, a predetermined number of representative sound source pulses for representing the audio signal are calculated for one representative pitch section, The amplitude and position of this pulse, spectral parameters representing the spectral envelope characteristics of the audio signal, and pitch period information are transmitted.

On the receiving side, using representative pulse amplitude, position, and pitch cycle information, the previous and subsequent frames are processed by v)
A driving sound source signal is restored by changing this representative pulse smoothly over time, and this driving sound source signal is combined with a spectral parameter obtained by changing smoothly over time through interpolation processing. It is characterized by the fact that it synthesizes high-quality audio signals using .

(Example) Hereinafter, an example of the present invention will be described in detail with reference to the drawings. FIG. 1(a) is a block diagram showing an embodiment of the transmitting side of the high-efficiency speech encoding system according to the present invention.
Figure (b) is a block diagram showing an embodiment of the receiving side.

In FIG. 1(a), an audio signal X(n) is input,
A predetermined number of samples are stored in the buffer memory circuit 110. Next, the parameter calculation circuit 14
0 inputs a predetermined number of samples of the audio signal from the buffer memory circuit 110 and calculates parameters representing the spectral envelope of the audio signal. The parameters here are the same as the PARCOR coefficients.

The autocorrelation method is well known as a method for calculating the K parameter. For more information on this method, please refer to ``Quantization Properties Op Transmission Parameters in Linear Predictive Systems'' by JOHNMAKHOUL et al.
RTIES OF TRANS-MISSION PAR
AMETER8INLINEARPREDIC-TIV
E SYSTEMS moon and moon paper (IEEE TRA
N.S., A.

S, S, P, , p, p, 309-321.
1983; Reference 5), so the explanation will be omitted here.

Returning to FIG. 1(a), the K parameter Ki is output to the K parameter encoding circuit 160. K parameter encoding circuit 160 encodes Ki based on a predetermined number of quantization bits, and outputs code ffK+ to multiplexer 260. Further, the K parameter encoding circuit 160 uses the parameter decoded value Ki' obtained by decoding the eKI, converts it into a prediction coefficient value ai', and outputs it to the impulse response calculation circuit 170 and the weighting circuit 200. . Furthermore, the parameter decoded value is output to the interpolation circuit 255.

Pitch analysis circuit 130 calculates pitch period P using the output of buffer memory circuit 110. The calculation method for P is
For example, "REAL-TIME IMPLE-M
ENTATION OF TIME DOMAIN H
ARMONIC8CALING OF 5PEECH5
IGNALS) (IEEE TRANS,
A, S, S, P, , p, p, 258-
272.1983; Reference 6) and the like can be used.

The pitch encoding circuit 150 quantizes and encodes the pitch period Pd using a preset number of quantization bits, and outputs a code ed to the multiplexer 260. Also decryption,
The obtained Pd' is output to the drive signal calculation circuit 220 and the interpolation circuit 255.

The impulse response calculation circuit 170 inputs the prediction system value ai' from the K parameter encoding circuit 160, and calculates an impulse response representing the transfer function of the weighted synthesis filter.
Calculate (n). Here, to calculate (n), for example, see FIG.
The same method as the impulse response calculation circuit 210 mounted on the G can be used. The impulse response hw(n) is
Autocorrelation function calculation circuit 180 and cross-correlation function calculation circuit 2
10.

The autocorrelation function calculation circuit 180 receives the impulse response hw(n) from the impulse response calculation circuit 170 and calculates the autocorrelation function according to the following equation.

The autocorrelation function Rhh (·) is the drive signal calculation circuit 220
Output to.

Next, the subtracter 120 subtracts the output of the synthesis filter circuit 250 by one frame from the audio signal X(n) of the buffer memory circuit 110, and sends the result e(n) to the weighting circuit 200. The weighting circuit 200 inputs e(n) and also inputs the prediction coefficient ai' from the parameter encoding circuit 160, and outputs eV(n) obtained by weighting e(n). do. Here, to calculate ew(n), for example, the same method as that of the weight scale circuit 410 shown in FIG. 4(a) of Japanese Patent Application No. 59-042305 can be used.

The cross-correlation function calculation circuit 210 receives e and -(n) from the weighting circuit 200 and inputs e and -(n) from the impulse response calculation circuit 17.
Input the impulse response hw (n) from 0 and calculate the cross-correlation function according to the following formula. ).

