JPH1011093A - Signal encoding device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To obtain excellent sound quality for voices from more than one speaker and musical sound signals generated by more than one musical instrument by calculating a sound source signal after finding delays from input signals. SOLUTION: A spectrum parameter calculating circuit 200 performs calculation in predetermined degree of a spectrum parameter by providing a window which is longer than subframe length for a voice signal of at least one subframe and segmenting a voice. A pitch extracting circuit 390 finds delays and gains corresponding to a pitch cycle by using the output of an auditory weighting circuit 230. Then a sound source quantizing circuit 350 performs vector quantization for a sound source signal by using a sound source code book 351. The output of a subtracter 236 and the output of an impulse response calculating circuit 310 are used to retrieve a sound source code vector cj (n) from the sound source code book 351. Further, the index of the selected sound source vector is outputted to a multiplexer 400.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a signal encoding apparatus for encoding a speech or music signal at a low bit rate with high quality.

[0002]

2. Description of the Related Art As a method for encoding a speech signal with high efficiency, for example, M.I. Schroeder and B.S.
"Code-excited line by Atal
arprediction: High quality
speech at very low bitr
ates "(Proc. ICASPS, pp. 937-
940, 1985), Kl.
"Improved speech" by Eijn et al.
quality and efficiencyintectect
or quantification in SELP "
(Proc. ICASSP, pp. 155-158, 1
1988), and a CELP (Code Excited Linear) described in a paper (Reference 2) and the like.
Predictive Coding) is known. In this conventional example, the transmitting side extracts a spectral parameter representing a spectral characteristic of an audio signal from the audio signal for each frame (for example, 20 ms) by using linear prediction (LPC) analysis. The frame is further divided into subframes (for example, 5 ms), and parameters (a delay parameter and a gain parameter corresponding to a pitch period) in the adaptive codebook are extracted for each subframe based on a past sound source signal. Pitch prediction of the audio signal of the subframe. For an excitation signal obtained by pitch prediction, an optimal excitation code vector is selected from an excitation codebook (vector quantization codebook) composed of predetermined types of noise signals, and an optimal gain is calculated. , Quantize the sound source signal. The excitation code vector is selected so as to minimize the error power between the signal synthesized from the selected noise signal and the residual signal. Then, the index and gain indicating the type of the selected code vector, the spectrum parameter and the parameter of the adaptive codebook are combined and transmitted by the multiplexer unit. Description on the receiving side is omitted.

[0003]

The conventional method has a problem that a large amount of calculation is required to select an optimal excitation code vector from an excitation codebook. In this method, in order to select a sound source code vector, filtering or convolution operation is once performed on each code vector, and this operation is repeated by the number of code vectors stored in the code book. Due to that. For example, when the number of bits in the codebook is B and the number of dimensions is N, the amount of operation is N × K × 2 per second, where K is the filter or impulse response length in the filtering or convolution operation.
Only ^B × 8000 / N is required. As an example, B = 1
If 0, N = 40 and K = 10, 81,9 per second
20,000 operations are required, which is extremely large. This problem becomes more serious as the input signal band is wider than the telephone band and the sampling frequency becomes higher.

[0004] Various methods have been proposed as a method for reducing the amount of calculation required for searching the sound source codebook. For example, ACELP (Argebraic Code Ex)
Citated Linear Prediction) has been proposed. This is, for example, C.I. Lafla
"16 kbps widebands by Mme et al.
peech coding technique ba
Sed on algebric CELP "(Proc. ICASP, pp. 13-16, pp. 13-16).
1991) (Literature 3). According to the method of Document 3, the sound source signal is represented by a plurality of pulses,
The position of each pulse is represented by a predetermined number of bits and transmitted. Here, since the amplitude of each pulse is limited to +1.0 or -1.0, the amount of calculation for the pulse search can be significantly reduced.

[0005] However, in any of the above-mentioned methods, although relatively good sound quality can be obtained for a single-pitch sound, voices of a plurality of speakers or a plurality of types of musical instruments are used for a conference or the like. For a music signal containing a plurality of pitches due to the coexistence, the sound quality is significantly deteriorated at a low bit rate.

