JPH0263335A - Stereo voice transmission system - Google Patents

Abstract

PURPOSE:To transmit a stereo sound with high quality by controlling the title system such that the sound of plural input channels with a maximum power is inputted to a main channel at all times. CONSTITUTION:The power of an input sound on a main channel 1003 and a sub channel 1004 is monitored by a power comparator 1005 to change over the correspondence between the output of left/right microphones 1R, 1L and the main and sub channels 1003, 1004 by using switches 1001, 1002 when the power on the sub channel 1004 exceeds the power of the main channel 1003. Moreover, the power of the main and sub channels 1003, 1004 after decoding by a decoding section is monitored by a power comparator 1006 even at the receiver side to change over the correspondence between the output of left/right microphones 2R, 2L and the main and sub channels 1003, 1004 by using switches 1007, 1008 when the power on the sub channel 1004 exceeds the power of the main channel 1003. Thus, quantization noise is reduced and the stereo sound with high quality is obtained.

Description

DETAILED DESCRIPTION OF THE INVENTION [Object of the Invention] (Field of Industrial Application) The present invention relates to a stereo audio transmission system applied to, for example, a remote conference system.

(Prior Art) The present inventor has disclosed a transmission system capable of low-cost, high-quality stereo audio transmission in Japanese Patent Laid-Open No. 61-82840. To summarize the invention, only the main audio signal and the residual signal are transmitted in stereo audio transmission,
The most popular method is to perform equivalent speech estimation on the receiving and transmitting sides, and then add a residual signal to this estimated speech signal on the receiving side.

Further, according to the present invention, a single speech is encoded in high-quality stereo, and even when multiple speeches are made, although the level of the residual signal of the subchannel increases, the adaptive quantizer prevents the increase in the audio level. Since it can be encoded following
It is possible to compensate for the deterioration in the inter-channel correlation component removal amount by increasing the quantization noise. Therefore, even when multiple speakers are speaking, the problem that the sound image is distorted does not occur.

However, in this method, the sub-channel that encodes the residual signal from which the estimated signal, which is the output of the inter-channel predictor, is removed, and the main channel that encodes the input No. 12 of the inter-channel predictor are connected to the left and right microphones, respectively. Because the allocation is fixed, for example, if a person close to the microphone on the sub-channel side speaks, the power on the sub-channel side to which a small number of bits of about 2 bits are allocated is greater than the power on the main channel side. Occur.

In a method such as ADPCM that assigns a quantization step size during adaptation according to the power of the input signal, it is advantageous to allocate a larger bit length to the one with higher power, so in the above case, the inter-channel correlation component is removed. However, since the power on the sub-channel side is originally large, there is a problem that quantization noise increases.

(Problems to be Solved by the Invention) The stereo audio transmission system disclosed by the present inventor is capable of transmitting high-quality stereo audio at low cost, but the main channel and sub-channel are connected to left and right microphones, respectively. As a result, the power on the sub-channel side is larger than the power on the main channel side, resulting in a problem of increased quantization noise.

The present invention was made based on the above-mentioned circumstances, and an object of the present invention is to provide a system that can transmit high-quality stereo sound with reduced quantization noise.

[Structure of the Invention] (Means for Solving the Problems) The present invention encodes a plurality of input signals, transmits them via two types of channels with different bit rates, decodes them, and outputs them from a plurality of speakers. In the stereo audio transmission system, the side that encodes the input signal includes first level detection means for detecting each level of the plurality of input signals;
Depending on the detection result by the first level detection means, at least one input signal among the plurality of input signals is encoded and assigned to one of the two types of channels, and one of the plurality of input signals is assigned to one of the two types of channels. a signal encoding/channel assignment means for encoding the remaining input signal and assigning it to the other of the two types of channels, and the side that decodes the encoded input signal is configured to
signal decoding means for respectively decoding encoded input signals transmitted from different channels; and second level detection means for detecting the level of each input signal decoded by the signal decoding stage f. and, according to the detection result by the second level detection means, at least t) one input signal among the input signals decoded by the signal decoding means is assigned to at least one of the speakers, and the signal decoding is performed. and signal allocation means for allocating the remaining input signals of the input signals decoded by the decoding means to the remaining speakers.

