JPH02143700A - Howling preventing device - Google Patents

Abstract

PURPOSE:To prevent howling by applying frequency analysis to an input voice signal, detecting the frequency component consecutive for a prescribed time and suppressing and outputting it thereby detecting howling automatically and surely. CONSTITUTION:An input voice signal is subjected to A/D converter 13, the result is fed to a FET transformation 15 with a buffer memory 14 for a prescribed period altogether and a PSD arithmetic unit 16 obtains the square sum of the real part and the imaginary part. The output and a PSD smoothed output in the past in a memory 28 are multiplied by a multiplier 17 for each frequency component. Then the component being the result of multiplication is decided by a discriminator 18 depending on the threshold level and a coefficient pattern is prepared depending on the quantity and the exchanged frequency component is multiplied with the pattern at a multiplier 19 to obtain a frequency pattern in which the frequency component over a prescribed time or over is suppressed is obtained. The pattern is subjected to an inverse FET transform 20, and a voice signal from which an oscillated frequency component is eliminated from the input voice signal is obtained via a buffer memory 21, a D/A converter 22 and an LPF 23.

Description

DETAILED DESCRIPTION OF THE INVENTION [Object of the Invention] (Industrial Application Field) This invention relates to a howling prevention device that prevents howling caused by electrical or acoustic feedback action.

(Prior Art) As is well known, when audio input is reproduced using a microphone, an amplifier, or a loudspeaker, howling, which is an oscillation sound emitted from the loudspeaker, becomes a problem.

The cause of this feedback is that the microphone is located close to the loudspeaker, or the amplifier has high sensitivity, and the sound reproduced by the loudspeaker is recorded by the microphone, creating an electrical and acoustic positive feedback effect. . Conventionally, there is no effective preventive measure against howling, and for example, the microphone 1, audio amplifier 2, and speaker 3 shown in FIG.
If howling occurs in the loudspeaker system, the howling can be suppressed by manually reducing the gain of the audio amplifier 2.

(Problems to be Solved by the Invention) As described above, there has been no effective method for preventing howling in the past.

This invention was made to solve the above-mentioned problems, and an object thereof is to provide a howling prevention device that can effectively prevent howling.

[Structure of the Invention] (Means for Solving the Problems) In order to achieve the above object, the howling prevention device according to the present invention includes an FFT transform means for performing frequency analysis on an audio input signal by fast Fourier transform processing, and this means. a detection means for detecting a constant subtraction from the analysis result, and a frequency component detected by this means is subjected to the FFT.
The apparatus includes a converting means and an audio reproducing means for reproducing the frequency components detected by the converting means. or
A variable gain amplifier that amplifies an audio input signal with an arbitrary gain, an FFT transformer that performs frequency analysis on the audio input signal by fast Fourier transform processing, and detects constantly continuous frequency components from the analysis results of this means. detection means; control means for controlling the gain of the variable gain amplifier so that the frequency component detected by the means is below a reference value; and sound reproduction means for acoustically reproducing the frequency component detected by the variable gain amplifier. It is composed of:

(Function) The howling prevention device with the above configuration performs frequency analysis on the audio input signal by FFT conversion processing,
A frequency component that continues steadily is detected, and the former subtracts the detected frequency component from the analysis result and performs inverse FFT transformation to generate an audio output signal, thereby reducing positive feedback from the audio output to the audio input. This prevents howling.

The latter uses a variable gain amplifier as an amplifier for the audio input signal, and prevents howling by automatically controlling the gain of the variable gain amplifier so that the detected frequency component is below a certain value.

(Example) The present invention will be described in detail below with reference to FIGS. 1 and 2.

FIG. 1 shows the configuration of a first embodiment according to the present invention, in which audio signals recorded by a microphone 11 and a microphone amplifier 12 are transmitted to a
/D is input to the (analog/digital) converter 13,
20, 48 [K! Iz] are converted into data by sampling, and then stored in the first buffer memory 14, and when 1024 pieces of data are complete, they are input to the FFT (Fast Fourier Transform) converter 15.

