JPS61106000A - Automatic anti-howling device - Google Patents

Abstract

PURPOSE:To suppress oscillation, i.e. singing of a loudspeaker system by automatically detecting howling caused in the loudspeaker device and controlling the loudspeaker gain. CONSTITUTION:Among a pair of signals (a) and (b) branched from the signal transmission line by the branching means 1, the signal (b) is delayed 5 to make a certain time difference between both signals. Furthermore, the signal (a) is branched 2 and between a pair of signals branched by phase shifting means 3, 4 and a phase difference of 90 deg. or about 90 deg. is given and then each is multiplied by the other signal branched 1 by the multipliers 6 and 7. The output signals of the multipliers 10 and 11 are integrated at a specified attenuation factor by the integrators 8 and 9. And with the arithmetic units 10 and 11, the absolute value or squared value of the output signal of the integrator is outputted. Then, with the adder 12, the outputs of the arithmetic units are added. By controlling the gain of the gain control circuit Q inserted in the signal transmission line by means of the output signal of the adder 12, the singing generated in the electroacoustic system including the signal transmission line L is suppressed.

Description

[Detailed Description of the Invention] When amplifying sound using electroacoustic equipment in a hall or conference room, the transmission system consisting of microphones, amplifiers, speakers, etc. creates a feedback loop due to acoustic coupling between the speaker and microphone. Howling occurs when the gain of the transmission system exceeds a certain amount. In order to avoid such howling, the directional characteristics of microphones and speakers are set so that howling is less likely to occur, and electrical methods are also used to lower the gain at frequencies where howling occurs.

However, even though it is an electrical method, the detection of noise ringing relies on the human ear, and the gain of the frequency that seems to be affected must be manually lowered using a graphic equalizer or the like.
During actual sound amplification, the gain between the microphone and speaker is manually lowered according to the occurrence of howling, regardless of frequency, and these operations require a certain degree of specialized knowledge. Since this method requires a lot of experience and experience, there has been a desire for a more efficient method that does not rely on manual methods.

The present invention has been made in response to the above needs.

That is, an object of the present invention is to provide an automatic howling prevention device that suppresses oscillation, that is, ringing, of a public address system by automatically detecting howling occurring in a public address system and controlling the public address gain.

Embodiments of the present invention will be described in detail below with reference to the drawings.

FIG. 1 is a block diagram showing a first embodiment of an automatic howling prevention device according to the present invention. In FIG. 1, (1) is a first branching means for taking out a pair of signals a and b from a signal transmission path. Therefore, signals a and b are the same. (
2) is a second branching means for branching the signal a. (8)
and (4) is a phase shift means that provides a phase difference of 90 degrees or approximately 90 degrees between a pair of branched signals, where the signal transmission path is 1/3 to 1 in one channel with a graphic equalizer. If an octave band-limited signal is to be transmitted, a pair of bypass filters and low-pass ψ filters obtained from a primary OCR circuit are used as phase shifting means in (8) and (4), respectively. can do. However, in that case, it is necessary to match the frequency at which each filter is attenuated by 3 dB to the middle of the signal frequency band. (5) is a delay means that delays the signal in order to give a certain time difference between the signals a and b. If the delay means is an analog circuit, a charge transfer device (eg, BBD) can be used. (6) and (7) are multiplication means, which can be realized by a multiplier. (8) and (9) are integrating means, which can be realized by an integrator having a constant attenuation rate (time constant) in which the integrated signal attenuates in proportion to the passage of time. αQ and a force are
It is an arithmetic means for obtaining the square value of an input signal or an absolute value as a substitute thereof, and a multiplier, a squarer, or an absolute value circuit can be used. (121 is C0) and a
This is an addition means that adds together the outputs of degrees Celsius, and can be realized by an adder. Q is a gain control circuit whose gain is controlled depending on the magnitude of the output. In FIG. 1, after branching means (1) % (2)% (81, ..., (2)
The above constitutes a sine wave detection circuit that detects and outputs a sound.

Next, the principle of operation of the present invention will be explained according to FIG.

From a physical perspective, audio and music signals are signals that fluctuate irregularly over time, and a signal that occurs at a certain rate on the time axis and a signal that occurs at a sufficiently distant point are mutually exclusive. It becomes a correlation. Therefore, the result of multiplying and integrating the signals generated at different times is expressed as a fluctuation whose average value is O.

On the other hand, the oscillated sound can be considered as a sine wave whose frequency is constant and whose amplitude increases over time. , respectively A(t)sin(2πf t )
Formula IA(t-T) sin(2πf(t-
T)) If Equation 2 is used, the phase shifter (8'l, (4)
The waveforms of C and d, which have a phase difference of 90 degrees, are multiplied by the waveform of e using a multiplier (+5)% (te), and the obtained waveforms of g and h are integrated, respectively. The results of integration by the instruments (8) and (9) are respectively K(t)(Crs(2πfT)+5in(2yr, fT
)) Equation 3K(t)(as(2πfT)−si
n(2·πfT)) Equation 4 is obtained. However, in the above formula, t is time and T is delay time 1. f represents the frequency, K(t)=澁−A(t)A(t−T) d
t Equation 5 Therefore, arithmetic unit (10), C11)
If G(t) is a squarer, the output G(t) of adder (2) becomes G(tl=-K(t)2 26.G(t) increases with time, and if the sound is If it is a sine wave generated with a constant amplitude from time 1 = 0, then Equation 5
From Equation 6, G(t)=A4=hip/4 Equation 7. However, A=A(t)-A(t-T).

Note that for audio signals or music signals (not including sustained sine waves), G(t)=N(t)<
(A certain constant value) It is expressed as 28.