Further, the second invention provides a stereo audio transmission system in which a plurality of input signals are encoded, transmitted via two types of channels with different bit rates, decoded, and output from a plurality of speakers. The encoding side includes a first estimating means for estimating an input signal allocated to one channel of a plurality of input signals from an input signal allocated to the other channel, and an estimation result by the first estimating means. a signal generating means for generating a residual signal by removing a correlation component with the input signal assigned to the one channel from the input signal assigned to the other channel; and a residual signal generated by the signal generating means. a signal transmission means for encoding the difference signal and transmitting the encoded input signal to the decoding side; detecting the level of the input signal assigned to the one channel; A first clearing means for clearing the estimation result by the estimation means of No. 1, and a side for decoding the encoded input signal, decoding the residual signal transmitted from the signal transmission means, and decoding the residual signal transmitted from the signal transmission means. second estimating means for synthesizing a correlation component obtained by estimating the input f bait assigned to the other channel from the difference signal and the input signal assigned to the one channel; and a second clearing means for detecting the level of the input signal and clearing the estimation result by the second estimating means according to the detection result.

(Function) In the present invention, the input signal encoding side encodes at least one input signal among the plurality of input signals according to each level of the plurality of input signals, assigns it to one channel, encode the remaining input signal among the input signals,
The side that decodes the encoded input signal assigns at least one of the input signals to at least one speaker according to each level of the plurality of input signals, and the input signal is assigned to the other channel. Since the remaining input signals are allocated to the remaining speakers, quantization noise is reduced and high-quality stereo audio can be transmitted.

(Example) The present invention will be explained in detail below.

FIG. 1 is a diagram for explaining the principle of the stereo audio transmission system of the present invention.

As shown in the figure, on the transmitting side of this system, switches 1001 and 1002 are inserted between the left and right microphones 1.1 and the encoding section (1?L omitted, not shown).

Then, the main channel 1003 and the sub channel 10
By monitoring the power of the input audio of 04 with the power comparator 1005, when the power of the sub channel 1004 exceeds the power of the main channel 1003, the switch 1001
．． 1002 switches the correspondence between the output L of the left and right microphones 1.1 and the main and sub channels 1003 and 1004.

Also, on the receiving side, a power comparator 1006 monitors the power of the main and sub channels 1003 and 1004 after decoding by a decoding unit (not shown). When the decoded audio of the sub channel 1004 exceeds that of the main channel 1003, the switch 1007.100B switches the correspondence between the main and sub channels 1003.1004 and the left and right speakers 2.2 and L.

As described above, in the present invention, among the voices of a plurality of human-powered channels, the one with the highest power is always controlled to be input to the main channel.

In addition, in order to perform more accurate switching, switching in encoding may be performed by detecting the power of this code output if the encoder has a built-in decoder, such as in ADPCM. can.

Next, specific examples of the present invention will be described.

FIG. 2 is a diagram showing the configuration of a stereo audio transmission system according to this embodiment.

As shown in the figure, the transmitter 100 of this stereo audio transmission system includes one of the left and right microphones 1.
Main channel A that ADL PCM encodes the audio input from 1 and assigns it to the main channel.
A DPCM encoding unit 30 ADPCM encodes the voice L manually input from either the left or right microphone 1.1 and assigns it to a sub channel.A sub channel ADPCM encoder 31 predicts the sub channel voice based on the main channel voice. an inter-channel correlation removal unit 34 that removes inter-channel correlation from the sub-channel audio;
Detects the levels of the main channel audio and sub-channel audio, and compares the output of the left and right microphones 1.1 with the main,
Level detection unit 300 that controls switching of correspondence with the sub-channel L channel
It becomes more.

