JP3218679B2 - High efficiency coding method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a high-efficiency encoding method which encodes data on a frequency axis obtained by dividing an input audio signal into blocks and converting the signal into a frequency axis.

[0002]

2. Description of the Related Art There are known various encoding methods for compressing a signal using a statistical property of a sound signal in a time domain or a frequency domain and a characteristic of human perception. This encoding method is roughly classified into encoding in the time domain, encoding in the frequency domain, and analysis-synthesis encoding.

[0003] As an example of high-efficiency encoding of a speech signal or the like, M
BE (Multiband Excitation) coding, SBE (Singleband Excitation) coding, Harmonics (Harmonic) coding, SBC
(Sub-band Coding: band division coding), LPC (Linear
Predictive Coding) or DCT
(Discrete Cosine Transform), MDCT (Modified DC
T), FFT (Fast Fourier Transform) and the like.

[0004] In the MBE coding and the like, a band (band) is divided into a plurality of bands with respect to the sound in one block (frame), and voiced / unvoiced sound is determined for each band, thereby improving sound quality. Is recognized.

[0005]

By the way, the above MBE
In encoding and the like, phase information of the amplitude of a voice spectrum is transmitted separately or phase information is predicted from a pitch and a past phase on a synthesis side (decoding side) without transmitting it to reduce the amount of transmission information. . In the prediction of the phase information on the synthesis side (hereinafter referred to as phase prediction), the initial phase of the frame is fixed to 0 or π / 2, and the cosine wave is synthesized according to the pitch transmitted from the analysis side. Phase addition has been performed. Then, on the synthesis side, a random factor is added to the phase to give a natural feeling according to the content rate of the unvoiced sound band determined by the voiced sound / unvoiced sound determination on the encoding side. However, a strong vowel part with a low pitch voice such as a male voice has an unnatural synthesized sound.

SUMMARY OF THE INVENTION The present invention has been made in view of such circumstances, and has as its object to provide a high-efficiency encoding method capable of obtaining high-quality speech without any unnatural feeling on the synthesis side. You .

[0007]

A high-efficiency encoding method according to the present invention comprises the steps of dividing an input audio signal into blocks and converting the signal into a frequency axis to obtain data on the frequency axis; Dividing the upper data into a plurality of bands, detecting a block initial phase for each of the divided bands, determining whether voiced or unvoiced for each of the bands, and determining a phase for each of the bands. A step of calculating a predicted value, a step of calculating an error between the actually measured phase and the predicted phase for each band, and an area width of a voiced sound area and an unvoiced sound area on a frequency axis according to the magnitude of the error. The above object is attained by having a changing step.

[0008] Further, by transmitting information indicating the magnitude of the error calculated from the measured phase and the predicted phase to the combining side, the combining side may correct the phase prediction according to the information.

Here, this high efficiency coding method does not transmit the phase information of the spectrum amplitude to the receiving side (combining side). The synthesis side predicts the block end phase based on the past and current pitches and the past phases transmitted from the encoding side.

[0010]

According to the present invention, a high-quality synthesized sound is obtained on the synthesis side by changing the widths of the voiced sound area and the unvoiced sound area on the frequency axis according to the magnitude of the error between the actually measured phase and the predicted phase for each band. be able to. Further, by transmitting the information indicating the magnitude of the error to the synthesis side, it is possible to correct the phase by phase prediction on the synthesis side to obtain a synthesized sound of high sound quality.

[0011]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the high efficiency coding method according to the present invention will be described below. The high-efficiency coding method includes converting a signal for each block on the frequency axis, dividing the signal into a plurality of bands, and applying V (V) for each band, as in the case of MBE (Multiband Excitation) described later. Voiced sound) or UV
(Unvoiced sound) can be used.

That is, as a general high-efficiency encoding method to which the present invention is applied, an audio signal is divided into blocks every predetermined number of samples (for example, 256 samples), and FFT is performed.
At the same time as transforming into spectrum data on the frequency axis by orthogonal transformation such as, the pitch of the voice in the block is extracted,
The spectrum on the frequency axis is band-divided at intervals according to this pitch, and V (voiced sound) / U
V (unvoiced sound) is determined. This V / UV discrimination information is encoded and transmitted together with the spectrum amplitude data.

