JPH08125990A - Encoding device and decoding device - Google Patents

Abstract

PURPOSE: To provide an encoding device and decoding device with which cost can be reduced by reducing hardware scale and signal degradation can be reduced. CONSTITUTION: At an encoding system, a signal SA is converted into a linear PCM signal by an A/D converter 11 and parallelly supplied to a telephone voice encoder 12 and low bit rate voice encoder 15 later. At a selector 13, either the output of the encoder 12 or the output of the encoder 15 is selected and defined as transmitting data DA based on the information of the decoding ability of an opposite side terminal or the like. At a decoding system, the data DA are parallelly supplied to a telephone voice decoder 24 and a low bit rate voice decoder 22. At a selector 21, either the output of the decoder 24 or the output of the decoder 22 is selected and the signal SA is provided based on the information of the encoding system of the data DA. When selecting any one encoding system at the encoding system or the decoding system, the processing of encoding or decoding related to the other encoding system is not required, cost can be reduced by reducing the hardware scale by reducing the arithmetic amount of a DSP and sound quality degradation can be reduced.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an encoding device and a decoding device suitable for application to, for example, an audio data processing unit (voice codec) corresponding to a plurality of encoding methods of a video conference apparatus.

[0002]

2. Description of the Related Art FIG. 6 shows a configuration example of a conventional video conference apparatus. In the figure, a video signal (luminance signal, color difference signal) SV output from the image pickup apparatus 1 is supplied to the image data processing unit 2. The operation of the image data processing unit 2 is controlled by the CPU 3 which constitutes the system controller. In the image data processing unit 2, after conversion from an analog signal to a digital signal, CCITT (Comite
Consultatif International Telegraphique et Teleph
onque), H. The data is compressed in the format defined by H.261 and the error correction code is added to form the image data DV.

The audio signal SA from the microphone 4 is supplied to an audio data processing section (voice codec) 5. The operation of the audio data processing unit 5 is performed by the CPU 3
Controlled by. In the audio data processing unit 5,
After being converted from an analog signal to a digital signal (linear PCM signal), CCITT, G.G. 711 and the like are data-compressed in a specified format to form audio data DA.

The image data DV output from the image data processing unit 2 and the audio data output from the audio data processing unit 5 are supplied to the demultiplexing unit 6. The operation of the demultiplexing unit 6 is controlled by the CPU 3. In the demultiplexing unit 6, the image data DV and the audio data DA are CCITT, H.264 and H.264. Multiplexed data DMU is formed by being multiplexed in the format defined by H.221.
This multiplexed data DMU is sent to the other terminal which is the communication target through the line interface 7 and the line 8.

The line 8 is, for example, an INS net 64 (Informat) which is one of ISDN (Service Integrated Digital Network).
64 [kbps] line of ion Network System Net 64), or ISDN 384 [kbps], 1536
The lines are [kbps] and 1920 [kbps].

Here, CCITT, H.M. The details of the format defined in 221 will be described. This format is specified according to the transmission speed. Transmission speed is 64
When a plurality of [kbps] lines are used, the channel of each line is defined as the B channel. Transmission speed is 384
When a plurality of [kbps] lines are used, the channel of each line is defined as H ₀ , and the transmission rate is 1536.
When the lines of [kbps] and 1920 [kbps] are used, they are defined as H ₁₁ channel and H ₁₂ channel, respectively.

In this format, in each channel, two consecutive frames are collected to form one sub-multiframe, and eight consecutive subframes (16 frames) are collected to form one multiframe. FIG. 7 shows the relationship between the frame structure of the B channel and the bit string, and FIG. 8 shows the frame structure of the B channel. In this B channel, 8-bit (125 μs) serial data is continuously formed for 10 ms to form one frame of data (640 bits).

In this case, 8-bit unit data is represented by an octet number, and each 8-bit bit string is represented by a sub-channel. The eighth sub-channel is called a service channel (SC), and a frame synchronization signal FAS (Frame Alignment Signal) is arranged in the first to eighth bits and a bit rate allocation signal BA in the ninth to sixteenth bits.
S (Bitrate Allocation Signal) is arranged.

FIG. 9 shows the structure of the frame synchronization signal FAS. The synchronization code is 7 bits "00" of the even frame.
It is composed of a total of 8 bits, 11011 "and 1 bit of the odd frame. The first sync bit of the even frame and the sync bit of the odd frame are inverted to emulate the sync code by another signal. This is to avoid

Since the sync code is configured as described above, the timing of the sync signal of the even frame is detected by sampling the continuous data string at the 8-bit cycle and detecting the bit pattern of "0011011". It is possible to detect the byte boundary of the data in each frame based on the detection result. Further, the phase shift between the lines can be corrected up to 10 ms based on the timing detection result of the synchronization signal of the even frames.

The A-bit is used to inform the other party of whether or not there is a probability of synchronization of multi-frames received by itself, and is used in the communication procedure. C ₁ -C ₄ bits, a bit for CRC (Cyclic Redundancy Check), in addition to being used to monitor line quality, frame synchronization to the position the same pattern as the frame synchronization in a communication signal is present It is also used to get out of the accidentally retracted state. E bit is used to display the C ₁ -C ₄ bit usage.