ADPCM signal D (
k) and F (k) are transmitted via the transmission line 25.26,
The signal is input to the receiving section 200.

On the other hand, the receiving section 200 of this stereo transmission system decodes the main channel ADPCM code D (k) and reproduces the main channel audio x 1 (t).
PCM decoding section 32, sub-channel ADPCM code F
The sub-channel ADPCM decoding unit 33 decodes the left channel audio y1 (k) and reproduces the left channel audio y1(t). a channel correlation adding section 35 that adds correlation between channels, and a level detection L section 400 that detects the levels of main channel audio and sub channel audio and controls switching of the correspondence between the main and sub channels and the left and right speakers 2.2. It becomes more.

Next, each part will be explained in detail.

Main channel ADPCM encoding section 30 The main channel ADPCM encoding section 30 is an A/D converter 1
01, subtracter 103, adaptive quantizer 105, adaptive inverse quantizer 107, adder 109, predictor 111, switch 1
Consists of 81.

The main channel sound x (k) inputted from either of the left and right microphones 1.1 and digitized and selected by the switch 181 is subtracted from the main channel predicted sound value (k) predicted by the predictor 111, and the result is calculated by prediction. Prediction residual signal d whose power is less than x (k)
(k), for example, 32 kb, by an adaptive quantizer 105 that adaptively changes the quantization step according to the amplitude.
It is encoded into an ADPCM code of about ps.

In addition, in the pre-AM unit 111, the reproduced prediction residual signal d1 (k) decoded by the adaptive inverse quantizer 107 and the output signal (k) of the predictor are added together in the adder 109. The same main channel reproduced audio x 1 (k) is input to the channel ADPCM decoding unit 32, and adaptive filter processing is performed to minimize the power of the reproduced prediction residual signal d1 (k).

Sub-channel ADPCM encoding unit 31 The sub-channel ADPCM encoding unit 31 is
13, a delay unit 115, 180, a subtracter 117, an adaptive quantizer 119, an adaptive inverse quantizer 121, an adder 123, a pre-7111j unit 127, and a switch 182.

The inter-channel correlation remover 3 receives R and L signals input from either of the left and right microphones 1.1, digitizes the signals, delays them by the delay circuit 115 or 180, and selects them by the switch 182.
The inter-channel correlation removed signal e (k
) is subjected to ADPCM encoding processing similar to the main channel.

The above-mentioned delay guarantees the causality of the main and sub-channel sounds x (k) and y (k) in the inter-channel correlation removal unit 34 (if the direct sound reaches the left microphone earlier than the right microphone) However, the inter-channel correlation remover 3
The input to 4 is such that the left audio is always delayed.

Here, the auto-correlation component of the audio is removed by the predictor 127, and the correlation component with the main channel audio mixed in the sub-channel audio is removed by the inter-channel correlation removal unit 34, so there is less information than the main channel. amount of ADPC
M code, for example, can be compressed to about 16 kbps.

Inter-channel correlation remover 34 The inter-channel correlation remover 34 is the inter-channel predictor 12
9. Consisting of a subtracter 131, main channel playback audio xt
(k) is input, and an inter-channel correlation component 9(k) is generated by an inter-channel predictor 129. The inter-channel predictor 129 performs adaptive filter processing and adaptively changes the characteristics of the filter so as to minimize the power of the sub-channel prediction residual signal e 1 (k) in the receiving section 200.

Level Detection Unit 300 The level detection unit 300 includes level detectors 161 and 164, comparators 162 and 163, a hangover processing circuit 165, an OR circuit 166, a multiplier 167, and an adder 168.

The level detection unit 300 inputs the main channel playback audio X l (k) of the adder 109 and also inputs the inter-channel correlation component 9 (
k) and the sub-channel prediction residual signal e 1 of the adder 123
(k) in an adder 168, the sub-channel reproduction signal yt(k) is input.

The levels of these input signals are detected by level detectors 161 and 164, and are input to comparators 162 and 163.