Here, for example, in a speech synthesis analysis system such as an MBE vocoder, the phase of a speech signal is predicted on the synthesis side, and an error in this phase prediction (hereinafter, a phase prediction error amount) is calculated on the encoding side ( Analysis side), and changing the width of the voiced and unvoiced regions on the frequency axis in accordance with the phase prediction error amount, or transmitting the information indicating the magnitude of the error to the synthesis side, and synthesizing. In this embodiment, a high-quality synthesized sound is obtained by correcting the phase by phase prediction on the side.

Hereinafter, a high-efficiency encoding method according to the present invention will be described.
A specific example applied to an MBE (Multiband Excitation) vocoder, which is one type of a speech signal synthesis analysis coding device (so-called vocoder), will be described with reference to the drawings. This MBE vocoder is a DW Gr
iffin and JS Lim, "Multiband Excitation Vocoder,"
IEEE Trans.Acoustics, Speech, and Signal Processin
g, vol. 36, No. 8, pp. 1223-1235, Aug. 1988, and a conventional PARCOR (PARtial au).
In a vocoder or the like, a voiced section and an unvoiced section are switched for each block or frame at the time of speech modeling, whereas the MBE vocoder uses the same time (the same block or frame). Modeling is based on the assumption that a voiced section (Voiced) section and an unvoiced section (Unvoiced) section exist in the frequency domain of (in).

FIG. 1 is a block diagram showing an overall schematic configuration of an embodiment in which the present invention is applied to the MBE vocoder. In FIG. 1, an audio signal is supplied to an input terminal 11, and this input audio signal is
F (High Pass Filter)
Removal of at least a low-frequency component (200 Hz or less) for removal of a so-called DC (direct current) offset or band limitation (for example, limited to 200 to 3400 Hz) is performed. A signal obtained through this filter 12 is output to a pitch extraction unit 13
And the windowing processing unit 14. In the pitch extracting unit 13, the input audio signal data is
The block is divided into units (for example, N = 256) (or cutout is performed using a rectangular window), and pitch extraction is performed on the audio signal in this block. Such a cut-out block (256 samples) is moved in the time axis direction at a frame interval of L samples (for example, L = 160) as shown in FIG. 2A, and the overlap between the blocks is N-. There are L samples (for example, 96 samples). The windowing processing unit 14 applies a predetermined window function, for example, a Hamming window, to one block N samples, and sequentially moves the windowed blocks in the time axis direction at intervals of one frame L samples.

If such a windowing process is expressed by a mathematical formula, _xw (k, q) = x (q) w (kL-q) (1) In the equation (1), k is a block number,
q represents a time index (sample number) of the data. The q-th data x (q) of the input signal before processing is windowed by a window function w (kL-q) of the k-th block. This shows that data x _w (k, q) can be obtained. The window function w _r (r) in the case of a rectangular window as shown in FIG. 2A in the pitch extraction unit 13 is given by: w _r (r) = 1 0 ≦ r <N (2) = 0 r <0, N ≦ r Further, the window function w _h (r) in the case of the Hamming window as shown in FIG. 2B in the windowing processing unit 14 is represented by w _h (r) = 0.54−0.46 cos (2πr / (N-1)) 0 ≦ r <N (3) = 0 r <0, N ≦ r. When such a window function w _r (r) or w _h (r) is used, the window function w (r) (= w (kL−
The non-zero section of q)) is expressed as follows: 0 ≦ kL−q <N By modifying this, kL−N <q ≦ kL Therefore, for example, in the case of the rectangular window, the window function w _r (kL−q) =
As shown in FIG. 3, kL-N <q ≦ kL
It is time of. Further, the above equations (1) to (3) indicate that the length N
(= 256) indicates that the sample window advances by L (= 160) samples. Hereinafter, the above (2)
Each N point (0 ≦ r) extracted by each window function of the equation (3)
<N) are represented by x _wr (k, r) and x _wh , respectively.
(k, r).