FIG. 10 shows the structure of the frame synchronization signal FAS in multi-frame. The first bit of the odd frame is "0010" from the first sub-multiframe.
11 ". Therefore, the first bit of the service channel SC of the odd-numbered frame is sampled and" 00101 "is sampled.
The timing of each sub-multiframe in the multiframe can be detected by detecting the bit pattern of 1 ". Based on this detection result, the phase shift between the lines can be corrected up to 80 ms.

Incidentally, L₁~ L₃Bit is for multiple channels
It is used to display the channel number. N ₁~ N_FourBit is compound
When configuring several channels, each multiframe has a 4-bit number.
This is for attaching the number (1111-0000).
Maximum 1.28 seconds (= 160 × 16 / 2ms)
Can be corrected for relative delays up to. N_FiveBit is multi-frame
Used to indicate whether the system is numbered or not.
You. The R bit is a bit for future use.
The bit indicates whether there is a failure inside the terminal
Used for.

FIG. 11 shows the structure of the bit rate allocation signal BAS. The physical code of the bit rate allocation signal BAS is composed of 8 bits, and the first 3 bits (b ₀ ~
b ₂ ) represents an attribute, and the remaining 5 bits (b _{3 to} b ₇ ) are expressed. Audio encoding commands, transfer rate commands, video and other commands, data commands, terminal capabilities 1 and 2 are defined as the attributes. For example, the audio encoding command transmits the information about the encoding method of the transmission data, and the information about the decoding ability indicating which encoding method can be decoded depending on the terminal ability. If the BAS code is erroneous, the operation of the entire system becomes inconsistent in transmission and reception. Therefore, protection is performed by adding an 8-bit parity to the 8-bit code and performing 2-bit error correction. In this case, information is sent in even frames and parity in odd frames.

In the case of data transmission using the B channel, the video conferencing apparatus connects as many lines as the desired number of lines within the maximum of 6 lines, and simultaneously connects the connected lines in parallel with 64 [kbps] data. Is sent. FIG. 12 shows B
An example of a transmission format when two channels (1B, 2B) are used is shown. In this case, H.264. Areas of audio data and image data (video data) are allocated according to the format defined by 221.
The symbol “CPU” represents data transmitted and received between the CPU 3 which is the system controller of the own terminal and the system controller (CPU) of the other terminal.

FIG. 13 shows the frame structure of the H ₀ channel. The H ₀ channel is configured such that one frame corresponds to 6 frames of the B channel and has 48 bits (1
25 μs) of serial data is continuously formed for 10 ms to form one frame of data. Each 48-bit bit string is represented by a sub-channel. Then, the eighth sub-channel is assigned to the service channel (SC). The frame synchronization signal FAS is arranged in the first to eighth bits, and the ninth to
The bit rate allocation signal BAS is arranged in the 16th bit. In the H ₀ channel, desired data transmission is performed by connecting a plurality of lines as in the case of the B channel.

The H ₁₁ channel and the H ₁₂ channel are H
Similar to the case of ₀ channel, one frame is configured to correspond to 24 frames and 30 frames of B channel, respectively, and 125 μs of 192 bits and 240 bits of serial data are continuously formed for 1 ms to form one frame of data. It This results in 1536 [kbp
[s], 1920 [kbps].

Returning to FIG. 6, the multiplexed data DMU transmitted from the partner terminal on the line 8 is the line interface 7
Is supplied to the demultiplexing unit 6 via. The demultiplexing unit 6 correctly detects the bit boundary in each frame by the frame synchronization signal FAS described above, and based on this, the image data DV and the audio data DA are separated from the multiplexed data DMU. Also, multiplexed data DMU
The bit rate allocation signal BAS is also separated, and this bit rate allocation signal BAS is supplied to the CPU 3 and used for controlling the image data processing unit 2, the audio data processing unit 5, and the like.

Image data DV separated by the demultiplexer 6
Is supplied to the image data processing unit 2 to be subjected to decoding processing, error correction processing and the like to form a video signal SV. Then, the video signal SV is supplied to the monitor television 9 and the image on the partner terminal side is displayed. The audio data DA separated by the demultiplexing unit 6 is supplied to the audio data processing unit 5 and subjected to decoding processing or the like to form an audio signal SA. And this audio signal S
A is supplied to the speaker 10 and the voice of the partner terminal is output.

FIG. 14 shows an example of the configuration of the audio data processing section 5. First, the configuration of the coding system will be described.
The audio signal SA from the microphone 4 is A / D converter 11
Is converted to a linear PCM signal (digital signal) by the telephone voice encoder 12. As a telephone voice encoding method, for example, a maximum bit rate is 64 [kbps],
TTC standard J whose reproduction frequency upper limit is 3.4 [kHz]
T-G711 (hereinafter referred to as "G.711") is adopted. The audio data (code code) output from the encoder 12 is supplied to the selector 13 and also to the telephone voice decoder (G.711 decoder) 14 to be decoded.

