JP2003323199A - Encoding device, decoding device, encoding method, and decoding method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To perform high quality encoding and decoding with a low bit rate. <P>SOLUTION: A downsampler 101 downsamples the sampling rate of an input signal from a sampling rate FH to a sampling rate FL. A basic layer encoder 102 encodes a sound signal of the sampling rate FL. A local decoder 103 decodes an encoded code outputted from the basic layer encoder 102. An upsampler 104 increases the sampling rate of a decoded signal to the FH. A subtracter 106 subtracts the decoded signal from the sound signal of the sampling rate FH. An expansion layer encoder 107 uses a parameter of a decoded result outputted from the local decoder 103 to encode a signal outputted from the subtracter 106. <P>COPYRIGHT: (C)2004,JPO

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a coding device, a decoding device, a coding method, and a decoding method for compressing and coding an acoustic signal such as a tone signal or a voice signal with high efficiency, and more particularly, The present invention relates to a coding device, a decoding device, a coding method, and a decoding method suitable for scalable coding and decoding capable of decoding a musical sound or voice even from a part of a coded code.

[0002]

2. Description of the Related Art An acoustic coding technique for compressing a tone signal or a voice signal at a low bit rate is important for the capacity of transmission lines such as radio waves in mobile communication and effective use of a recording medium. The ITU (Inte
There are methods such as G726 and G729 standardized by rnational Telecommunication Union). These methods are
Targeting narrow band signals (300Hz to 3.4kHz), 8kbit / s to 32kb
It can be encoded with high quality at a bit rate of it / s.

In addition, ITU G722, G722.1 and 3GPP (The 3rd) are used as standard methods for wide band signals (50 Hz to 7 kHz).
Generation Partnership Project) AMR-WB exists. These systems have bit rates of 6.6kbit / s to 64k.
A wide-band speech signal can be encoded with high quality at bit / s.

Here, as an effective method for encoding a voice signal at a low bit rate and with high efficiency, CELP (Code Excite) is used.
d Linear Prediction). CELP is a method of performing encoding based on a model of a human voice generation model that is simulated by engineering. Specifically, in CELP, an excitation signal represented by a random number is passed through a pitch filter corresponding to the strength of periodicity and a synthesis filter corresponding to the vocal tract characteristic, and the squared error between the output signal and the input signal is an auditory characteristic. The coding parameters are determined so as to be the minimum under the weighting of.

Most of the recent standard speech coding systems perform coding based on CELP. For example, G7
The 29 can encode a narrow band signal at 8 kbit / s, and the AMR-WB can encode a wide band signal at 6.6 kbit / s to 23.85 kbit / s.

On the other hand, in the case of the musical tone coding for coding the musical tone signal, the musical tone signal is converted into the frequency domain like the layer III system or the AAC system standardized by MPEG (Moving Picture Expert Group). Generally, a method of encoding using a psychoacoustic model is used. It is known that these systems hardly deteriorate at a sampling rate of 44.1 kHz at 64 kbit / s to 96 kbit / s per channel.

This tone encoding is a method for encoding music with high quality. The musical tone encoding can perform high quality encoding of a voice signal having music or environmental sound in the background described above. Also, the band of the target signal can be up to about 22 kHz which is CD quality.

[0008]

However, in the case where a signal mainly composed of a voice signal is encoded using a voice encoding method on a signal in which music or environmental sound is superposed on the background, the music or environmental sound of the background portion is As a result, there is a problem that not only the signal in the background portion but also the audio signal is deteriorated and the overall quality is deteriorated.

The problem is that the speech coding method is CELP.
It occurs because it is based on a method specialized for the voice model. In addition, the signal band that can be supported by the voice encoding method is up to 7 kHz, and there is a problem in that it is not possible to sufficiently support a signal having a component in a band higher than that due to the configuration.

Further, in the tone coding system, it is necessary to use a high bit rate in order to realize high quality coding. In the tone encoding system, the bit rate
There is a problem that the quality of the decoded signal is greatly deteriorated when the encoding is performed while keeping the value as low as 32 kbit / s. for that reason,
There is a problem that it cannot be used in a communication network with a low transmission rate.

The present invention has been made in view of the above points, and it is possible to encode and decode a signal having a low bit rate and a high quality even if the signal is mainly voice and music or environmental sound is superimposed on the background. It is an object of the present invention to provide an encoding device, a decoding device, an encoding method, and a decoding method that can be converted.

[0012]

The encoding device of the present invention is
Down-sampling means for reducing the sampling rate of the input signal, base layer encoding means for encoding the input signal with the reduced sampling rate to obtain a first encoded code, and a decoded signal based on the first encoded code Decoding means, up-sampling means for increasing the sampling rate of the decoded signal to the same rate as the input signal, and parameters generated in the decoding processing of the decoding means, An enhancement layer coding means for coding a difference value with a decoded signal with an increased sampling rate to obtain a second coded code, and a multiplexing means for multiplexing the first coded code and the second coded code. Adopt a configuration that has.

In the coding apparatus of the present invention, the base layer coding means adopts a configuration for coding the input signal using the code excitation linear prediction method.

In the coding apparatus of the present invention, the enhancement layer coding means employs a configuration for coding the input signal using orthogonal transform.

In the coding apparatus of the present invention, the enhancement layer coding means adopts a configuration for coding the input signal using MDCT transform.

According to these configurations, a component having a frequency equal to or lower than a predetermined frequency is extracted from the input signal, encoded suitable for speech encoding, and the result of decoding the obtained encoded code is used for musical tone encoding. By performing suitable encoding, it is possible to perform encoding with high quality at a low bit rate.

In the coding apparatus of the present invention, the enhancement layer coding means employs a configuration in which the LPC coefficient of the base layer generated in the decoding processing of the decoding means is used to perform the coding processing.

In the coding apparatus of the present invention, the enhancement layer coding means transforms the LPC coefficient of the base layer into the LPC coefficient of the enhancement layer based on a preset conversion table, and the spectrum based on the LPC coefficient of the enhancement layer. A configuration is used in which an envelope is calculated and the spectrum envelope is used for at least one of spectrum normalization and vector quantization in encoding processing.

According to these configurations, the LPC coefficient of the enhancement layer is quantized by using the LPC coefficient quantized by the base layer encoder.
By obtaining the coefficient and calculating the spectrum envelope from the LPC analysis of the enhancement layer, there is no need for LPC analysis and quantization, and the number of quantization bits can be reduced.

The coding apparatus of the present invention employs a configuration in which the enhancement layer coding means performs the coding processing by using the pitch period and the pitch gain generated in the decoding processing of the decoding means.

In the coding apparatus of the present invention, the enhancement layer coding means calculates the spectrum fine structure using the pitch period and the pitch gain, and uses the spectrum fine structure for spectrum normalization and vector quantization in the coding process. Adopt a configuration to utilize.

According to these configurations, the spectrum fine structure is calculated using the pitch period coded by the base layer encoder and decoded by the local decoder, and the spectrum fine structure is subjected to spectrum normalization and vectorization. Quantization performance can be improved by utilizing it for quantization.

In the coding apparatus of the present invention, the enhancement layer coding means employs a structure in which the power of the decoded signal generated by the decoding means is used to perform coding processing.

In the coding apparatus of the present invention, the enhancement layer coding means quantizes the fluctuation amount of the power of the MDCT transform coefficient based on the power of the decoded signal, and quantizes the MDCT for power normalization in the coding process. A configuration that utilizes the fluctuation amount of the power of the conversion coefficient is adopted.

According to these configurations, the correlation between the power of the decoded signal of the base layer and the power of the MDCT coefficient of the enhancement layer is used, and the M of the decoded signal of the base layer is used.
By predicting the power of the DCT coefficient and encoding the amount of change from the predicted value, it is possible to reduce the number of bits required to quantize the power of the MDCT coefficient.

The decoding apparatus of the present invention comprises a base layer decoding means for decoding a first coded code to obtain a first decoded signal, and an enhancement layer for decoding a second coded code to obtain a second decoded signal. Decoding means, up-sampling means for increasing the sampling rate of the first decoded signal to the same rate as the second decoded signal, and addition means for adding the first signal and the second signal with the increased sampling rate. And a configuration including

In the decoding device of the present invention, the base layer decoding means adopts a configuration for decoding the first coded code by using the code excitation linear prediction method.

In the decoding device of the present invention, the enhancement layer decoding means employs a configuration for decoding the second coded code using orthogonal transform.

In the decoding device of the present invention, the enhancement layer decoding means employs a configuration for decoding the second coded code by using IMDCT conversion.

According to these configurations, by decoding the enhancement layer decoder using the parameters decoded by the base layer decoder, the enhancement layer coding is performed using the decoding parameters in the base layer coding. The decoded signal can be generated from the coded code of the acoustic coding means for performing.

In the decoding apparatus of the present invention, the enhancement layer decoding means employs a structure for decoding the second coded code by using the LPC coefficient of the base layer.

In the decoding device of the present invention, the enhancement layer decoding means transforms the LPC coefficient of the base layer into the LPC coefficient of the enhancement layer based on a preset conversion table, and the spectrum based on the LPC coefficient of the enhancement layer. The envelope is calculated, and the spectrum envelope is used for vector decoding in the decoding process.