As shown in FIG. 4, the windowing processing section 14 _applies 179 to the sample sequence x _wh (k, r) of 256 samples per block to which the Hamming window of the above equation (3) is applied.
Two samples of 0 data are added (so-called zero-filled) to make 2048 samples, and the orthogonal transform unit 105 performs orthogonal transform such as FFT (fast Fourier transform) on the time axis data sequence of 2048 samples. Processing is performed.

The pitch extraction section 13 performs pitch extraction based on the sample sequence of x _wr (k, r) (N samples in one block). As the pitch extraction method, a method using a periodicity of a time waveform, a periodic frequency structure of a spectrum, an autocorrelation function, and the like are known. In this embodiment, an autocorrelation method of a center clip waveform is employed. are doing. For the center clip level in the block at this time,
One clip level may be set for one block. However, when a peak level of a signal of each section (each sub-block) obtained by subdividing the block is detected and a difference between the peak levels of these sub-blocks is large. Then, the clip level is changed stepwise or continuously in the block. The pitch period is determined based on the peak position of the autocorrelation data of the center clip waveform.
At this time, a plurality of peaks are obtained from the autocorrelation data belonging to the current frame (the autocorrelation is obtained from data of one block N samples), and the maximum peak among the plurality of peaks is equal to or larger than a predetermined threshold. In this case, the maximum peak position is used as the pitch cycle, and in other cases, a pitch within a pitch range that satisfies a predetermined relationship with a pitch obtained in a frame other than the current frame, for example, the previous and next frames, for example, the center of the pitch of the previous frame. ± 2
A peak within the range of 0% is obtained, and the pitch of the current frame is determined based on the peak position. In the pitch extraction unit 13, a relatively rough pitch search is performed by an open loop, and the extracted pitch data is sent to a high-precision (fine) pitch search unit 16, and a high-precision pitch search (pitch) by a closed loop is performed. Fine search) is performed.

The high-precision (fine) pitch search unit 16 stores coarse (rough) pitch data of an integer value extracted by the pitch extraction unit 13 on the frequency axis, for example, subjected to FFT by the orthogonal transformation unit 15. Data is supplied. In this high-precision pitch search section 16, an optimum decimal point (floating) is obtained by swinging ± several samples at intervals of 0.2 to 0.5 around the coarse pitch data value.
To the value of fine pitch data. As a method of fine search at this time, analysis by so-called synthesis
Using the (Analysis by Synthesis) method, the pitch is selected so that the synthesized power spectrum is closest to the power spectrum of the original sound.

The fine search of the pitch will be described. First, in the MBE vocoder, the FF
S (j) as spectral data on the frequency axis orthogonally transformed by T or the like is expressed as S (j) = H (j) | E (j) | 0 <j <J (4) Model is assumed. Where J is
_{ω s / 4π = f s /} 2 in response, the sampling frequency f
_s = when ω _s / 2π is, for example, 8kHz corresponding to 4kHz. In the above equation (4), when the spectrum data S (j) on the frequency axis has a waveform as shown in FIG.
(j) represents the original spectrum data S as shown in FIG.
(j) shows the spectral envelope,
(j) shows the spectrum of the excitation signal (excitation) at the same level and periodic as shown in FIG. 5C. That is, the FFT spectrum S (j) has a spectrum envelope H (j) and a power spectrum | E of the excitation signal.
(j) Modeled as the product with |.

The power spectrum of the excitation signal | E (j)
| Takes into account the periodicity (pitch structure) of the waveform on the frequency axis determined according to the pitch, and converts the spectrum waveform corresponding to the waveform of one band (band) for each band on the frequency axis. It is formed by arranging it repeatedly. The waveform for one band is obtained by performing a FFT by regarding a waveform obtained by adding 0 data for 1792 samples (padding) to a Hamming window function of 256 samples as shown in FIG. 4 as a time axis signal, for example. It can be formed by cutting out an impulse waveform having a certain bandwidth on the frequency axis according to the pitch.

