JPH05183523A - Voice/music sound identification circuit - Google Patents

Abstract

PURPOSE:To attain coding processing of a voice signal and a music sound signal efficiently with high quality by identifying the voice signal and the music sound signal from the digitized input signal. CONSTITUTION:A digitized input signal Xi is divided into a low frequency signal and a high frequency signal by a band division filter 11 and quantized and coded respectively by low frequency and high frequency quantization coding circuits 12, 13, and multiplexed by a multiplexer circuit 14. When a voice/music tone identification circuit 20 discriminates the input signal Xi to be a voice signal, a multiplexer inhibit means 14a is active by a voice identification signal S20 and multiplexing of the high frequency quantization coding signal S13 in the multiplexer circuit 14 is inhibited.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a voice / tone encoding device for encoding a voice signal spoken by a human and a tone signal generated by a musical instrument.

[0002]

2. Description of the Related Art Conventionally, as a technique in such a field,
For example, some documents were described in the following documents. References: 1991 IEICE Spring National Conference Proceedings B-577 Shimizu / Takeo / Horiguchi "Acoustic Communication Codec for ATM Fixed Rate" p. 3-129 Conventionally, there is a codec circuit described in the above-mentioned document as one which encodes a voice signal and a musical tone signal to perform communication.
In this codec circuit, the voice signal and the tone signal are
It is designed to be encoded and transmitted with high quality in the Hz and 20 KHz bands. FIG. 2 shows a speech / tone encoding apparatus used in this codec circuit. FIG. 2 is a functional block diagram showing an example of the configuration of the conventional speech / tone encoding apparatus described in the above document.

In this voice / tone encoding apparatus, a digitized input signal Xi is converted into a low frequency signal S1a and a high frequency signal S1.
It has a band division filter 1 for dividing into b. The band division filter 1 is composed of a low band band pass filter 1a and a high band band pass filter 1b, and a low band quantization coding circuit 2 and a high band quantization coding are provided on each output side thereof. Circuits 3 are respectively connected. The low-frequency quantization coding circuit 2 is a circuit that quantizes and codes the low-frequency signal S1a with the allocated number of bits and outputs a low-frequency quantized and coded signal S2. The high-frequency quantization coding circuit 3 is a circuit that quantizes and codes the high-frequency signal S1b with the allocated number of bits and outputs a high-frequency quantized coded signal S3. As the coding algorithm in the quantization coding circuits 2 and 3,
ADPCM (adaptive differential pulse-code modul)
ation) method. The multiplexing circuit 4 is connected to the output sides of the quantization coding circuits 2 and 3. Multiplexing circuit 4
Is a circuit that multiplexes the low-frequency quantized coded signal S2 and the high-frequency quantized coded signal S3 to output an output signal Zi.

In this type of speech / musical sound coding apparatus, an analog / digital converter (hereinafter referred to as an A / D converter) is used.
The input signal Xi digitized by, for example, is divided by the band division filter 1 into a low band signal S1a and a high band signal S1b. The divided low-frequency signal S1a and high-frequency signal S1b
Are coded by the low-frequency quantization coding circuit 2 and the high-frequency quantization coding circuit 3, respectively, to obtain the low-frequency quantization coded signal S2.
And the high-frequency quantized coded signal S3 is output. The low band quantized coded signal S2 and the high band quantized coded signal S3 are multiplexed by the multiplexing circuit 4 and output as the output signal Zi. In this voice / musical sound coding apparatus, in the quantization coding circuits 2 and 3, the high band signal S1b is quantized by allocating fewer bits as compared with the low band signal S1a.
It realizes high-quality coding while suppressing the transmission rate.

[0005]

However, the conventional speech / tone encoding apparatus has the following problems when encoding speech signals and tone signals. Here, a musical tone means a sound that can be separated from an audio signal spoken by humans. A voice signal is a means of communication for realizing a conversation, and is a sound that can ensure a certain level of quality. Speech is generated from the human vocal tract and is given spectral characteristics mainly by the formants. Further, the musical tone is a musical tone that gives a rich musical feeling, and the musical tone has a characteristic that vibrations are periodic and harmonic frequencies form a positive ratio. As an example of these, FIG. 3 shows a time-series waveform diagram of an audio signal, and FIG. 4 shows a time-series waveform diagram of a musical tone signal. The tone signal of FIG. 3 is a waveform diagram by the orchestra. The waveform diagrams drawn below the waveform diagrams of FIGS. 3 and 4 are partially enlarged views of the above diagrams.