SUMMARY OF THE INVENTION An object of the present invention is to solve the above-mentioned problems and to provide a signal code with a relatively small amount of operation and a small deterioration in sound quality not only for a wideband voice but also for a music signal even when the bit rate is low. It is to provide a generalization method.

[0007]

According to a first aspect of the present invention, there is provided a spectrum parameter calculator for obtaining and quantizing a spectrum parameter from an input signal, and a spectrum parameter calculator corresponding to a first pitch period from the input signal. 1 delay, a pitch prediction residual signal is determined, a second delay corresponding to a second pitch period is determined from the pitch prediction residual signal, and a plurality of predetermined delays including at least the first and second delays are determined. A delay calculation unit that calculates and outputs the number of delays, and a sound source quantization unit that obtains and quantizes and outputs a sound source signal for a pitch-predicted residual signal using the plurality of delays. Thus, a signal encoding device that performs

According to a second aspect of the present invention, there is provided a signal encoding apparatus characterized in that an excitation signal of the signal is composed of a plurality of pulses having a non-zero amplitude.

According to a third aspect of the present invention, there is provided a spectrum parameter calculator for obtaining and quantizing a spectrum parameter from an input signal, a mode discriminator for extracting a feature quantity from the input signal and discriminating a mode, A first delay corresponding to a first pitch period is obtained from the input signal in a predetermined mode, a pitch prediction residual signal is obtained, and a second delay corresponding to a second pitch period is obtained from the pitch prediction residual signal. And a delay calculator for calculating and outputting a plurality of predetermined delays including at least the first and second delays; and a sound source for the residual signal pitch-predicted using the plurality of delays. A signal encoding device having a sound source quantization unit for obtaining a signal, quantizing the signal, and outputting the signal is obtained.

According to a fourth aspect of the present invention, there is provided a signal encoding apparatus characterized in that an excitation signal of the signal is constituted by a plurality of non-zero amplitude pulses.

According to a fifth aspect of the present invention, there is provided a spectrum parameter calculator for obtaining and quantizing a spectrum parameter from an input signal, and obtaining a first delay corresponding to a first pitch period from the input signal. A pitch prediction unit for obtaining a prediction residual signal, calculating a first pitch prediction gain, obtaining a second delay corresponding to a second pitch period from the pitch prediction residual signal, and repeating these processes; A determination unit that determines whether the gain satisfies a predetermined condition, and a sound source signal for a residual signal pitch-predicted using the delay when the pitch prediction gain does not satisfy a predetermined condition. A signal encoding apparatus characterized by having a sound source quantization unit for obtaining, quantizing and outputting the obtained signal is obtained.

According to a sixth aspect of the present invention, there is provided a signal encoding apparatus characterized in that an excitation signal of the signal is composed of a plurality of pulses having a non-zero amplitude.

According to a seventh aspect of the present invention, there is provided a spectrum parameter calculation unit for obtaining and quantizing a spectrum parameter from an input signal, a mode discrimination unit for extracting a feature amount from the input signal and discriminating a mode, In a predetermined mode, a first delay corresponding to a first pitch period is obtained from the input signal, a pitch prediction residual signal is obtained, a first pitch prediction gain is calculated, and a second pitch prediction gain is calculated from the pitch prediction residual signal. A pitch prediction unit that determines a second delay corresponding to the pitch period and repeats these processes; a determination unit that determines whether the pitch prediction gain satisfies a predetermined condition; If the given condition is not satisfied, a sound source quantum for obtaining and quantizing a sound source signal for the residual signal pitch-predicted using the delay and outputting the sound source signal. Signal encoding device is obtained, characterized in that it comprises a part.

According to an eighth aspect of the present invention, there is provided a signal encoding apparatus characterized in that an excitation signal of the signal is composed of a plurality of pulses having a non-zero amplitude.

In the first aspect of the present invention, the calculation of a delay corresponding to a pitch period and the calculation of a pitch prediction residual signal are repeated from an input signal to obtain a predetermined number of delays, and the plurality of delays are calculated. A sound source signal is obtained from the residual signal whose pitch has been predicted using the obtained signal, quantized, and output.

According to a second aspect of the present invention, in the first aspect, the sound source signal is composed of a pulse train having a number M of non-zero amplitudes, and the amplitude and position of the pulse are obtained to quantize the sound source signal. .