Next, for each band divided according to the pitch, a value (a kind of amplitude) | A representative of the above H (j) (to minimize the error for each band) | A _m
| Here, for example, when the lower and upper points of the m-th band (band of the m-th harmonic) are a _m and b _m , respectively, the error ε _m of the m-th band is

[0023]

(Equation 1)

Can be expressed by | A _m | that minimizes this error ε _m is

[0025]

(Equation 2)

When | A _m | in equation (6),
Minimize the error ε _m . The amplitude | A _m | is obtained for each band, and the error ε _m for each band defined by the above equation (5) is obtained using the obtained amplitude | A _m |. Next, a total value Σε _m of all the bands of the error ε _m for each band is obtained. Further, the error sum Shigumaipushiron _m of all such bands, calculated for different pitches to some small, obtaining the pitch as an error sum Shigumaipushiron _m is minimized.

That is, several types are prepared above and below, for example, in increments of 0.25 with the rough pitch obtained by the pitch extracting unit 13 as a center. Each respective pitches of different pitches to these plurality of types of fine finding the error sum Σε _m. In this case, when the pitch is determined, the bandwidth is determined. From the above equation (6), the power spectrum | S (j) | of the data on the frequency axis and the excitation signal spectrum | E
(j) | seek error epsilon _m of equation (5) using a, can be obtained sum Shigumaipushiron _m of all the bands. Obtains the error sum Shigumaipushiron _m for each pitch, is not to determine the pitch corresponding to error sum total value which is the smallest as the optimal pitch. As described above, the optimum fine (for example, in increments of 0.25) pitch is obtained by the high-precision pitch search unit 16, and the amplitude | Am | corresponding to the optimum pitch is determined.

In the above description of the fine search of the pitch, in order to simplify the description, all the bands are voiced (Vo).
iced), but the MBE
In a vocoder, an unvoiced sound (Un
Since the model in which a voiced) region exists is used, it is necessary to discriminate voiced / unvoiced sounds for each band.

The data of the optimum pitch and amplitude | A _m | from the high-precision pitch search section 16 is sent to a voiced / unvoiced sound discriminating section 17, where the voiced / unvoiced sound is discriminated for each band. For this determination, NSR (noise t
o signal ratio). That is, N of the m-th band
SR is

[0030]

(Equation 3)

When this NSR value is larger than a predetermined threshold value (for example, 0.3) (error is large), | S (j) by | A _m || E (j) | | Is not good (the excitation signal | E (j) | is inappropriate as a basis), and the band is identified by UV (Unvoice
d, unvoiced sound). In other cases, it can be determined that the approximation has been performed to some extent, and the band is
(Voiced, voiced sound).

Next, the amplitude re-evaluation unit 18 stores the data on the frequency axis from the orthogonal transformation unit 15 and the high-precision pitch search unit 1
6 and the evaluated amplitude | A _m | and the V from the voiced / unvoiced sound discriminating unit 17.
/ UV (voiced / unvoiced) discrimination data is supplied. The amplitude re-evaluation unit 18 obtains the amplitude again for the band determined to be unvoiced (UV) by the voiced / unvoiced sound determination unit 17. The amplitude | A _m | _UV for this UV band is

[0033]

(Equation 4)

Is obtained by

The data on the frequency axis from the orthogonal transform unit 15 and the data of the fine pitch and the estimated amplitude | A _m | from the high-precision pitch search unit 16 are also supplied to the phase evaluation unit 19. You. The phase evaluation unit 19 obtains the phase of the voiced sound band from each of the data. Here, for example, assuming that the lower and upper points of the m-th band (the band of the m-th harmonic) are a _m and b _m , respectively, the error ε _cm of the m-th band is

[0036]

(Equation 5)

Can be expressed by Where S (i), _Am and E
(I) are both complex numbers, and when rewritten in polar coordinate format, they are expressed by the following equation (10).