When such a voice signal and a musical tone signal are coded, the conventional voice / musical tone coding apparatus processes the voice signal and the musical tone signal by the same coding method. The bandwidth was too wide and not efficient. Since the voice signal and the tone signal have different characteristics, they are distinguished from each other, and the voice signal secures the quality of the band corresponding to the voice signal and the efficient transmission form to improve the clarity and intelligibility. In addition, it is difficult to provide a high-quality and efficient voice / tone encoding device by handling the tone signal in a wide band corresponding to the tone signal. DISCLOSURE OF THE INVENTION The present invention has, as a problem that the above-mentioned conventional art has, distinguishes a voice signal and a musical tone signal, improves the intelligibility and the intelligibility of the voice signal, and provides a high-quality and efficient voice signal and musical tone signal. It is an object of the present invention to provide a speech / musical sound coding device that solves the problem that coding is difficult.

[0007]

In order to solve the above-mentioned problems, a first invention is a band division filter for dividing a digitized input signal into a low band signal and a high band signal, and an allocated number of bits. A low-band quantization coding circuit that quantizes and codes the low-band signal to output a low-band quantized code; and a high-band signal that quantizes and codes the high-band signal with an assigned number of bits. A voice / tone code including a high frequency quantization coding circuit for outputting a high frequency quantization coded signal and a multiplexing circuit for multiplexing the low frequency quantization coded signal and the high frequency quantization coded signal. The following means are provided in the chemical conversion device. That is, in the conventional voice / musical sound encoding device, a voice / musical sound discrimination circuit for discriminating a voice signal and a musical tone signal from the input signal and outputting a voice discrimination signal when the discrimination result is a voice signal, and the voice discrimination signal. Based on the above, there is provided a multiplexing inhibiting means for inhibiting the multiplexing of the high frequency quantized coded signal in the multiplexing circuit.

In the second invention, bit allocation reducing means is provided in place of the multiplexing inhibiting means of the first invention. The bit allocation reducing means has a function of controlling the bit allocation of the multiplexing circuit based on the voice identification signal output from the voice / tone identification circuit. In the third invention, a bit allocation control means is provided in place of the multiplexing prohibition means of the first invention. This bit allocation control means is
It has a function of controlling the ratio of the bit allocation of the high frequency quantization coding circuit and the low frequency quantization coding circuit based on the voice identification signal output from the voice / musical sound discrimination circuit.

[0009]

According to the first aspect of the present invention, since the voice / musical tone coding apparatus is configured as described above, when the digitized input signal is input to the band division filter and the voice / musical tone discrimination circuit, the band The division filter divides the input signal into a low frequency signal and a high frequency signal. In the low-frequency quantization coding circuit and the high-frequency quantization coding circuit, the divided low-frequency signal and high-frequency signal are quantized and coded with a predetermined number of bits, and the low-frequency quantized coded signal and the high-frequency signal are quantized. The quantized coded signals are output. These signals are multiplexed by a multiplexing circuit. On the other hand, voice
The tone discrimination circuit discriminates the voice signal and the tone signal from the input signal, and outputs the voice discrimination signal when the voice signal.
As a result, the multiplexing prohibiting means prohibits the multiplexing of the high frequency quantization coded signal in the multiplexing circuit.

According to the second aspect of the invention, when the voice / musical sound discrimination circuit determines that the signal is a voice signal, the bit allocation reducing means reduces the bit allocation of the high frequency quantization encoding circuit based on the voice identification signal. Control in the direction. The high band quantized coding signal quantized and coded by the high band quantized coding circuit and the low band quantized coded signal which is coded by the low band quantized coding circuit are multiplexed. The circuit is multiplexed.