This FFT converter 15 performs frequency analysis by obtaining a frequency spectrum for a certain period (corresponding to the number of data of 1024) by fast Fourier transform, and the analysis result is converted into PSD (Power Spectrum Density).
ity) is input to the arithmetic unit 16. This PSD calculator 1B calculates the power ratio of the frequency spectrum by squaring and adding the real and imaginary parts of the frequency components detected by the FFT 15, and the frequency components detected by the PSD calculator 16. is input to the first multiplier 17.

This first multiplier 17 converts current information from the PSD calculator 16 and past information from a memory 28, which will be described later, into 20
[11Z Multiplication is performed for each frequency at 1 intervals, and the multiplication results are input to the determiner 18. This determiner 18 compares the input value with a threshold value, and if the input value is larger than the threshold value, the first
If the coefficient is smaller than 1, it is set to 1 to generate a coefficient pattern, and the coefficient pattern is manually input to the second multiplier 19. This second multiplier 19 multiplies the coefficient pattern from the determiner 18 by the frequency component from the FFT converter 15, and the multiplication result is the inverse FFT.
The FT converter 21 is manually operated.

The inverse FFT converter 20 performs inverse FFT processing on input data to convert it into the original signal pattern, and the converted signal is stored in the second buffer memory 21. Then, from this buffer memory 21, 1 at intervals of 20.48 [KHz]
After being extracted one by one and converted into an analog voltage signal by a D/A (digital/analog) converter 22, the LPF (
The stepped waveform is converted into a smooth waveform signal by a low-pass filter 23, and acoustically reproduced by an audio amplifier 24 and a speaker 25 as an output signal.

On the other hand, the frequency component detected by the PSD calculator 16 is also input to the third multiplier 2B. This third multiplier 2
B is for multiplying the input value by a second coefficient value (approximately 0.01), and the multiplication result is input to the adder 27.

This adder 27 multiplies the past spectrum data stored in the memory 28 by a third coefficient value (approximately 0.99), and the frequency component value detected by the fourth multiplier 29 is combined with the frequency component value detected by the third multiplier 26. The result of the addition is stored in the memory 28 as new spectrum data.

The operation of the above configuration will be explained below.

First, in a normal state where no howling occurs, it can be said that a specific frequency component hardly continues for a long time in the audio signal recorded by the microphone 11. Also,
When the speaker 25 and the microphone 11 are located close to each other, and the sound reproduced by the speaker 25 is inputted to the microphone 11 to form a positive feedback loop, and an abnormal state occurs such as howling, the specific frequency of the audio signal is continuously It continues to oscillate. That is, when an audio signal obtained by the microphone 11 is manually inputted to this howling prevention device via the microphone amplifier 12, the result is as follows.

First, the audio signal is sent to the A/D converter 13 at 20.48[KH
It is converted into digital data by sampling zl, and is processed by FFT for a certain period of time by the buffer memory 14.
After being input to the converter 15 and undergoing frequency analysis there,
The PSD calculator 16 calculates the sum of the squares of the real part and imaginary part of the FFT processing result.

The frequency component values detected by this PSD calculator 16 are as follows for each frequency (20 [Hzl ~ 10, 24 [KIIz
Therefore, the frequency component detected by this PSD calculator 16 and the past PSD smoothed output stored in the memory 28 are first multiplied for each frequency component. When multiplied by the multiplication unit 17, components with matching frequencies will have large values, and components without matching frequencies will have small values.

Therefore, the component obtained by this first multiplier 17 is compared with a threshold value by a determiner 18, and if it is larger than the threshold value, the first coefficient value is made smaller than 1, and if it is smaller, it is set to 1, and the coefficient pattern is created. do. The coefficient pattern obtained by this and the frequency component obtained by the FFT converter 15 are
If multiplied by multiplier I9, the coefficient will be smaller than 1 for frequency components over a certain period of time, so
A frequency pattern in which only that portion is suppressed is obtained. This frequency pattern is inverse FFT-transformed by an inverse FFT converter 20, and after being stored in a buffer memory 21, 20, 48
By extracting the signal at [KHz] intervals, converting it into an analog voltage signal by the D/A converter 22, and passing it through the LPF 23, an audio signal is obtained by removing the oscillation frequency component from the input audio signal. Even if this audio signal is amplified by the audio amplifier 24 and reproduced by the speaker 25, of course no
Uring does not occur.