According to these configurations, the LPC coefficient of the enhancement layer is quantized by using the LPC coefficient quantized by the base layer decoder.
By obtaining the coefficient and calculating the spectrum envelope from the LPC analysis of the enhancement layer, there is no need for LPC analysis and quantization, and the number of quantization bits can be reduced.

In the decoding apparatus of the present invention, the enhancement layer decoding means employs a structure in which at least one of the pitch period and the pitch gain is used to perform the decoding process.

In the decoding device of the present invention, the enhancement layer decoding means is configured to calculate the spectrum fine structure using the pitch period and the pitch gain and utilize the spectrum fine structure for vector decoding in the decoding process. take.

According to these configurations, the spectral fine structure is calculated using the pitch period coded by the base layer coder and decoded by the local decoder, and the spectral fine structure is normalized by the spectrum and vectorized. By utilizing it for quantization, it is possible to perform acoustic decoding corresponding to acoustic encoding with improved quantization performance.

In the decoding apparatus of the present invention, the enhancement layer decoding means employs a structure in which the decoding processing is performed by using the power of the decoded signal generated by the decoding means.

In the decoding device of the present invention, the enhancement layer decoding means decodes the variation amount of the power of the MDCT transform coefficient based on the power of the decoded signal, and the decoded MDCT is subjected to the power normalization in the decoding process. A configuration that utilizes the fluctuation amount of the power of the conversion coefficient is adopted.

According to these configurations, the decoder corresponding to the encoder that predicts the power of the MDCT coefficient by using the decoded signal of the base layer and encodes the amount of change from the predicted value is configured. By doing so, the number of bits required to quantize the power of the MDCT coefficient can be reduced.

The acoustic signal transmitting apparatus of the present invention comprises acoustic input means for converting an acoustic signal into an electrical signal and A / A for converting the signal output from the acoustic input means into a digital signal.
D conversion means, the above-mentioned coding device for coding the digital signal output from this A / D conversion means, and RF modulation means for modulating the coded code output from this coding device into a radio frequency signal. And a transmitting antenna for converting the signal output from the RF modulator into a radio wave for transmission.

According to this structure, it is possible to provide an acoustic signal transmitting apparatus which efficiently encodes an acoustic signal with a small number of bits.

The acoustic signal receiving apparatus of the present invention comprises a receiving antenna for receiving a radio wave, an RF demodulating means for demodulating a signal received by the receiving antenna, and the information obtained by the RF demodulating means. Decoding device and D / which converts the signal output from this decoding device into an analog signal
A configuration is provided that includes an A conversion unit and an acoustic output unit that converts an electrical signal output from the D / A conversion unit into an acoustic signal.

According to this structure, since the acoustic signal encoded with a small number of bits can be efficiently decoded, a good acoustic signal can be output.

The communication terminal device of the present invention has a configuration including at least one of the above acoustic signal transmitting device and the above acoustic signal receiving device. A base station apparatus of the present invention has a configuration including at least one of the above acoustic signal transmitting apparatus and the above acoustic signal receiving apparatus.

According to this structure, it is possible to provide an acoustic encoding device which efficiently encodes an acoustic signal with a small number of bits. Further, according to this configuration, since the acoustic signal encoded with a small number of bits can be efficiently decoded, a good acoustic signal can be output.

The encoding method of the present invention comprises the steps of reducing the sampling rate of the input signal, encoding the input signal having the reduced sampling rate to obtain a first encoded code, and based on the first encoded code. Generating a decoded signal by using the parameters obtained in the step of increasing the sampling rate of the decoded signal to the same rate as the input signal, and the parameter obtained in the process of generating the decoded signal, Encoding a difference value with the decoded signal having an increased sampling rate to obtain a second encoded code;
The method further comprises the step of multiplexing the first coded code and the second coded code.

According to this method, a component having a frequency equal to or lower than a predetermined frequency is extracted from the input signal, encoded suitable for voice encoding is performed, and the result of decoding the obtained encoded code is used for suitable tone encoding. By performing the encoding described above, it is possible to perform encoding with high quality at a low bit rate.

The decoding method of the present invention comprises the steps of decoding the first coded code to obtain the first decoded signal, decoding the second coded code to obtain the second decoded signal, and the first step.
The method further comprises a step of increasing the sampling rate of the decoded signal to the same rate as the second decoded signal, and a step of adding the first signal and the second signal with the increased sampling rate.

According to this method, the enhancement layer decoder is decoded using the parameters decoded by the base layer decoder, so that the enhancement layer is coded using the decoding parameters in the base layer coding. The decoded signal can be generated from the coded code of the acoustic coding means.

[0050]

BEST MODE FOR CARRYING OUT THE INVENTION The essence of the present invention has two layers, a base layer and an enhancement layer, and the base layer is based on CELP and has a high quality at a low bit rate in a narrow band or wide band frequency region of an input signal. To be encoded. Next, background music and environmental sounds that cannot be represented by the base layer, and signals with frequency components higher than the frequency domain covered by the base layer are coded in the enhancement layer. It is to have a structure that can handle all signals.

This makes it possible to efficiently encode background music and environmental sounds that cannot be fully expressed by the base layer, and signals having frequency components higher than the frequency range covered by the base layer. At this time, it is a feature of the present invention that the enhancement layer is encoded by using the information obtained from the encoding code of the base layer. By this means, it is possible to obtain the effect that the number of encoded bits in the enhancement layer can be kept low.

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. (Embodiment 1) FIG. 1 is a block diagram showing a configuration of a signal processing apparatus according to Embodiment 1 of the present invention. The signal processing device 100 of FIG. 1 includes a downsampling device 101 and a base layer encoder 102.
, Local decoder 103, and upsampler 10
4, a delay unit 105, a subtractor 106, an enhancement layer encoder 107, and a multiplexer 108.

The down-sampler 101 down-samples the sampling rate of the input signal from the sampling rate FH to the sampling rate FL, and the acoustic signal of the sampling rate FL is base layer encoder 10
Output to 2. Here, the sampling rate FL is a frequency lower than the sampling rate FH.

The base layer encoder 102 encodes the acoustic signal of the sampling rate FL and outputs the encoded code to the local decoder 103 and the multiplexer 108.

Local decoder 103 decodes the coded code output from base layer coder 102 and outputs the decoded signal to upsampling unit 104 and enhancement layer coder 107.

The up-sampler 104 raises the sampling rate of the decoded signal to FH and outputs it to the subtractor 106.

The delay device 105 delays the input acoustic signal of the sampling rate FH for a predetermined time and then subtracts it. This delay time is reduced to 1
01, the base layer encoder 102, and the upsampler 104 have the same value as the time delay, thereby preventing a phase shift in the next subtraction process.

The subtractor 106 has a sampling rate FH.
The decoded signal is subtracted from the acoustic signal of 1 and the subtraction result is output to enhancement layer encoder 107.

The enhancement layer encoder 107 has the subtractor 10
The signal output from 6 is encoded using the parameter of the decoding result output from the local decoder 103, and output to the multiplexer 108. The multiplexer 108 multiplexes the signals encoded by the base layer encoder 102 and the enhancement layer encoder 107, and outputs the multiplexed signals.

Next, the base layer coding and the enhancement layer coding will be described. FIG. 2 is a diagram showing an example of components of an input signal. In FIG. 2, the vertical axis represents the information amount of the signal component, and the horizontal axis represents the frequency. FIG. 2 shows in which frequency band the voice information included in the input signal and the background music / background noise information exist.

A large amount of voice information exists in a low frequency region, and the amount of information decreases toward higher frequencies. On the other hand, the background music / background noise information has a relatively small amount of information in the low frequency range and a large amount of information included in the high frequency range as compared with the voice information.

Therefore, the signal processing apparatus of the present invention uses a plurality of encoding methods and performs different encoding for each area to which each encoding method is suitable.

FIG. 3 is a diagram showing an example of a signal processing method of the signal processing apparatus according to this embodiment. In FIG.
The vertical axis represents the information amount of the component of the signal, and the horizontal axis represents the frequency.

The base layer encoder 102 is designed to efficiently represent the voice information in the frequency band between 0 and FL, and the voice information in this area can be encoded with good quality. However, the background music in the frequency band between 0 and FL
The coding quality of the background noise information is not high. The enhancement layer encoder 107 encodes a signal in a frequency band between FL and FH and a portion that cannot be encoded by the base layer encoder 102.

Therefore, by combining the base layer encoder 102 and the enhancement layer encoder 107, high quality encoding can be realized in a wide band. Furthermore, it is possible to realize a scalable function in which the voice information can be decoded only with the coded code of the base layer coding means.

As described above, useful parameters among the parameters generated by the encoding in the local decoder 103 are given to the enhancement layer encoder 107, and the enhancement layer encoder 10
7 performs encoding using this parameter.

Since this parameter is generated from the coded code, when decoding the signal coded by the signal processing apparatus of the present embodiment, the same parameter can be obtained during the acoustic decoding process. It is not necessary to add a parameter and transmit it to the decoding side. Therefore, the enhancement layer coding means can improve the efficiency of the coding process without increasing the additional information.

For example, among the parameters decoded in the local decoder 103, as the parameters used in the enhancement layer encoder 107, the input signal is a signal having a strong periodicity such as a vowel or a noise characteristic such as a consonant. There is a configuration that uses a voiced / unvoiced flag that indicates a strong signal. Using the voiced / unvoiced flag, in the voiced section, the enhancement layer allocates bits by emphasizing the low range over the high range, and in the unvoiced section, distributing bits by emphasizing the high range over the low range.
Can be adapted.

