JPH0646754B2 - Level control conference call system - Google Patents

Description

The present invention relates to a system for performing level control for suppressing distortion and the like due to voice addition in a conference call.

In the conventional conference call in which the voice samples are simply added, when the number of participants is increased, the overload distortion due to the addition occurs, so that the scale of the conference cannot be increased. Also, there is a method in which the capacity of the conference is assumed in advance and the attenuation amount corresponding to the capacity is fixedly provided, but there is a problem that it is not versatile.

An object of the present invention is to eliminate the above-mentioned drawbacks of the prior art and to provide a highly versatile conference call system capable of performing level control according to the scale of the conference.

For the above purpose, the present invention provides control information indicating the amount of attenuation to be given to each conference participant before or after a circuit for adding audio samples of conference participants to create a voice sample for a conference call. And a level control means for attenuating and outputting the input voice sample according to the content of the control information, and a means for appropriately changing the value of the control information in the storage means according to the scale of the conference. Are to be taken.

An embodiment of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram of a conference trunk according to the present invention, and FIG.
FIG. 1 is a circuit diagram showing an embodiment of the level control circuit shown in FIG. In FIG. 1, reference numeral 1 is an input highway to the conference trunk according to the present invention, and 2 is an opposite output highway.
For example, 2 ^N multiplexed audio samples (quantized PCM signals) are transmitted and received serially in time division. Reference numeral 3 is a serial-parallel conversion circuit, 4 is a parallel-serial conversion circuit, 5 is a decompression circuit for converting a quantized PCM signal into a linear PCM signal, and 6 is a compression circuit for performing the reverse conversion. 7 and 8 at the level control circuit according to the present invention, and outputs the attenuated by an amount indicated by the fineness of about 1d _B pitch input audio samples in each time slot. Reference numeral 9 denotes a voice synthesizing circuit that adds voice samples of participants for each group of conferences to create a voice sample for a conference call. Reference numeral 10 is a control unit of the conference trunk, and via the control bus 11, level control circuits 7 and 8 and a voice synthesis circuit 9
To control.

The audio sample time-division multiplexed on the incoming highway 1
The serial-parallel conversion circuit 3 converts each time slot into a parallel signal, and the decompression circuit 5 converts it into a linear PCM signal. As a linear PCM signal, μ as a quantized PCM signal
Considering the signal of the law, 14 bits including the polarity bit are sufficient, but in order to have redundancy for the addition process, an extra 2 bits are added to form a 16-bit configuration. Now, assume a conference call of eight-party configuration using time slots 1-8.
In an eight-party conference, the voice level per maximum channel becomes eight times when simply added. In the present embodiment, since the number of bits for designating the audio level has a redundancy of 2 bits,
Up to 2 ² = 4 times, no overflow occurs even if the voices of each speaker are added. However, in the above, the remaining 8/4 = 2 times overflows. In order to deal with this, if the level of the signal is attenuated by 1/2, that is, 6 dB in the level conversion circuit 7, overflow does not occur even if the voices of the respective speakers are added. Audio signals of time slots 1-8 attenuated by 6d _B is added as conference call voice synthesizer circuit 9, and each time slot outputs in the form obtained by subtracting the own voice from the added voice. The sound level was processed with 16 bits at the time of addition, but in order to return the addition result to the original quantized PCM signal,
Since this embodiment has 14 bits, 2 bits are removed. That is, in the level control circuit 8, 1/2 ² = 1/4)
12 dB) Attenuate. The output of the level control circuit 8 is converted into a quantized PCM signal in the compression circuit 6, converted into the original serial signal in the parallel-serial conversion circuit 4, and output to the output highway 2. The operation of the level control circuit will be described below with reference to FIG.

In FIG. 2, reference numerals 20 and 21 respectively denote an input highway and an output highway of the level control circuit, and the linear PCM signals are transmitted and received in parallel. 22 to linear PCM signal on input highway 20 (hereinafter abbreviated as P), P, 1 / 2P , ..., 1/2 7 P, i.e. 6d _B pitch 0d _B ～42D output is attenuated in a range of _B A straight line PCM signal input P is directly applied to the input D ₀ and a straight line PC is input to the input unit D _i (D ₁ --- D ₇ ).
The M signal input P is supplied in the form of i (i = 1 to 7) bits shifted in the direction of decreasing the level (lower order). Reference numeral 23 is a memory that stores control information b _{5 to} b ₀ that indicates the amount of attenuation to be applied to the time slot using the time slot as an address. The upper 3 bits b _{5 to} b ₃ are supplied to the selector 22, Are the respective straight line PCs from the 8 speakers supplied to the selector 22 with their 3 bits b ₅ to b _3.
M signal (0 --- 7-bit shifted linear PCM signal)
Is selectively selected and used as the output. 24, 25, and 26 are gate circuits, 27 is an addition circuit, and 28 is a subtraction circuit. The output Y of the selector 22 is branched into four and one is a subtraction circuit as it is.
One of the inputs of the gate circuits 24, 25, and 26 becomes one input of 28, and the remaining three branches are shifted to the lower side by 2 bits, 3 bits, and 4 bits, respectively. Gate circuit 24,25,26
One of the lower 3 bits b _{2 to} b ₀ of the memory 23 is connected to the other input of the memory 23, and if the bit is “1”, the output selected by the selector 22 is passed. If it is "0", the output is forcibly cut off (all "0"). The adder circuit 27 adds the outputs of the gate circuits 24, 25 and 26, and the subtraction circuit 28 subtracts the addition result from the output of the selector 22. In this way, the gates 24, 25, and 26 are controlled according to the output of the selector 22, and the gate output is passed through the adder circuit 27 and the subtractor circuit in the subsequent stage, and the output of the subtractor circuit is adjusted by adjusting the amount of addition and subtraction. You can change the level. That is, the output of the selector 22 (tentatively P ′
Abbreviated), the lower 3 bits b _{2 to} b _{0 of} the memory 23 are “00”.
0 ', "001", ..., "111" depending on P', 15 / 16P ',
..., output of the form 9 / 16P 'is obtained. Therefore, the level of the output can be controlled. 29 is a counter that generates a sequential address in a 125 μS cycle of voice sampling in synchronization with the input highway 20 time slot,
Reference numeral 30 is a selector for selecting the address of the memory 23. 31
Is a write control circuit for controlling the writing operation of control information to the memory 23, and 32 is its R / W signal line. 33 and 34 are address lines and data lines of the control bus 11, respectively.

