JPS63123257A - Telephone conference system - Google Patents

Abstract

PURPOSE:To always maintain a satisfactory calling level without a distorsion due to an overrange by performing a level adjustment with an attenuator and an amplifier so that the signal level of an N number of adding outputs can be always housed in a prescribed range. CONSTITUTION:An N-1 number of sound signals to exclude the sound signal of respective subscribers is added from sound signals A, B, C and D of an N number of subscribers, an N number of adding outputs 10 is obtained and a sum 11 of the sound signal of the N number of the subscribers is obtained. Based on the sum 11, the signal level of the N number of the adding outputs is adjusted by a level adjusting means 13 and the N number of the level-adjusted adding outputs is multiplexed 17. Thus, since the level adjustment is executed so that the signal level of the N number of adding outputs can be always housed in a prescribed range, the coping is executed to the change of the number of speakers and the change of the intensity of a voice from moment to moment and a satisfactory calling level can be always maintained without a distorsion due to the overrange.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a teleconferencing device realized by addition such as an N-1 adder type.

[Prior Art] A conventional conference telephone system using addition, for example, an N-1 adder type, is configured as shown in the block diagram of FIG. 8, for example.

FIG. 8 shows a case of a call between four parties (A, B, C, and D).

Input codes A, B, C, and D from the four parties are time-sequentially converted into linear codes by an expander 1, and added to the output from the shift register 3 in an adder 2. The linear code from decompressor 1 is also supplied to shift register 4. Adder 2
The addition output from the shift register 3 is supplied to the shift register 3, and is also supplied to the subtracter 5, where the output from the shift register 4 is subtracted. The subtracted output is supplied to a compressor 7 via an attenuator 6, and is converted back into a PCM code and sent out.

In the case of a four-party (A, B, C, D) call, the human code A is first converted into a linear code by the decompressor 1, and then transferred to the shift register 3.
and stacked at 4. The next input code B is similarly code-converted by the decompressor 1, then added to the data A stacked in the shift register 3 by the adder 2, and then transferred to the shift register 3 again. In parallel, the code-converted output is also transferred from the decompressor 1 to the shift register 4, where it is successively stacked after the data A already stacked there.

When such an operation is performed for input codes C and D, the shift register 3 has the sum of A, [1, C, and D (A+
B+C+D) are stacked, and shift register 4 has (
A, B, C, D) are stacked consecutively in this order.

Finally, the contents of the shift register 3 (A
+B+C+D) to the contents of shift register 4 (A, B,
By sequentially subtracting the data of f;, D), the corresponding 3
The calls made by the two people are synthesized. Since these data are simply added, they may overrange, so in order to keep them within a predetermined range, they are attenuated by an attenuator 6, and then converted back to a PCM code (nonlinear) by a compressor 7. Perform compression.

[Problems to be Solved by the Invention] Here, the attenuation amount of the attenuator 6 is a fixed amount predetermined according to the number of people participating in the conference. If the amount of attenuation is fixed in this manner, the audio signal passing through this telephone conference device will undergo a certain amount of attenuation in total. Furthermore, since the amount of attenuation of the attenuator 6 is determined in proportion to the number of participants in the conference, the amount of attenuation for the voice signal of each subscriber is independent of the number of subscribers who are actually speaking. constant. As a result, a multi-person conference call system using the N-1 addition method has a narrow range for each subscriber, resulting in poor call quality. In particular, in normal conference situations, there are many cases where one person's pronunciation is heard by a number of other people, and in that case, uniform attenuation based on a narrow range based on a fixed attenuation amount, assuming that all subscribers are talking at the same time. Low call levels are not enough.

SUMMARY OF THE INVENTION Therefore, an object of the present invention is to provide a teleconferencing device in which the audio signals of N-1 people are added from the audio signals of N subscribers to obtain N addition outputs. To provide a teleconference device which is appropriately constructed and arranged so that the level is always adjusted within a predetermined range and a sufficient speech level can be obtained without distortion due to overrange.