As described above, according to the signal processing apparatus of the present embodiment, the components having a frequency equal to or lower than the predetermined frequency are extracted from the input signal, the encoding suitable for the speech encoding is performed, and the obtained encoded code is decoded. By performing the encoding suitable for the musical tone encoding using the result, it is possible to perform the encoding with high quality at a low bit rate.

Further, the sampling rates FH and FL only need to satisfy FL> FL, and the values are not limited. For example, the sampling rate is FH = 24kHz, FL = 16k
And can be encoded.

(Embodiment 2) In this embodiment, of the parameters decoded by local decoder 103 of Embodiment 1, the spectrum of the input signal is the parameter used in enhancement layer encoder 107. An example using the represented LPC coefficient will be described.

The signal processing apparatus according to the present embodiment performs coding using CELP in base layer coder 102 in FIG. 1 and uses LPC coefficients representing the spectrum of the input signal in enhancement layer coder 107. Encode.

Here, first, the base layer encoder 10
After the detailed operation of No. 2 is described, the basic configuration of the enhancement layer encoder 107 will be described. The basic configuration here is for simplifying the description of the future embodiments, and refers to a configuration that does not use the encoding parameter of the local decoder 103. After that, the local decoder 103, which is a feature of the present embodiment, decodes the LPC coefficient, and the enhancement layer encoder 107 using this LPC coefficient will be described.

FIG. 4 is a diagram showing an example of the configuration of base layer encoder 102. Base layer encoder 10 of FIG.
2 is an LPC analyzer 401 and a perceptual weighting unit 402.
, Adaptive codebook searcher 403, and adaptive gain quantizer 4
04, a target vector generator 405, a random codebook searcher 406, a noise gain quantizer 407, and a multiplexer 408.

The LPC analyzer 401 finds the LPC coefficient from the input signal sampled at the sampling rate FL in the down-sampling unit 101, and outputs it to the perceptual weighting unit 402.

The perceptual weighting unit 402 includes the LPC analyzer 4
The input signal is weighted based on the LPC coefficient obtained in 01, and the weighted input signal is applied to the adaptive codebook searcher 4
03, adaptive gain quantizer 404, and target vector generator 405.

Adaptive codebook searcher 403 searches the adaptive codebook using the perceptually weighted input signal as a target signal, and searches the adaptive vector searched for by adaptive gain quantizer 404.
To the target vector generator 405. Then, adaptive codebook searcher 403 outputs to adaptive multiplexer 408 the code of the adaptive vector that has the least quantization distortion.

The adaptive gain quantizer 404 quantizes the adaptive gain by which the adaptive vector output from the adaptive codebook searcher 403 is multiplied, and outputs it to the target vector generator 405. Then, the code is output to the multiplexer 408.

The target vector generator 405 vector-subtracts the input signal output from the perceptual weighting unit 402 by the result of multiplying the adaptive vector by the adaptive gain, and the subtraction result is used as the target vector for the noise codebook searcher 406 and the noise gain. Output to the quantizer 407.

The random codebook searcher 406 searches the random codebook for a noise vector having the minimum distortion with respect to the target vector output from the target vector generator 405. Then, the noise codebook searcher 406 gives the searched noise vector to the noise gain quantizer 407 and outputs the code to the multiplexer 408.

Noise gain quantizer 407 quantizes the noise gain by which the noise vector searched by noise codebook searcher 406 is quantized, and outputs the code to multiplexer 408.

The multiplexer 408 multiplexes the LPC coefficient, the adaptive vector, the adaptive gain, the noise vector, and the coded code of the noise gain and outputs them to the local decoder 103 and the multiplexer 108.

Next, the operation of base layer coder 102 in FIG. 4 will be described. First, the signal of the sampling rate FL output from the down sampler 101 is input, and the LPC analyzer 401 obtains the LPC coefficient. The LPC coefficient is converted into a parameter suitable for quantization such as an LSP coefficient and quantized. The coded code obtained by this quantization is given to the multiplexer 408, and the quantized LSP coefficient is calculated from the coded code and converted into the LPC coefficient.

By this conversion, the quantized LPC coefficient is obtained. Using the quantized LPC coefficient, the adaptive codebook, adaptive gain, noise codebook, and noise gain are encoded.

Next, the perceptual weighting unit 402 weights the input signal based on the LPC coefficient obtained by the LPC analyzer 401. This weighting is performed for the purpose of performing spectrum shaping so that the spectrum of the quantization distortion is masked by the spectrum envelope of the input signal.

Next, the adaptive codebook searcher 403 searches the adaptive codebook using the perceptually weighted input signal as the target signal. A signal obtained by repeating a past sound source sequence in a pitch cycle is called an adaptive vector, and the adaptive codebook is configured by the adaptive vector generated in the pitch cycle in a predetermined range.

Let t (n) be the perceptually weighted input signal and pi (n) be the signal obtained by convolving the adaptive vector of the pitch period i with the impulse response of the synthesis filter composed of LPC coefficients. The pitch period i of the adaptive vector that minimizes the evaluation function D of (1) is sent to the multiplexer 408 as a parameter.

[Equation 1] Here, N indicates the vector length.

Next, the adaptive gain quantizer 404 quantizes the adaptive gain by which the adaptive vector is multiplied.
The adaptive gain β is expressed by the following equation (2), and this β is scalar-quantized and the code thereof is sent to the multiplexer 408.

[Equation 2]

Next, the target vector generator 405 subtracts the influence of the adaptive vector from the input signal to generate the target vector used by the noise codebook searcher 406 and the noise gain quantizer 407. Here, pi (n) is a signal obtained by convolving a synthesis filter with an adaptive vector when the evaluation function D represented by Formula 1 is minimized, and βq is a scalar quantization of the adaptive vector β represented by Formula 2. The target vector t2 (n) is represented by the following equation (3), where the quantized value is

[Equation 3] The target vector t2 (n) and the LPC coefficient are given to the random codebook searcher 406, and a random codebook search is performed.

Here, a typical configuration of the random codebook included in the random codebook searcher 406 is an algebraic codebook. The algebraic codebook is represented by a vector having a predetermined very small number of pulses having an amplitude of 1. Further, in the algebraic codebook, possible positions for each pulse are predetermined without overlapping. And the algebraic codebook is the position of the pulse and the sign (polarity) of the pulse.
The feature is that the optimal combination of can be determined with a small amount of calculation.

When the target vector is t2 (n) and the noise vector corresponding to the code j is cj (n), the noise vector index j that minimizes the evaluation function D of the following equation (4) is used as a parameter for the multiplexer. Sent to 408.

[Equation 4]

Next, the noise gain quantizer 407 quantizes the noise gain multiplied by the noise vector.
The noise gain γ is represented by the following equation (5), and this γ is scalar-quantized and the code thereof is sent to the multiplexer 408.

[Equation 5] The multiplexer 408 multiplexes the LPC coefficient, the adaptive codebook, the adaptive gain, the noise codebook, and the coded code of the noise gain that have been sent, and outputs the multiplexed code to the local decoder 103 and the multiplexer 108.

Then, the above processing is repeated while a new input signal is present. If there is no new input signal, the process ends.

Next, the enhancement layer encoder 107 will be described. FIG. 5 is a diagram showing an example of the configuration of enhancement layer encoder 107. Enhancement layer encoder 107 of FIG.
Is an LPC analyzer 501 and a spectrum envelope calculator 50.
2, an MDCT unit 503, a power calculator 504, a power normalizer 505, a spectrum normalizer 506,
It mainly includes a Bark scale normalizer 508, a Bark scale shape calculator 507, a vector quantizer 509, and a multiplexer 510.

The LPC analyzer 501 applies the LPC to the input signal.
Analysis is performed, and the obtained LPC analysis coefficient is output to the spectrum envelope calculator 502 and the multiplexer 510. The spectrum envelope calculator 502 calculates the spectrum envelope from the LPC coefficient and outputs it to the vector quantizer 509.

The MDCT section 503 receives the MDCT for the input signal.
Transform (Modified Discrete Cosine Transform) is performed, and the obtained MDCT coefficient is output to the power calculator 504 and the power normalizer 505. The power calculator 504 calculates the power of the MDCT coefficient, quantizes it, and then outputs it to the power normalizer 505 and the multiplexer 510.

The power normalizer 505 normalizes the MDCT coefficient with the quantized power and outputs the normalized power to the spectrum normalizer 506. The spectrum normalizer 506 normalizes the MDCT coefficient normalized by the power using the spectrum envelope, and outputs it to the Bark scale shape calculator 507 and the Bark scale normalizer 508.

The Bark scale shape calculator 507 calculates the shape of the spectrum which is band-divided at equal intervals on the Bark scale, quantizes the spectrum shape, and outputs the quantized spectrum shape to the Bark scale normalizer 508. It is output to the vector quantizer 509 and the multiplexer 510.

The Bark scale normalizer 508 calculates the
The Bark scale shape B (k) is quantized and the encoded code is output to the multiplexer 510. Then, the Bark scale normalizer 508 decodes the Bark scale shape to generate a normalized MDCT coefficient, and outputs it to the vector quantizer 509.