The linear PCM signal output from the decoder 14 is supplied to the low bit rate speech encoder 15 and encoded. As the low bit rate audio encoding method, for example, TTC standard JT-G728 (hereinafter referred to as “G.728”) having a bit rate of 16 [kbps] and a reproduction frequency upper limit of 3.4 [kHz] is adopted. The audio data (code code) output from the encoder 15 is supplied to the selector 13. C for selector 13
The control signal SCa is supplied from the PU3, and one of the output audio data of the encoders 12 and 15 is selected based on the decoding capability information included in the bit rate allocation signal BAS of the partner terminal (call target) to select the audio data. DA
Is output as

Next, the structure of the decoding system will be described. The audio data DA from the demultiplexer 6 is supplied to the selector 21. The audio data DA from the demultiplexing unit 6 is a low bit rate audio decoder (G.728 decoder).
22 and the output signal of the decoder 22 is supplied to a telephone speech coder (G.711 coder) 23. In this case, when the audio data DA is encoded by the low bit rate speech encoder (G.728 encoder), it is decoded by the decoder 22 to obtain a linear PCM signal, and this linear PCM signal is obtained. It is supplied to the selector 21 after being encoded by the encoder 23.

A control signal S is sent from the CPU 3 to the selector 21.
Cb is supplied and the bit rate allocation signal BA of the partner terminal
Either the audio data DA from the demultiplexing unit 6 or the output audio data of the encoder 23 is selected based on the information on the encoding system of the audio data DA included in S. That is, when the audio data DA is encoded by the telephone voice encoder (G.711 encoder), the audio data DA itself is selected, while the audio data DA is the low bit rate voice encoder (G.711). 728 encoder), the output audio data of the encoder 23 is selected.
The audio data selected by the selector 21 is supplied to the telephone voice decoder (G.711 decoder) 24 and decoded, and then converted into an analog signal by the D / A converter 25 and output as an audio signal SA. .

FIG. 15 is a diagram illustrating a G.264 as a telephone voice encoder 12. The structural example of a 711 encoder is shown. In the figure,
The linear PCM signal is supplied to the A-law logarithmic quantizer 121 and the μ-law logarithmic quantizer 122, respectively, logarithmically quantized, and then supplied to the selector 123. A control signal SCc is supplied from the CPU 3 to the selector 123 (a control line is not shown in FIG. 14), and the quantizer 121, based on the decoding capability information included in the bit rate allocation signal BAS of the other terminal. One of the output data of G.122 is selected, and G.122. It is output as a 711 code.

Further, FIG. 11 shows a configuration example of a 711 decoder. In the figure, the G. The 711 code is supplied to the A-law antilogarithmic quantizer 141 and the μ-law antilogarithmic quantizer 142, respectively, and subjected to antilogarithmic quantization (linear reproduction) and then supplied to the selector 143. A control signal SCd is supplied from the CPU 3 to the selector 143 (a control line is not shown in FIG. 14), one of the output data of the quantizers 141 and 142 is selected and output as a linear PCM signal.
That is, in the selector 143, the G. When the 711 code is according to the A-law, the output data of the quantizer 141 is selected, while the G.711. When the 711 code is μ-law, the output data of the quantizer 142 is selected.

Although not described in detail, the decoding system telephone voice encoder (G.711 encoder) 23 has the same configuration as the telephone voice encoder 12 shown in FIG. G.711 decoder) 24 is configured similarly to the telephone voice decoder 14 shown in FIG.

Next, FIG. 17 shows the G.264 as the low bit rate speech encoder 15. The structural example of the 728 encoder (LD-CELP encoder) is shown. In the figure, 151 is a vector buffer to which a linear PCM signal is supplied. In this vector buffer 151, five consecutive audio samples are buffered to form a five-dimensional audio vector. Reference numeral 152 is a perceptual weighting filter for weighting the frequency of the input signal vector and the output vector of the synthesis filter described later.

153 is the difference between the perceptually weighted input signal vector and the output vector of the synthesis filter,
Of the 1024 (10-bit) quantized error signal vector candidates, the one having the smallest root mean square error is selected,
G. VQ index (16 [kb
ps]) is the least mean square error unit. Reference numeral 154 is an excitation VQ codebook as a table of 1024 quantized error signal vectors. Reference numeral 155 is a gain adjustment unit that adjusts the gain of the selected quantization error signal vector.

Reference numeral 156 is a backward gain adaptor for updating the gain coefficient of the gain adjusting unit 155 by backward adaptation. Reference numeral 157 is a synthesis filter for obtaining a quantized speech vector from the quantized error signal vector whose gain has been adjusted. Reference numeral 158 denotes a backward synthesis filter adaptive unit that updates the coefficient of the synthesis filter 157 by backward adaptation from the quantized speech vector obtained by the synthesis filter 157. Reference numeral 159 is a subtractor for calculating the difference between the input vector and the output vector of the synthesis filter 157.

FIG. 18 shows a low bit rate speech decoder 22.
As G. The structural example of the 728 decoder (LD-CELP decoder) is shown. In the figure, reference numeral 221 denotes an excitation VQ similar to the excitation VQ codebook 154 of the encoder 15 described above.
This is the Q code book 221. This codebook 221
To G. VQ index (16 [k
bps]) is supplied, the corresponding quantization error signal vector is output. Gain adjustment unit 22
2, backward gain adaptor 223, synthesis filter 22
4 and the backward synthesis filter adaptor 225 respectively include the gain adjustment unit 155 of the encoder 15 described above,
It is similar to the backward gain adaptor 156, the synthesis filter 157, and the backward synthesis filter adaptor 158.