Another object of the present invention is to determine, based on the instantaneous peak values of the voice signals of N subscribers, whether there is a single speaker, a low signal level or a signal level that results in an overrange. , N summation output signals are kept within a predetermined range, so that the best speech level can always be maintained according to the number of speakers.

Still another object of the present invention is to adjust the signal levels of the N addition outputs to always fall within a predetermined range by hardware or software using a digital logic circuit or a one-chip microcomputer, and thereby To provide a small and inexpensive telephone conference device that does not cause a drop in speech level.

[Means for Solving the Problems] In order to achieve such an object, the first form of the present invention eliminates the voice signal of each subscriber from the voice signals of N subscribers. an addition means for adding N audio signals to obtain an N addition output; a means for calculating the sum of the N subscribers' audio signals; and adjusting the signal level of the N addition outputs based on the sum. and means for multiplexing the N level-adjusted addition outputs.

A second form of the present invention calculates the total sum of the voice signals of N subscribers, subtracts the voice number 43 of each subscriber from the total sum, multiplexes the N signals, and extracts the N-1 addition output. N-1
The present invention is characterized by comprising an adding means and a level adjusting means for adjusting the signal level of the N-1 addition output based on the summation.

[Function] According to the present invention, in a conference telephone device that obtains N summed outputs from the voice signals of N subscribers, the signal level of the N summed outputs is always within a predetermined range. In addition, level adjustment is performed using attenuators and amplifiers, which prevents overrange caused by simple addition in the N-1 addition method, for example, and prevents momentary changes in the number of speakers and voice It can deal with changes in loudness, etc., and can always maintain a good call level without distortion due to overrange.

Furthermore, the present invention uses a digital logic circuit or a one-chip microcomputer, etc. to adjust the signal level of the N addition outputs by hardware or software so that it always falls within a predetermined range. Furthermore, it is possible to provide a teleconference device that does not cause a drop in call level, and is effective as a one-chip LSI dedicated to a teleconference trunk that can be easily inserted into a local exchange or the like.

[Example] The present invention will be described in detail below with reference to the drawings.

FIG. 1 shows an embodiment of the present invention applied to the N-1 addition force formula, where lO is the shift register 3 shown in FIG.
and 4, an N-1 circuit consisting of an adder 2 and a subtracter 5. The N-1 adder circuit 10 has four subscribers A',
B', C', D' expanded audio signals a,
Input b, c, d and their sum (a+b+c+d)
is determined by the shift register (Fig. 7), and its absolute value la
+b+c+d l is determined by the absolute value circuit 11. A signal indicating this absolute value is supplied to the peak hold circuit 12, where the peak value P of the absolute value is held for a predetermined period, for example, a period corresponding to one syllable, and the peak value P is stored in a variable gain amplifier or the like. The signal is supplied to the level control input terminal of the level adjustment means 13. This level adjustment means 13 has an N
−1 addition output from addition circuit lO, i.e. subscriber A
, data a'=b+c+d for subscriber B', data b'=,1+(+d, data l=a+l)+d for subscriber C', and data d'=a+b+c for subscriber D' are supplied in time sequence. , by the peak value P, for example in Fig. 2 and the following (1
) The output level is adjusted using a level coefficient G such as a gain determined by the equation.

However, the maximum amplitude level per subscriber is 1.

Here, outputs a'', b'' addressed to each subscriber ^', B', Ni', and D' are taken out from the level adjustment means 13.

C″, d″ are the coefficients G as a′ as shown in the following equation (2).
, b', c', and d'. In the case of a PCM system, these outputs are supplied to a compressor 7, where nonlinear compression is performed.

The level adjustment means 13 can be composed of only a variable gain amplifier or a combination of an attenuator and a variable gain amplifier. For example, when the level adjustment means 13 consists only of a variable gain amplifier, when there is only one speaker, the gain G is set to 1 to obtain a normal speech level, and on the other hand,
In the case of multiple speakers, the gain G is reduced to 1/P so that the maximum amplitude level can be maintained without overrange.