The vector quantizer 509 vector-quantizes the normalized MDCT coefficient output from the Bark scale normalizer 508 to obtain a representative value with the smallest distortion, and the multiplexer 510 uses this index as an encoding code.
Output to.

The multiplexer 510 multiplexes the coded code and outputs it to the multiplexer 108.

Next, the operation of enhancement layer encoder 107 of FIG. 5 will be described. The subtracted signal obtained by the subtractor 106 of FIG. 1 is LPC analyzed by the LPC analyzer 501. Then, the LPC coefficient is calculated by the LPC analysis. Quantization is performed after converting the LPC coefficient into a parameter suitable for quantization such as an LSP coefficient. The encoding code relating to the LPC coefficient obtained here is the multiplexer 5
Given to 10.

The spectrum envelope calculator 502 calculates the spectrum envelope according to the following equation (6) based on the decoded LPC coefficient.

[Equation 6] Here, α _q is the decoded LPC coefficient, NP is the order of the LPC coefficient, and M is the spectral resolution. The spectrum envelope env (m) obtained by the equation (6) is used by the spectrum normalizer 506 and the vector quantizer 509 described later.

Next, the MDCT section 503 performs MDCT conversion on the input signal to obtain MDCT coefficients. The MDCT transform completely overlaps the adjacent adjacent frames and the analysis frame by half, and uses an orthogonal basis of an odd function in the first half of the analysis frame and an even function in the second half of the analysis frame. is there.
When performing MDCT, a window function such as a sin window is multiplied on the input signal. When the MDCT coefficient is X (m), the MDCT coefficient is calculated according to the following equation (7).

[Equation 7] Here, x (n) represents a signal obtained by multiplying the input signal by the window function.

Next, in the power calculator 504, the MDCT
The power of the coefficient X (m) is obtained and quantized. Then, the power normalizer 505 uses Expression (8) to normalize the MDCT coefficient with the quantized power.

[Equation 8] Here, M indicates the order of the MDCT coefficient. After quantizing the power pow of the MDCT coefficients, this coded code is sent to the multiplexer 510. MD using encoding code
After decoding the power of the CT coefficient, the value is used to MD
The CT coefficient is normalized according to the following equation (9).

[Equation 9] Here, X1 (m) represents the MDCT coefficient after power normalization, and powq represents the power of the MDCT coefficient after quantization.

Next, the spectrum normalizer 506 calculates the MDCT normalized by the power using the spectrum envelope.
Normalize the coefficients. The spectrum normalizer 506 performs normalization according to the following equation (10).

[Equation 10]

Next, the Bark scale shape calculator 507
The spectrum shape is quantized after calculating the shape of the spectrum divided into equal intervals on the Bark scale. The Bark scale shape calculator 507 sends this encoded code to the multiplexer 510 and uses the decoded value thereof to output the MDC which is the output signal of the spectrum normalizer 506.
The T coefficient X2 (m) is normalized. The Bark scale and the Herz scale are associated with each other by the conversion formula represented by the following formula (11).

[Equation 11] Here, B indicates the Bark scale and f indicates the Herz scale. B
The ark scale shape calculator 507 calculates the shape according to the following equation (12) for each of the subbands that are band-divided at equal intervals on the Bark scale.

[Equation 12] Where fl (k) is the lowest frequency of the kth subband and fh (k) is the kth subband.
The maximum frequency of a subband is shown, and K shows the number of subbands.

Then, the Bark scale normalizer 508 is
The Bark scale shape B (k) of each band is quantized, the encoded code is sent to the multiplexer 510, and the Bark scale shape is decoded to generate the normalized MDCT coefficient X3 (m) according to the following expression (13). .

[Equation 13] Here, Bq (k) represents the Bark scale shape after quantization of the k-th subband.

Next, in the vector quantizer 509, Bark
Vector quantization of the output X3 (m) of the scale normalizer 508 is performed. The vector quantizer 509 divides X3 (m) into a plurality of vectors, obtains the representative value with the smallest distortion using the codebook corresponding to each vector, and sends this index to the multiplexer 510 as an encoded code.

The vector quantizer 509 determines two important parameters when performing vector quantization by using the spectrum information of the input signal. The parameters are
One is quantized bit allocation, and the other is weighting in codebook search. The quantized bit allocation is determined using the spectrum envelope env (m) obtained by the spectrum envelope calculator 502.

When the quantized bit allocation is determined using the spectrum envelope env (m), the number of bits allocated to the spectrum corresponding to frequencies 0 to FL can be set to be small.

As one implementation example thereof, frequencies 0 to FL
Set the maximum number of bits MAX_LOWBAND_BIT that can be allocated to
The maximum number of bits allocated to this band is MAX_LOWB
There is a way to set a limit so that AND_BIT is not exceeded.

In this example of implementation, since coding is already performed in the base layer in frequencies 0 to FL, it is not necessary to allocate a large number of bits, and the quantization in this band is intentionally coarsened to allocate bits. Can be reduced, and the extra bits there can be allocated to the frequencies FL to FH and quantized to improve the overall quality. Further, the bit allocation may be determined by combining the spectrum envelope env (m) and the Bark scale shape Bq (k) described above.

The spectrum envelope env (m) obtained by the spectrum envelope calculator 502 and the quantized Bark scale shape Bq (k) obtained by the Bark scale shape calculator 507.
Vector quantization is performed using the distortion measure using the weighting calculated from Vector quantization is realized by obtaining the index j of the code vector C that minimizes the distortion D defined by the following equation (14).

[Equation 14] Here, w (m) indicates a weighting coefficient.

The weighting function w (m) is the spectral envelope env
It can be expressed as the following Expression (15) using (m) and the Bark scale shape Bq (k).

[Equation 15] Here, p is a constant between 0 and 1, and Herz_to_Bark () is a function for converting the Herz scale to the Bark scale.

When determining the weighting function w (m),
It is also possible to set the weighting function to be distributed to the spectrum corresponding to frequencies 0 to FL to be small. As one implementation example thereof, the maximum value of the weighting function w (m) corresponding to the frequencies 0 to FL is preset as MAX_LOWBAND_WGT, and the value of the weighting function w (m) in this band is MA.
There is a way to set a limit so that it does not exceed X_LOWBAND_WGT. In this implementation example, coding is already performed in the base layer in frequencies 0 to FL, and the precision of quantization in this band is intentionally lowered to relatively improve the precision of quantization in frequencies FL to FH. Can improve the overall quality.

Finally, the multiplexer 510 multiplexes the coded code and outputs it to the multiplexer 108. Then, the above processing is repeated while a new input signal is present. If there is no new input signal, the process ends.

As described above, according to the signal processing apparatus of the present embodiment, a component having a frequency equal to or lower than a predetermined frequency is extracted from the input signal, encoding is performed using the code excitation linear prediction method, and the obtained encoded code is obtained. It is possible to perform encoding with high quality at a low bit rate by performing encoding by MDCT conversion using the result of decoding.

In the above, an example in which the LPC analysis coefficient is analyzed from the subtraction signal obtained by the subtractor 106 has been described. However, the signal processing apparatus of the present invention uses the local decoder 10
You may encode using the LPC coefficient decoded in 3.

FIG. 6 is a diagram showing an example of the configuration of enhancement layer encoder 107. However, components having the same configurations as those in FIG. 5 are denoted by the same reference numerals as those in FIG. 5, and detailed description thereof will be omitted.

The enhancement layer encoder 107 of FIG. 6 includes a conversion table 601, an LPC coefficient mapping unit 602,
5 is different from enhancement layer encoder 107 in FIG. 5 in that it includes spectrum envelope calculator 603 and transformation unit 604 and performs encoding using the LPC coefficient decoded by local decoder 103.

The conversion table 601 is the LP of the base layer.
The C coefficient and the LPC coefficient of the enhancement layer are stored in association with each other.

The LPC coefficient mapping unit 602 refers to the conversion table 601 and sets the LPC coefficient of the base layer input from the base layer encoder 102 to the LP of the enhancement layer.
It is converted into a C coefficient and output to the spectrum envelope calculator 603.

The spectrum envelope calculator 603 obtains the spectrum envelope based on the LPC coefficient of the enhancement layer and outputs it to the transforming unit 604. The transformation unit 604 transforms the spectrum envelope and outputs it to the spectrum normalizer 506 and the vector quantizer 509.

Next, the operation of enhancement layer encoder 107 of FIG. 6 will be described. The LPC coefficient of the base layer is obtained for a signal having a signal band of 0 to FL, and does not match the LPC coefficient used for the target signal of the enhancement layer (signal band of 0 to FH). However, there is a strong correlation between the two. Therefore, the LPC coefficient mapping unit 602 uses this correlation to advance the signal band 0
A conversion table 601 representing the correspondence between the LPC coefficients for signals of FL to the signals and the LPC coefficients for signals of signal bands 0 to FH.
Is designed separately. Using this conversion table 601, the LPC coefficient of the enhancement layer is obtained from the LPC coefficient of the base layer.

FIG. 7 is a diagram showing an example of the extended LPC coefficient calculation. The conversion table 601 is the LPC of the enhancement layer.
J candidates {Yj (m)} representing the coefficient (order M) and the same order (= K) as the LPC coefficient of the base layer associated with [Yj (m)]
It is composed of candidates {yj (k)} with. {Yj (m)} and {yj {k}} are designed and prepared in advance from large-scale musical sounds and voice data. When the LPC coefficient x (k) of the base layer is input, the LPC coefficient most similar to x (k) is obtained from {yj (k)}. By outputting the LPC coefficient Yj (m) of the enhancement layer corresponding to the index j of the LPC coefficient determined to be most similar, it is possible to realize mapping of the LPC coefficient of the enhancement layer from the LPC coefficient of the base layer. it can.