According to the third aspect of the present invention, the mode is determined by extracting the characteristic amount from the input signal, and the same operation as in the first aspect is performed only in a predetermined mode.

According to a fourth aspect of the present invention, in the third aspect, the sound source signal is composed of a pulse train of a number M having a non-zero amplitude, and the sound source signal is quantized by obtaining the amplitude and position of the pulse. .

According to a fifth aspect of the present invention, the calculation of the delay corresponding to the pitch period, the calculation of the pitch prediction residual signal, and the calculation of the pitch prediction gain are repeated from the input signal, and the pitch prediction gain is set to a predetermined condition. A sound source quantizing unit that determines whether the pitch prediction gain does not satisfy a predetermined condition and obtains and quantizes a sound source signal for a residual signal pitch-predicted using the delay when the pitch prediction gain does not satisfy a predetermined condition. Having.

According to a sixth aspect of the present invention, in the fifth aspect, the sound source signal is constituted by a pulse train of a number M having a non-zero amplitude, and the sound source signal is quantized by obtaining the amplitude and position of the pulse. .

In the seventh aspect of the present invention, the mode is determined by extracting the characteristic amount from the input signal, and the same operation as in the fifth aspect is performed only in a predetermined mode.

According to an eighth aspect of the present invention, in the seventh aspect, the sound source signal is constituted by a pulse train having a number M of non-zero amplitudes, and the amplitude and position of the pulse are obtained to quantize the sound source signal. .

[0023]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Next, embodiments of the present invention will be described with reference to the drawings.

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

In the figure, an audio signal is input from an input terminal 100, a frame dividing circuit 110 divides the audio signal for each frame (for example, 10 ms), and a sub-frame dividing circuit 120 divides the audio signal of the frame into a frame shorter than the frame. It is divided into subframes (for example, 5 ms).

The spectrum parameter calculation circuit 200 cuts out the speech signal by applying a window (for example, 24 ms) longer than the subframe length to the speech signal of at least one subframe, and sets the spectrum parameter to a predetermined order (for example, (P = 10th order) is calculated. Here, a well-known LPC analysis, Burg analysis, or the like can be used for calculating the spectrum parameters. Here, Burg analysis is used. For details of the Burg analysis, see the book entitled "Signal Analysis and System Identification" written by Nakamizo (Corona Publishing Co., 1988), 82-.
Since it is described on page 87 (Document 4) and the like, the description is omitted. Further, in the spectrum parameter calculation unit, Bung
Linear prediction coefficient α _i (i = 1,..., 1)
0) is converted into LSP parameters suitable for quantization and interpolation. Here, the conversion from the linear prediction coefficient to the LSP is performed by a paper entitled "Speech Information Compression by Line Spectrum Pair (LSP) Speech Analysis / Synthesis Method" by Sugamura et al. -606, 1981)
(Reference 5). For example, the linear prediction coefficient obtained by the Burg method in the second subframe is represented by L
The LSP of the first sub-frame is converted to SP parameters, the LSP of the first sub-frame is obtained by linear interpolation, and the LSP of the first sub-frame is inversely converted to a linear prediction coefficient, and the linear prediction coefficient α _il (i = 1, ..., 10, l = 1, ...,
2) is output to the auditory weighting circuit 230. Also, the second
The LSP of the subframe is output to spectrum parameter quantization circuit 210.

The spectrum parameter quantization circuit 210 efficiently quantizes LSP parameters of a predetermined subframe. It is assumed that vector quantization is used as a quantization method, and LSP parameters of the second subframe are quantized. A well-known method can be used for the method of vector quantization of LSP parameters. Specific methods are described in, for example, JP-A-4-171500 (Japanese Patent Application No. 2-297600) (Reference 6), JP-A-4-36300 (Japanese Patent Application No. 3-261925) Reference 7), and JP-A-5-6199 (Japanese Patent Application No. 3-15)
No. 5049) (Reference 8) and T.I. Nomura et
al. "LSP CodingUsing V
Q-AVQ With Interpolation
in4.075kbps M-LCELP Spec
h Coder "(Proc. Mobil
e Multimedia Communicatio
ns, pp. B. 2.5, 1993) (Reference 9).