[0038]

(Equation 6)

That is, the error ε _cm of the m-th band is
From the above equation (10),

[0040]

(Equation 7)

Can be rewritten. And minimizing this equation (11),

[0042]

(Equation 8)

Is as follows. When rearranging this equation (12),

[0044]

(Equation 9)

Is obtained. Therefore, the phase φ _m for optimizing the evaluation function of the above equation (9) is

[0046]

(Equation 10)

Is as follows.

The phase from the phase evaluation unit 19 (actually measured value)
Is supplied to the phase prediction calculation & error evaluation unit 20. The phase prediction detection & error evaluation section 20 can perform phase prediction performed on the synthesis side. That is, the amplitude | A evaluated as a fine pitch from the high-precision pitch search unit 16
Predict the phase based on _m |. Then, the phase evaluation unit 1
9 and the error between the predicted phase and the predicted phase.

The calculation of the phase prediction and the evaluation of the prediction error in the phase prediction calculation & error evaluation section 20 will be described. In the synthesis side, time 0 as in the (beginning of frame) the m-th harmonic of the phase (frame initial phase) and [psi _{0 m,} the phase [psi _Lm of the last _{frame, ψ Lm = mod2π (ψ 0m} + m ( ω ₀₁ + ω _L1 ) L / 2) (15) In this equation (15), mod2π (x) is a function that returns the principal value of x between -π and + π. For example, x = 1.
When 3π, mod2π (x) =-0.7π, When x = 2.3π
mod2π (x) = 0.3π, x = −1.3π mod2π
(X) = 0.7π, and so on. L is a frame interval, ω ₀₁ is a basic angular frequency at the leading end (n = 0) of the combined frame, and ω _L1 is a basic angular frequency at the end of the combined frame (n = L: leading end of the next combined frame). It is.

The ψ _Lm obtained by the above equation (15) includes a prediction error. Therefore, the encoding side detects the difference between the predicted phases obtained by the above equation (15) from the actually measured phases of the phase evaluation unit 19, and estimates the prediction accuracy. This is the phase prediction calculation & error evaluation unit 20.

The measured phase φ _lm obtained by the above equation (14) can be expressed as follows: φ _lm = mod2π (φ _0m + m (ω ₀₁ + ω _L1 ) L / 2) + ε _m (16) Here, φ _0m is the measured phase of the m-th harmonic, and ε _m is the error between the measured phase and the predicted phase. Therefore, it is possible to determine the prediction accuracy of the phase displacement by m (ω ₀₁ + ω _L1 ) L / 2 from the error ε _m . _{That, ε m = mod2π (φ lm} -mod2π (φ 0m + m (ω 01 + ω L1) L / 2)) = mod2π (φ lm -φ 0m -m (ω 01 + ω L1) L / 2) ··· ( 17) This ε _m is obtained for each band (m bands).

Next, for the error ε _{m, the} variance σ ² is given by
determined in the range of a _m of b _m. First, a of the error ε _m
The average ε from _m to b _m is

[0053]

[Equation 11]

Is determined by Here, a _m = M / 4, b
_m = M, where M is the number of harmonics up to π. M represents the angular frequency of the pitch as ω ₀ (= 2πf /
f _s : f is a pitch frequency, and f _s is a sample frequency), and is an integer closest to π / ω ₀ .

Therefore, the variance σ ² is

[0056]

(Equation 12)

Is obtained by When this variance σ ² is large, that is, as shown in FIG. 6A, it means that the phase prediction is generally off. On the other hand, when the variance σ ² is small, that is, as shown in FIG. 6B, it means that the phase prediction is fairly accurate.

The amplitude reevaluating section 18 and the phase evaluating section 1
9 and the respective data from the high-precision pitch search unit 16 are sent to a data number conversion (a kind of sampling rate conversion) unit 21. The data number conversion unit 21 is used to make the number constant, taking into account that the number of divided bands on the frequency axis differs according to the pitch and the number of data (particularly the number of amplitude data) differs. is there. That is, for example, if the effective band is up to 3400 Hz, this effective band is divided into 8 to 63 bands in accordance with the pitch, and the amplitude | Am | obtained for each of these bands is obtained.
(Including UV band amplitude | _Am | _UV ) Number of data m
_MX + 1 also changes from 8 to 63. Therefore, the data number converter 21 converts the variable number m _MX +1 of amplitude data into a fixed number N _C (for example, 44) data.