Next, the spectrum envelope calculator 603 obtains the spectrum envelope based on the LPC coefficient of the enhancement layer determined in this way. Then, the spectrum envelope is transformed by the transformation unit 604. Then, the modified spectrum envelope is regarded as the spectrum envelope of the above-described embodiment and is processed.

As one implementation example of the transforming unit 604 for transforming the spectrum envelope, there is a process for reducing the influence of the spectrum envelope corresponding to the signal bands 0 to FL that are the object of coding of the base layer. When the spectral envelope is env (m), the modified spectral envelope env '(m) is given by the following equation (1)
It is represented by 6).

[Equation 16] Here, p represents a constant between 0 and 1.

In frequencies 0 to FL, the base layer has already been encoded, and the spectrum of frequencies 0 to FL of the subtraction signal to be encoded in the enhancement layer is close to flat. Nevertheless, the LP as described in this embodiment
Such effects are not taken into consideration in the mapping of the C coefficient. Therefore, it is possible to improve the quality by using the method of correcting the spectrum envelope using the equation (16).

As described above, according to the signal processing apparatus of the present embodiment, the LPC coefficient of the enhancement layer is obtained using the LPC coefficient quantized by the base layer encoder, and the spectrum envelope is calculated from the LPC analysis of the enhancement layer. By calculating, L
The need for PC analysis and quantization is eliminated, and the number of quantization bits can be reduced.

(Embodiment 3) FIG.8 is a block diagram showing a configuration of an enhancement layer encoder of a signal processing apparatus according to Embodiment 3 of the present invention. However, components having the same configurations as those in FIG. 5 are denoted by the same reference numerals as those in FIG. 5, and detailed description thereof will be omitted.

The enhancement layer encoder 107 of FIG. 8 comprises a spectrum fine structure calculator 801 and uses the pitch period encoded by the base layer encoder 102 and decoded by the local decoder 103. The difference from the enhancement layer encoder of FIG. 5 is that the structure is calculated and the spectral fine structure is utilized for normalization and vector quantization of the spectrum.

The spectrum fine structure calculator 801 calculates the spectrum fine structure from the pitch period T and pitch gain β encoded in the base layer, and the spectrum normalizer 5
It outputs to 06.

Specifically, the pitch period T and the pitch gain β are a part of the coded code, and the same information can be obtained in an acoustic decoder not shown here. Therefore, the bit rate does not increase even if the encoding is performed using the pitch period T and the pitch gain β.

The spectrum fine structure calculator 801 calculates the spectrum fine structure har (m) according to the following equation (17) using the pitch period T and the pitch gain β.

[Equation 17] Here, M indicates the spectral resolution. Equation (17) is β
Since it becomes an oscillation filter when the absolute value of is 1 or more, the range of the absolute value of β can be set in advance.
There is also a method of setting a limit so that the set value is less than 1 (eg 0.8) or less.

The spectrum normalizer 506 calculates the spectrum envelope env (m) calculated by the spectrum envelope calculator 502.
And the spectrum fine structure har (m) obtained by the spectrum fine structure calculator 801 are used to perform normalization according to the following equation (18).

[Equation 18]

The distribution of quantized bits in the vector quantizer 509 is such that the spectrum envelope env (m) obtained by the spectrum envelope calculator 502 and the spectrum fine structure har (m) obtained by the spectrum fine structure calculator 801. Both are used to determine. Also, the weighting function for vector quantization
Spectral fine structure is also used to determine w (m). Specifically, the weighting function w (m) is defined according to the following equation (19).

[Formula 19] Here, p is a constant between 0 and 1, and Herz_to_Bark () is a function for converting the Herz scale to the Bark scale.

As described above, the signal processing apparatus of the present embodiment calculates the spectrum fine structure using the pitch period coded by the base layer coder and decoded by the local decoder, and the spectrum fine structure is calculated. Can be used for normalization of the spectrum and vector quantization to improve the quantization performance.

(Embodiment 4) FIG. 9 is a block diagram showing the configuration of an enhancement layer encoder of a signal processing apparatus according to Embodiment 4 of the present invention. However, components having the same configurations as those in FIG. 5 are denoted by the same reference numerals as those in FIG. 5, and detailed description thereof will be omitted.

The enhancement layer coder 107 of FIG. 9 comprises a power estimator 901 and a power fluctuation amount quantizer 902, and performs local decoding using the coding code obtained by the base layer coder 102. 5 is that the decoded signal is generated in the device 103, the power of the MDCT coefficient is predicted from the decoded signal of the base layer using the decoded signal, and the amount of change from the predicted value is coded. different.

The signal sl (n) decoded by the local decoder 103 in FIG. 5 is input to the power estimator 901. Then, the power estimator 901 estimates the power of the MDCT coefficient from the decoded signal sl (n). MDC
When the estimated value of the power of the T coefficient is powp, powp is expressed by the following equation (20).

[Equation 20] Here, N is the length of the decoded signal sl (n), and α is a predetermined constant for correction. Further, in another method using the spectrum slope obtained from the LPC coefficient of the base layer, the estimated power value of the MDCT coefficient is expressed by the following equation (21).

[Equation 21] Here, β represents a variable depending on the spectrum slope obtained from the LPC coefficient of the base layer. When the spectrum slope is large (power is relatively in the low frequency band), β approaches zero and the spectrum slope is small ( Β has a property of approaching 1 when the power is relatively high.

Next, in the power fluctuation amount quantizer 902,
The power of the MDCT coefficient obtained by the MCDT unit 503 is normalized by the power estimation value powp obtained by the power estimator 901, and the fluctuation amount thereof is quantized. The fluctuation amount r is expressed by the following equation (22).

[Equation 22] Here, pow represents the power of the MDCT coefficient, and is expressed by equation (23).
Calculated at.

[Equation 23] Here, X (m) indicates the MDCT coefficient, and M indicates the frame length. The power fluctuation amount quantizer 902 quantizes the fluctuation amount r,
The encoded code is sent to the multiplexer 510, and the quantized fluctuation amount rq is decoded. In the power normalizer 505, the MDCT coefficient is normalized using the following expression (24) using the quantized change amount rq.

[Equation 24] Here, X1 (m) represents the MDCT coefficient after power normalization.

As described above, the signal processing apparatus according to the present embodiment has the power of the decoded signal of the base layer and the MD of the enhancement layer.
By using the correlation with the power of the CT coefficient, the power of the MDCT coefficient is predicted using the decoded signal of the base layer, and the variation amount from the predicted value is coded to obtain the MDC.
The number of bits required to quantize the power of the T coefficient can be reduced.

(Embodiment 5) FIG. 10 is a block diagram showing a configuration of a signal processing apparatus according to Embodiment 5 of the present invention. The signal processing apparatus 1000 of FIG. 10 includes a demultiplexer 1001, a base layer decoder 1002, an upsampler 1003, and an enhancement layer decoder 10.
04 and an adder 1005.

The demultiplexer 1001 separates the coded code to generate a coded code for the base layer and a coded code for the enhancement layer. Then, the demultiplexer 1001 outputs the coding code for the base layer to the base layer decoder 1002, and outputs the coding code for the enhancement layer to the enhancement layer decoder 1004.

The base layer decoder 1002 decodes the decoded signal of the sampling rate FL using the base layer coding code obtained by the demultiplexer 1001 and outputs the decoded signal to the up sampler 1003. At the same time, the parameters decoded by the base layer decoder 1002 are output to the enhancement layer decoder 1004. The up-sampler 1003 raises the sampling frequency of the decoded signal to FH and outputs it to the adder 1005.

The enhancement layer decoder 1004 decodes the decoded signal of the sampling rate FH using the enhancement layer coding code obtained in the demultiplexer 1001 and the parameters decoded in the base layer decoder 1002, Output to the adder 1005.

Adder 1005 vector-adds the decoded signal output from up-sampler 1003 and the decoded signal output from enhancement layer decoder 1004.

Next, the operation of the signal processing apparatus of this embodiment will be described. First, a code encoded by the signal processing device according to any of the first to fourth embodiments is input, and the demultiplexer 1001 separates the code to encode a base layer encoding code and an enhancement layer encoding. Generate code.

Next, in the base layer decoder 1002,
The decoded signal of the sampling rate FL is decoded by using the base layer coding code obtained by the demultiplexer 1001. Then, the upsampler 1003
Raises the sampling frequency of the decoded signal to FH.

The enhancement layer decoder 1004 decodes the decoded signal of the sampling rate FH using the enhancement layer coding code obtained by the demultiplexer 1001 and the parameters decoded by the base layer decoder 1002. .

An adder 1005 adds the decoded signal of the base layer and the decoded signal of the enhancement layer which are up-sampled by the up-sampling device 1003. Then, the above processing is repeated while a new input signal is present. If there is no new input signal,
The process ends.

As described above, the signal processing apparatus according to the present embodiment performs decoding in base layer coding by performing decoding in enhancement layer decoder 1004 using the parameters decoded in base layer decoder 1002. The decoded signal can be generated from the coded code of the acoustic coding means for coding the enhancement layer using the parameter.