Using the LSP codebook 211, a code vector that minimizes distortion represented by the following equation 1 is selected and output.

[0029]

(Equation 1) Here, LSP (i), QLSP (i) _j , and W (i) are the i-th LSP before quantization, the j-th result after quantization, and the weight coefficient, respectively.

The spectrum parameter quantization circuit 2
At 10, the LSP parameters of the first sub-frame are restored based on the LSP parameters quantized in the second sub-frame. Here, the LSP of the first subframe is restored by linearly interpolating the quantized LSP parameter of the second subframe of the current frame and the quantized LSP of the second subframe of the previous frame. Here, LSP before quantization
After selecting one type of code vector that minimizes the error power between the LSP and the quantized LSP, the LSP of the first subframe can be restored by linear interpolation.

The L of the first subframe restored as described above
The SP and the quantized LSP of the second subframe are converted into linear prediction coefficients α ′ _i (i = 1,..., 10) for each subframe and output to the impulse response calculation circuit 310. Also,
An index representing the code vector of the quantized LSP of the second subframe is output to the multiplexer 400.

The perceptual weighting circuit 230 inputs the linear prediction coefficients α _i (i = 1,..., P) before quantization from the spectrum parameter calculation circuit 200 for each subframe,
Based on Document 1, perceptual weighting is performed on the audio signal of the subframe, and a perceptual weighting signal is output.

The response signal calculation circuit 240 receives the linear prediction coefficient α _i for each subframe from the spectrum parameter calculation circuit 200, and quantizes, interpolates, and restores the linear prediction coefficient α _i from the spectrum parameter quantization circuit 210. α ′ _i is input for each sub-frame, and a response signal with the input signal set to zero d (n) = 0 is calculated for one sub-frame using the stored value of the filter memory, and the subtractor 235
Output to Here, the response signal x _z (n) is represented by Equation 2 below.

[0034]

(Equation 2) However, when ni ≦ 0, y (ni) = p (N + (ni)) (3) _xz (ni) = _sw (N + (ni)) (4) Here, N indicates a subframe length. γ is a weight coefficient for controlling the perceptual weighting amount, and is the same value as the following equation (6). s _w (n) and p (n) denote the output signal of the weighting signal calculation circuit and the output signal of the denominator term of the filter on the right-hand side first term in equation (6) described later, respectively.

The pitch extraction circuit 390 obtains a plurality of delays and gains corresponding to the pitch period using the output of the auditory weighting circuit 230. Hereinafter, this number is set to two. FIG. 2 shows the configuration of the pitch extraction circuit 390.

In FIG. 2, a perceptual weighting signal x _w (n) is input from a terminal 391. The first delay / gain calculation circuit 392 obtains the first delay T ₁ so as to minimize the equation (5) represented by the following equation 3, and obtains the first delay T ₁ from the equation (6) represented by the following equation 4. A gain β ₁ of ₁ is obtained.

[0037]

(Equation 3)

[0038]

(Equation 4) Further, a first pitch prediction signal y ₁ (n) is obtained according to the following equation and output to the subtractor 394.

Y ₁ (n) = β _{1 xw} (n−T ₁ ) (7) The subtractor 394 obtains the first pitch prediction residual signal e ₁ (n) by the following equation.

E ₁ (n) = x _w (n) −y ₁ (n) (8) The second delay / gain calculation circuit 393 calculates the second delay T ₂ (n) from e ₁ (n), determine the gain β _2. For these, in equations (5) and (6), e ₁ (n) may be used instead of x _w (n).

T ₂ and T ₁ are terminals 397 and 3 respectively.
98.

The impulse response calculation circuit 310 calculates the impulse response h _w (n) of the perceptual weighting filter whose z-transform is expressed by the following equation 5 by a predetermined point L, and the adaptive codebook circuit 300 and the sound source Output to the quantization circuit 350.

[0043]

(Equation 5) The adaptive codebook circuit 300 calculates a delay T _c1 that minimizes the following Expression 6 for samples near T ₁ . Here, the order of the adaptive codebook is 1.

[0044]

(Equation 6) Here, y _w (n-T _c1 ) = v ₁ (n-T _c1 ) * h _w (n) (11), and the symbol * represents a convolution operation.