9h, (ffil=Σe, (nl・h, (n−rn)
+ (1≦m≦M . . . (21wrl) The cross-correlation function % (m) is output to the drive signal calculation circuit 220.

Next, the drive signal calculation circuit 220 calculates a pulse sequence for a typical pitch section as a sound source signal representing an audio signal. This procedure is shown below. First, a frame is divided into subframes each having a pitch period P d'. For this division, it is necessary to know the pitch excitation position, which can be found by finding the Nokruste representing the sound source. In other words, ffl of the first pulse? From iE, the excitation position of the tonochi can be known. Here, the method shown in equation (21) of the specification of Japanese Patent Application No. 57-231606 can be used to calculate the pulse train (↓). , Figure 2 (
b) shows the state of subframes divided using the Noculus g1 found first and the position of this pulse. For the next process and for each subframe, a predetermined number of /<ruses are calculated. As a typical method for selecting a pitch section, for example, a method can be used in which a subframe including a pulse with a large absolute value is set as a representative pitch section, and a pulse included in this section is set as a representative pulse. Figure 2 (c) shows the representative and Tsuchinokurus obtained in this way.
Shown below. The amplitude and position of the pulse are output to encoder 230. Further, the subframe phase T and the subframe number of the representative pitch section (3 in FIG. 2(c)) are encoded with a predetermined number of bits as the representative pitch position, and are output to the multiplexer 260.

The encoding circuit 230 encodes the amplitude and position of the input pulse. Then, the amplitude and position sign of the pulse in the representative pitch section are output to the multiplexer.

In addition, the decoded values gi of the amplitude and position of all input pulses are
', mi' are output to the drive signal restoration circuit 240. Here, the pulse encoding method includes, for example, the patent application 1986-23.
The same method as the encoding circuit 250 described in the '1605 specification can be used. As other encoding methods, various methods such as a variable length encoding method or a differential encoding method can be considered.

The drive signal restoration circuit 240 generates a pulse for one frame using the decoded values of the amplitude and position of the pulse inputted from the encoding circuit 230, and outputs this as a drive sound source signal to the synthesis filter circuit 250. .

The interpolation circuit 255 inputs the pitch period Pd', the subframe phase T, and the representative pitch position. Then, the frame is divided into subframes for each pitch period Pd', and the K parameter is interpolated for each subframe. Here, the interpolation is linear interpolation, and parameter values for one frame past and one frame ahead are used. This interpolation is shown in FIG. In the figure, the parameter Ki,j of the i-th frame of the j-th frame is interpolated for each subframe using the value Ki,j-1 of one frame past and the value Ki,j+1 of one frame ahead. The decoded parameter values input from the K-parameter encoding circuit 160 are used for the representative and 7,000-thousand sections. The parameters obtained through interpolation in this manner are output to the synthesis filter circuit 250.

The synthesis filter 250 inputs the driving sound source signal and the interpolated parameters, and generates a response signal X(n
). Here, for this calculation, for example,
Synthetic fill circuit 32 described in -231605 specification
0 can be used.

The multiplexer circuit 260 uses the code 'Ki of the K parameter encoding circuit 160 and the code e of the pitch encoding circuit 150.
, and the code, subframe phase, and representative pitch position of the encoding circuit 230 are inputted, and these are combined and outputted from the transmission side output terminal 270. This concludes the explanation of the transmission side of the high-efficiency speech encoding system according to the present invention.

Next, the configuration of the receiving side of the audio encoding system according to the present invention will be explained with reference to FIG. 1(b).

The demultiplexer 290 converts the K parameter from the code input from the receiving side input terminal 280 into a hail code, and
A parameter decoding circuit 330, a pitch decoding circuit 320, a pulse decoding circuit 300, and a parameter decoding circuit 330, a pitch decoding circuit 320, a pulse decoding circuit 300, and separate codes representing the pitch period, pulse amplitude and position, and subframe phase and representative pitch position are separated. It is output to the drive signal restoration circuit 340.

K parameter decoding circuit 330 decodes the K parameter and outputs a decoded value Ki' to interpolation circuit 335.

The pitch decoding circuit 320 decodes the pitch period P d' and outputs it to the drive signal restoration circuit 340 and the interpolation circuit 335.

The pulse decoding circuit 300 has a pulse amplitude gi′ and a pulse position mi′.
is decoded and output to the drive signal restoration circuit 340.

