JPS60107933A - Adpcm encoding device - Google Patents

Abstract

PURPOSE:To eliminate the unnaturalness of a hearing sensation on the reception side in a soundless section when ADPCM codes only in a sound section are transmitted by sending information on the length of the soundless section and a mean noise level from a transmission side to the reception side. CONSTITUTION:An ADPCM encoder 13 is controlled with the output of a sound detector 10 to perform ADPCM encoding only in the sound section. A noise analyzer 12 generates representative values of soundless-section length information and soundless-section level information, and they are inputted to and encoded by an encoder 15. The output of the encoder 15 is multiplexed by a multiplexing circuit 16 with the output of the encoder 13 and outputted. A demultiplexing circuit 7 on the reception side demultiplexes the receive signal into ADPCM codes and the codes from the encoder 15, and the representative value from the encoder 15 are decoded by a decoder 18 and supplied to a pseudo random noise generator 20 to obtain a random noise with a level based upon the representative value in the soundless section, thereby outputting it from a terminal 4 in the soundless section.

Description

DETAILED DESCRIPTION OF THE INVENTION [Technical Field of the Invention] The present invention relates to an ADPCM encoding device configured to remove redundancy of audio and perform communication by compressing a band.

[Prior art]

Digital audio usually has a strong correlation between each sample value, and ADPCM encoding is one of the encoding methods designed to remove this correlation. A CM No. 1 converter is being considered.

A normal ADPCM encoding device performs ADPCM encoding regardless of the presence or absence of voice, but the amount of information generated can be further reduced by detecting only the voiced section of voice, performing ADPCM encoding, and transmitting it. , such ADPCM
Encoding devices are also being considered.

In this way, the ADPCM code 1hi device that transmits only the sound section uses a voice detector to determine whether there is a sound or no sound on the transmitting side, and then combines the ADPCM encoded information only with the sound section and the sound/no sound. The receiving side receives information regarding voice/silence and information regarding ADPCMff coding, reproduces the ADPCM code only in the voice section, and outputs an audio signal. In this case, conventionally, it was only possible for the receiving side to output silence or a constant noise level while there was no audio signal, which had the disadvantage of creating an unnatural sound. .

[Object and structure of the invention]

The present invention has been made in view of the above circumstances, and its purpose is to provide an ADPCM code that eliminates the unnaturalness in the receiver's auditory sense during silent periods when ADPCM coding is performed and transmitted only in voiced periods. The objective is to provide a device for converting

In order to achieve such an object, the present invention detects the length and average noise level of a silent interval on the transmitting side and transmits this information to the receiving side, and the receiving side detects the silent interval length and average noise level. The noise level that is the same as that of the transmitting side noise is inserted into the silent interval based on the information regarding the length and average noise level. Hereinafter, the present invention will be explained in detail using Examples.

〔Example〕

FIG. 1 is a block diagram showing an embodiment of the present invention. In the figure, a voice signal input manually from a transmission side input terminal 1 is given to a voice detector 1G, a delay circuit 11, and a noise analyzer 12. The delay circuit 11 is a circuit that compensates for the processing time in the audio detector 10. The voice detector 10 detects the sound section of the input voice, and converts the voice detector output 101 into A.
It is applied to the DPCM encoder 13 and the inversion circuit 14. ADPC
The M encoder 13 is controlled by the audio detector output 101 and converts only the sound section of the audio signal input from the delay circuit 11 into A.
DPCM encode. The noise analyzer 12 is composed of a silent interval length counter that counts the length of a silent interval and a silent interval level detector that detects the average signal level of the silent interval, and is activated by the voice detector inverted output 102. Under the control, representative values of silent interval length information and silent interval level information of the silent interval noise component are created and inputted to the encoder 15. encoder 1
5 encodes the silent interval length information and silent interval level information and supplies the encoded result to the multi-IT conversion circuit 16. The multiplex conversion circuit 16 performs multiple conversion on the output of the ADPCM encoder 13 and the output 10-3 of the encoder 15, and sends out a code to the transmitter output terminal 2.