The gain β _c1 is obtained according to the following equation (7).

[0046]

(Equation 7) Here, in order to improve the extraction accuracy of the delay for the female sound and the voice of the child, the delay may be obtained by a decimal sample value instead of the integer sample. The specific method is, for example,
P. "Pitch predic," by Kron et al.
tors with high temporal r
esolution ”(Proc. ICA
SSP, pp. 661-664, 1990) (Reference 1).
1) etc. can be referred to.

In a similar manner, a delay T _c2 and a gain β _c2 are searched for in a sample near T ₂ .

Next, a pitch prediction signal is calculated and output to the excitation quantization circuit 350.

Q _w1 (n) = β _c1 v (n−T _c1 ) * h _w (n) + β _c2 v (n−T _c2 ) * h _w (n) (13) The delays T _c1 and T _c2 are multiplexers. Output to 400.

The subtractor 236 calculates the following equation and outputs the subtraction result to the sound source quantization circuit 350.

Z _w (n) = x ′ _w (n) −q _w (n) (14) In the excitation quantization circuit 350, the excitation signal is vector-quantized using the excitation codebook 351. Subtractor 236
Using the output of the impulse response calculation circuit 310 and the output of the impulse response calculation circuit 310, a sound source code vector c _j (n) is searched from the sound source codebook 351 so as to minimize the following Expression 8.

[0052]

(Equation 8) Here, the symbol * indicates a convolution operation.

The index of the selected sound source code vector is output to the multiplexer 400.

The gain quantization circuit 365 reads a gain code vector from the gain code book 355, and selects a gain code vector that minimizes the following equation 9 for the selected excitation code vector.

Here, an example in which the gain of the excitation code vector is vector-quantized will be described.

[0056]

(Equation 9) Here, G ′ _t is the t-th code vector in the gain codebook stored in the gain codebook 355.

An index representing the selected gain code vector is output to the multiplexer 400.

The weighting signal calculation circuit 360 receives the respective indexes, reads out the corresponding code vectors from the indexes, and obtains the driving sound source signal v (n) based on the following equation.

V (n) = g (n) + G ′ _t c _j (n) (19) Outputs v (n) to the adaptive codebook circuit 300.

Next, the spectrum parameter calculation circuit 20
Using the output parameter of 0 and the output parameter of the spectrum parameter quantization circuit 210, the response signal s _w (n) is calculated for each subframe by the following equation 10, and output to the response signal calculation circuit 240.

[0061]

(Equation 10) This concludes the description of the embodiment corresponding to the first invention.

FIG. 3 is a block diagram showing the configuration of the second embodiment. In the drawing, what is different from FIG. 1 is a sound source quantization circuit 500, an amplitude codebook 540, a gain quantization circuit 550, and a gain codebook 560.

The sound source quantization circuit 500 calculates the positions and the amplitudes of the M non-zero amplitude pulse trains.

FIG. 4 is a block diagram showing the configuration of the sound source quantization circuit 500. In FIG. 4, a correlation coefficient calculation circuit 5
10 is x from the terminals 501 and 502, respectively.
_w (n) and h _w (n) are input, and two types of correlation coefficients d (n) and φ are calculated according to the following Expressions 11 and 12.
Output to the position calculation circuit 520 and the amplitude quantization circuit 530.

[0065]

[Equation 11]

[0066]

(Equation 12) The position calculation circuit 520 calculates the positions of a predetermined number M of non-zero amplitude pulses. For this, as in Reference 3, for each pulse, the position of the pulse that maximizes the following equation is determined for a predetermined position candidate.

For example, as an example of a position candidate, if the subframe length is N = 40 and the number of pulses is M = 5, it can be expressed as shown in Table 1 below.

[0068]

[Table 1] For each pulse, a candidate position is examined and the position that maximizes the following equation is selected.

D = C _k ² / E _k (23) Here, C _k and E _k are represented by the following equations (13) and (14).

[0070]

(Equation 13)

[0071]

[Equation 14] Here, m _k is, sgn indicating the position of the k-th pulse
(K) is the polarity of the k-th pulse.

The positions of the M pulses are determined by the amplitude quantization circuit 53.
Output to 0.

The amplitude quantization circuit 530 quantizes the pulse amplitude using the amplitude codebook 540. The amplitude code vector that maximizes the following equation is selected.