A post filter 226 filters the quantized voice vector output from the synthesis filter 224 in order to improve the perceptual quality. Two
Reference numeral 27 denotes a PCM conversion unit that converts the five-dimensional quantized speech vector filtered by the post filter 226 into a corresponding five-sample linear PCM signal and outputs the linear PCM signal.

[0032]

The conventional audio data processing unit 5 is constructed as shown in FIG. 14, and when a low bit rate audio encoding system is selected as an audio encoding system, an encoding system is used. Since the encoding and decoding processes relating to the telephone voice encoding system are required in each of the audio and decoding systems, the DS for performing the encoding and decoding processes is required.
There has been a problem that the computing power of P (digital signal processor) is wasted. Then, depending on the DSP, the amount of calculation required for encoding becomes a decisive factor, the scale of hardware required for implementation becomes large, and there is a problem that the cost increases. Further, even when the low bit rate audio encoding method is selected, there is a problem that audio distortion occurs due to encoding and decoding according to the telephone audio encoding method, resulting in deterioration of sound quality.

Therefore, according to the present invention, the amount of operations of the DSP for encoding and decoding is reduced to reduce the cost by reducing the hardware scale, and the encoding device and the decoding device capable of reducing the signal deterioration. Is provided.

[0034]

According to a first aspect of the present invention, there is provided a plurality of coding means having different coding capabilities for supplying input signals in parallel, and a plurality of coding means. And a selector for selecting and outputting one code code from the code codes to be output.

An encoding apparatus according to a second aspect of the present invention is the encoding apparatus according to the first aspect of the present invention, which has selector control means for determining the decoding capability of the partner terminal and controlling the operation of the selector according to the decoding capability. is there.

According to a third aspect of the present invention, in the first aspect of the present invention, the plurality of encoding means are a telephone voice coder and a low bit rate voice coder. A linear PCM signal is supplied in parallel to the rate speech coder as an input signal.

According to a fourth aspect of the present invention, in the first aspect of the present invention, the plurality of encoding means are a telephone voice coder, a low bit rate voice coder and a high quality voice coder. The speech encoder, the low bit rate speech encoder, and the high-quality speech encoder have a linear P as an input signal.
CM signals are supplied in parallel.

According to a fifth aspect of the present invention, there is provided a decoding device in which a plurality of decoding means having different decoding capabilities are supplied with the input code code in parallel, and the decoding means outputs the decoding means. A selector for selecting and outputting one decoded signal from the signals and a selector control means for controlling the operation of the selector according to the encoding system of the input code are provided.

A decoding device according to a sixth aspect of the present invention is the decoding device according to the fifth aspect, wherein the plurality of decoding means are a telephone voice decoder and a low bit rate voice decoder. An input code code is supplied in parallel to the rate speech decoder.

In the decoding device according to the invention of claim 7, in the invention of claim 6, the plurality of decoding means are a telephone voice decoder, a low bit rate voice decoder and a high quality voice decoder. The input code code is supplied in parallel to the speech decoder, the low bit rate speech decoder, and the high-quality speech decoder.

[0041]

According to the present invention, an input signal is supplied in parallel to a plurality of coding means having different coding capabilities, and one code code is selected by a selector from output code codes of the plurality of coding means. It is selected and output. Therefore, when a certain encoding method is selected, it is not necessary to perform the encoding and decoding processing according to the other encoding method accordingly, and the operation amount of the DSP encoding and decoding is reduced to reduce the hardware. It is possible to reduce the cost by reducing the scale and reduce the signal deterioration.

According to the second aspect of the present invention, the decoding capability of the partner terminal is discriminated and the operation of the selector is controlled in accordance with the decoding capability. Therefore, an appropriate encoding method according to the decoding capability of the partner terminal is automatically selected. Can be selected.

According to the third aspect of the present invention, it is possible to select either the telephone voice encoding system or the low bit rate voice encoding system, and at that time, the encoding or decoding of the other encoding system is selected. It is possible to reduce the amount of calculation of the DSP to reduce the cost by reducing the hardware scale and reduce the deterioration of the sound quality.

In the invention of claim 4, it is possible to select one of the telephone voice coding system, the low bit rate voice coding system, and the coding relating to the high-quality voice coding system, and at that time, another coding system is selected. It is possible to reduce the amount of calculation for DSP encoding and decoding, reduce the cost by reducing the hardware scale, and reduce the deterioration of sound quality. Become.

According to the invention of claim 5, the input code code is supplied in parallel to a plurality of decoding means having different decoding capabilities, and the input code code is encoded from the output signals of the plurality of decoding means. One output signal is selected and output according to the above. Therefore, when the decoding according to a certain encoding method is selected, it is not necessary to perform the processing of encoding or decoding according to the other encoding method accordingly, which reduces the operation amount of the DSP and reduces the hardware scale. It is possible to reduce the cost due to the reduction and reduce the signal deterioration.

According to the sixth aspect of the present invention, it is possible to select either of the telephone voice encoding system and the low bit rate voice encoding system, and at that time, the encoding and decoding of other encoding systems are selected. It is possible to reduce the amount of calculation of the DSP to reduce the cost by reducing the hardware scale and reduce the deterioration of the sound quality.