In this embodiment, dummy data is used to interpolate values from the last data in the block to the first data in the block with respect to the amplitude data of one effective band on the frequency axis. Is added to expand the number of data to N _F , and then the band-limited K _OS times (for example, 8
Obtain an amplitude data of K _OS times the number by performing oversampling multiplied), the K _OS times the number ((m _MX +
1) × K _OS amplitude data is linearly interpolated and expanded to more N _M (for example, 2048), and this N _M data is thinned to obtain the above-mentioned fixed number N _C (for example, 44). Convert to data.

The data from the data number conversion unit 21 (the fixed number N _C of amplitude data) is converted to a vector quantization unit 22.
, And are grouped into a vector for each of a predetermined number of data, and subjected to vector quantization. Quantized output data from the vector quantization unit 22 is extracted via an output terminal 25. Also, the high-precision pitch search unit 1
The high-precision (fine) pitch data from No. 6 is encoded by the pitch encoding unit 23 and extracted via the output terminal 27. Here, the prediction error from the phase prediction calculation & error evaluation unit 20 is supplied to the movable contact 24a of the switching unit 24, and when the movable contact 24a comes into contact with the fixed contact 24b, the magnitude of the prediction error is increased. (Variance σ ² ) is supplied to the voiced / unvoiced sound discriminating unit 17. Then, the voiced sound / unvoiced sound discriminating unit 17 compares the predetermined threshold value with the variance σ ² , and if the variance σ ² is large, the V as shown in FIG.
Is reduced as shown in FIG. 7B to extend the UV bandwidth. Such voiced / unvoiced sound (V / UV) discrimination data from the voiced / unvoiced sound discriminating unit 17 is extracted via the output terminal 26. On the other hand, when the movable contact piece 24a comes into contact with the fixed contact 24c, information (variance σ ² ) indicating the magnitude of the prediction error is extracted via the output terminal 28.

Each of these data is obtained by processing the data in the block of N samples (for example, 256 samples), and the block is represented on the time axis by the frame of L samples. , The data to be transmitted is obtained in the frame unit. That is, the pitch data, V / UV discrimination data, amplitude data, and information indicating the magnitude of the prediction error are updated in the frame cycle.

Next, a synthesizing side (decoding side) for synthesizing an audio signal based on each of the data obtained by transmission.
Will be described with reference to FIG. 8, the input terminal 31 receives the vector-quantized amplitude data, the input terminal 32 receives the encoded pitch data, the input terminal 33 receives the V / UV discrimination data, and The terminal 34 is supplied with information indicating the magnitude of the prediction error. The quantized amplitude data from the input terminal 31 is sent to an inverse vector quantizer 35 where it is inversely quantized and sent to an inverse data number converter 36 where it is inversely transformed. 37 and an unvoiced sound synthesizer 38. The encoded pitch data from the input terminal 32 is decoded by the pitch decoding unit 39,
It is sent to the data number inverse conversion unit 36, voiced sound synthesis unit 37, unvoiced sound synthesis unit 38, and phase correction unit 40. Input terminal 3
The V / UV discrimination data from 3 is sent to the voiced sound synthesizer 37 and the unvoiced sound synthesizer 38. The information indicating the magnitude of the prediction error from the input terminal 34 is sent to the phase correction unit 40.

In the voiced sound synthesizer 37, for example, a cosine
A voiced sound waveform on the time axis is synthesized by wave synthesis, and the unvoiced sound synthesizer 38 synthesizes an unvoiced sound waveform on the time axis by filtering, for example, white noise with a band-pass filter. The adder 41 adds and synthesizes the waveform and the waveform, and extracts the resultant from the output terminal 42.