C _j ² / E _j (26) Here, C _j and E _j are represented by the following Expressions 15 and 16.

[0075]

(Equation 15)

[0076]

(Equation 16) Here, g ′ _kj indicates the amplitude of the k-th pulse in the i-th amplitude code vector.

Note that an amplitude codebook for quantizing the pulse amplitude can be learned and stored in advance using an audio signal. Codebook learning methods are described, for example, by Linde et al., “An algor.
ism for vector quantizat
ion design, "(IEEE T
rans. Commun. Pp. 84-95, Jan
uary, 1980) (Literature 12).

The index and position information of the amplitude code vector are output from terminals 503 and 504, respectively.

The gain quantization circuit 550 quantizes the pulse gain using the gain codebook 560. A gain code vector that minimizes the following Expression 17 is selected, and the index is output to the multiplexer 400.

[0080]

[Equation 17] The weighting signal calculation circuit 570 inputs the respective indexes, reads out the corresponding code vectors from the indexes, and firstly obtains the driving sound source signal v based on the following equation.
Find (n).

[0081] v (n) = g (n ) + G 't g' kj h w (n-m k) (30) outputs v (n) to the adaptive codebook circuit 300.

Next, the spectrum parameter calculation circuit 20
Using the output parameter of 0 and the output parameter of the spectrum parameter quantization circuit 210, the response signal s _w (n) is calculated for each subframe by the following Expression 18, and is output to the response signal calculation circuit 240.

[0083]

(Equation 18) FIG. 5 is a block diagram showing the configuration of the third embodiment.

The mode discriminating circuit 900 receives the perceptual weighting signal from the perceptual weighting circuit 230 in frame units, and outputs the mode discriminating information to the pitch extracting circuit 600 and the multiplexer 400.

Here, the feature amount of the current frame is used for mode determination. As the characteristic amount, for example, a pitch prediction gain averaged in a frame is used. For example, the following equation 19 is used to calculate the pitch prediction gain.

[0086]

[Equation 19] Here, L is the number of subframes included in the frame. P _i and E _i represent the speech power and the pitch prediction error power in the i-th subframe, respectively, and are represented by the following equations 20 and 21.

[0087]

(Equation 20)

[0088]

(Equation 21) Here, T is an optimal delay for maximizing the prediction gain.

The frame average pitch prediction gain G is compared with a plurality of predetermined thresholds, and classified into a plurality of types of modes. As the number of modes, for example, 4 can be used.

Pitch extraction circuit 600 receives the mode discrimination information, performs the same processing as in FIG. 2 in the case of a predetermined mode, and outputs a plurality of delays. In other modes, no delay is output.

FIG. 6 is a block diagram showing the configuration of the fourth embodiment. The mode discriminating circuit 900 in FIG.
, And the pitch extraction circuit 600 is used, so that the description is omitted.

FIG. 7 is a block diagram showing the configuration of the fifth embodiment. In the figure, the configuration different from that of FIG.
Therefore, these will be described.

FIG. 8 is a block diagram showing a configuration of the pitch extraction circuit 700.

The first pitch prediction gain calculation circuit 710 obtains the first pitch prediction gain from the following equation 22 using the first delay and the delay obtained by the gain calculation circuit 392.

[0095]

(Equation 22) Discriminating circuit 730, if the pitch prediction gain G ₁ is greater than the predetermined threshold is a second delay, to continue the process on the gain calculating circuit.

If G ₁ is smaller than the threshold, the second
, And outputs the first delay T ₁ without performing the processing of the gain calculation circuit 393.

When the processing is continued, the second delay / gain calculation circuit 393 outputs x _w (n) of the equations (5) and (6) to e ₁ (n ) And calculate the second delay and gain.

The second pitch prediction gain calculation circuit 720
In the equation (35), the second pitch prediction gain G ₂ is calculated in place of x _w (n) e ₁ (n).

The discriminating circuit 730 discriminates G ₂ as a threshold value, and outputs a first delay and a second delay when it is larger than a predetermined threshold value. When advance G ₂ is smaller than the threshold value defined outputs only the first delay. Further, the number of delays is output from a terminal 399.