On the other hand, the frequency component detected by the PSD calculator 16 is
is multiplied by the second coefficient by the multiplier 26 of the memory 2
The smoothed PSD output of 8 is multiplied by a third coefficient by a fourth multiplier 29. Third and fourth power generator 26
If the frequency components detected at .degree. 29 are added by an adder 27, new smoothed PSD information can be obtained. The PSD information obtained here is stored in the memory 28 as past information. Note that by increasing the second coefficient and decreasing the third coefficient, the information in the memory 28 can be updated more quickly, and by increasing the third coefficient, the past information in the memory 28 can be retained for a longer time. Can be done.

Therefore, the howling prevention device with the above configuration analyzes the frequency of the input audio signal, detects frequency components that continue for a certain period of time, and suppresses and outputs the detected frequency components, thereby automatically and reliably detecting howling. can be prevented.

FIG. 2 shows a second embodiment of the invention. However, in FIG. 2, the same parts as in FIG. 1 are given the same reference numerals, and the explanation thereof will be omitted.

In FIG. 2, an audio signal obtained through a microphone 11 and a microphone amplifier 12 is transmitted to a variable gain amplifier 30.
The signal is amplified by the speaker 25, and the sound is reproduced by the speaker 25. Such a howling prevention device for a public address system performs a series of processing on the input audio signal by the A/D converter 13, the puffer memory 14, the FFT converter 15, and the PSD calculator 16, and then processes the input audio signal by the first multiplier 17. is entered into
It is multiplied by the past PSD information stored in the memory 28 and input to the determiner 31. This determiner 31 determines the maximum value of the input value, and determines whether the value is within the vicinity of a constant value. If so, a negative (-) command signal is sent to the attenuation amount controller 32. This attenuation amount controller 32 controls to lower the gain of the variable gain amplifier 30 if the command signal from the determiner 31 is (+), and to increase the gain at that time if it is (-).

In other words, the howling prevention device configured as described above utilizes the fact that when howling occurs, positive feedback is applied and components of a constant frequency are amplified.The frequency analysis of the audio signal is performed using FFT, and the components are Since the constant occurrence of howling is detected and the gain of the variable gain controller 30 is controlled to prevent it from occurring, it is possible to reliably and easily prevent howling.

It goes without saying that the present invention is not limited to the above-mentioned audio amplification device, but can also exhibit excellent effects when used in a handset or receiver of a television conference system or the like.

[Effects of the Invention] As described above, according to the present invention, it is possible to provide a howling prevention device that can effectively prevent howling.

[Brief explanation of the drawing]

Fig. 1 is a block circuit diagram showing a first embodiment of a howling prevention device according to the present invention, Fig. 2 is a block circuit diagram showing a second embodiment of the howling prevention device according to the invention, and Fig. 3 shows howling is likely to occur. FIG. 2 is a block circuit diagram showing the configuration of a loudspeaker. 11...Microphone, 12...Mic amplifier,
13... A/D converter, 14. 21... Buffer memory, 15... FFT converter, 16... PSD calculator, 17. 19.26°29... Multiplier, 18. 31
...Determiner, 20...Inverse FFT converter, 22...
D/A converter, 23... LPF (low pass filter), 24... Audio amplifier, 25... Speaker, 27
...Adder, 28...Memory, 30...Variable gain amplifier, 32...Attenuation amount controller. Applicant's agent Patent attorney Takehiko Suzue

Claims (2)

[Claims] (1) FFT transform means for frequency-analyzing an audio input signal by fast Fourier transform processing; detection means for detecting constantly continuous frequency components from the analysis results of this means; and frequency components detected by this means. A howling prevention device comprising: a subtraction means for subtracting from the analysis result of the FFT transformation means; an inverse FFT transformation means for performing an inverse fast Fourier transform on the output of this means; and a sound reproduction means for acoustically reproducing the output of this means. (2) a variable gain amplifier that amplifies the audio input signal with an arbitrary gain; an FFT transformer that performs frequency analysis on the audio input signal by fast Fourier transform processing; and a frequency component that constantly continues from the analysis result of this means. a detection means for detecting the frequency component; a control means for controlling the gain of the variable gain amplifier so that the frequency component detected by the means is below a reference value; and an audio reproduction means for acoustically reproducing the output of the variable gain amplifier. Howling prevention device.