Note that the peak hold circuit 12 refreshes the peak value P at predetermined intervals. In other words, if the current peak value is larger than the previous peak value, that larger value will be stored as the new peak value, but if the current peak value is smaller than the previous peak value, it will be stored for a certain period of time. By holding the previous peak value and then gradually decreasing the peak value, vibrations caused by low-frequency manual power are prevented.

By refreshing the peak value in this manner, the peak value is maintained for the N-1 added audio signal, as shown in FIG. 3, for example.

FIG. 4 shows another embodiment of the invention. Here, voice signals from each subscriber are received in the form of μmLAN PCM signals, and the signals are converted into parallel signals by a serial-parallel converter 21. Since the PCM signal uses compressed representation to reduce the number of required bits, applying N-1 addition to the input PCM signal as it is will exceed the allowable range for N-1 addition. Therefore, the parallel signal is
LAN-2 complement converter 22 and μmLAN
Convert the PCM signal to two's complement representation, then attenuator 2
3 and then supplied to the N-1 adder circuit IO.

The addition output from the N-1 adder circuit 1O is supplied to the two's complement-μm LAN converter 25 via the variable gain amplifier 24, where it is converted into a μm LAN PCM signal, and further transmitted to the PC.
The M signal is converted into a serial signal by a parallel-to-serial converter 26 to obtain a μmLAN PCM.

Here, the peak value obtained by supplying the output of the adder 2 to the peak hold circuit 27 is supplied to the gain control circuit 28, and the gain of the amplifier 24 is controlled by the output.
4, an output with the speech level adjusted according to the peak value is taken out.

In this embodiment, the human input data whose code has been linearly converted by the converter 22 is passed through the attenuator 23 to be attenuated in advance. N-1 additions can be performed without overflowing.

By constantly detecting the speech level of the N-1 addition output in the form of a peak value, the speech level is prevented from over-range, so that a sufficient speech level can be obtained without distortion due to over-range. I can do it.

FIG. 5 shows another embodiment of the invention, in which:
The functions of the block diagram shown in FIG. 1 are realized using a signal processor made of a one-chip microcomputer dedicated to digital signal processing, and FIG. 6 shows a flowchart of an example of a software program for its control procedure.

The signal processor shown in FIG. 5 employs a barbard architecture in which program memory and data memory are separated;
Ptl), 32 is a data storage RAM, 33 is a data storage ROM, 34 is a program storage ROM, 3
5 is a multiplier, 36 is a serial interface, 37
is a parallel interface, and 38 is a bus.

The program storage n0M34 stores software that follows a control procedure as shown in FIG. 6, for example. Here, it is assumed that the maximum number of conference participants to the conference telephone device of this embodiment is six, and that six input/output time slots are allocated on the PCM data highway. Further, in the case of a conference at 5 Å or less, an unused channel is specified by an external control input, and the data of the unused channel is treated as 0.

First, in the first input/output calchin S1 of the program, output of PCM data to each time slot stored in the data RAM 32 in the previous processing and manual input of the PCM data are performed. Input data a, b, c, d, e, f is converted into 13-bit linear data A, B, C, D, E, F by decompression using data RAM 32, and then data R
It is stored in AM32. Since the internal calculations of the signal processor are performed using 16 bits, overflow will not occur even if the inputs of 8 time slots are simply added together.

Next, in step S2, all the linearly converted input data A, B, C, D, E, and F from each time slot are added and stored in the data RAM 32 as Xn=A+B+C+D+E+F.

In the next step S3, N- is obtained by subtracting the human data A, B, C, D, E, F from each time slot from Xn.
1 addition data A', B', C', D',
E' and F' are stored in the data nAM32.

Next, in step S4, Xn is compared with the previous peak value P, and if Xn is larger, Xn is stored as a new peak value in the data RAM 32, and a hold time is set in a soft timer in the CPU 31. On the other hand, if the previous peak value P is larger than Xn, the previous peak value is kept as is and the soft timer is decremented.

After this, when the timer reaches 0, it is assumed that the hold time has ended and the peak value is decreased at a certain rate. Note that at this time, 1 is set in the timer.

