JPS63260209A - Automatic gain controller - Google Patents

Abstract

PURPOSE:To design the titled controller in such a manner that it hardly follows quickly the noise at a non-sound period but follows a sound signal in the sound period by providing a means detecting the sound period in an input signal so as to control the trailing time constant shorter in the sound period and longer in the no sound period. CONSTITUTION:The input signal 21 inputted to an input terminal 11 is rectified and smoothed by an amplitude calculation section 12 and becomes an amplitude signal 22. The signal 22 is inputted to a sound period detection section 13 and a maximum value hold section 14. The detection section 13 analyzes the signal 22 to discriminate whether or not the period is the sound period and outputs the result of discrimination to a hold section 14 as a sound detection signal 23. The trailing time constant of the hold section 14 is controlled shorter in the sound period and longer in the no sound period by the signal 23. The hold section 14 obtains a maximum signal 24 being a maximum value of the signal 22 at the condition of the time constant as above. A variable gain amplifier section 15 is controlled to have a gain 25 by a signal 24 and an output signal 26 is outputted from an output terminal 16.

Description

[Detailed Description of the Invention] Industrial Field of Application The present invention relates to an automatic gain control device built into equipment for recording, amplifying, and transmitting audio signals, such as recording devices, loudspeaker devices, and telephones. .

Conventional technology In recent years, cassette tape recorders and video cameras.

The types of audio devices used by general users, such as telephones, are increasing, and their functions are also becoming more sophisticated. In such devices, various adjustment functions are automated as much as possible in order to reduce the burden on the user during use.

One of these automated adjustment functions is an automatic gain control function. Let's discuss the role and necessity of this automatic gain control function for recording.

-The most important thing when recording audio is adjusting the recording level. The sound level of the audio signal picked up by the microphone varies greatly depending on the speaker. This is because the distance between the speaker and the microphone differs more than the voice power of the speaker differs. Even if the sound pickup level differs greatly, in order to avoid distortion due to excessive input or deterioration of S/H due to insufficient input, constantly monitor the sound pickup level and adjust the recording level with a variable amplifier. There is a need.

This level adjustment problem is a common problem in equipment that handles audio signals. For this purpose, automatic gain controls, or AGOs, have been developed and incorporated into equipment.

Hereinafter, the above-mentioned conventional automatic gain control device will be explained with reference to the drawings.

FIG. 3 shows a block diagram of a conventional automatic gain control device.

In FIG. 3, 31 is a signal input terminal, 32 is an amplitude calculation unit that rectifies and smoothes this signal to obtain an amplitude value, 33 is a maximum value hold unit that determines the maximum value of this amplitude value, and 34 is a maximum value hold unit. A variable gain amplification section whose gain is controlled by the output of section 33, and 35 are signal output terminals.

The operation of the automatic gain control device configured as described above will be described below.

FIG. 4 shows signal waveforms at each part of FIG. 3, and the operation of FIG. 3 will be explained using this diagram.

In FIG. 4, 41 is the waveform of the input signal of the mermaid in FIG. 45 shows the waveform of the output signal at point D, showing how the gain changes.

First, the input signal 41 shown in FIG. 4 is input to the input terminal 31 shown in FIG. In this input signal 41, time TO~T1
The three sections T2-T3 and T4-T5 are silent sections in which only noise is collected, and no audio signals are included in these sections. Two sections from time T1 to T2 and from T3 to T4 are audio sections, and are sections in which the voices of speaker M1 and speaker B are collected, respectively. In this example of the input signal 41, the speaker's position is closer to the microphone than speaker B, so the amplitude of the speaker's signal is larger than that of speaker B.

This input signal 41 is rectified and smoothed by the amplitude calculation section 32 shown in FIG. 3, and becomes an amplitude value 42 shown in FIG. Next, this amplitude value 42
is input to the maximum value holding section 33 to obtain a maximum value signal 43. This maximum value signal 43 traces the maximum value of the noise signal amplitude at time ToNT1. When the voice of speaker M is input at time T1, the value rises rapidly. After that, time T
2i, the maximum value of the audio signal amplitude of the speaker is traced. After the time T2 when the voice of the speaker disappears, the value gradually decreases according to the falling time constant set by the maximum value holding section 33, and the maximum value of the noise signal amplitude is traced again. become. When the voice of speaker B is input at time T3, the maximum value signal 43 rises rapidly again, and traces the maximum value of the amplitude of the voice signal of speaker B until time T4.