The drive signal restoration circuit 340 first inputs the subframe phase, the code representing the representative pitch position, and the pitch period Pd', and divides the frame into subframes for each pitch period Pd'. Then, a pulse of amplitude g' is generated at position m' for the subframe section represented by the representative pitch position. Next, pulses are interpolated and determined for each subframe using the representative pitch pulse and representative pulses from one frame past and one frame ahead. In this way, pulses are generated for all subframes to restore the drive sound source signal, and the drive sound source signal for one frame is output to the synthesis filter circuit 350.

The interpolation circuit 335 operates in the same way as the interpolation circuit 255 on the transmitting side, interpolates the decoded parameters for each pitch period, and applies the interpolated parameters to the synthesis filter circuit 350.
Output to.

The synthesis filter circuit 350 inputs the driving sound source signal and the interpolated parameters, and sends the synthesis 7 filter circuit 2 on the transmitting side.
The same operation as 50 is performed to calculate the synthesized audio signal X(n) for one frame and output it from the receiving side output terminal 360.

This is according to the present invention. I will explain the receiving side of the high-efficiency speech encoding method.

As the pulse calculation method in the drive signal calculation circuit 220, various methods can be used in addition to the method described in this embodiment. For example, a method may be used in which the amplitude of previously determined pulses is adjusted each time one pulse is determined.

For details of this method, please refer to the paper entitled "Study of sound source pulse search method in multi-pulse driven speech coding method" by Otsuta Ono et al.
3; Document 7), so the explanation will be omitted here.

In addition, when determining pulses in the drive signal calculation circuit 220, pulses were determined for each subframe after dividing the frame into 7 subframes, but instead of dividing the frame into subframes, Alternatively, a predetermined number of pulses may be obtained. Furthermore, on the transmitting side of this embodiment, pulses were obtained for each subframe within this frame assuming that the value of the K parameter was constant within the frame (that is, the characteristics of the synthesis filter did not change within the frame). The pulse may be determined while smoothly changing the value of . Specifically, K
The parameter values are interpolated for each subframe using the parameter values of the previous and subsequent frames, and this value is converted into a prediction coefficient and output to the weighting circuit 200, impulse response calculation circuit 170, and synthesis filter circuit 250. , pulses are calculated using the cross-correlation and autocorrelation obtained by updating the coefficients for each subframe. It is better to do it like this,
Better speech can be synthesized. In addition, when interpolating parameter values for pulses and pulses, the transmitting side determines the position of a representative pitch section and transmits it, and interpolates in synchronization with the pitch period using this section as a reference. Either one or both of the parameters may be interpolated using a predetermined pitch interval (for example, a pitch interval near the center of the frame) as a reference. If both of these interpolation methods are used, it is not necessary to transmit the code representing the position of the representative pitch section, and the transmission bit rate can be reduced. Furthermore, although a subframe including a pulse with a large absolute value is selected as the representative pitch section, other methods may be used, such as selecting a subframe at the center of the frame. Furthermore, although the pitch period is constant when performing subframe division, this value may also be interpolated using the pitch period of the previous and subsequent frames. On the other hand, a method of capturing pulses and parameters without synchronizing them with the pitch period may also be considered.

In this case, the frame is cut into predetermined time intervals (for example, about 2.5 msec) and interpolation processing is performed for each interval. In this case, the subframe phase does not need to be transmitted. As the reference interval, a representative interval may be determined on the transmitting side or may be determined in advance (for example, near the center of the frame).In the latter case, the subframe phase and representative pitch position need not be transmitted. The bitrate can be further reduced.

Next, as an interpolation method for pulses, synthesis filter parameters, and pitch periods, methods other than linear interpolation may be considered.

For example, logarithmic interpolation and the like can be considered for pulse and pitch periods. In addition, when interpolating the parameters of the synthesis filter, in this example, the parameters are interpolated, but for example, it is possible to interpolate the prediction coefficients (however, in this case, it is necessary to check the stability of the filter), or to A method of interpolating an area function, a method of interpolating an autocorrelation function, etc. can also be used. These specific methods are:
``By B, S, ATAL et al.
SPEECHANALYSIS AND 5YNTH
ESISBY LINEAR PREDICTION O
F THE 5PEECHWAVE)
J, ACOUST, SOC, AM,, p, p
.

637-655.1971; Reference 8), etc., so the explanation will be omitted.

In this embodiment, the parameters are analyzed and the sound source pulse train is calculated assuming that the frame length is constant, but the frame length may be variable. In this case, the frame length can be shortened in the changing part of the audio, and the frame length can be made long in the constant part, so that the transmission bit rate can be reduced.