Next, base layer decoder 1002 will be described. FIG. 11 is a block diagram showing an example of base layer decoder 1002. The base layer decoder 1002 of FIG. 11 is mainly composed of a demultiplexer 1101, an excitation generator 1102, and a synthesis filter 1103, and performs CELP decoding processing.

The demultiplexer 1101 separates various parameters from the coded code for the base layer output from the demultiplexer 1001, and the excitation generator 1102.
And output to the synthesis filter 1103.

The sound source generator 1102 decodes the adaptive vector, the adaptive vector gain, the noise vector, and the noise vector gain, generates a sound source signal using these, and outputs the sound source signal to the synthesis filter 1103. The synthesis filter 1103 generates a synthesized signal using the decoded LPC coefficient.

Next, the base layer decoder 100 shown in FIG.
The operation of No. 2 will be described. First, the demultiplexer 1101 separates various parameters from the coded code for the base layer.

Next, the sound source generator 1102 decodes the adaptive vector, adaptive vector gain, noise vector, and noise vector gain. Then, the sound source generator 1102 generates a sound source vector ex (n) according to the following equation (25).

[Equation 25] Here, q (n) is an adaptive vector, β _q is an adaptive vector gain, c (n) is a noise vector, and γ _q is a noise vector gain.

Next, the synthesizing filter 1103 uses the decoded LPC coefficient to synthesize the synthesized signal syn (n) with the following equation (26).
According to.

[Equation 26] Here, α _q is the decoded LPC coefficient, and NP is the order of the LPC coefficient.

The decoded signal syn (n) thus decoded is
It is output to the upsampler 1003 and the enhancement layer decoder 1004. Then, the above processing is repeated while a new input signal is present. If there is no new input signal, the process ends. Depending on the configuration of CELP, there may be a form in which the combined signal is output after passing through a post filter. The post filter mentioned here has a post-processing function that makes it difficult to perceive the coding distortion.

Next, enhancement layer decoder 1004 will be described. FIG. 12 is a block diagram showing an example of enhancement layer decoder 1004. The enhancement layer decoder 1004 of FIG. 12 includes a demultiplexer 1201 and an LPC.
Coefficient decoder 1202 and spectrum envelope calculator 120
3, a vector decoder 1204, a Bark scale shape decoder 1205, a multiplier 1206, and a multiplier 120.
7, a power decoder 1208, a multiplier 1209,
It is mainly composed of the IMDCT unit 1210.

The demultiplexer 1201 separates various parameters from the coding code for the enhancement layer output from the demultiplexer 1001. The LPC coefficient decoder 1202 decodes the LPC coefficient using the coding code related to the LPC coefficient, and the spectrum envelope calculator 1203.
Output to.

The spectrum envelope calculator 1203 uses the decoded LPC coefficient to calculate the spectrum envelope according to equation (6).
env (m) is calculated and output to the vector decoder 1204 and the multiplier 1207.

The vector decoder 1204 has a spectrum envelope calculated by the spectrum envelope calculator 1203.
The quantized bit allocation is determined based on env (m), and the normalized MDCT coefficient X3q (m) is decoded from the coded code obtained from the demultiplexer 1201 and the quantized bit allocation. The method of quantized bit allocation is the same as the method used in the enhancement layer coding in any of the coding methods of the first to fourth embodiments.

The Bark scale shape decoder 1205 decodes the Bark scale shape Bq (k) based on the coded code obtained from the demultiplexer 1201 and outputs it to the multiplier 1206.

The multiplier 1206 calculates the normalized MDCT coefficient X3q (m) and the Bark scale shape Bq according to the following equation (27).
(k) is multiplied and the multiplication result is output to the multiplier 1207.

[Equation 27] Where fl (k) is the lowest frequency of the kth subband and fh (k) is the kth subband.
It represents the highest frequency of a subband, and K represents the number of subbands.

The multiplier 1207 calculates the normalized MDCT coefficient X2 obtained from the multiplier 1206 according to the following equation (28).
(m) is multiplied by the spectrum envelope env (m) obtained by the spectrum envelope calculator 1203, and the multiplication result is multiplied by the multiplier 1
Output to 209.

[Equation 28] The power decoder 1208 includes a demultiplexer 1201.
Decode the power powq based on the obtained code,
The decoding result is output to the multiplier 1209.

The multiplier 1209 multiplies the normalized MDCT coefficient X1 (m) by the decoding power powq according to the following equation (29), and outputs the multiplication result to the IMDCT section 1210.

[Equation 29] The IMDCT unit 1210 transforms the decoded MDCT coefficients obtained in this way into IMDCT transform (Modified Discrete).
Cosine Transform: Inverse modified cosine transform) is applied, the signal decoded in the previous frame is overlapped with the half of the analysis frame and added to generate an output signal, and this output signal is output to the adder 1005. Then, the above processing is repeated while a new input signal is present. If there is no new input signal, the process ends.

As described above, according to the signal processing apparatus of the present embodiment, the decoding of the enhancement layer decoder is performed using the parameters decoded by the base layer decoder,
A decoded signal can be generated from the coding code of the acoustic coding means for coding the enhancement layer using the decoding parameter in the base layer coding.

(Embodiment 6) FIG. 13 is a diagram showing an example of the configuration of enhancement layer decoder 1004. However, the same components as those in FIG. 12 are designated by the same reference numerals as those in FIG. 12, and detailed description thereof will be omitted.

The enhancement layer decoder 1004 of FIG.
Conversion table 1301 and LPC coefficient mapping unit 13
02, the spectrum envelope calculator 1303, and the transformation unit 13
04, and the LP decoded by the local decoder 103.
It differs from enhancement layer decoder 1004 in FIG. 12 in that decoding is performed using C coefficient.

The conversion table 1301 is the L of the base layer.
The PC coefficient and the LPC coefficient of the enhancement layer are stored in association with each other.

The LPC coefficient mapping unit 1302 refers to the conversion table 1301 and refers to the base layer decoder 100.
The LPC coefficient of the base layer input from 2 is converted into the LPC coefficient of the enhancement layer, and the spectrum envelope calculator 1303
Output to.

Spectral envelope calculator 1303 finds a spectral envelope based on the LPC coefficient of the enhancement layer, and outputs it to transforming section 1304. The transformation unit 1304 transforms the spectrum envelope, and the multiplier 1207 and the vector decoder 1
Output to 204. For example, as a modification method, there is a method represented by the equation (16) of the second embodiment.

Next, the enhancement layer decoder 100 shown in FIG.
The operation of No. 4 will be described. The LPC coefficient of the base layer is obtained for a signal having a signal band of 0 to FL, and is a target signal of the enhancement layer (signal band of 0 to F).
It does not match the LPC coefficient used in H). However, there is a strong correlation between the two. Therefore, the LPC coefficient mapping unit 1302 uses this correlation to preliminarily use signal LPC coefficients for signal bands 0 to FL and signal bands 0 to F.
The conversion table 1301 representing the correspondence with the H signal LPC coefficient is separately designed. This conversion table 13
01 is used to find the LPC coefficient of the enhancement layer from the LPC coefficient of the base layer.

Details of the conversion table 1301 are the same as those of the conversion table 601 of the second embodiment.

As described above, according to the signal processing apparatus of the present embodiment, the LPC coefficient of the enhancement layer is obtained using the LPC coefficient quantized by the base layer decoder, and the spectrum envelope is calculated from the LPC analysis of the enhancement layer. By calculating, L
The need for PC analysis and quantization is eliminated, and the number of quantization bits can be reduced.

(Embodiment 7) FIG.14 is a block diagram showing the configuration of an enhancement layer decoder of a signal processing apparatus according to Embodiment 7 of the present invention. However, the same components as those in FIG. 12 are designated by the same reference numerals as those in FIG. 12, and detailed description thereof will be omitted.

The enhancement layer decoder 1004 of FIG.
The spectrum fine structure calculator 1401 is provided, the spectrum fine structure is calculated using the pitch period decoded by the base layer decoder 1002, the spectrum fine structure is utilized for decoding, and the quantization performance is improved. The difference from the enhancement layer encoder in FIG. 12 is that acoustic decoding corresponding to the acoustic encoding is performed.

The spectrum fine structure calculator 1401 has the pitch period T decoded by the base layer decoder 1002.
And the spectrum fine structure is calculated from the pitch gain β and output to the vector decoder 1204 and the multiplier 1207.

Specifically, the pitch period T and the pitch gain β are a part of the coded code, and the same information can be obtained in an acoustic decoder not shown here. Therefore, the bit rate does not increase even if the encoding is performed using the pitch period T and the pitch gain β.

In the spectrum fine structure calculator 1401,
Using pitch period T and pitch gain β, the following equation (17)
The spectral fine structure har (m) is calculated according to

[Equation 30] Here, M indicates the spectral resolution. Equation (17) is β
Since it becomes an oscillation filter when the absolute value of is 1 or more, the range of the absolute value of β can be set in advance.
A limit may be set so that the set value is less than 1 (for example, 0.8) or less.