According to the third aspect of the invention, when the voice / musical sound discrimination circuit determines that the signal is a voice signal, the bit allocation control means operates the bit of the quantization coding circuit for the high frequency band and the low frequency band based on the voice signal. The allocation ratio is controlled, and the allocated bits are respectively quantized and encoded and multiplexed by a multiplexing circuit. Therefore, the above problem can be solved.

[0012]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS First Embodiment FIG. 1 is a functional block diagram of a speech / tone encoding apparatus showing a first embodiment of the present invention. This speech / musical sound coding apparatus converts a digitized input signal Xi into a low-frequency signal S1.
1a and a band division filter 1 for dividing into a high frequency signal S11b
Have one. The band division filter 11 has a low-frequency signal S.
The low-pass band-pass filter 11a which outputs 11a and the high-pass band-pass filter 11b which outputs the high-frequency signal S11b.
And a low-band quantization coding circuit 12 and a high-band quantization coding circuit 13 are respectively connected to their outputs. The low-frequency quantization coding circuit 12 is a circuit that quantizes and codes the low-frequency signal S11a with the allocated number of bits and outputs the low-frequency quantized coded signal S12. The high-frequency quantization coding circuit 13 is a circuit that quantizes and codes the high-frequency signal S11b with the allocated number of bits and outputs a high-frequency quantized coded signal S13. A multiplexing circuit 14 is connected to the output sides of the quantization coding circuits 12 and 13.

The multiplexing circuit 14 is a circuit that multiplexes the low band quantized coded signal S12 and the high band quantized coded signal S13 and outputs an output signal Zi. The multiplexing circuit 14 is provided with a multiplexing inhibiting means 14a including a gate circuit for inhibiting the multiplexing of the high frequency quantization coded signal S13 based on the voice identification signal S20. On the input side of the multiplexing circuit 14, a voice / tone identification circuit 20 for providing a voice identification signal S20 is provided. The voice / tone identification circuit 20 is a circuit that identifies a voice signal and a tone signal from the input signal Xi and outputs a voice identification signal S20 to the multiplexing circuit 14 when the identification result is a voice signal. voice·
Voice identification signal S20 output from the tone identification circuit 20
Is input to the multiplexing circuit 14 as auxiliary data S20a.

FIG. 5 shows a voice / tone identification circuit 20 in FIG.
3 is a functional block diagram showing a configuration example of FIG. The voice / tone identification circuit 20 calculates the power of the digitized input signal Xi to determine whether there is voice or no voice, and the input signal based on the voice determination of the power calculator 21. And an autocorrelation coefficient calculating means 22 for calculating the autocorrelation coefficient R (t) from Xi. The output side of the autocorrelation coefficient calculation means 22 is connected to first and second threshold value judgment integration means 23 and 24.

The first threshold value judgment integrating means 23 has a function of integrating the number of times that the value of the autocorrelation coefficient R (t) is larger than the positive threshold value PK1, and the autocorrelation coefficient R (t) is calculated. The threshold value judging means 23a which outputs when the autocorrelation coefficient R (t) is larger than the positive threshold value PK1 and the number of outputs of the threshold value judging means 23a is integrated (counted) and the integrated value M is obtained.
It is composed of a judgment number accumulating means 23b for outputting 1. Similarly, the second threshold judgment integrating means 24 has a function of integrating the number of times the autocorrelation coefficient R (t) is smaller than the negative threshold value PK2, and the autocorrelation coefficient R (t).
Is compared with a negative threshold value PK2 to output when the autocorrelation coefficient R (t) is small, and a determination to output the integrated value M2 by integrating the number of outputs of the threshold value determining means 24a. It is composed of the number-of-times accumulating means 24b.

First and second threshold value judgment integrating means 23, 24
The number-of-accumulations determination means 25 is connected to the output side of, and the count identification means 26 is connected to the output side thereof. The integrated number determination means 25 determines whether or not the integrated values M1 and M2 exceed the thresholds K1 and K2, and outputs the determination result of the number when the conditions of M1> K1 and M2> K2 are satisfied. have.