Returning to FIG. 7, the sound source quantization circuit 850
First, the number of delays output from the pitch extraction circuit 700 is checked, and if the number of delays is 2, the sound source is generated using the first sound source codebook 851, which is a normal number of bits (B ₁ bit). The signal is quantized. If the number of delays is 1, the second excitation codebook 852 having the same number of bits (B ₂ ) as the number of bits representing the delay is transmitted to the excitation codebook 85.
Use together with 1.

FIG. 9 is a block diagram showing the configuration of the sixth embodiment. 3, the pitch extraction circuit 700 of FIG.
Is used.

The sound source quantization circuit 860 inputs the number of delays from the pitch extraction circuit 700, switches the number of pulses to M ₁ and M ₂ (M ₁ <M ₂ ) according to the number of delays, and Switching between amplitude codebooks having different numbers of bits. The number of pulses when the number is two of delay is set to M _1,
The first amplitude codebook 861 is used. When the number of delays is 1, the number of pulses is M ₂ and the second amplitude codebook 862 is used.

FIG. 10 is a block diagram showing the configuration of the seventh embodiment.

In FIG. 7, the mode discriminating circuit 90 shown in FIG.
0 is added, and the pitch extraction circuit 800 performs the same operation as the pitch extraction circuit 700 in FIG. 7 in a predetermined mode.

FIG. 11 is a block diagram showing the configuration of the eighth embodiment. FIG. 11 is obtained by adding the mode determining circuit 900 and the pitch extracting circuit 800 shown in FIG. 10 to FIG.

The present invention is not limited to the above-described embodiment, and various modifications are possible.

Using the mode information, a sound source quantization circuit,
A configuration in which the gain codebook is switched may be adopted.

When the sound source codebook is used, the expression (1)
A plurality of code vectors are selected in ascending order of the distortion shown in 5), and while the gain is quantized by the gain quantization circuit, a combination of the excitation code vector and the gain code vector that minimizes the expression (18) is selected. May be.

When the sound source is represented by a pulse train, when quantizing the pulse amplitude, a plurality of sets of pulse positions are obtained, an amplitude codebook is searched for each of these positions, and equation (26) is maximized. May be selected.
Alternatively, a combination of a position, an amplitude code vector, and a gain code vector that minimizes the expression (29) may be selected by outputting these combinations to a plurality of types of gain quantization circuits and performing gain quantization.

Instead of the amplitude codebook 540, a polarity codebook having a predetermined number of bits may be used.

A plurality of delays obtained by the pitch extraction circuit are as follows:
By performing differential coding, the number of quantization bits can be reduced.

[0112]

As described above, according to the present invention,
By calculating the sound source signal after obtaining a plurality of delays from the input signal, it is possible to obtain better sound quality than before for voices composed of multiple speakers or music signals composed of a plurality of musical instruments. This has the effect.

Further, the pitch prediction gain is obtained while obtaining the delay, and it is determined whether or not the pitch prediction gain satisfies a predetermined condition, thereby making the number of delays variable. The number of delays can be appropriately selected, and the input signal can be satisfactorily encoded.

[Brief description of the drawings]

FIG. 1 is a block diagram of a signal encoding device according to a first embodiment of the present invention.

2 is a pitch extraction circuit 390 of the signal encoding device of FIG.
It is a block diagram of.

FIG. 3 is a block diagram of a signal encoding device according to a second embodiment of the present invention.

FIG. 4 is an excitation quantization circuit 500 of the signal encoding apparatus of FIG. 3;
It is a block diagram of.

FIG. 5 is a block diagram of a signal encoding device according to a third embodiment of the present invention.

FIG. 6 is a block diagram of a signal encoding device according to a fourth embodiment of the present invention.

FIG. 7 is a block diagram of a signal encoding device according to a fifth embodiment of the present invention.

8 is a block diagram of a pitch extracting circuit 700 of the signal encoding device of FIG.

FIG. 9 is a block diagram of a signal encoding device according to a sixth embodiment of the present invention.

FIG. 10 is a block diagram of a signal encoding device according to a seventh embodiment of the present invention.

FIG. 11 is a block diagram of a signal encoding device according to an eighth embodiment of the present invention.