Then, the spectrum envelope env (m) obtained by the spectrum envelope calculator 1203 and the spectrum fine structure har (m) obtained by the spectrum fine structure calculator 1401.
Is used to determine the quantized bit allocation in vector decoder 1204. Then, the normalized MDCT coefficient X3 (m) is decoded from the quantized bit allocation and the coded code obtained from the demultiplexer 1201. Further, in the multiplier 1207, the normalized MD is calculated according to the following equation (30).
The normalized MDCT coefficient X1 (m) is obtained by multiplying the CT coefficient X2 (m) by the spectral envelope env (m) and the spectral fine structure har (m).

[Equation 31] As described above, the signal processing apparatus according to the present embodiment calculates the spectrum fine structure using the pitch period coded by the base layer encoder and decoded by the local decoder, and the spectrum fine structure By utilizing it for normalization and vector quantization, it is possible to perform acoustic decoding corresponding to acoustic encoding with improved quantization performance.

(Embodiment 8) FIG.15 is a block diagram showing the configuration of an enhancement layer decoder of a signal processing apparatus according to Embodiment 8 of the present invention. However, the same components as those in FIG. 12 are designated by the same reference numerals as those in FIG. 12, and detailed description thereof will be omitted.

The enhancement layer decoder 1004 of FIG.
Power estimator 1501 and power variation decoder 150
2 and a power generator 1503, a decoder corresponding to an encoder that predicts the power of the MDCT coefficient using the decoded signal of the base layer and encodes the amount of change from the predicted value. The configuration is different from the enhancement layer decoder of FIG.

Further, in FIG. 10, the parameters decoded from base layer decoder 1002 to enhancement layer decoder 1004 are output, but in this embodiment,
Further, the decoded signal obtained in base layer decoder 1002 is output to enhancement layer decoder 1004.

The power estimator 1501 estimates the power of the MDCT coefficient from the decoded signal sl (n) decoded by the base layer decoder 1002 using the equation (20) or the equation (21).

The power change amount decoder 1502 decodes the power change amount from the coded code obtained from the demultiplexer 1201 and outputs it to the power generator 1503. The power generator 1503 calculates power from the power change amount.

The multiplier 1209 calculates the MDCT coefficient according to the following equation (31).

[Equation 32] Here, rq is the decoded value of the power change amount, and powp is the estimated power value. Further, X1 (m) represents the output signal of the multiplier 1207.

As described above, according to the signal processing apparatus of the present embodiment, the encoder that predicts the power of the MDCT coefficient using the decoded signal of the base layer and encodes the amount of change from the predicted value. By configuring the decoder corresponding to, it is possible to reduce the number of bits required to quantize the power of the MDCT coefficient.

(Ninth Embodiment) Next, a ninth embodiment of the present invention will be described with reference to the drawings. FIG. 16 is a block diagram showing the configuration of the acoustic signal coding apparatus according to Embodiment 9 of the present invention. Signal processing device 1 in FIG.
A feature of the present embodiment is that 603 is configured by one of the signal processing devices described in the first to fourth embodiments.

As shown in FIG. 16, communication apparatus 1600 according to Embodiment 9 of the present invention has input apparatus 1601, A /
The signal processing apparatus 1603 connected to the D conversion apparatus 1602 and the network 1604 is provided.

The A / D converter 1602 is the input device 16
01 output terminal. Signal processing device 160
The input terminal of No. 3 is connected to the output terminal of the A / D converter 1602. The output terminal of the signal processing device 1603 is connected to the network 1604.

The input device 1601 converts the sound wave audible to the human ear into an analog signal which is an electrical signal and A / D
It is given to the conversion device 1602. The A / D converter 1602 converts an analog signal into a digital signal and gives it to the signal processor 1603. The signal processing device 1603 encodes the input digital signal to generate a code and outputs it to the network 1604.

As described above, according to the communication device of the embodiment of the present invention, the effects as shown in the above-described first to fourth embodiments can be enjoyed in communication, and the acoustic signal can be coded efficiently with a small number of bits. It is possible to provide an audio encoding device for encoding.

(Tenth Embodiment) Next, a tenth embodiment of the present invention will be described with reference to the drawings. FIG. 17
FIG. 11 is a block diagram showing the configuration of an acoustic signal decoding device according to Embodiment 10 of the present invention. The signal processing device 1703 in FIG. 17 is characterized in that it is configured by one of the signal processing devices shown in the fifth to eighth embodiments.

As shown in FIG. 17, communication apparatus 1700 according to Embodiment 10 of the present invention has network 1701.
1702 and signal processing device 17 connected to the
03, and D / A conversion device 1704 and output device 170
It is equipped with 5.

The input terminal of the receiving device 1702 is connected to the network 1701. Signal processing device 1703
The input terminal of is connected to the output terminal of the receiving device 1702. The input terminal of the D / A conversion device 1704 is connected to the output terminal of the signal processing device 1703. The input terminal of the output device 1705 is connected to the output terminal of the D / A conversion device 1704.

The receiving device 1702 is connected to the network 170.
It receives the digital encoded acoustic signal from 1 to generate a digital received acoustic signal and gives it to the signal processing device 1703. The signal processing device 1703 receives the reception acoustic signal from the reception device 1702, performs decoding processing on the reception acoustic signal, generates a digital decoding acoustic signal, and outputs the D / A signal.
It is given to the conversion device 1704. D / A converter 1704
Converts the digital decoded audio signal from the signal processing device 1703 to generate an analog decoded audio signal and gives it to the output device 1705. The output device 1705 converts an analog decoded acoustic signal which is an electrical signal into vibration of air and outputs it as a sound wave so that it can be heard by the human ear.

As described above, according to the communication apparatus of the present embodiment, it is possible to enjoy the effects as shown in the above-described fifth to eighth embodiments in communication, and the acoustic signal efficiently encoded with a small number of bits. Can be decoded, so that a good acoustic signal can be output.

(Eleventh Embodiment) Next, an eleventh embodiment of the present invention will be described with reference to the drawings. FIG.
FIG. 16 is a block diagram showing the configuration of an acoustic signal transmission encoding apparatus according to Embodiment 11 of the present invention. In Embodiment 11 of the present invention, the signal processing device 1803 in FIG.
Is characterized in that this embodiment is configured by one of the acoustic coding means shown in the first to fourth embodiments.

As shown in FIG. 18, communication apparatus 1800 according to Embodiment 11 of the present invention has input apparatus 1801, A
The D / D converter 1802, the signal processor 1803, the RF modulator 1804, and the antenna 1805 are provided.

The input device 1801 converts a sound wave audible to the human ear into an analog signal which is an electric signal and supplies it to the A / D conversion device 1802. The A / D conversion device 1802 converts an analog signal into a digital signal to convert the signal processing device 1
Give to 803. The signal processing device 1803 encodes the input digital signal to generate a coded acoustic signal, and supplies it to the RF modulator 1804. RF modulator 18
04 modulates the coded acoustic signal to generate a modulation coded acoustic signal, and supplies it to the antenna 1805. Antenna 180
5 transmits the modulation coded acoustic signal as a radio wave.

As described above, according to the communication apparatus of the present embodiment, it is possible to enjoy the effects as shown in the above-described first to fourth embodiments in wireless communication, and efficiently encode an acoustic signal with a small number of bits. can do.

The present invention can be applied to a transmitter, a transmission encoder, or an acoustic signal encoder that uses an audio signal. The present invention can also be applied to a mobile station device or a base station device.

(Twelfth Embodiment) Next, a twelfth embodiment of the present invention will be described with reference to the drawings. FIG. 19
[Fig. 13] is a block diagram showing a configuration of an acoustic signal reception decoding apparatus according to Embodiment 12 of the present invention. In Embodiment 12 of the present invention, the signal processing device 1903 in FIG.
Is characterized in that it is configured by one of the acoustic decoding means shown in the above-described fifth to eighth embodiments.

As shown in FIG. 19, communication apparatus 1900 according to Embodiment 12 of the present invention includes antennas 1901 and R.
The F demodulator 1902, the signal processor 1903, the D / A converter 1904, and the output device 1905 are provided.

The antenna 1901 receives a digital encoded acoustic signal as a radio wave, generates a digital received encoded acoustic signal of an electric signal, and gives it to the RF demodulator 1902. The RF demodulation device 1902 demodulates the reception coded acoustic signal from the antenna 1901 to generate a demodulation coded acoustic signal and gives it to the signal processing device 1903.

The signal processing device 1903 is the RF demodulation device 1
The digital demodulated coded acoustic signal from 902 is received and subjected to a decoding process to generate a digital decoded acoustic signal, which is given to the D / A converter 1904. The D / A conversion device 1904 converts the digital decoded audio signal from the signal processing device 1903 to generate an analog decoded audio signal and gives it to the output device 1905. Output device 1905
Converts an analog decoded audio signal, which is an electrical signal, into vibration of air and outputs it as a sound wave so that it can be heard by the human ear.

As described above, according to the communication apparatus of the present embodiment, it is possible to enjoy the effects shown in the above-described fifth to eighth embodiments in wireless communication, and the audio coded efficiently with a small number of bits. Since the signal can be decoded, a good acoustic signal can be output.

The present invention can be applied to a receiving device, a receiving / decoding device or a voice signal decoding device which uses an audio signal. The present invention can also be applied to a mobile station device or a base station device.

Further, the present invention is not limited to the above-mentioned embodiment, but can be implemented with various modifications. For example,
In the above-described embodiment, the case where the signal processing device is used is described, but the present invention is not limited to this, and the signal processing method can be performed as software.