The counting identification means 26 is an integrated number determination means 2
It has a function to count the continuity of the number of times occurred in 5,
Two count-up conditions (CNT1, CNT2) are connected so that they are input to other resets R1, R2.
Two counters 26a and 26b, and an integrated number determination means 2
An inverter 26c that inverts the output of 5 and a counter 26
A flip-flop (hereinafter, F) which is reset by the carrier CR1 of a and reset by the carrier CR2 of the counter 26b and outputs the voice identification signal S20 from the negative phase output terminal.
26d) (referred to as F).

The operation ((1) quantization / encoding processing, (2) speech / tone identification processing, and (3) multiplexing processing) of the speech / tone encoding apparatus configured as described above will be described. (1) Quantization Coding Processing When the input signal Xi digitized by the A / D converter or the like is input to the band division filter 11 and the voice / tone discrimination circuit 20, the band division filter 11 outputs the low frequency band. The pass filter 11a and the high pass band pass filter 11b divide the low pass signal S11a and the high pass signal S11b.
The divided low-frequency signal S11a and high-frequency signal S11b are
The low-frequency quantization coding circuit 12 and the high-frequency quantization coding circuit 13 perform quantization coding. High frequency quantization coding circuit 1
In 3, a smaller number of bits are assigned as compared with the low-frequency quantization encoding circuit 12, and the bits are quantized.
High quality encoding is performed while suppressing the transmission rate. The low-frequency quantized coded signal S12 and the high-frequency quantized coded signal S13 quantized by these quantized coding circuits 12 and 13.
Are sent to the multiplexing circuit 14.

(2) Speech / Musical Sound Discrimination Processing When the digitized input signal Xi is input to the speech / musical sound discrimination circuit 20, in FIG.
It is stored in a memory (not shown) in block units (frame units). The block-by-block input signal Xi stored in the memory is input to the power calculation means 21 and the autocorrelation coefficient calculation means 22. The power calculation means 21 calculates the power of the input signal Xi, determines whether or not the signal power is larger than the threshold value Pth. When the signal power is larger than the threshold value Pth, it is considered as voiced, and the autocorrelation coefficient calculation means 2
Give the calculation instructions to 2.

The autocorrelation coefficient calculating means 22 calculates the autocorrelation coefficient R (t) by the following calculation formula.

[0021]

[Equation 1] The calculated autocorrelation coefficient R (t) is sent to the first and second threshold value judgment integrating means 23, 24. Threshold value judging means 23
In a, the autocorrelation coefficient R (t) is compared with the positive threshold value PK1 and when the autocorrelation coefficient R (t) is larger than the positive threshold value PK1, the result is output to the judgment number integrating means 23b. As a result, the judgment count integrating means 23b counts up. On the other hand, the threshold value judging means 24a compares the autocorrelation coefficient R (t) with the negative threshold value PK2, and when the autocorrelation coefficient R (t) is smaller than the negative threshold value PK2, the judgment number integrating means 24b. Output to. As a result, the determination count integrating means 24b counts up. The integration of these determination times is repeated by the number N of signals in the block.

The integrated values M1 and M2 of the judgment number integrating means 23b and 24b are sent to the integration number judging means 25. In the cumulative number judging means 25, M1> K at the end of one block.
It is determined whether or not the condition of 1 and M2> K2 is satisfied, and the determination result is given to the counter 26a in the count identifying means 26, and is inverted by the inverter 26c to be counter 26.
give to b. When the conditions of M1> K1 and M2> K2 are satisfied, the counter 26a counts up and the reset R1 resets the counter 26b to zero. On the contrary, when the conditions of M1> K1 and M2> K2 are not satisfied, the counter 26b counts up and the reset R2 resets the counter 26a to zero.