The gain 44 of the variable gain amplification section 34 is based on this maximum value signal 43.
controlled by The larger the maximum value signal 43, the higher the gain 4.
4, and the smaller the gain 44 is, the higher the gain 44 is controlled. As a result, the output signal 46 output from the output terminal 36 to the outside includes the speaker's audio signal at times T1 to T2 as well as the voice signals from time T3 to T4.
The audio signal of speaker B is also adjusted to have approximately the same amplitude level.

Problems to be Solved by the Invention However, in the above configuration, as shown at times T2 to T3 of the output signal 45, when the speaker remains silent for a long time, the gain gradually increases, and the noise is reduced. This has the problem of amplification.

To prevent this, the falling time constant of the local maximum value hold section 33 is made longer, and the local maximum value signal 4 in the silent section is
It is necessary to slow down the rate of decrease of 3, that is, the rate of gain increase. However, if it is too long, the gain control speed for the audio signal will drop too much. Once a signal with a very large amplitude is input and the gain is once reduced, the gain does not return to the original level easily, resulting in a disadvantage that the output signal level remains insufficient for some time.

In the conventional configuration, it has been impossible to achieve both gain stability in a non-speech section and gain followability with respect to a voice signal in a speech section.

In view of the above-mentioned problems, the present invention controls the time constant of the fall of the local maximum value holding section to a different value for the voice section and the no-voice section, so that the voice signal of the voice section can be quickly followed, and The present invention provides an automatic gain control device that is difficult to follow noise caused by noise.

Means for Solving the Problems To achieve this object, the automatic gain control device of the present invention includes an amplitude calculation section that calculates the amplitude value of an input signal, and a speech section detection section that detects a speech section in the input signal. and a maximum value hold section for determining the maximum value of the amplitude value, which controls the falling time constant to be short in voice sections and long in non-voice sections; and a variable gain amplification section that amplifies or attenuates the input signal and outputs it to the outside.

Effect: With this configuration, it is possible to realize an automatic gain control device in which the gain accurately and quickly follows the audio signal in the audio section, but does not easily follow the noise signal in the non-speech section. In this way, gain control can be performed using only the audio signal in the input signal, so the gain does not increase excessively even if there is a relatively long silent period, and the noise is over-amplified and output to the outside. The phenomenon can be prevented.

EXAMPLE An example of the present invention will be described below with reference to the drawings.

FIG. 1 shows a block diagram of an automatic gain control device in one embodiment of the present invention.

In FIG. 1, 11 is a signal input terminal, 12 is an amplitude calculation unit that calculates the amplitude value of the input signal, and 13 is an audio unit that monitors temporal changes in the amplitude value to detect a voice section in the input signal. A section detecting section 14 controls the falling time constant to be short in voice sections and long in non-voice sections, and a maximum value holding section 15 for determining the maximum value of the amplitude value; A variable gain amplifying section 16 is a signal output terminal that amplifies or attenuates the input signal and outputs it to the outside, where the gain is controlled to be higher as the gain is lower.

The operation of the automatic gain control device configured as described above will be described below.

FIG. 2 shows signal waveforms at each part of FIG. 1, and the operation of FIG. 1 will be explained using this diagram.

In Figure 2, 21 is the waveform of the input signal at point M in Figure 1, 22 is the waveform of the amplitude value signal at the first point B, 23 is the waveform of the voice detection signal at point 0, and 24 is the maximum at point D waveform of the value signal,
25 shows how the gain of the variable gain amplification section 16 changes; 26
shows the waveform of the output signal at two points.

First, the input signal 21 shown in FIG. 2 is input to the input terminal 11 shown in FIG. In this input signal 21, time TO~T1
, T2-T3, and T4-T5 are silent sections in which only noise is collected, and no audio signals are included in these sections. Two sections from time T1 to T2 and from T3 to T4 are audio sections, and are sections in which the voices of speaker M and speaker B are collected, respectively. In this example of the input signal 21, since the position of the speaker M is closer to the microphone than the position of the speaker B, the amplitude of the speaker M is larger than the amplitude of the speaker B.