Therefore, if the gain of the gain control circuit is lowered when G(t) exceeds a certain value, if a sound occurs in the loudspeaker system including the signal transmission path shown in Figure 1, it will automatically be suppressed. The gain of the system decreases and the noise is suppressed.

In the above explanation, the time constant of the integrator was set to infinity in order to avoid the complexity of calculation, but from a practical standpoint, it is better to give a time constant of about several seconds. This is because when the microphone moves, the degree of acoustic coupling between the speaker and the microphone may change from moment to moment, and it is not necessary to keep the gain of the amplification system at a lower level at the first sound. Furthermore, if some kind of sine wave is added as a signal, regardless of the sound, the gain of the system that has been lowered by it can be restored to its original value with the passage of time according to the time constant.

When an absolute value circuit is used for the arithmetic unit (101, αη) in FIG. /2 becomes Equation 9. k depends on fT and % 1 <, k < ?2
It is a coefficient that takes values in the range of .

For audio signals, even in the case of vowels with many pure tone components,
Since the time period during which the fundamental frequency can be considered constant is several tens of milliseconds at most, T may be at least several tens of milliseconds. In the case of music (although it is a secondary target for howling prevention),
Depending on the instrument being played, a stationary sine wave may be generated, so it is necessary to set T larger than that for voice.

In particular, when an electronic musical instrument plays a certain pitch, it becomes physically indistinguishable from a ringing sound, and the automatic howling prevention device according to the present invention determines this to be a ringing sound and lowers the gain of the amplification system.

Figure 2 shows the actual measured value R of the output waveform obtained by multiplying and integrating two waveforms at different times using one octave irregular noise with a center frequency of IKHz, and the output waveform obtained by squaring and integrating one waveform. The actual measured value S is shown. R corresponds to 28, and it can be seen that the output waveform obtained by multiplying and integrating two uncorrelated signals has a fluctuation with an average value of O. S
is not the result of multiplying and integrating sine waves, so it corresponds to Equation 9, although it is not direct. If it is a sine wave, S will have a waveform that increases smoothly over time. (Refer to Proceedings of the Acoustical Society of Japan, March 1981, Ni 1-1-17)
.

FIG. 3 is a schematic diagram showing a second embodiment of the automatic howling prevention device according to the present invention. In FIG. 3, (1) shows the first branching means, and (2) shows the second branching means. (1
5) represents an oscillator that outputs a pair of sine waves C and d with a phase difference of 90 degrees, (4), (5), and (6') represent multipliers, and a pair of signals branched at (2). is a phase difference of 90
By multiplying them with a pair of sine waves having different angles, they are converted into modulated signals g and h having a phase difference of 90 degrees. On the other hand, the signal branched by the first branching means is
The signal is delayed by the delay means (3), multiplied by the output C of the oscillator (oscillator section) by the multiplier (4), and converted into a modulated signal e, and further extracted by the bandpass filter (γ) into a single sideband signal. Here, (4) to (7) and @ear, the processing of (3) and (4) in FIG. 1, which is the first embodiment, is to obtain a pair of signals with a phase difference of 90 degrees. There is, but Figure 3
This method is suitable for providing a 90 degree phase difference over the entire signal band. (8) and (9) are multiplication means;
(10) and (11J are integrating means, (2) and 03
) is an arithmetic means, and (14) is an addition means, which correspond to (6) to (1) in FIG. 1, respectively.

The output signals C and d of the oscillator (b) are respectively fB
(2πFt), (B sin (2πFt), and if the signal a (and b) can be expressed by Equation 1, then Fig. 3
The output of the adder (14) in is represented by 26.

Therefore, the principle of suppressing noise is the same as that explained in FIG. 1, so the explanation thereof will be omitted.

In addition, in FIG. 3, the sine wave detection circuit is connected to the first branching means (
1) Thereafter, it consists of (2), (3)...(15),
A gain control circuit inserted into the signal transmission path is indicated by Q. In the figure, a bandpass filter (7) may be used for signals g and h. Further, the delay circuit (3) may be used for the signal a, which is also the case in FIG.

As described above, according to the present invention, it is possible to automatically detect and suppress noise generated in a public address system.

Note that the present invention is not limited to the above-described embodiment, and for example, the entire or part of the configuration shown in FIG. 1 may be realized by a digital circuit. In that case, the phase shifter can be obtained using a pair of low-pass filters and a filter in bypass formed by series connection of -order cyclic and acyclic filters, similar to analog circuits. Further, the integrator can be realized by a first-order recursive filter.

[Brief explanation of drawings]

FIG. 1 is a diagram showing a first embodiment of the present invention. FIG. 2 is a measured example for explaining the present invention in detail, and FIG. 3 is a block diagram showing a second embodiment of the present invention. Figure 1
And in FIG. 2, (1) is the branching means (branching part), L
is a signal transmission path, M is a sine wave detection circuit, and Q is a gain control circuit.

Claims (1)

[Claims] a delay means for providing a certain time difference between the pair of signals branched from the signal transmission path by the first branching means; a second branching means for further branching one of the pair of signals;
phase shifting means for providing a phase difference of 90 degrees or about 90 degrees between the pair of signals branched by the branching means; and the first branching means for each of the pair of signals to which the phase difference is given by the phase shifting means. a pair of integrating means that integrates the output signal of the multiplying means at a constant attenuation rate; and a pair of integrating means that outputs the absolute value or square value of the output signal of the integrating means. calculation means,
Controlling the gain of a gain control circuit inserted in the signal transmission path by the output signal of the sine wave detection circuit, using a sine wave detection circuit comprising at least addition means for adding together the respective outputs of the calculation means. An automatic howling prevention device characterized by suppressing noise generated in an electroacoustic system including the signal transmission path.