JPH02272814A - Voice amplifier - Google Patents

Abstract

PURPOSE:To output a voice which can easily be heard by a man even if a peripheral sound is large and offensive by controlling the gain of an amplifier which amplifies an input voice in accordance with the inference result of a fuzzy inference means. CONSTITUTION:A microphone 1 inputting the sound, a low pass filter 3 deriving only voice components, a band pass filter 6 deriving only the signal of a frequency equivalent to the frequency characteristic of a man's ear, a collection microphone 9 taking in the peripheral sound, a low pass filter 11 and a band pass filter 14 with the frequency characteristic of the man's ear are provided. Furthermore, the inference means 17 executing fuzzy inference, a band pass filter 18 having the frequency characteristic of the man's ear, the amplifier I 19 and the amplifier II 21 are provided, and the amplifier I 19 and the amplifier II 21 are gain-controlled by the inference result of the inference means 17. Thus, the input voice with respect to the peripheral sound can be emphasized by an area equivalent to the frequency characteristic of the man's ear, and comfortable voice can be heard.

Description

DETAILED DESCRIPTION OF THE INVENTION (a) Field of Industrial Application This invention relates to an audio amplifier, and particularly to an audio amplifier that automatically adjusts the volume using fuzzy inference means to make it easier for people to hear.

(B) Conventional technology In general, when the audio output is large, an audio amplifier controls the output by feeding it back to the input side to reduce the output, and conversely, when the audio output is small, the output increases. Some models are equipped with an auto level control function to control the level.

(c) Problems to be solved by the invention The conventional audio amplifiers with the auto level control function described above control the sound within a certain range, but do not take into account the loudness of surrounding sounds, ease of hearing, etc. Not there,
Therefore, when the surrounding sound is loud, especially when the surrounding sound has a frequency close to the frequency characteristics of the human ear, there is a problem that it is difficult to arrange the space because it is harsh on the ears.

This invention has been made in view of the above-mentioned problems, and aims to provide an audio amplifier that can output audio that is easy for people to hear even if the surrounding sound is loud or harsh. There is.

(d) Means and Effects for Solving the Problems The audio amplifier of the present invention includes a sound detection means for detecting the loudness of input sound, an ambient sound detecting means for detecting the loudness of surrounding sounds, and a human ear. a first filter having a frequency characteristic and receiving the input sound as an input; a second filter having a frequency characteristic of a human ear and receiving the ambient sound manually; the voice detecting means; the ambient sound detecting means; fuzzy inference means for performing fuzzy inference using each output of the first filter and the second filter as input, and the gain of an amplifier for amplifying input audio is controlled according to the inference result of the fuzzy inference means. ing.

In this audio amplifier, input audio and ambient sound are detected by the audio detection means and the ambient sound detection means, respectively.

Further, the first filter derives the frequency characteristic component of the human ear included in the audio input, and the second filter derives the frequency characteristic component of the human ear included in the ambient sound. Then, the input voice, the ambient sound, the frequency characteristic component of the human ear included in the voice input, and the magnitude of the frequency characteristic component of the human ear included in the ambient sound are input to the fuzzy inference means.

The fuzzy inference means performs fuzzy inference on the input data based on a predetermined rule, and outputs a signal for adjusting the gain of the amplifier as the inference result. For example, if the frequency characteristics of the human ear in the surrounding sound are large, this will be very harsh to the ears, so the gain of the audio input amplifier in the frequency area corresponding to the frequency characteristics of the human ear is adjusted so that the level of the sound matches that of the surrounding sound. Compare it and raise it so that it becomes relatively large.

(E) Examples The present invention will be explained in more detail with reference to Examples below.

FIG. 1 is a block diagram showing an audio amplifier according to an embodiment of the present invention. The audio amplifier of this embodiment includes a microphone for inputting audio, a preamplifier 2 for amplifying the manually inputted audio from the microphone, a low-pass filter 3 for deriving only audio components, and rectifying the output of the low-pass filter 3. A rectifier 4 that smooths and outputs the average level of the input, X, a smoothing circuit 5, and a bandpass filter 6 that derives only a signal with a frequency corresponding to the frequency characteristics of the human ear from the input signal. , a combing device 7, a smoothing circuit 8, an integrated microphone 9 that captures ambient sound, a preamplifier 10, a low-pass filter 11, a rectifier 12, a smoothing circuit 13,
A bandpass filter 14 having frequency characteristics of the human ear, a rectifier 15, a smoothing circuit 16, and a smoothing circuit 13.16.
an inference means 17 that receives each output of 5.8 as inputs XI, Xi, X3, and X4 and executes fuzzy inference; a bandpass filter 18 having frequency characteristics of the human ear;
It consists of an amplifier t19, an adder 20 that adds the output of the preamplifier 2 and the output of the amplifier 119, an amplifier lI21 that amplifies the output of the adder 20, and a speaker 22. Further, the gains of the amplifier [19 and the amplifier [21] are adjusted according to the inference results yI and y2 of the inference means 17.

The reasoning means 17 is shown in more detail in FIG.

In the figure, the ambient sound level 31 inputted by the smoothing circuit vr13 passes through the sample hold circuit 35, and
It is input to the fuzzy inference device 39.

In addition, the surrounding sound magnitude 32 after being filtered manually by the smoothing circuit 16 passes through the sample hold circuit 36, and
It is input to the fuzzy tll logic device 39.