[Explanation of symbols]

 Reference Signs List 110 frame division circuit 120 subframe division circuit 200 spectrum parameter calculation circuit 210 spectrum parameter quantization circuit 230 auditory weighting circuit 235,236 subtraction circuit 240 response signal calculation circuit 300 adaptive codebook circuit 310 impulse response calculation circuit 350, 500, 850, 860 Sound source quantization circuit 351 Sound source codebook 355, 560 Gain codebook 365, 550 Gain quantization circuit 390, 600, 700, 800 Pitch extraction circuit 392 First delay calculation circuit 393 Second delay calculation circuit 400 Multiplexer 510 phase Relation number calculation circuit 520 Position calculation circuit 530 Amplitude quantization circuit 540 Amplitude codebook 710 First pitch prediction gain calculation circuit 720 Second pitch prediction gain calculation circuit 730 Another circuit 851 first excitation codebook, 852 second excitation codebook 861 first amplitude codebook 862 second amplitude codebook 900 mode discriminating circuit

Claims (8)

[Claims] 1. A spectrum parameter calculation unit for obtaining and quantizing a spectrum parameter from an input signal, obtaining a first delay corresponding to a first pitch period from the input signal, obtaining a pitch prediction residual signal, and obtaining the pitch prediction residual signal. A delay calculating unit that calculates a second delay corresponding to a second pitch period from the residual signal, and calculates and outputs a plurality of predetermined delays including at least the first and second delays;
A signal encoding apparatus comprising: an excitation quantization unit that obtains an excitation signal for a residual signal whose pitch has been predicted using the plurality of delays, quantizes and outputs the excitation signal. 2. The signal encoding apparatus according to claim 1, wherein the excitation signal of the signal is composed of a plurality of pulses having a non-zero amplitude. 3. A spectrum parameter calculation unit for obtaining and quantizing a spectrum parameter from an input signal, a mode discrimination unit for extracting a feature quantity from the input signal and discriminating a mode, and the input signal in a predetermined mode. , A first delay corresponding to a first pitch period is obtained, a pitch prediction residual signal is obtained, a second delay corresponding to a second pitch period is obtained from the pitch prediction residual signal, and the first and second delays are obtained. A delay calculation unit for obtaining and outputting a plurality of predetermined delays, and obtaining and quantizing an excitation signal for a pitch-predicted residual signal using the plurality of delays. A signal encoding device, comprising: 4. The signal encoding apparatus according to claim 3, wherein the excitation signal of the signal is composed of a plurality of non-zero amplitude pulses. 5. A spectrum parameter calculator for obtaining and quantizing a spectrum parameter from an input signal, a first delay corresponding to a first pitch period is obtained from the input signal, and a pitch prediction residual signal is obtained. A pitch prediction unit that calculates a pitch prediction gain, obtains a second delay corresponding to a second pitch period from the pitch prediction residual signal, and repeats these processes; and the pitch prediction gain satisfies a predetermined condition. A determining unit that determines whether the pitch prediction gain does not satisfy a predetermined condition, and obtains and quantizes a source signal with respect to a residual signal pitch-predicted using the delay when the pitch prediction gain does not satisfy a predetermined condition. A signal encoding device comprising an encoding unit. 6. The signal encoding apparatus according to claim 5, wherein the excitation signal of the signal is composed of a plurality of pulses having a non-zero amplitude. 7. A spectrum parameter calculation unit for obtaining and quantizing a spectrum parameter from an input signal, a mode discrimination unit for extracting a feature amount from the input signal and discriminating a mode, the input signal in a predetermined mode , A first delay corresponding to a first pitch period is obtained, a pitch prediction residual signal is obtained, a first pitch prediction gain is calculated, and a second pitch corresponding to a second pitch period is calculated from the pitch prediction residual signal. A pitch prediction unit that repeats these processes to determine the delay, a determination unit that determines whether the pitch prediction gain satisfies a predetermined condition, and a case where the pitch prediction gain does not satisfy a predetermined condition. A sound source quantizing unit that obtains a sound source signal for the residual signal whose pitch is predicted using the delay, and quantizes and outputs the sound source signal. Signal encoding device. 8. The signal encoding apparatus according to claim 7, wherein the excitation signal of the signal includes a plurality of non-zero amplitude pulses.