The counters 26a and 26b output the carriers CR1 and CR2 as overflows when they have counted a predetermined number of times. This counter 26
In the case of a and 26b, the number-of-accumulations determination means 2 is usually about 2 to 4.
Count the continuity of the number of occurrences in 5. That is,
The counter 26a counts up when the determination by the integration number determination means 25 is true, outputs the carrier CR1 due to the count overflow, sets the FF 26d, and the FF 26d.
It is determined to be a musical tone signal by the positive phase output of 26d. Similarly, the counter 26b counts up when the determination made by the number-of-accumulations determination means 25 is false, and FF is caused by a count overflow.
26d is reset, a voice signal is determined by the reverse phase output of the FF 26d, and the voice identification signal S2 is given to the multiplexing circuit 14.
Activate 0.

When the identification process within one block is completed in this way, the identification process for the next block is performed.

(3) Multiplexing Process In FIG. 1, the low-frequency quantized coded signal S12 and the high-frequency quantized coded signal S13 output from the quantized coding circuits 12 and 13 are input to the multiplexing circuit 14. It Further, the voice / musical sound discrimination circuit 20 discriminates between the voice signal and the musical tone signal, and when it is judged that the signal is a musical tone signal, the voice discrimination signal S20 given to the multiplexing circuit 14 is made non-active, and a part of the auxiliary data S20a of the signal. To the multiplexing circuit 14. In the multiplexing circuit 14, the low frequency quantization coded signal S12
And the high-frequency quantized coded signal S13 are multiplexed, auxiliary data S20a is added, and an output signal Zi is output. on the other hand,
When the voice / musical sound identification circuit 20 determines that the signal is a voice signal, the voice identification signal S20 becomes active. Then,
The multiplex prohibiting means 14a in the multiplex circuit 14 receives the high-frequency quantized coded signal S13 input to the multiplex circuit 14.
Prohibit multiplexing. Also, this voice identification signal S
Since 20 is input to the multiplexing circuit 14 as auxiliary data S20a, the output signal Zi of the multiplexing circuit 14 is data in which only the low-frequency quantized coded signal S12 and the auxiliary data S20a are added.

An example of the frame structure of this output signal Zi is shown in FIG.
Shown in. In the present embodiment, 1 bit of the auxiliary data S20a is used, and the sound mode and the sound mode are switched by the sound identification signal S20. That is, when the voice / musical sound identification circuit 20 determines that it is a musical sound, the audio identification signal S2
0 becomes non-active, and auxiliary data S20 of "0"
Since a is given to the multiplexing circuit 14, the multiplexing circuit 1
4, the low-frequency quantized coded signal S12 and the high-frequency quantized coded signal S13 are multiplexed, and the output signal Zi1 having a frame structure including auxiliary data “0”, low-frequency coded data, and high-frequency coded data. Is output. On the other hand, when the voice / musical sound identification circuit 20 determines that the signal is a voice signal, the voice identification signal S20
Becomes active and supplies the auxiliary data S20a of "1" to the multiplexing circuit 14. Then, in the multiplexing circuit 14,
The multiplexing of the high-frequency quantized coded signal S13 is prohibited, and the output signal Zi2 having the frame structure including the auxiliary data "1" and the low-frequency coded data is output.

As described above, when it is determined that the signal is a voice signal, the band is only a low frequency band signal, and the processing for ensuring the quality of the band corresponding to the voice signal ensures an efficient transmission mode and, at the same time, provides clarity. And, it becomes possible to improve the intelligibility. Also, in this first embodiment, the autocorrelation coefficient R
(T) counts the number of times that a value larger than the positive threshold value PK1 occurs and counts the number of times that a value smaller than the negative threshold value PK2 occurs, and determines whether the positive and negative count values exceed the threshold value. Since it is determined by the integrated number determination means 25, the musical tone signal and the voice signal can be clearly detected. Moreover, the discrimination between the musical tone signal and the voice signal is, for example, 3
It can be identified in a short time of 4 block time. Furthermore, since the amount of calculation is small, the circuit scale can be reduced.