For example, a program for executing the above signal processing method is stored in advance in a ROM (Read Only Memory), and the program is stored in a CPU (Central Processor Uni
You may make it operate | moved by t).

Further, a program for executing the above signal processing method is stored in a computer-readable storage medium, the program stored in the storage medium is recorded in a RAM (Random Access Memory) of the computer, and the computer is stored in the program. You may make it operate | move according to.

Further, in the above description, the MDCT transform is used for the enhancement layer coding and the IMDCT transform is used for the enhancement layer decoding, but the present invention is not limited to this, and any orthogonal transform method can be applied.

[0216]

As described above, according to the encoding device, the decoding device, the encoding method, and the decoding method of the present invention,
By encoding the enhancement layer by using the information obtained from the coding code of the base layer, even if the signal is mainly voice and music or noise is superimposed on the background, it is possible to improve at a low bit rate. Quality can be encoded.

[Brief description of drawings]

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

FIG. 2 is a diagram showing an example of components of an input signal.

FIG. 3 is a diagram showing an example of a signal processing method of the signal processing device according to the above embodiment.

FIG. 4 is a diagram showing an example of the configuration of a base layer encoder.

FIG. 5 is a diagram showing an example of a configuration of an enhancement layer encoder.

FIG. 6 is a diagram showing an example of the configuration of an enhancement layer encoder.

FIG. 7 is a diagram showing an example of extended LPC coefficient calculation.

FIG. 8 is a block diagram showing the configuration of an enhancement layer encoder of the signal processing apparatus according to Embodiment 3 of the present invention.

FIG. 9 is a block diagram showing the configuration of an enhancement layer encoder of the signal processing apparatus according to Embodiment 4 of the present invention.

FIG. 10 is a block diagram showing a configuration of a signal processing device according to a fifth embodiment of the present invention.

FIG. 11 is a block diagram showing an example of a base layer decoder.

FIG. 12 is a block diagram showing an example of an enhancement layer decoder.

FIG. 13 is a diagram showing an example of a configuration of an enhancement layer decoder.

FIG. 14 is a block diagram showing the configuration of an enhancement layer decoder of the signal processing apparatus according to Embodiment 7 of the present invention.

FIG. 15 is a block diagram showing the configuration of an enhancement layer decoder of the signal processing apparatus according to Embodiment 8 of the present invention.

FIG. 16 is a block diagram showing the configuration of an acoustic signal coding apparatus according to Embodiment 9 of the present invention.

FIG. 17 is a block diagram showing a configuration of an acoustic signal decoding apparatus according to Embodiment 10 of the present invention.

FIG. 18 is a block diagram showing the configuration of an acoustic signal transmission encoding apparatus according to Embodiment 11 of the present invention.

FIG. 19 is a block diagram showing the configuration of an acoustic signal receiving / decoding apparatus according to Embodiment 12 of the present invention.

[Explanation of symbols]

101 Downsampler 102 base layer encoder 103 local decoder 104, 1003 Upsampling device 105 delay device 106 Subtractor 107 enhancement layer encoder 1002 Base layer decoder 1004 enhancement layer decoder 1005 adder

─────────────────────────────────────────────────── ─── Continuation of front page (51) Int.Cl. ⁷ identification code FI theme code (reference) H03M 7/36 G10L 7/04 G

Claims (26)

[Claims] 1. Down-sampling means for reducing a sampling rate of an input signal, base layer encoding means for obtaining a first encoded code by encoding the input signal with a reduced sampling rate, and based on the first encoded code. Decoding means for generating a decoded signal by using, up-sampling means for increasing the sampling rate of the decoded signal to the same rate as the input signal, using the parameters generated in the decoding process of the decoding means, An enhancement layer coding means for coding a difference value between the input signal and the decoded signal with the increased sampling rate to obtain a second coded code, and the first coded code and the second coded code are multiplexed. An encoding device comprising: 2. The encoding apparatus according to claim 1, wherein the base layer encoding means encodes the input signal using a code excitation linear prediction method. 3. The encoding apparatus according to claim 1, wherein the enhancement layer encoding means encodes the input signal using orthogonal transform. 4. The encoding apparatus according to claim 3, wherein the enhancement layer encoding means encodes the input signal using MDCT transform. 5. The enhancement layer coding means performs the coding processing by using the LPC coefficient of the base layer generated in the decoding processing of the decoding means.
5. The encoding device according to claim 4. 6. The enhancement layer coding means transforms an LPC coefficient of a base layer into an LPC coefficient of an enhancement layer based on a preset conversion table, calculates a spectrum envelope based on the LPC coefficient of the enhancement layer, and encodes the code. The coding apparatus according to claim 5, wherein the spectrum envelope is utilized for at least one of spectrum normalization and vector quantization in the coding processing. 7. The enhancement layer coding means performs the coding processing by utilizing the pitch period and the pitch gain generated in the decoding processing of the decoding means. 6. The encoding device according to any one of 6. 8. The enhancement layer coding means calculates a spectrum fine structure using a pitch period and a pitch gain, and utilizes the spectrum fine structure for spectrum normalization and vector quantization in a coding process. The encoding device according to claim 7. 9. The enhancement layer coding means performs the coding process by utilizing the power of the decoded signal generated by the decoding means. Encoding device. 10. The enhancement layer encoding means quantizes the variation amount of the power of the MDCT transform coefficient based on the power of the decoded signal, and the power variation of the quantized MDCT transform coefficient in the power normalization in the encoding process. The encoding device according to claim 9, wherein the encoding device utilizes a quantity. 11. A base layer decoding means for decoding a first coded code to obtain a first decoded signal, an enhancement layer decoding means for decoding a second coded code to obtain a second decoded signal, and Up-sampling means for increasing the sampling rate of the first decoded signal to the same rate as the second decoded signal, and addition means for adding the first signal and the second signal with the increased sampling rate are provided. Characterizing decoding device. 12. The decoding device according to claim 11, wherein the base layer decoding means decodes the first coded code using a code-excited linear prediction method. 13. The decoding apparatus according to claim 11 or 12, wherein the enhancement layer decoding means decodes the second coded code by using orthogonal transformation. 14. The enhancement layer decoding means is IMDCT.
The decoding device according to claim 13, wherein the second coded code is decoded by using conversion. 15. The decoding according to claim 11, wherein the enhancement layer decoding means decodes the second coded code by using the LPC coefficient of the base layer. apparatus. 16. The enhancement layer decoding means transforms the LPC coefficient of the base layer into an LPC coefficient of the enhancement layer based on a preset conversion table, calculates a spectrum envelope based on the LPC coefficient of the enhancement layer, and decodes the spectrum envelope. The decoding apparatus according to claim 15, wherein the spectrum envelope is utilized for vector decoding in the coding processing. 17. The enhancement layer decoding means performs the decoding process using at least one of the pitch period and the pitch gain.
The decoding device according to any one of 1. 18. The enhancement layer decoding means calculates a spectral fine structure using a pitch period and a pitch gain,
18. The decoding device according to claim 17, wherein the spectral fine structure is utilized for vector decoding in the decoding process. 19. The enhancement layer decoding means uses the power of the decoded signal generated by the decoding means to perform the decoding process. Decryption device. 20. The enhancement layer decoding means decodes the variation amount of the power of the MDCT transform coefficient based on the power of the decoded signal, and the variation of the power of the decoded MDCT transform coefficient for power normalization in the decoding process. 20. Decoding device according to claim 19, characterized in that it utilizes quantity. 21. Acoustic input means for converting an acoustic signal into an electrical signal, A / D conversion means for converting the signal output from this acoustic input means into a digital signal, and this A / D
An encoding device according to any one of claims 1 to 10 for encoding the digital signal output from the converting means, and an RF for modulating the encoding code output from the encoding device into a radio frequency signal. An acoustic signal transmitting apparatus comprising: a modulating unit and a transmitting antenna that converts a signal output from the RF modulating unit into an electric wave and transmits the electric wave. 22. A receiving antenna for receiving a radio wave, an RF demodulating means for demodulating a signal received by the receiving antenna, and decoding the information obtained by the RF demodulating means. Any one of the decoding device, a D / A conversion unit for converting a signal output from the decoding device into an analog signal, and an electric signal output from the D / A conversion unit into an acoustic signal A sound signal receiving device comprising: 23. A communication terminal device comprising at least one of the acoustic signal transmitting device according to claim 21 and the acoustic signal receiving device according to claim 22. 24. A base station apparatus comprising at least one of the acoustic signal transmitting apparatus according to claim 21 and the acoustic signal receiving apparatus according to claim 22. 25. A step of reducing a sampling rate of an input signal, a step of encoding the input signal having a reduced sampling rate to obtain a first encoded code, and a decoded signal generated based on the first encoded code. Using the parameters obtained in the steps of increasing the sampling rate of the decoded signal to the same rate as the input signal and the processing of generating the decoded signal, the input signal and the sampling rate are increased. An encoding method comprising: a step of encoding a difference value with a decoded signal to obtain a second encoded code; and a step of multiplexing the first encoded code and the second encoded code. . 26. Decoding the first coded code to obtain a first decoded signal; decoding the second coded code to obtain a second decoded signal; and sampling rate of the first decoded signal. A decoding method comprising: a step of increasing the same rate as that of the second decoded signal; and a step of adding the first signal and the second signal whose sampling rate is increased.