In the invention of claim 7, it is possible to select any one of the telephone speech coding method, the low bit rate speech coding method and the decoding relating to the high-quality speech coding method, and at that time, another coding method is selected. Since it is not necessary to perform the encoding and decoding processing according to the above, it is possible to reduce the DSP calculation amount, reduce the cost by reducing the hardware scale, and reduce the sound quality deterioration.

[0048]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The first embodiment of the present invention will be described below with reference to FIG.
Examples will be described. This example is an example applied to the audio data processing unit 5 (see FIG. 6) of the video conference apparatus. In FIG. 1, parts corresponding to those in FIG. 14 are designated by the same reference numerals, and detailed description thereof will be omitted.

First, the configuration of the coding system (transmission side) will be described. The audio signal SA from the microphone 4 is A / D
After being converted into a linear PCM signal (digital signal) by the converter 11, a telephone voice encoder (G.711 encoder) 12 and a low bit rate voice encoder (G.728 encoder) 15
Are supplied in parallel to and encoded. And encoder 1
The audio data (code code) output from 2 and 15 are respectively supplied to the selector 13. Selector 1
The control signal SCa is supplied from the CPU 3 to the CPU 3, and the outputs of the encoders 12 and 15 are based on the transmission encoding method determined by the decoding capability information included in the bit rate allocation signal BAS of the partner terminal (call target). One of the audio data is selected and output as audio data DA.

Next, the configuration of the decoding system (reception side) will be described. The audio data DA from the demultiplexing unit 6 is supplied in parallel to a telephone voice decoder (G.711 decoder) 24 and a low bit rate voice decoder (G.728 decoder) 22. The output signals of the decoders 24 and 22 are supplied to the selector 21, respectively. Here, when the audio data DA is encoded by the telephone voice encoder (G.711 encoder), it is decoded by the decoder 24 to obtain a linear PCM signal. Also, audio data DA
Is encoded by a low bit rate speech encoder (G.728 encoder), it is decoded by the decoder 22 to obtain a linear PCM signal.

A control signal S from the CPU 3 is sent to the selector 21.
Cb is supplied and the bit rate allocation signal BA of the partner terminal
Decoders 24, 2 based on the reception encoding system determined by the encoding system information of audio data DA included in S
One of the two output signals is selected. That is, when the audio data DA is encoded by the telephone voice encoding system, the linear PC output from the decoder 24 is output.
When the M signal is selected, on the other hand, when the audio data DA is encoded by the low bit rate audio encoding system, the linear PCM signal output from the decoder 22 is selected. The linear PCM signal output from the selector 21 is D
The signal is converted into an analog signal by the / A converter 25 and output as an audio signal SA.

Here, a method of determining the transmission coding method (the encoder side of the own terminal = the decoder side of the partner terminal) will be described.

1. By receiving the bit rate allocation signal BAS from the partner terminal, the decoding capability of the partner terminal is obtained. For example, it is expressed as follows. Here, G. What is the decoding capability of 722-48? Ability to decode all modes of G.722. What is the decoding capability of 722-64. 722 of 64
Ability to decode only [kbps] mode, G. 711-
What is the decoding capability of A.G. Ability to decode A-rule of 711,
G. What is the decoding capability of 711-μ? The ability to decode 711 μ-law. G. Decoding capability of 728: Yes (1) No (0) 722-48 decoding ability: Yes (1) No (0) 722-64 decoding ability: Yes (1) No (0) Decoding ability of 711-A: Yes (1) No (0) Decoding ability of 711-μ: Yes (1) No (0)

2. The same applies to the coding capability of the terminal itself. Here, regarding the encoder, the G. There is no 722 mode distinction. That is, G. If 722-64 is 1, G. 722-48 is also set to 1. G. 728 code capability: Yes (1) No (0) 722-48 coding capability: Yes (1) No (0) 722-64 coding capability: Yes (1) No (0) 711-A coding capability: Yes (1) No (0) 711-μ coding ability: Yes (1) No (0)

3. Exchange capabilities on both terminals. That is,
Performs a logical product between the corresponding encoding methods.

4. G. 728, G.I. 722-48, G.I.
722-64, G.I. 711-A and G.G. When 711-μ is associated with bit 4, bit 3, bit 2, bit 1 and bit 0, respectively, the encoding method is determined based on the logical product result as shown in FIG. Here, G.
728 and G.H. When both terminals share the capability of 722, that is, in the case of 1 * 1 **, it is originally desirable that the decoder side has the selection right. This is because the quality of code decoding affects only the receiving side. On the decoder side,
If priority is given to image quality, G.I. If the sound quality is prioritized, G. 722 will be selected. In order to realize this, it is sufficient to perform the capability exchange again, assuming that the encoding method which is not the capability desired to be selected on the decoder side does not have the capability. For example, if the image quality is prioritized, the decoding capability is set to 0 * 1 ** and the capability exchange is requested again to the encoder side.

5. The encoding method uniquely determined is transmitted to the partner terminal (decoder side) using the bit rate allocation signal BAS (audio encoding command), and the transmission of the audio data DA is started by the encoding method.

G. 722 is an encoding method adopted as a high-quality audio encoding method, in which the maximum bit rate is 64 [kbps] and the reproduction frequency upper limit is 7.0 [kHz.
z], which is the TTC standard JT-G722.