Here, the phase correcting section 40 corrects the phase of the phase prediction based on the information indicating the magnitude of the prediction error and the decoded pitch data from the pitch decoding section 39. That is, the phase Φ _Lm at the end of the m-th harmonic frame on the synthesis side is _calculated as ψ _Lm expressed by the above equation (15).
Using, obtained from _{Φ Lm = ψ Lm + S c} R dp ··· (20). Here, S _c represents the scale value, R _dp is 2πf (m, M) × (Gaussian random number). Also,
f (m, M) is a function for changing the σ value according to the frequency. When the above _Rdp exceeds the range of ± π, ± π is set to the clipping level and _Rdp is clipped. Specifically, the scale value S _c of the expression (20) for example, 0.2 and 0.
7, the flag indicating the magnitude of the prediction error transmitted from the analysis side indicates that the variance σ ² is larger than a certain threshold.
In this case , the phase is corrected using 0.7 as the scale value. Then, the phase corrected by the phase correction unit 40 is supplied to the voiced sound synthesis unit 37.

Hereinafter, the synthesizing process in the voiced sound synthesizing section 37 will be described in detail. The m-th discriminated as V (voiced sound)
1 on the time axis in the band (band of the m-th harmonic)
Synthetic frame (L samples, e.g., 160 samples) when the amount of the voiced and V _m (n), using a time index (sample number) n in the synthetic _{frame, V m (n) = A} m (n ) cos (θ _m (n)) 0 ≦ n <L (21) The final voiced sound V (n) is synthesized by adding (ΣV _m (n)) the voiced sounds of all the bands determined as V (voiced sound) in all the bands.

A _m (n) in the equation (21) is the amplitude of the m-th harmonic interpolated from the top to the end of the composite frame. In the simplest case, the value of the m-th harmonic of the amplitude data updated for each frame may be linearly interpolated.
That is, the m-th position at the end (n = 0) of the composite frame
The amplitude value of the harmonic is A _0m , and the end of the synthesized frame (n =
When the amplitude value of the m-th harmonic at the end of the next synthesized frame is A _Lm , A _m (n) = (Ln) A _{0 m} / L + nA _Lm / L (22) A _m (n) may be calculated.

FIG. 9A shows an example of the spectrum of an audio signal. Each band having a band number (harmonics number) m of 8, 9, and 10 is set to UV (unvoiced sound), and the other bands are set to UV. Is V (voiced sound). This V
The time axis signal of the (voiced sound) band is the voiced sound synthesis unit 3
7 and the time axis signal of the UV (unvoiced sound) band is synthesized by the unvoiced sound synthesis unit 38.

Hereinafter, the unvoiced sound synthesizing process in the unvoiced sound synthesizing section 38 will be described. The white noise signal waveform on the time axis from the white noise generating unit 51 is windowed with a predetermined length (for example, 256 samples) using an appropriate window function (for example, a Hamming window), and the STFT processing unit 52 performs ST processing.
By performing FT (Short Term Fourier Transform) processing, a power spectrum on the frequency axis of white noise as shown in FIG. 9B is obtained. This STFT processing unit 52
Is transmitted to the band amplitude processing unit 53, and as shown in FIG. 9C, the amplitude | A _m | _UV is calculated for the UV (unvoiced) bands (for example, m = 8, 9, 10). The multiplication is performed, and the amplitudes of the other V (voiced sound) bands are set to zero. The band amplitude processing unit 53 is supplied with the amplitude data, the pitch data, and the V / UV discrimination data. The output from the band amplitude processing unit 53 is IS
The phase is sent to the TFT processing unit 54, and the phase is converted into a signal on the time axis by performing an inverse STFT process using the phase of the original white noise. The output from the ISTFT processing unit 54 is sent to an overlap addition unit 55, which repeats overlap and addition while weighting appropriately (to restore the original continuous noise waveform) on the time axis.
Synthesize a continuous time axis waveform. An output signal from the overlap addition unit 55 is sent to the addition unit 41.

As described above, the respective signals of the voiced sound portion and the unvoiced sound portion that have been synthesized in the synthesis portions 37 and 38 and returned on the time axis are added by the addition portion 41 at an appropriate fixed mixing ratio. The reproduced audio signal is extracted from the output terminal 42.