The control information is set in the level control circuit by writing in the memory 23. When the control unit 10 determines that level control is necessary, such as when the conference is large, it activates the write control circuit 31 via the control bus 11 and at the same time gives a time slot number for level control to the address line 33 and the time slot. Control information indicating the amount of attenuation to be transmitted is transmitted to the data line 34. The write control circuit 31 switches the output of the selector 30 to the B input address line 33, generates a write signal on the R / W signal line 32, and writes the control information on the data line 34 in the memory 23.
When the writing is completed, the write control circuit 31 causes the selector 30
Output is switched to the A input side. As a result, memory 23
The address of is sequentially changed by the time slot number synchronized with the input highway 20 by the counter 29.
The voice signal time-division multiplexed on the incoming highway 20 is a memory
An attenuation amount corresponding to the content of the control information read from the time slot position of 23 is given and output to the output highway 21.

How given of attenuation in two stages, initially containing the significant attenuation of at 6d _B pitch selector 22 by the contents of high-order three bits b ₅ ~b ₃ of the control information in the memory 23, then the lower of the control information Gate circuit 24, 25, 26, adder circuit depending on the contents of 3 bits b _{2 to} b ₀
27 and the configuration of the subtraction circuit 28 adds the amount of attenuation with fine about 1d _B pitch than 6d _B. Accordingly to the audio signal on the input highway 20, it outputs to the highway 21 exit giving fine attenuation at about 1d _B pitch.

As described above with reference to the embodiments, according to the present invention, since the level of each participant participating in the conference can be finely controlled, it is possible to perform the level control according to the scale of the conference (the number of conference participants, etc.). effective. Further, since the level control can be performed at any time depending on the number of participants who have a conversation at the same time, it is possible to provide a conference call service with extremely high flexibility and versatility.

[Brief description of drawings]

FIG. 1 is a block diagram of a conference trunk according to the present invention, and FIG. 2 is a detailed circuit diagram of the level control circuit of FIG. 1 ... Input highway 2 ... Output highway 3 ... Serial-parallel conversion circuit 4 ... Parallel-serial conversion circuit 5 ... Expansion circuit, 6 ... Compression circuit 7, 8 ... Level control circuit, 9 ... Voice synthesis circuit 10 ... Control section, 11 Control bus 20 Input highway 21, Output highway 22,30 Selector, 23 Memory 24, 25, 26 Gate circuit 27 Addition circuit 28 Subtraction circuit 29 Counter 31 Write control circuit 32 R / W signal line, 33 ... Address line 34 ... Data line

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Hirohiko Sato 1-2356 Takeshi, Yokosuka City, Kanagawa, Japan Yokosuka Telecommunications Research Institute (56) Reference JP-A-56-52960 (JP, A) 57-185757 (JP, A)

Claims (1)

[Claims] 1. A conference call system for adding time-division-multiplexed voice samples to create a voice sample for a conference call, wherein the voice samples are converted into a linear PCM signal or a linear PCM signal is converted into the voice sample signal. A conversion means for converting into a voice call sample for the conference call, a means for converting the linear PCM signal into a plurality of parallel signals each having a predetermined bit shift amount, and vice versa, and the conference call. A level control means for attenuating and outputting the conference call input voice sample is provided before or after the means for creating a conference voice sample, and the level control means determines the amount of attenuation to be given to each conference participant. Storage means for storing control information to be instructed, and the parallel signals are received, and the parallel signals are selectively selected to obtain a shift amount of 0 and each. A selector that outputs a plurality of signals that have been bit-shifted by a predetermined amount, a plurality of gates that receive the plurality of signals that have been bit-shifted by the predetermined amount among the signals selected by the selector, and based on control information of the storage means. Means for controlling the selector and the plurality of gates; addition means for adding outputs of the plurality of gates; subtraction means for receiving and subtracting the output of the addition means and the shift amount 0 of the selector; A level-controlled conference call system comprising: means for setting control information in a storage means, and the control information can be changed according to the size of the conference.