Next, a method of determining the reception coding method (decoder side of own terminal = coder side of partner terminal) will be described. 1. The decoding capability of the own terminal is set to the bit rate allocation signal BAS.
(Terminal capability 2) is used to transmit to the partner terminal. 2. The partner terminal uniquely determines the transmission coding scheme in the same procedure as the method of determining the transmission coding scheme of the own terminal. 3. The uniquely determined encoding method is transmitted to the terminal itself (decoder side) using the bit rate allocation signal BAS from the partner terminal, and the encoding data is used to decode the audio data DA from the partner terminal.

In the example of FIG. 1, in the coding system,
Telephone audio encoder 12 and low bit rate audio encoder 15 after audio signal SA has been converted to a digital signal.
Are supplied in parallel, and one of the output code codes of the encoders 12 and 15 is selected by the selector 13 and output as audio data DA. Also, in the decoding system, the audio data DA is the telephone voice decoder 24.
And the low bit rate audio decoder 22 are supplied in parallel, and one of the output signals of the decoders 24 and 22 is selected by the selector 21 according to the encoding system of the audio data DA and output as the audio signal SA. . Therefore, when any one of the telephone speech coding scheme and the low bit rate speech coding scheme is selected in the coding system and the decoding system, the coding relating to the other coding system as in the past is performed. Therefore, it is possible to reduce the amount of calculation for DSP encoding and decoding, to reduce the cost due to the reduction in the hardware scale, and to reduce the sound quality deterioration.

Since the selecting operation of the selector 13 of the coding system is controlled based on the information such as the decoding capability of the partner terminal, the decoding capability of the partner terminal among the code codes output from the encoders 12 and 15 is determined. There is a benefit that an appropriate code code corresponding to the selected data is automatically selected and output as the audio data DA.

Next, a second embodiment of the present invention will be described with reference to FIG. This example is also an example applied to the audio data processing unit 5 (see FIG. 6) of the video conference apparatus. 3, parts corresponding to those in FIG. 1 are designated by the same reference numerals, and detailed description thereof will be omitted.

First, the configuration of the coding system (transmission side) will be described. The audio signal SA from the microphone 4 is A / D
After being converted into a linear PCM signal (digital signal) by the converter 11A, the linear PCM signal is supplied to the high-quality speech encoder 16 and encoded. Then, the audio data (code code) output from the encoder 16 is supplied to the selector 17. As a high-quality audio encoding method, for example, a maximum bit rate of 6
TTC standard JT-G722 (hereinafter, referred to as "G.722") having 4 [kbps] and a reproduction frequency upper limit of 7.0 [kHz].
Will be adopted. In the A / D converter 11A, the G. The audio signal SA is sampled at the sampling frequency corresponding to the reproduction upper limit frequency of the coding method 722.

Further, the linear PCM signal output from the A / D converter 11A is band-limited by the low-pass filter 18 and then supplied to the down converter 19, and G. G.711 as the encoding system and the low bit rate audio encoding system. After being converted to a sample rate corresponding to the reproduction upper limit frequency of the encoding system of G.728, the telephone voice encoder (G.711 encoder) 12 and the low bit rate voice encoder (G.728 encoder) 15
Are supplied in parallel to and encoded. And encoder 1
The audio data (code code) output from 2 and 15 are supplied to the selector 17, respectively.

A control signal S is sent from the CPU 3 to the selector 17.
One of the audio data output from the encoders 16, 12, and 15 based on the transmission encoding method that is supplied with Ce and is determined by the decoding capability information included in the bit rate allocation signal BAS of the other terminal (call target). Is selected and output as audio data DA. The method of determining the transmission coding method is performed in the same manner as described in the example of FIG.

Next, the configuration of the decoding system (reception side) will be described. The audio data DA from the demultiplexing unit 6 includes a high-quality speech decoder (G.722 decoder) 26, a telephone speech decoder (G.711 decoder) 24, and a low bit rate speech decoder (G.728 decoder). ) 22 in parallel. Here, the audio data DA has been encoded by a high-quality speech encoder (G.722 encoder), a telephone speech encoder (G.711 encoder) and a low bit rate speech encoder (G.728 encoder). If so, it is decoded by decoders 26, 24 and 22 respectively to obtain a linear PCM signal.

The output signal of the decoder 26 is supplied to the selector 27. The output signals of the decoders 24 and 22 are supplied to the selector 21, respectively. CPU for the selector 21
3, the control signal SCf is supplied, and the audio data DA is transmitted by the telephone voice encoding system based on the reception encoding system determined by the encoding system information of the audio data DA included in the bit rate allocation signal BAS of the partner terminal. Output linear PCM of decoder 24 when encoded
The signal is selected, while the output linear PCM signal of the decoder 22 is selected when the audio data DA has been encoded by a low bit rate speech coding scheme. The method of determining the reception coding method is performed in the same manner as described in the example of FIG.

The linear PCM signal output from the selector 21 is supplied to the up-converter 28, and the G.264 standard as a high-quality voice encoding system is used. After being converted to a sample rate corresponding to the reproduction upper limit frequency of the encoding method 722, the band is limited by the low-pass filter 29 and supplied to the selector 27. The control signal SCg is supplied from the CPU 3 to the selector 27, and the selection operation is controlled based on the reception coding system determined by the coding system information of the audio data DA included in the bit rate allocation signal BAS of the partner terminal. .