The comparator 162 sets the level of the main channel reproduced audio x 1 (k) to a predetermined threshold level TH.
When the prediction coefficient clear signal HCL is output from the OR circuit 166, the transmission section 100 and the reception section 2
00, it simultaneously clears the pren1 coefficient (specifically, the tap coefficient of the adaptive filter) of the inter-channel predictors 121 and 151, thereby achieving synchronization between the transmitter 100 and the receiver 200.

On the other hand, the comparator 163 compares the levels of the reproduced signals of the main and auxiliary channels, and the level of the reproduced signal of the auxiliary channel is higher than the level of the reproduced signal of the main channel by a margin determined by the multiplier 167 (in this embodiment, the coefficient is 0). .5 so 6dB
) minutes or more, both left and right microphones 1.1
A switching signal RLFLAG is generated to switch the correspondence between the main and subchannel L. Note that the hangover processing circuit 165 performs hangover processing to prevent switching from occurring frequently, and once a switching occurs, the next switching is prohibited for a time T. Also, at the time of this switching, the adaptive filter is cleared by the OR circuit 166 to alleviate the temporary deterioration of the inter-channel correlation removal performance due to a change in the speaker.

The correspondence between the left and right microphones 1.1 and the main and subchannel channels is switched using the switch 181 described above.
．． 182 under the control of RLFLAG.

Main channel ADPCM decoding section 32 The main channel ADPCM decoding section 32 includes an adaptive inverse quantizer 133, an adder 135, a predictor 137, a delay device 139 .
184, a D/A converter 141, and a switch 183.

The received ADPCM code D (k) is converted into main channel reproduction prediction residual signal d, (k
), and the main channel pre-Pj signal parent (k) output from the predictor 137 is added by the adder 135, resulting in the main channel reproduced audio X l (k).

Thereafter, in order to compensate for the delay added to the subchannel by the transmitter 100 by the delay device 139, the same amount of delay is added to the subchannel reproduced audio, and then the D
/A converter 141 from Sbee power 2R or D/A
It is output from the speaker 2R via the converter 149.

The predictor 137 inputs the main channel reproduced audio x1(k) and performs adaptive filter processing to generate the main channel predicted signal R(k), and the learning of this filter is based on the main channel reproduced predicted residual signal d1( k) by minimizing the power.

This predictor 137 is the same as the predictor 111 of the main channel ADPCM encoder 30, and also receives the same signal, so that the transmitter 100 and the receiver 200 can output the same prediction signal x (k). .

The delay device 184 is 1. A delay device 18 inserted into the main channel to provide causality to the inter-channel prediction filter on the transmitter 100 side when the sub-channel becomes the main channel.
This is to cancel the newly generated left and right delay difference due to the delay processing of 0, and the same delay as that for the subchannel is inserted.

Sub-channel ADPCM decoding unit 33 The sub-channel ADPCM decoding unit 33 includes an adaptive inverse quantizer 143, an adder 145, a predictor 147, and a D/A converter 1.
49 and a switch 185.

Similarly to the main channel, a receiving side sub-channel prediction residual signal e(k) is generated from the received ADPCM code F(k).

After that, the correlation component 9(k) from the inter-channel correlation adding section 35 is added to this signal, and the sub-channel reproduced audio y
t(k), and then from the speaker 2R via the switch 185 to the D/A converter 141 or from the D/A converter 1.
The signal is output from the speaker 2R via 49.

Inter-channel correlation addition unit 35 The inter-channel correlation addition unit 35
It consists of an adder 151 and an adder 153, which inputs the main channel reproduced audio x 1 (k) and calculates the inter-channel correlation component 9 (
k) is generated by the inter-channel predictor 151.

This inter-channel predictor 151 is the same as the inter-channel predictor 129 of the transmitting section 100, and is an adaptive filter that learns to minimize the power of the receiving side sub-channel prediction residual signal e1 (k) like the transmitting section 100. , the same predicted value 9(k) as the transmitter 100 is obtained.