Second Embodiment FIG. 7 is a functional block diagram of a speech / tone encoding apparatus showing a second embodiment of the present invention, in which elements common to those in FIG. It is attached. In this voice / musical sound coding apparatus, instead of the high frequency quantizing coding circuit 13 and the multiplexing circuit 14 in FIG. 1 showing the first embodiment, a high frequency quantizing coding circuit 13 having a different configuration is used. -1 and multiplexing circuit 1
4-1 is provided. High frequency quantization coding circuit 13-1
Has a function of quantizing and encoding the high frequency signal S11b output from the high frequency band pass filter 11b with a predetermined number of bits, and based on the voice identification signal S20 output from the voice / tone identification circuit 20. The quantization coding circuit 1
3-1 has a bit allocation reducing means 13a for controlling the bit allocation of 3-1 to reduce, and the high-frequency quantized coded signal S13 is provided.
It is a circuit for giving -1 to the multiplexing circuit 14-1. The multiplexing circuit 14-1 has a function of multiplexing the low-frequency quantized coded signal S12 and the high-frequency quantized coded signal S13-1 and outputting an output signal Zi.

In this speech / tone encoding apparatus, when the speech / tone discrimination circuit 20 determines that the speech signal is a speech signal, the speech discrimination signal S20 determines the bit allocation in the high frequency quantizing / encoding circuit 13-1. The reducing means 13a controls the bit allocation in the quantization coding circuit 13-1 so as to reduce it, and operates so as to reduce the transmission rate. As a result, in the frame structure of the output signal Zi output from the multiplexing circuit 14-1, for example, in the output signal Zi2 shown in FIG. 6, high frequency encoded data having a small number of bits is added after low frequency encoded data. It will be a frame structure. Therefore,
The transmission efficiency of the audio signal is improved as compared with the conventional case.

Third Embodiment FIG. 8 is a functional block diagram of a speech / tone encoding apparatus showing a third embodiment of the present invention. Elements common to those in FIGS. 1 and 7 are common to those elements. Is attached. This voice
In the tone encoding apparatus, the multiplexing circuit 14 shown in FIG. 7 is provided in place of the multiplexing circuit 14 shown in FIG. 1, and the bit allocation control means 30 is connected to the output side of the voice / tone identification circuit 20. .. The bit allocation control means 30 is based on the voice identification signal S20 output from the voice / tone identification circuit 20.
It has a function of controlling the ratio of the bit allocation of the low frequency quantization coding circuit 12 to the bit allocation of the high frequency quantization coding circuit 13. When the voice / musical sound discrimination circuit 20 determines that it is a voice signal, the voice discrimination signal S20 becomes active, and the bit allocation control means 30 switches the bit allocation ratio of the quantization coding circuits 12 and 13. As the bit allocation ratio, for example, an operation of switching the bit allocation ratio between the low band and the high band from 5: 3 to 6: 2 is performed. As a result, the transmission rate remains unchanged, but the ratio of bit allocation in the low frequency band can be increased, so that the quantization accuracy on the low frequency band side can be improved and the quality can be further improved.

The present invention is not limited to the above embodiment,
Various modifications are possible. Examples of such modifications include the following. (A) The bit allocation reducing means 13a shown in FIG. 7 may be provided outside the high frequency quantization coding circuit 13-1. Similarly, the bit allocation control means 30 shown in FIG. 8 may be provided in each of the low band quantization coding circuit 12 and the high band quantization coding circuit 13. Further, the multiplexing circuits 14, 14-
The output signal Zi of 1 may have a frame configuration other than that shown in FIG. (B) The voice / tone identification circuit 20 may be changed to a circuit configuration other than that shown in FIG. (C) The circuits shown in FIGS. 1, 7, and 8 may be configured by individual circuits using integrated circuits or the like, or may be executed by program control using a digital signal processor (DSP) or the like. When executed by program control, the circuit scale can be made smaller than that of an individual circuit.

[0032]

As described in detail above, according to the first aspect of the present invention, when the voice signal and the musical tone signal are discriminated by the voice / tone discrimination circuit and the voice / tone discrimination circuit determines that they are voice signals, Based on the voice identification signal, the multiplexing prohibiting means prohibits the multiplexing of the output of the high frequency quantization coding circuit. Therefore, in the case of a voice signal, only a signal with a low band is provided, and the voice signal is processed to secure the quality of the band corresponding to the voice signal to secure an efficient transmission form, and at the same time, improve the intelligibility and intelligibility. It is possible to improve. Furthermore, by handling the musical tone signal in a wide band corresponding to the musical tone signal, high quality and efficient signal processing becomes possible. Therefore, the present invention can be applied to various signal processing devices that process voice signals and tone signals with digital signals.