That is, when the audio data DA is encoded by the high-quality voice encoding system, the linear PCM signal output from the decoder 26 is selected. On the other hand, when the audio data DA is encoded by the telephone voice encoder system or the low bit rate voice encoding system, the linear PCM signal output from the low pass filter filter 29 is selected. The linear PCM signal output from the selector 27 is converted into an analog signal by the D / A converter 25A and output as the audio signal SA.

FIG. 4 is a diagram showing an example of the G.264 as high-quality speech encoder 16. The example of a structure of the 722 encoder (SB-ADPCM encoder) is shown. In the figure, 161 is an input signal (linear P
CM signal) Xin is a signal of 4 [kHz] or higher (high range) XH
And a signal (low band) XL of 4 [kHz] or less, and a transmission quadrature mirror filter for down-sampling each. A high band ADPCM encoder 162 performs ADPCM (adaptive differential PCM) coding on the high band XH to obtain a signal IH of 16 [kbps]. 163
Is a low band ADPCM encoder for performing ADPCM (adaptive differential PCM) coding on the low band XL to obtain a signal IL of 48 [kbps]. 164 is the signals IH and IL
Are time-division multiplexed to obtain a transmission signal (audio data DA) I of 64 [kbps].

FIG. 5 is a diagram showing an example of the G.35 as the high-quality speech decoder 26. The example of a structure of a 722 decoder (SB-ADPCM decoder) is shown. In the figure, 261 is the received 64 [k
bps] or lower signal Ir and 16 [kbps] signal IH
And a signal ILr of 48 [kbps] or less. 262 is ADPCM for the signal IH
It is a high frequency ADPCM decoder for performing (adaptive differential PCM) decoding to obtain a high frequency audio signal rH. 26
3 performs ADPCM (adaptive differential PCM) decoding on the signal ILr and the line audio rate (64 kbps / 56).
It is a low band ADPCM decoder for obtaining a low band audio signal rL by performing a mode corresponding to kbps / 48 kbps). Reference numeral 264 is a reception quadrature mirror filter for interpolating the high-frequency and low-frequency audio signals rH and rL and up-sampling to obtain the audio signal (linear PCM signal) Xout of the entire band.

In the example of FIG. 3, in the encoding system,
After the audio signal SA is converted into a digital signal, the audio signal SA is supplied in parallel to the high-definition audio encoder 16, the telephone audio encoder 12 and the low bit rate audio encoder 15, and output code codes of the encoders 16, 12, 15 Is selected by the selector 13 and output as the audio data DA. In the decoding system, the audio data DA is the high-definition audio decoder 26 and the telephone audio decoder 24.
And the low bit rate audio decoder 22 are supplied in parallel, and one of the output signals of the decoders 26, 24 and 22 is supplied to the selector 2 according to the encoding system of the audio data DA.
It is selected by 1 and 27 and output as the audio signal SA. Therefore, when any one of the high-definition speech coding method, the telephone speech coding method, and the low bit rate speech coding method is selected in the coding system and the decoding system, the other coding method is selected as in the conventional method. It is not necessary to perform encoding or decoding processing related to the encoding method, the amount of calculation for DSP encoding or decoding is reduced, the cost is reduced by reducing the hardware scale, and the sound quality deterioration is reduced. You can

Since the selecting operation of the selector 17 of the encoding system is controlled based on the information such as the decoding capability of the partner terminal, the decoding of the partner terminal among the code codes output from the encoders 16, 12, 15 is performed. There is a benefit that an appropriate code code according to the capability is automatically selected and output as the audio data DA.

Although the above-described embodiment is applied to the audio data processing unit 5 of the video conference apparatus, the present invention is not limited to the encoding and decoding of audio signals, and other types of video signals and the like. It is also effective when applied to the case where a signal is encoded and decoded by a plurality of encoding methods.

[0075]

According to the first aspect of the present invention, the input signal is supplied in parallel to a plurality of coding means having different coding capabilities, and one of the output code codes of the plurality of coding means is selected by the selector. Since the code code is selected and output, when a certain coding method is selected, it is not necessary to perform the coding or decoding processing related to the other coding method, and the coding of the DSP or It is possible to reduce the calculation amount of decoding and reduce the cost due to the reduction of the hardware scale, and reduce the signal deterioration.

According to the second aspect of the present invention, the decoding capability of the partner terminal is discriminated and the operation of the selector is controlled in accordance with the decoding capability. Therefore, an appropriate encoding method according to the decoding capability of the partner terminal is automatically selected. Can be selected as desired.

According to the third aspect of the present invention, it is possible to select either the telephone voice encoding system or the low bit rate voice encoding system, and at that time, the encoding and decoding of other encoding systems are performed. It is possible to reduce the amount of calculation of the DSP to reduce the cost due to the reduction of the hardware scale and to reduce the deterioration of the sound quality.

According to the fourth aspect of the present invention, it is possible to select any one of the telephone speech coding method, the low bit rate speech coding method and the high-quality speech coding method, and at that time, another coding method is selected. It is not necessary to perform encoding and decoding processing according to the method, and it is possible to reduce the DSP calculation amount, reduce the cost by reducing the hardware scale, and reduce the sound quality deterioration.

