JPH0586720B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION Field of Industrial Application The present invention relates to a howling suppression device that uses a digital filter to suppress howling in a public address system.

Conventional Structure and Problems In recent years, with the development of conference systems and teleconference systems, howling suppression devices have been strongly desired in order to provide clear and natural sound amplification.

A conventional howling suppression device will be explained below. FIG. 1 is a block diagram of a howling suppression device using a conventional digital filter. 1st
In the figure, 1 is a microphone, 2 is a microphone amplifier, 3 is a subtraction unit, 4 is a transfer characteristic estimation unit,
A transfer characteristic estimator 5 stores the estimated transfer characteristic in the form of an impulse response, and 5 is a switch that controls the transfer characteristic estimator 4. When the switch 5 is closed, the transfer characteristic estimation operation is performed. and the estimation operation stops when it is open. 6 is a digital filter (hereinafter referred to as echo signal estimating section) that performs convolution of the transfer characteristic, that is, impulse response, stored in the transfer characteristic estimating section 4 and the input signal of the speaker amplifier;
7 is a speaker amplifier; 8 is a speaker; 9 is a switch for cutting off the amplification system and inputting white noise, which is a signal for estimating the transfer characteristic, to the speaker amplifier 7 when estimating the transfer characteristic; and 10 is a white noise generating section.

The operation of the conventional howling suppression device configured as described above will be described below.

First, the switch 5 is closed, the switch 9 is connected to the B side, and the transfer characteristic between the speaker 8 and the microphone 1, that is, the impulse response is estimated. The white noise, which is a signal for estimating the transfer characteristic, generated by the white noise generating section 10 is amplified by the speaker amplifier 7 and the speaker 8, and is then transmitted to the microphone 1.
The sound is recorded in The output of the white noise generator 10 is also input to the echo signal estimator 6. Echo signal estimator 6
Now, the convolution of this white noise signal and the estimated impulse response sent from the transfer characteristic estimating section 4 is calculated and used as an estimated echo signal, which is subtracted from the output of the microphone amplifier 2 in the subtracting section 3. The output of the subtraction section 3 is sent to the transfer characteristic estimating section 4, and the transfer characteristic estimating section 4 modifies the estimated impulse response stored internally in a direction such that the output of the subtraction section 3 decreases. Next, the estimated echo signal is calculated and subtracted again using this corrected estimated impulse response, a new output of the subtraction unit 3 is obtained, and the estimated impulse response is estimated again in the direction in which this output decreases. By repeating the above procedure a sufficient number of times, the output of the subtraction unit 3 becomes zero, and as a result, an impulse response with a high degree of approximation, that is, a transfer characteristic, is stored in the transfer characteristic estimation unit 4. After the estimation of the transfer characteristic is completed in this manner, the switch 5 is opened to prohibit modification of the estimated impulse response.

Next, switch 9 is connected to the A side for amplification.
During this amplification, the echo signal that is reflected from the speaker 8 out of the output of the microphone 1 is canceled by the subtraction unit 3 based on the same principle as the white noise signal when estimating the transfer characteristic, and the howlin loop is blocked. , the occurrence of howling is suppressed.

However, with the above configuration, in a room with poor soundproofing, a good S/N ratio cannot be secured unless the level of the white noise emitted from the speaker 8 when estimating the transfer characteristic, that is, the transfer characteristic estimation sound, is considerably high.
The drawback was that a sufficient howling suppression effect could not be obtained. Such a room with poor soundproofing effects has the problem that noise from outside the room mixes into the room and at the same time, sound from the room leaks to the outside.

Purpose of the invention The present invention solves the above-mentioned conventional problems.
It is an object of the present invention to provide a howling suppression device that can obtain a sufficient howling suppression effect with a transfer characteristic estimated sound having a small volume.

Structure of the Invention The present invention includes a frequency analysis section that performs frequency analysis of a microphone output signal based on indoor noise, a filter section whose frequency characteristics are controlled based on the analysis result of the frequency analysis section, and a speaker output signal through the filter section. A white noise generator inputs a transfer characteristic estimation signal, a transfer characteristic estimation section estimates the transfer characteristic of the echo path from the speaker to the microphone, and a transfer characteristic from the speaker to the microphone is calculated from the estimated transfer characteristic and the input signal of the speaker. This howling suppression device is composed of an echo signal estimating section that approximately synthesizes the reflected echo signals, and a subtraction section that subtracts the approximate synthesized echo signal from the output signal of the microphone, and includes a frequency analysis section, a filter section, and a white noise generation section. This function generates colored noise adjusted to the minimum necessary sound pressure level to obtain sufficient howling suppression effect at all frequencies, and by using this as the transfer characteristic estimated sound, the sound is transmitted at a low volume. This makes it possible to realize a howling suppression device whose characteristics can be estimated.

DESCRIPTION OF EMBODIMENTS An embodiment of the present invention will be described with reference to FIGS. 2 to 4.