In the second aspect of the invention, when the voice / musical sound discrimination circuit determines that the signal is a voice signal, the bit allocation reducing means controls the bit allocation of the high frequency quantization coding circuit in a direction of reducing the bit allocation. The transmission efficiency of the audio signal is improved as compared with the conventional case, and the same effect as that of the first invention can be obtained.

According to the third aspect of the invention, when the voice / musical sound discrimination circuit determines that the signal is a voice signal, the bit allocation control means controls the bit allocation ratio of the high frequency and low frequency quantization coding circuits. Since this is done, substantially the same effect as in the first aspect of the invention can be obtained, and the quantization accuracy on the low frequency side is improved and the quality is further improved.

[Brief description of drawings]

FIG. 1 is a functional block diagram of a voice / tone encoding apparatus showing a first embodiment of the present invention.

FIG. 2 is a functional block diagram of a conventional speech / tone encoding apparatus.

FIG. 3 is a time-series waveform diagram of an audio signal.

FIG. 4 is a time-series waveform diagram of a musical tone signal.

5 is a functional block diagram of a voice / tone discrimination circuit in FIG. 1. FIG.

FIG. 6 is a diagram showing a frame configuration example of an output signal in FIG.

FIG. 7 is a functional block diagram of a speech / tone encoding apparatus showing a second embodiment of the present invention.

FIG. 8 is a functional block diagram of a speech / tone encoding apparatus showing a third embodiment of the present invention.

[Explanation of symbols]

 11 Band Division Filter 11a Low Band Band Pass Filter 11b High Band Band Pass Filter 12 Low Band Quantization Coding Circuit 13, 13-1 High Band Quantization Coding Circuit 13a Bit Allocation Reduction Means 14, 14-1 Multiplexing circuit 14a Multiplexing prohibiting means 20 Voice / tone discrimination circuit S11a Low-frequency signal S11b High-frequency signal S12 Low-frequency quantized coded signal S13, S13-1 High-frequency quantized coded signal S20 Speech identification signal Xi Input signal Zi Output signal

Claims (3)

[Claims] 1. A band division filter for dividing a digitized input signal into a low band signal and a high band signal, and a low band quantized code signal by quantizing and coding the low band signal with an assigned number of bits. A low-band quantization coding circuit, a high-band quantization coding circuit for quantizing-coding the high-band signal with an allocated number of bits to output a high-band quantization coded signal, and In a voice / musical sound coding device comprising a low-frequency quantized coded signal and a multiplexing circuit for multiplexing the high-frequency quantized coded signal, a voice signal and a musical sound signal are identified from the input signal, and A voice / tone identification circuit that outputs a voice identification signal when the identification result is a voice signal; and a multiplexing inhibiting means that inhibits the multiplexing of the high-frequency quantized coded signal in the multiplexing circuit based on the voice identification signal. Is provided A voice / musical sound encoding device. 2. A voice / musical tone coding apparatus comprising the band division filter according to claim 1, a low frequency quantization coding circuit, a high frequency quantization coding circuit, and a multiplexing circuit, wherein the input signal A voice signal and a tone signal, and outputs a voice identification signal when the identification result is a voice signal; and a bit allocation of the high frequency quantization coding circuit based on the voice identification signal. And a bit allocation reducing means for controlling in a direction to reduce the voice / tone encoding apparatus. 3. A voice / musical sound coding apparatus comprising the band division filter according to claim 1, a low-frequency quantization coding circuit, a high-frequency quantization coding circuit, and a multiplexing circuit, wherein the input signal A voice signal and a tone signal, and outputs a voice identification signal when the identification result is a voice signal; and a bit allocation of the high frequency quantization coding circuit based on the voice identification signal. And a bit allocation control means for controlling a bit allocation ratio of the low-frequency quantization coding circuit.