Next, in step S5, the coefficient G shown in FIG. 2 is calculated using the peak value P. However, since it is difficult to perform division by 17P, the curve in FIG. demand. Here, G is approximately determined using a multiplier 35 using a fifth-order regression equation.

Next, in step S6, the multiplier 35 converts this G into Δ', B', [;', D', E
', F' are multiplied and data to be output in each time slot 8'', B'', C''.

D', E'', F- are obtained, and finally, in step S7, these data A'', B'', C', D'', E#,
After compressing F- and returning it to PCM code, data RAM
32 and waits for the next input/output timing. By repeating this series of operations, an N-1 additive type teleconference device in which the speech level is adjusted according to the present invention by the signal processor shown in FIG. 5 is realized.

In addition, although various embodiments in which the present invention is applied to the N-17JD calculation method have been described above, the present invention also applies to the N-1
This is not limited only to the case of the addition method, and the voice signals of N-1 people are added from the voice signals of N subscribers.
This can be widely applied to teleconferencing devices that obtain a summation output of .

An example of another form for obtaining N addition outputs is shown in FIG.

Figure 7 shows the case of four subscribers, where the audio signals A, B, C, and D of each subscriber are added and supplied to 41 to obtain the sum A+B+C+D of all the audio signals of the four subscribers. demand. Furthermore, the audio signals of three subscribers (A, B, C) obtained by removing the audio signals of each subscriber from the audio signals of the four subscribers.
; (A, B, D) : (A, ni, D) and (
11, C, D), adders 42, 43, and 44, respectively.
and 45, and add output A+B+C, to [1
+[1, we get ^+COO and B+B+D.

Sum A+B+ for 4 subscribers from adder 41
(: Absolute value l A obtained by supplying +D to the absolute value circuit 11
+n+C+D l is supplied to the peak hold circuit 12. The peak value P is supplied to the level adjustment means 13.

In this example, the level adjustment means 13 includes level adjustment means 13-1, 13-243-: ] and 13- for each addition output.
4, and the addition outputs of adders 42, 43, 44 and 45 are connected to these means 13-1.13-2.1.
3-3 and 13-4, and the signal levels of the four summed outputs are individually adjusted according to the peak value P.

The addition outputs whose levels have been adjusted by these level adjustment means 13-1, 13-2, 13-3 and 13-4 are supplied to the multiplexer 46 to be multiplexed, and multiple outputs are taken out.

This embodiment is effective not only when adding digital audio signals but also when adding analog audio signals.

Furthermore, the addition form of this embodiment is as shown in FIG. 4 or 5.
Of course, the invention can also be applied to the embodiment shown in the figure.

[Effects of the Invention] As is clear from the above, according to the present invention, in the conference telephone device M that obtains N summation outputs, the signal level of the N summation outputs is always within a predetermined range. To,
Since the level is adjusted using an attenuator or amplifier, it prevents overrange caused by simple addition in N-1 addition method, etc., and thereby prevents momentary changes in the number of speakers and the loudness of the sound. This allows you to maintain a good call level without distortion due to overrange.

Furthermore, the present invention uses a digital logic circuit or a one-chip microcomputer to adjust the signal level of the N addition outputs by hardware or software so that it always falls within a predetermined range. It is possible to provide a telephone conference device at low cost that does not cause a drop in speech level, and it is effective as a one-chip LSI dedicated to a telephone conference trunk that can be easily inserted into a local exchange or the like.

[Brief explanation of the drawing]