FIG. 2 shows a block diagram of a howling suppression device in one embodiment of the present invention. In FIG. 2, 1 is a microphone, 2 is a microphone amplifier, 3 is a subtraction section, 4 is a transfer characteristic estimation section, 5 is a switch that controls this transfer characteristic estimation section, 6 is a reverberation signal estimation section, 7 is a speaker amplifier, Reference numeral 8 represents a speaker, and reference numeral 10 represents a white noise generator, which have the same configuration as the conventional example. 11 is a switch that controls the operation of the public address system; 12 is a frequency analysis section; 13
14 is a filter, and 14 is a switch for controlling the filter.

The operation of the howling suppression device of this embodiment configured as described above will be described below.

First, prior to amplification, the switch 11 is connected to the B side, and frequency analysis of the noise in the room is performed in the absence of amplified sound from the speaker 8. Room noise is collected by a microphone 1, amplified by a microphone amplifier 2, and then subjected to frequency analysis by a frequency analyzer 12. The result of this frequency analysis is sent to filter 13. At this time, the switch 14 is kept in a closed state, allowing the frequency characteristics of the filter 13 to be changed. The frequency characteristics of the filter 13 are controlled so that the gain is higher in a frequency band with a higher noise level. Next, the switch 14 is opened to prohibit changing the frequency characteristics, the switch 11 is connected to the C side, the transfer characteristics are estimated, and then the switch 11 is connected to the A side to perform amplification. The estimation of the transfer characteristic and the subsequent amplification operation are exactly the same as in the conventional example, except that the signal for estimating the transfer characteristic is colored by the filter 13.

FIGS. 3 and 4 are frequency characteristic diagrams showing the effects of this embodiment. Figure 3 shows microphone amplifier 2
An example of the frequency analysis results of the output of This shows the level when noise is used as a signal for estimating transfer characteristics. Figure 4 shows the frequency characteristics of howling suppression amount during amplification, and
8 is the frequency characteristic of the conventional example, and 19 is the frequency characteristic of this embodiment. From the comparison between Fig. 3 and Fig. 4, it is clear from the comparison of characteristics 15 and 16 that in the conventional example, the S/N is extremely poor in the low frequency range, and therefore, as shown in characteristic 18, the amount of howling suppression in the low frequency range is was extremely bad. For this reason, it can be seen that howling is likely to occur in this low frequency range, making it impossible to increase the amount of amplification. In order to improve this, it would be better to further increase the level of the estimated sound of the transfer characteristic, but as mentioned earlier, in rooms with poor soundproofing effects, the noise from outside the room mixes into the room, and at the same time the sound inside the room increases. It also leaks to the outside. Therefore, in this example, the level of the estimated sound of the transfer characteristic is set such that the level of the frequency band with high room noise level 15 is high and the level of the frequency band with low noise level is set low, as shown in characteristic 17. Even if the level, or sense of volume, is low, sufficient S/N is ensured in all frequency bands.

When using the transfer characteristic estimated sound of characteristic 17, as shown in characteristic 19, from the low range to the high range,
A nearly satisfactory howling suppression effect can be obtained.

As described above, according to this embodiment, a sufficient howling suppression effect can be obtained with the transfer characteristic estimation sound having a small volume.

Effects of the Invention By providing a room noise frequency analysis section and a filter section that controls the frequency characteristics based on the result, the present invention converts colored noise having frequency characteristics similar to room noise into transfer characteristic estimated sound. Therefore, it is possible to realize a howling suppression device that has a sufficient howling suppression effect over all frequency bands with a low-volume transfer characteristic estimation sound.

[Brief explanation of the drawing]

Fig. 1 is a block diagram of a conventional howling suppression device, Fig. 2 is a block diagram of a howling suppression device according to an embodiment of the present invention, Fig. 3 is a frequency characteristic diagram of the output signal of the microphone amplifier, and Fig. 4 is a block diagram of a howling suppression device according to an embodiment of the present invention. It is a frequency characteristic diagram of the amount of howling suppression. DESCRIPTION OF SYMBOLS 1... Microphone, 2... Microphone amplifier, 3... Subtraction section, 4... Transfer characteristic estimation section, 6...
...Reverberant signal estimator, 7... Speaker amplifier, 8...
...Speaker, 10...White noise generator, 12...
Frequency analysis section, 13...filter section.

Claims (1)

[Claims] 1. A frequency analysis section that performs frequency analysis of an output signal based on the noise of the microphone; a filter section that controls the gain so that the frequency band with the higher noise level has a higher gain based on the analysis result of the frequency analysis section; a white noise generation section that inputs a signal for estimating a transfer characteristic to a speaker; a transfer characteristic estimating section that estimates a transfer characteristic of an echo path from the speaker to the microphone; an input signal of the speaker; A reverberant signal estimating section that approximately synthesizes a reverberant signal circulating from the speaker to the microphone based on the estimated transfer characteristic; and a subtracter that subtracts the output of the reverberant signal estimator from the output signal of the microphone. Featuring a howling suppression device.