Furthermore, if the frame length is determined according to the pitch period (for example, an integral multiple of the pitch period), it is not necessary to send the subframe phase described in this embodiment, so the transmission bit rate can be further reduced. .

As another configuration method of the present invention, the drive signal restoration circuit 240, synthesis filter circuit 250, interpolation circuit 255, and subtraction circuit 120 in FIG. 1(a) may be omitted. In this case, there is no need to synthesize audio signals on the transmitting side, and the device configuration can be simplified.

Note that, as is well known in the field of digital signal processing, the autocorrelation function can also be calculated from the power spectrum. Further, the cross-correlation function can also be calculated from the cross-power spectrum.

Regarding these correspondence relationships, please refer to “DIGITAL 5IGNAL 2Q” by Niebu Oppenheim (A, V, OPPENHEIM) and others
Since it is explained in detail in Chapter 8 of the book published by J. PROCESSING (Reference 9), etc., the explanation will be omitted here.

(Effects of the Invention) According to the present invention, a sound source signal is represented by a combination of a plurality of pulse trains (multipulses), and the periodicity of the sound signal is used to express the sound signal at a representative pitch so as to represent the sound signal well. The pulses in the interval are determined and transmitted, and the driving sound source signal is restored and synthesized while changing this representative pitch pulse smoothly over time, so high-quality audio can be synthesized even at low transmission bit rates. There is an effect that it can be done.

[Brief explanation of the drawing]

FIGS. 1(a) and 1(b) are block diagrams representing an embodiment of the high-efficiency speech encoding method according to the present invention, and FIGS. 2(a)-(
c) is a diagram showing an example of processing contents in the drive signal calculation circuit 220, FIG. 3 is a diagram showing an example of processing in the interpolation circuit, FIG. 4 is a block diagram showing the configuration of the synthesis side of the conventional method, and FIG. FIG. 3 is a diagram showing an example of a sound source signal expressed as multi-pulses. In the figure, 110...Buffer memory circuit 120...Subtraction circuit 250, 350...Synthesis filter circuit 200...
... Weighting circuit 170 ... Impulse response calculation circuit 180 ...
... Autocorrelation function calculation circuit 210 ... Cross correlation function calculation circuit 220 ... Drive signal calculation circuit 240, 340 ... Drive signal restoration circuit 130 ...
... Pitch analysis circuit 140 ... K parameter calculation circuit 150 ...
- Pitch encoding circuit 160...K parameter encoding circuit 230...
... Encoding circuits 255, 335 ... Interpolation circuit 260 ... Multiplexer 290 ... Demultiplexer 300 ... Pulse decoding circuit 320 ... Pitch decoding circuit 330- . . - show parameter decoding circuits, respectively. E 2 Diagram (b) O 3 Mouth O 4 Diagram O 5 Draft procedure amendment (Koshiv) Dear Commissioner of the Japan Patent Office Kan 1, Indication of the case 1982 Patent Application No. 2724
35 No. 2, Name of the invention Audio encoding system and its device 3, Relationship with the amended case Applicant 5-33-1 Shiba, Minato-ku, Tokyo (423) NEC Corporation Representative Tadahiro Sekimoto 4; Agent, 61,3. Subject of amendment (1) [Claims column of Book B (2) Detailed explanation column of the invention in the specification & contents of amendment +1) The scope of claims will be amended as shown in the attached sheet. (2) "Spiktor" and 6 on page 47, line 9 of the specification
``Spectrum'' and ``spectrum''. (3) In the 17th line of page 7 of the specification, the phrase "symbol and t" is corrected to "symbol and the symbol representing the representative sound source pulse train." (4) “The driving sound source” on page 9, line 13 of the specification
The error is corrected to ``the reconstructed driving sound source''. (5) “Pasuru” in the first line of page 17 of Akira i4v2ig
It is corrected as ``stop t-r pulse''. (6) In the second line of page 18 of the specification, rhi, jJ Stop O is corrected to rKijJ. (7) On page 18, line 3 of the specification, “Ki, j
−I J and bru are corrected as trKij−IJ. (8) On page 18, line 3 of the specification, “Ki, published by j”
The text has been corrected to read "Published by Ki)." (9) ``Subframe phase'' in the 6th line of page 19 of the specification is corrected to ``subframe phase J. αl11 ``Method, etc.'' should be changed to ``method, etc.'' in the 1st line of page 23 of the specification. :r method, method of selecting a section that can reproduce good audio, etc." , 1/11N Agent Patent Attorney Susumu Uchihara Patent Claims L On the transmitting side, a discrete audio signal is input, and from the audio signal, a spectral parameter representing the short-time spectrum envelope t- and a pitch parameter representing the pitch are extracted. A representative section of the time section in which a representative sound source pulse train for representing the audio signal is determined according to the pitch parameter by encoding the audio signal, the pitch parameter, and the spectrum parameter. A predetermined number of pulses are determined and encoded, and a code representing the pitch parameter, a code representing the spectrum parameter t-, a code representing the representative sound source pulse train t'', and t are obtained.
- combine and output, and in the reception example manually input the combined codes, separate and signal the code representing the pitch parameter, the code representing the spectrum parameter, and the code representing the representative sound source pulse tally; Based on the decoded pitch parameter and the decoded sound source pulse train, the sound source pulse train is subjected to processing to give a zero smooth change over time to restore the driving sound source signal, and the restored driving sound source signal and the sound source pulse train are A speech encoding method characterized in that the speech signal is synthesized based on restored spectrum parameters. 2. A parameter calculation circuit that extracts and encodes a spectral parameter representing a short-time spectral envelope from an input 7tf voice signal, a pitch parameter and a pitch parameter, and a A predetermined number of representative sound source pulse trains representing the first pitch signal are smoothly determined and encoded for a representative interval of the time interval determined according to the pitch parameter. A speech encoding device characterized by comprising: a drive signal calculation circuit; and a multiplexer circuit that combines and outputs an output code of the parameter calculation circuit and an output code of the drive signal calculation circuit. 1. A code sequence in which a code representing a pitch parameter, a code representing a spectral parameter, and a code representing an fm pulse train are combined is input. a multiplexer circuit that separates and decodes the code, and a driving sound source that performs processing to give a temporally smooth change to the sound source pulse train based on the decoded extended KL7C hitch parameter and the decoded sound source pulse train. Signal f: a drive sound source signal restoration circuit to restore, and i
An audio decoding device characterized by a synthesis filter circuit that synthesizes and outputs an audio signal based on a drive sound source signal restored from the Ir period and the decoded spectral parameters. Agent: Susumu Uchihara, Patent Attorney: ゛,ノ