Also, the magnitude of the sound input from the smoothing circuit 5.8 is 33
, the filtered sound magnitude 34 are also input to the fuzzy inference device 39 via sample and hold circuits 37 and 38, respectively. The fuzzy inference device 39 is a circuit device for executing fuzzy inference, and uses a dedicated device or a processor programmed to execute fuzzy inference. This fuzzy inference device 39 calculates the surrounding sound volume X+, the surrounding sound volume after filtering X2
, sound size X3, and filtered sound size X
4 as an input, performs fuzzy inference according to the rules stored in the fuzzy rule memory 40, and uses the inference result as a signal y5 to adjust the gain of the amplifier 119;
A signal y2 for adjusting the gain of the amplifier 1121 is output. As shown in FIG. 1, the signal y is input to the amplifier 19 via the amplifier 41, and the signal y2 is input to the amplifier II21 via the amplifier 42.

The following fuzzy rules are stored in the fuzzy rule memory 40.

*■if (if), x, = PL, and x, = P
Lthen (if) Yz =ZR*■if
Xi = pt,, and x3 = NLt
heny2=PL *■if Xi=ZR, and x3=
ZRthen y2=ZR *■if x,=NL,and x3=P
Lthen yz = N L*■i
f xl = NL, and x3 = NL
then y2=ZR *■if xz=PL, and x, =P
Lthen y, =Ps *■if XZ =PL, and x4=
NLthen y, = PI, *■if
XZ = ZR, and X4 = ZRthe
n y, =PS *■if Xz =Nt,, and x,
= PL then y, = NS
*(lif Xz =NL, and
x4=NLthen y
! =PM Tables of this rule are shown in FIGS. 3 and 4. In the table of FIG. 3, the group axis represents the ambient sound volume X1, the horizontal axis represents the sound volume X3, and the column where the two intersect represents the gain of the amplifier 1121. In the table in Figure 4, the vertical axis represents the ambient sound size after filtering, XZ.
, the horizontal axis represents the magnitude of the filtered sound X4, and the column where the two intersect represents the gain of the amplifier 19.

The input of the surrounding sound volume X7, the sound volume after filtering XZ, the sound volume X3, the sound volume after filtering X4 used in the above rules, and the gain yI of amplifier I,
Examples of membership functions of the output of gain y2 of amplifier (2) are shown in FIGS. 5 to 10.

Here, - for boat fishing, NL is negative and very large (very small), NS is negative and slightly large (slightly small), P
S represents positive and somewhat large, and PL represents positive and very large. For example, in the case of the ambient sound level xl, PL means very loud, ZR means normal, and NL means much smaller than normal.

In the audio amplifier of the above embodiment, it is assumed that all the audio inputs are at the level shown in a of FIG. 11, and the ambient sound is at the level b. The level itself is PL, assuming that it is very large, and even when filters with the frequency characteristics of the human ear are connected, voltages of large levels are derived, and X
Both Z and X4 are PL. In this case, in the table of Figure 3, X1 = PL, x, = PL and y2 = ZR, so
The y2 output from the inference means 17 causes the amplification H! The gain of I21 is l! ! (moderate). On the other hand, in the table of FIG. 4, x t= P L , X 4= P
Since yl=PS at L, the gain of the amplifier 119 is slightly higher than normal based on the yI ratio output from the inference means 17, and as shown in FIG. When the adder 20 adds this signal to the audio signal output from the preamplifier 2 and outputs it, the audio is output from the speaker 22 via the amplifier r[21.The audio characteristics are as shown in Figure 13. 300 as shown
An audio signal with a high level of Hz to 30 Q OHz, that is, an audio signal with characteristics almost similar to the frequency characteristics of the human ear, is output, and people can listen to it with ease.

(f) Effects of the invention According to this invention, the magnitude of the surrounding sound after passing through a filter that has the frequency characteristics of the human ear, the volume of the voice, and the frequency characteristics of the human ear. Fuzzy inference is executed using the loudness of the sound after passing through the filter as input, and the gain of the amplifier that amplifies the input sound is controlled based on the result of the inference. By emphasizing the area corresponding to the characteristic, you can listen to the sound with a pleasant sound.

[Brief explanation of drawings]

FIG. 1 is a circuit block diagram of an audio amplifier showing an embodiment of the present invention, FIG. 2 is a block diagram showing the configuration of an inference means of the audio amplifier, and FIGS. 3 and 4 are block diagrams of the audio amplifier. FIG. 3 is a diagram showing a table representing an inference rule executed in FIG. 3, where the inference result is the gain of amplifier r of the audio amplifier, and FIG. The gain diagrams, Figures 5, 6, 7, and 8, respectively show the loudness of the surrounding sound input to the fuzzy inference device, the loudness of the surrounding sound after filtering, and the loudness of the voice. Figures 9 and 10 are diagrams showing the membership functions of the loudness of the voice after filtering, and Figures 9 and 10 are diagrams showing the membership functions of the gain of the amplifier and the gain of the amplifier H output from the same fuzzy inference device. , FIG. 11 is a diagram illustrating a case where the voice and surrounding noise are close in level, and FIG.
The figure shows a case where only the audio level is relatively increased in the area corresponding to the frequency characteristics of the ear under the condition shown in Figure 11.
FIG. 13 is a diagram showing the frequency characteristics of the human ear. 1: Audio input microphone, 3.11: Low pass filter, 6.14.18: Band pass filter, 17: tl
V theory pitching means 9: Amplifier ■, 21: Amplifier ■.

Claims (1)

[Claims] (1) A sound detection means for detecting the loudness of the input sound, an ambient sound detecting means for detecting the loudness of the surrounding sound, and a first filter having frequency characteristics of the human ear and receiving the input sound as an input. a second filter having frequency characteristics of the human ear and receiving the ambient sound as input; and a fuzzy filter using each output of the voice detection means, the ambient sound detection means, the first filter, and the second filter as input. 1. An audio amplifier comprising: fuzzy inference means for performing inference; and in accordance with the inference result of said fuzzy inference means, the gain of an amplifier for amplifying input audio is controlled.