Therefore, in the embodiment of the high-efficiency encoding method according to the present invention, the accuracy of phase prediction on the synthesis side is calculated on the analysis side, and the number of unvoiced sound bands is increased or Controls the randomization factor of the phase at.

The components on the voice analyzing side (encoding side) in FIG. 1 and the voice synthesizing side (decoding side) in FIG. 8 are described in terms of hardware. It can also be realized by a software program using a signal processor) or the like.

[0072]

According to the high-efficiency encoding method of the present invention, the width of the voiced area and the unvoiced area on the frequency axis is changed according to the magnitude of the error between the measured phase and the predicted phase for each band. be able to. Further, by transmitting information indicating the magnitude of the error to the combining side, the phase is corrected on the combining side. Therefore, a high-quality reproduced sound can be obtained on the synthesis side.

[Brief description of the drawings]

FIG. 1 is a functional block diagram showing a schematic configuration of an analysis side (encoding side) of a speech signal synthesis analysis encoding apparatus as a specific example of an apparatus to which the high efficiency encoding method according to the present invention is applied.

FIG. 2 is a diagram for explaining windowing processing.

FIG. 3 is a diagram for explaining a relationship between a windowing process and a window function.

FIG. 4 is a diagram showing time axis data as an object of orthogonal transform (FFT) processing.

FIG. 5 is a diagram showing spectrum data on a frequency axis, a spectrum envelope (envelope), and a power spectrum of an excitation signal.

FIG. 6 is a waveform chart for explaining the accuracy of phase prediction.

FIG. 7 is a diagram for explaining extension of a UV band.

FIG. 8 is a functional block diagram showing a schematic configuration of a synthesis side (decoding side) of a speech signal synthesis analysis encoding apparatus as a specific example of an apparatus to which the high efficiency encoding method according to the present invention is applied.

FIG. 9 is a diagram for explaining unvoiced sound synthesis when synthesizing an audio signal.

[Explanation of symbols]

13: pitch extraction unit 14: windowing processing unit 15: orthogonal transform (FFT) unit 16: high-precision (fine) pitch search unit 17: Voiced (V) / unvoiced (UV) discriminating unit 18 Amplitude re-evaluating unit 19 Phase estimating unit 20 Phase estimating & error estimating unit 21 Data number conversion (data rate conversion) unit 22 ... Vector quantization unit

Continuation of the front page (56) References JP-A-63-273898 (JP, A) JP-A-3-259200 (JP, A) JP-A-2-280200 (JP, A) JP-A-2-282800 (JP, A) JP-A-2-84700 (JP, A) JP-A-3-53300 (JP, A) JP-A-2-7100 (JP, A) (58) Fields investigated (Int. Cl. ⁷ , DB Name) G10L 19/02 G10L 19/00

Claims (2)

(57) [Claims] 1. A high-efficiency coding method, comprising the steps of: dividing an input audio signal into blocks and converting the signal into a frequency axis to obtain data on the frequency axis; and dividing the data on the frequency axis into a plurality of bands. Performing a step of detecting a block initial phase for each of the divided bands; a step of determining whether voiced or unvoiced for each of the bands; and a step of calculating a predicted phase value for each of the bands. Calculating the error between the measured phase and the predicted phase for each band, and changing the width of the voiced and unvoiced regions on the frequency axis according to the magnitude of the error. High efficiency coding method. 2. A high-efficiency encoding method, comprising the steps of: dividing an input audio signal into blocks and converting the signal into a frequency axis to obtain data on the frequency axis; and dividing the data on the frequency axis into a plurality of bands. Performing a step of detecting a block initial phase for each of the divided bands; a step of determining whether voiced or unvoiced for each of the bands; and a step of calculating a predicted phase value for each of the bands. Calculating an error between the measured phase and the predicted phase for each band; transmitting information indicating the magnitude of the error; and correcting the phase of phase prediction according to the information. A high-efficiency encoding method characterized by the following.