Claims (1)

[Claims] 1. On the transmitting side, a discrete audio signal is input, and a spectral parameter representing a short-time spectrum envelope and a pitch parameter representing a pitch are extracted and encoded from the audio signal, and A predetermined number of representative sound source pulse sequences for representing the audio signal based on the pitch parameter and the spectrum parameter are provided for a representative period of the time period obtained according to the pitch parameter. The code representing the pitch parameter and the code representing the spectral parameter are combined and output, and the receiving side inputs the combined code and generates a code representing the pitch parameter and a code representing the spectral parameter. and a code representing the representative sound source pulse train, and perform a process of applying a temporally smooth change to the sound source pulse train based on the decoded pitch parameter and the decoded sound source pulse train. 1. A sound encoding method comprising: restoring a drive sound source signal by applying a signal to the sound source, and synthesizing the sound signal based on the restored drive sound source signal and the restored spectral parameter. 2. A parameter calculation circuit that extracts and encodes a spectral parameter representing a short-time spectral envelope and a pitch parameter representing a pitch from an input audio signal; a drive signal calculation circuit that obtains and encodes a predetermined number of representative sound source pulse trains for representing a sound signal in a representative section of the time sections obtained according to the pitch parameter; and the parameter A speech encoding device comprising: a multiplexer circuit that combines and outputs the output code of the calculation circuit and the output code of the drive signal calculation circuit. 3. A code sequence in which a code representing a pitch parameter, a code representing a spectral parameter, and a code representing an excitation pulse train are combined is input, and a code representing the pitch parameter, a code representing the spectral parameter, and a code representing the excitation pulse train are input. a multiplexer circuit that separates and decodes the sound source pulse train; and a drive sound source signal that performs processing to give a temporally smooth change to the sound source pulse train based on the decoded pitch parameter and the decoded sound source pulse train. An audio decoding device comprising: a driving excitation signal restoration circuit for restoring; and a synthesis filter circuit for synthesizing and outputting an audio signal based on the driving excitation signal and the decoded spectrum parameter.