Level Detection Unit 400 The level detection unit 400 includes level detectors 171 and 174, comparators 172 and 173, a hangover processing circuit 175, an OR circuit 176, and a multiplier 177.

This level detection section 400 is similar to the level detection section 30 described above.
Performs the same function as 0. That is, this level detection section 4
00, the switching signal RLPLAG and the prediction coefficient clear signal HCI are set to the inter-channel reservation a'PI at the same timing as the transmitter 100 based on the levels of the reproduced signals of the main and sub channels detected by the level detectors 171 and 174. Switching the correspondence between the device 151, main and sub channels, and the left and right speakers 2.2 1? L is applied to switches 183 and 185. Further, the hangover processing is also performed by the hangover processing circuit 175 similarly to the transmitter 100.

Note that the clearing process of the adaptive filter based on the level of the main channel can be omitted by employing leak processing (processing of gently forgetting past coefficients) for the tap coefficients of the adaptive filter.

In this way, according to this embodiment, it is possible to allocate the optimal main and sub channels based on the output level of the microphone depending on the position of the speaker. Since the long quantizer is fed a high power signal, the quantization noise is reduced and the audio quality of the conventional stereo audio coding scheme is significantly improved.

[Effects of the Invention] As explained above, according to the present invention, quantization noise is reduced and high quality stereo sound can be provided.

[Brief explanation of the drawing]

FIG. 1 is a diagram for explaining the present invention in detail, and FIG. 2 is a configuration diagram of a stereo audio transmission system according to an embodiment of the present invention. 30... Main channel ADPCM encoding unit 31... Sub channel ADPCM encoding unit 32... Main channel ADPCM decoding unit 33.・・・・・・・・・Subchannel ADPCM
Decoding unit 34... Inter-channel correlation removing unit 35... Inter-channel correlation adding unit 30
0.400・・・・・・Level detection unit applicant
Toshiba Corporation

Claims (2)

[Claims] (1) In a stereo audio transmission system that encodes multiple input signals, transmits them via two types of channels with different bit rates, decodes them, and outputs them from multiple speakers, a) Encodes the input signals. a first level detection means for detecting each level of the plurality of input signals; and a first level detection means for encoding at least one input signal among the plurality of input signals according to a detection result by the first level detection means. , signal encoding/channel assignment means for assigning to one of the two types of channels, encoding the remaining input signal of the plurality of input signals, and assigning it to the other of the two types of channels; b) The side that decodes the encoded input signal includes a signal decoding means for respectively decoding the encoded input signals transmitted from the two types of channels, and decoding by the signal decoding means. a second level detecting means for detecting the level of each of the input signals decoded by the signal decoding means; assigning a signal to at least one of the speakers;
A stereo audio transmission system comprising: signal allocation means for allocating the remaining input signals among the input signals decoded by the signal decoding means to the remaining speakers. (2) In a stereo audio transmission system that encodes multiple input signals, transmits them via two types of channels with different bit rates, decodes them, and outputs them from multiple speakers, a) Encode the input signals. a first estimating means for estimating an input signal allocated to one channel of the plurality of input signals from an input signal allocated to the other channel; and based on the estimation result by the first estimating means, Signal generating means for generating a residual signal by removing a correlation component with the input signal assigned to the one channel from the input signal assigned to the other channel; and a residual signal generated by the signal generating means. a signal transmission means for encoding and transmitting the encoded input signal to a decoding side; detecting the level of the input signal assigned to the one channel; and detecting the level of the input signal assigned to the one channel; a first clearing means for clearing the estimation result by the means; b) a side that decodes the encoded input signal decodes the residual signal transmitted from the signal transmission means, and the decoded residual second estimating means for synthesizing a correlation component obtained by estimating the input signal assigned to the other channel from the signal and the input signal assigned to the one channel; A stereo audio transmission system comprising: second clearing means for detecting the level of an input signal and clearing the estimation result by the second estimating means according to the detection result.