The codes outputted from the output terminal 2 are inputted to the receiving side input terminal 3 and given to the separation conversion circuit 17. Is the separation conversion circuit 1T ADPCM? The rgf code and the signal output from the encoder 15 are separated and sent to the ADPCM decoder 19.
A DPCM code output 104 is provided while a decoder 18 is provided with the encoder output 105. The decoder 18 decodes the representative values of the silent section information and the silent section level information and provides these two pieces of information to the pseudorandom noise generator 2°, while outputting the silent section inversion information 10 to the ADPCM decoder 19.
Give 6.

It decodes the ADPCM code under the control of the power 106 and outputs an audio signal. On the other hand, pseudorandom noise generator 2
0 generates random noise at a level determined by the representative value of the silent section level information for the silent section using the representative value of the silent section information.

These audio signals from the ADPCM decoder 19 and the random 168 from the pseudorandom noise generator 20 are sent to the adder 2
1 and outputted as an audio signal from the receiving side output terminal 4.

Next, the data format in this embodiment will be shown using FIG. Since the data in this example is only data of a sound section, 20
The head flags of the sound section indicated by 0 and 300 are added. These 7 jigs 200.300 are AD in Figure 1.
It is inserted before the audio code in the PCM encoder 13, and the ADP
The CM decoder 19 detects this flag and acquires word synchronization. On the other hand, 201.301 is a representative value of silent section length information and silent section level information. Also 202
, 302 is written. Therefore, on the receiving side, the leading flag was detected and word synchronization was obtained, so the silent interval length information and the silent interval level had representative value data of 201°30.
1, and then reproduces the audio data 202, 302.

〔Effect of the invention〕

As explained above, according to the present invention, the length of the silent section and the average noise level of the silent section are detected on the transmitting side, and the representative value thereof is encoded and transmitted, and the receiving side receives the encoded signal. By decoding and controlling the pseudo-random noise generator based on the length of the silent interval and the representative value of the average noise level of the silent interval, the noise level is generated during the silent interval at the same level as that on the transmitting side. It is possible to eliminate the discomfort of

[Brief explanation of drawings]

FIG. 1 is a block diagram showing one embodiment of the present invention, and FIG. 2 is a diagram showing an example of a data format in that case. 1... Sending side input terminal, 2... Sending side output terminal, 3φ... Receiving side input terminal, 4... Receiving side output terminal, 10...φ・Audio detector, 11φ・・Delay circuit,
12φ...Noise analyzer, 13...ADPCM encoder, 14...Inverting circuit, 15...Encoder, 16
...Multiple conversion circuit, 1T... Separation conversion circuit, 1
8...Decoder, 19...ADPCM decoder, 2
0...Pseudorandom noise generator, 21...Conflict adder.

Claims (1)

[Claims] In an ADPCM encoding device that ADPCM encodes an audio signal, detects the presence or absence of the audio signal, classifies it into a sound section and a silent section, and transmits a signal of only the sound section, the transmitter side:
A silent section length counter that counts the length of a silent section and a silent section level detector that detects an average signal level of the silent section, and an output code of the silent section length counter and an output code of the silent section level detector. After inserting the flag indicating the start of the sound section into the next sound section signal, the ADPCM encoded data is sent out, and the receiving side receives a pseudo-random signal in which the noise level generated according to the section level signal is controlled. After acquiring word synchronization by detecting a flag representing the beginning of a voiced period from the signal string of the transmitted voiced period, the pseudo-random noise generator generates a silent period on the transmitting side. The series number, Shicho Kaiya 1st novel, or χ calculation ←■
ptfGtorRJ + Shijurokukai Yuki 2☆ADPC for Tachi section
An ADPCM encoding device characterized in that an ADPCM decoder output signal obtained by decoding Mi code is outputted, and the pseudorandom noise is outputted in a silent period.