Fig. 1 is a block diagram showing an embodiment of the present invention, Fig. 2 is a characteristic diagram showing an example of the level coefficient G, and Fig. 3 shows the relationship between the N-1 added audio signal and the peak hold value. 4 is a block diagram showing another embodiment of the present invention, FIG. 5 is a block diagram showing still another embodiment of the present invention, and FIG. 6 is a block diagram showing a control program in the embodiment of FIG. FIG. 7 is a block diagram showing still another embodiment of the present invention, and FIG. 8 is an explanatory diagram of a general N-1 addition method. DESCRIPTION OF SYMBOLS 1...Extender, 2...Adder, 3.4...Shift register, 5...Subtractor, 6...Attenuator, 7...Compressor, 1O...N- 1 addition circuit, 11...absolute value circuit, 12...peak hold circuit, 13.13-1 to 13-4...level adjustment means (variable gain amplifier), 21...serial-parallel converter , 22...μm LAW-complement converter, 23...attenuator, 24...variable gain i1 width converter, 25...2's complement-μm LAW converter, 26...parallel-series converter IA , 27... Peak hold circuit, 28... Gain control circuit, 31... Central processing unit ((:PII), 32... M1 to data storage R 33... Data storage ROM, 34...・Program storage I (OM. 35... Multiplier, 36... Serial interface, 37... Parallel interface, 38... Bus, 41... Adder for summation, 42.43.44 .45...Adder for adding N-1 audio signals, 46...Multiplexer. Rehearsal thread pattern G level Louis thread count's death chart Figure 2. Flowchart Figure 6 Procedural Amendments December 16, 1986

Claims (1)

[Scope of Claims] 1) Adding means for adding N-1 audio signals excluding the audio signal of each subscriber from the audio signals of N subscribers to obtain N addition outputs; means for determining the sum of the voice signals of the human subscribers; level adjustment means for adjusting the signal level of the N addition outputs based on the sum; and multiplexing the level-adjusted N addition outputs. A telephone conference device characterized by comprising means. 2) The level adjustment means has a holding means for holding the signal level of the peak value of the sum at a sampling period with a time width corresponding to one syllable of the voice signals of the N subscribers, and the holding output thereof is 2. The telephone conference device according to claim 1, wherein the signal level of the N summation outputs is adjusted by . 3) A patent characterized in that the holding means is a peak hold circuit, which compares the signal level of the held output with a new total signal level every sampling period, and holds the peak value of the higher signal level. A telephone conference device according to claim 2. 4) The level adjustment means, when the maximum amplitude level per subscriber in the addition means is 1,
When the peak value is 0 or more and less than 1, the signal level of the N-1 addition output is output as is, and when the peak value is 1 or more and less than N, the signal level of the N addition output is divided by 1/(peak The telephone conference device according to claim 2 or 3, wherein the telephone conference device doubles the value (value) and outputs the multiplied value. 5) The level adjusting means includes an attenuator placed before the adding means and a variable gain amplifier placed after the adding means, and adjusts the gain of the variable gain amplifier to the signal level of the peak value of the sum. The telephone conference device according to any one of claims 2, 3, and 4, characterized in that the information is changed accordingly. 6) N-1 addition means for calculating the sum of the voice signals of N subscribers, subtracting the voice signals of each subscriber from the sum and multiplexing the N signals to obtain an N-1 addition output; A telephone conference device comprising: level adjustment means for adjusting the signal level of the N-1 addition output based on the summation. 7) The level adjustment means has a holding means for holding the signal level of the peak value of the sum at a sampling period with a time width corresponding to one syllable of the voice signals of the N subscribers, and the holding output thereof is 7. The telephone conference apparatus according to claim 6, wherein the signal level of the N-1 addition output is adjusted by. 8) The holding means is a peak hold circuit, which compares the signal level of the holding output with a new total signal level at each sampling period, and holds the peak value of the higher signal level. A teleconferencing device according to claim 7. 9) When the maximum amplitude level per subscriber in the N-1 adding means is 1, the level adjusting means adjusts the N when the peak value is 0 or more and less than 1.
A patent characterized in that the signal level of the -1 addition output is output as is, and when the peak value is 1 or more and less than N, the signal level of the N-1 addition output is multiplied by 1/(peak value) and output. A teleconferencing device according to claim 7 or 8. 10) The level adjustment means has an attenuator placed before the N-1 addition means and a variable gain amplifier placed after the N-1 addition means, and the gain of the variable gain amplifier is adjusted to the sum of the gains. Claims 7 and 8 are characterized in that the signal level is changed according to the signal level of the peak value.
The teleconferencing device according to any one of paragraphs 1 and 9.