According to the invention of claim 5, the input code code is supplied in parallel to a plurality of decoding means having different decoding capabilities, and the input code code is encoded from the output signals of the plurality of decoding means. Since one output signal is selected and output according to the system, when the decoding according to a certain coding system is selected, the encoding or decoding processing according to another coding system is performed accordingly. It is not necessary to reduce the amount of calculation for DSP encoding and decoding to reduce the cost by reducing the hardware scale and reduce the signal deterioration.

According to the sixth aspect of the present invention, it is possible to select either the decoding according to the telephone speech coding system or the low bit rate speech coding system, and at that time, the coding or decoding according to the other coding system is performed. It is possible to reduce the amount of calculation of the DSP to reduce the cost due to the reduction of the hardware scale and to reduce the deterioration of the sound quality.

According to the invention of claim 7, it is possible to select any one of the decoding according to the telephone voice encoding system, the low bit rate voice encoding system and the high-definition voice encoding system, and at that time, another encoding is selected. It is not necessary to perform encoding and decoding processing according to the method, and it is possible to reduce the DSP calculation amount, reduce the cost by reducing the hardware scale, and reduce the sound quality deterioration.

[Brief description of drawings]

FIG. 1 is a configuration diagram showing a first embodiment of the present invention.

FIG. 2 is a diagram for explaining a method of determining a transmission coding method.

FIG. 3 is a configuration diagram showing a second embodiment of the present invention.

FIG. 4 G. 722 encoder (SB-ADPCM encoder)
It is a figure which shows the structural example.

FIG. 5 G. 722 decoder (SB-ADPCM decoder)
It is a figure which shows the structural example.

FIG. 6 is a diagram illustrating a configuration example of a video conference device.

FIG. 7 is a diagram showing a relationship between a frame structure of a B channel and a bit string.

FIG. 8 is a diagram showing a frame structure of a B channel.

FIG. 9 is a diagram showing a structure of a frame synchronization signal FAS.

FIG. 10 is a frame synchronization signal F in multi-frame.
It is a figure which shows the structure of AS.

FIG. 11 is a diagram showing a configuration of a bit rate allocation signal BAS.

FIG. 12 is a diagram showing an example of a transmission format when two channels of B channel are used.

FIG. 13 is a diagram showing a frame structure of an H ₀ channel.

FIG. 14 is a diagram showing a configuration example of a conventional audio data processing unit (voice codec).

FIG. 15 G. It is a figure which shows the structural example of a 711 encoder.

FIG. 16 G. It is a figure which shows the structural example of a 711 decoder.

FIG. 17: G. 728 encoder (LD-CELP encoder)
It is a figure which shows the structural example.

FIG. 18 G. 728 decoder (LD-CELP decoder)
It is a figure which shows the structural example.

[Explanation of symbols]

 1 Imaging Device 2 Image Data Processing Unit 3 CPU 4 Microphone 5 Audio Data Processing Unit 6 Demultiplexing Unit 7 Line Interface 8 Line 9 Monitor Television 10 Speaker 11, 11A A / D Converter 12 Telephone Voice Encoder (G.711 Encoder) ) 13, 17, 21, 27 selector 15 low bit rate speech encoder (G.728 encoder) 16 high-quality speech encoder (G.722 encoder) 18, 29 low-pass filter 19 down converter 22 low bit rate speech decoding Device (G.728 decoder) 24 telephone speech decoder (G.711 decoder) 25,25 A D / A converter 26 high-definition speech decoder (G.722 decoder) 28 up-converter

Claims (7)

[Claims] 1. A plurality of encoding means for supplying input signals in parallel and having different encoding capabilities, and one code code selected from the code codes output from the plurality of encoding means and output. An encoding device comprising: a selector. 2. The encoding apparatus according to claim 1, further comprising selector control means for determining the decoding capability of the partner terminal and controlling the operation of the selector according to the decoding capability. 3. The plurality of coding means are a telephone speech coder and a low bit rate speech coder, and a linear PCM signal is parallel to the telephone speech coder and the low bit rate speech coder as the input signal. The encoding device according to claim 1, wherein the encoding device is supplied to 4. The plurality of encoding means are a telephone speech coder, a low bit rate speech coder and a high quality speech coder, and these telephone speech coder, low bit rate speech coder and high quality speech coder. The coding apparatus according to claim 1, wherein a linear PCM signal is supplied in parallel to the input terminal as the input signal. 5. A decoding signal selected from a plurality of decoding means which are supplied with input code codes in parallel and have different decoding capabilities, and decoded signals output from the plurality of decoding means. And a selector for controlling the operation of the selector according to the encoding method of the input code code. 6. The plurality of decoding means are a telephone voice decoder and a low bit rate voice decoder, and the input code code is supplied in parallel to the telephone voice decoder and the low bit rate voice decoder. The decoding device according to claim 5, wherein: 7. The plurality of decoding means are a telephone speech decoder, a low bit rate speech decoder and a high quality speech decoder, and the telephone speech decoder, the low bit rate speech decoder and the high quality speech decoder. The decoding device according to claim 5, wherein the input code code is supplied in parallel to.