This input signal 21 is rectified and smoothed by the amplitude calculation section 12 shown in FIG. 1 to become an amplitude value signal 22 shown in FIG. This amplitude value signal 2
2 is input to the voice section detecting section 13 and the maximum value holding section 14. The voice section detection section 13 analyzes the amplitude value signal 22, determines whether it is a voice section or not, and outputs the determination result to the local maximum value holding section 14 as a voice detection signal 23. The falling time constant of the local maximum value holding section 14 is controlled by this audio detection signal 23. That is, the time constant is controlled to be short in a section where it is determined that there is voice, and to be controlled to be long in a no-voice section. Under such time constant conditions, the local maximum value holding unit 14 obtains the local maximum value signal 24 which is the local maximum value of the amplitude value signal 22.

This time constant is determined from the following viewpoints. That is, time T
In the speaking section of speaker M from 1 to T2 and the speech section of speaker B speaking from T3 to T4, the maximum value signal 24 quickly follows the maximum value of each speaker's voice, and this maximum value , but at times T ONT 1 , T 2 NT
3. T4 to T5 (7) In the silent section, the maximum value signal 2
4 is a value that decreases extremely slowly and does not decrease to the noise amplitude value during the length of a silent section that occurs in normal conversation.

The variable gain amplification section 15 uses this local maximum value signal 24 to
The gain is controlled to be as shown in 5. As a result, the output signal 26 outputted to the outside from the output terminal 16 is adjusted to approximately the same amplitude level for the audio signal of speaker M at time T1-T2 and the audio signal of speaker B at time T3-T4, and moreover, Noise is not over-amplified.

As described above, according to the present embodiment, the voice section detecting section 13 is introduced, and based on the determination result of the voice section detecting section 13, the time constant of the fall of the local maximum value holding section is set to be short in the voice section and short in the non-voice section. Therefore, it is possible to realize an automatic gain control device in which the gain accurately and quickly follows the voice signal in the voice section, but does not easily follow the noise signal in the non-voice section. In this way, gain control operation can be performed using only the audio signal in the input signal, so even if a relatively long silent period continues, the gain will not increase excessively.
It is possible to prevent the phenomenon of excessively amplifying noise and outputting it to the outside.

In addition, in this embodiment, the time constant of the fall in the no-speech section is set to a finite value, but an infinitely large value may be selected, that is, the falling operation in the no-speech section may be prohibited. Of course.

Effects of the Invention The present invention includes an amplitude calculation unit that calculates an amplitude value of an input signal;
A voice section detecting section detects a voice section in the input signal, and a determination result of the voice section detecting section determines the amplitude value, which controls the falling time constant to be short in a voice section and long in a non-voice section. A maximum value hold section that determines the maximum value, and a variable gain amplification section that amplifies or attenuates the input signal and outputs it to the outside, which controls the gain to be lower as the maximum value is larger and higher as the maximum value is smaller. By configuring the gain control device, it is possible to realize an automatic gain control device in which the gain accurately and quickly follows the voice signal in the voice section, but does not easily follow the noise signal in the non-voice section. In this way, the gain control operation can be performed using only the audio signal in the input signal, so the gain does not increase excessively even if a relatively long silent period continues, and the noise can be over-amplified and output to the outside. It is possible to provide an automatic gain control device without

[Brief explanation of the drawing]

FIG. 1 is a block diagram of an automatic gain control device according to an embodiment of the present invention, FIG. 2 is a waveform diagram of signals in each part of FIG. 1, FIG. 3 is a block diagram of a conventional automatic gain control device, and FIG. The figure is a waveform diagram of signals at each part in FIG. 3. 12...Amplitude calculation section, 13...Speech section detection section. 14... Maximum value hold section, 16...
Variable gain amplifier. Name of agent: Patent attorney Toshio Nakao and 1 other person No. 1
figure

Claims (1)

[Claims] an amplitude calculation section that calculates the amplitude value of an input signal; a speech section detection section that detects a speech section in the input signal; a local maximum value holding section that determines the local maximum value of the amplitude value; and a variable gain amplification section that amplifies or attenuates the input signal and outputs it to the outside, where the larger the maximum value is, the lower the gain is, and the smaller the maximum value is, the higher the gain is. An automatic gain control device comprising: