JPH06334457A - Automatic sound volume controller - Google Patents

Abstract

PURPOSE:To properly maintain a sound volume in the sense of hearing over the entire frequency region of a voice signal by calculating a correction quantity as to plural frequency bands so as to form a filter providing a gain frequency characteristic and allowing the voice signal to pass through the filter. CONSTITUTION:A level detector 7 divides an output of a FET 6 into proper frequency bands to provide a level of a voice signal included in each frequency band thereby obtaining an effective value by the entire data entering the set frequency band. A surrounding signal obtained by a microphone 11 is converted into level data at an equal frequency interval by a FET 14 and the level detector 15 divides the data into a proper frequency bands to provide a level of the surrounding sound signal included for each band. A comparator 21 compares an output of the detector 15 with an output of a hearing sense characteristic latch means 20 for each band and provides an output which is larger to a gain calculation means 8, in which the gain for each band is decided. A filter coefficient calculation means 10 applies inverse Fourier transformation to an output of an interpolation device 9 to obtain an FIR filter coefficient, which is sent to an FIR filter 4.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to automatic volume control of a sound reproducing device.

[0002]

2. Description of the Related Art In a sound reproducing device used in a noisy environment such as a car radio and a car stereo, conventionally, the reproduction volume is automatically increased according to the level of ambient noise to maintain a listening condition in the presence of noise. Attempts have been made to try.

FIG. 8 is a block diagram showing a conventional volume automatic adjusting device disclosed in, for example, Japanese Patent Publication No. 3-13726. In the figure, 300 is a voltage control amplifier, 301 is a power amplifier, 302 is a speaker, 303 is a noise detection circuit, 304 is a volume adjusting means, 305 is an adder, 306 is a volume operating means, and 307 is a volume setting means.

In the above configuration, the audio signal is a voltage controlled amplifier (hereinafter referred to as VCA) 300 and a power amplifier 301.
Through the speaker 302. VCA30
The amplification factor of 0 is controlled by the output of the adder 6. The adder receives a noise level signal whose level is adjusted by the volume adjusting means 304 after being detected by the noise detecting circuit 303 and a volume increasing / decreasing operation signal from the volume operating means 304, and a volume level signal corresponding to a desired volume of the user. The output of the volume setting means 307 for outputting is input. With this configuration, the volume is controlled by the noise detected by the noise detection circuit. Further, the volume adjusting means 304 is capable of receiving the output of the volume adjusting means 304 and changing the amplification degree for the noise level signal according to the volume setting at that time. Specifically, as shown in FIG. 9, when the volume level is high, the amplification degree for the noise level is lowered, and the volume increase for the noise increase is reduced, and when the volume level is low, the amplification degree for the noise level is increased, and the volume increase for the noise increase is made. Is to make the volume control for ambient noise highly audible.

As described above, the contents that have been conventionally performed are as follows.
Although there was consideration for the volume level, the playback volume was simply increased according to the noise level.

[0006]

However, the lack of sound volume caused by noise is mainly due to the masking of the voice signal due to noise. Therefore, the masked voice signal component is a frequency component having a large noise component. It is large in the frequency range and small in the frequency range where the noise component is small.

Therefore, if a uniform process unrelated to the frequency of the audio signal is performed to control the reproduction volume as in the conventional device, the effect becomes insufficient in the frequency range where the noise component is large, There is a problem that the effect is insufficient in the frequency range where the noise component is small. In fact, the running noise of an automobile has a large bass component and a small treble component. Therefore, a mere gain increase process inevitably causes a lack of bass and an excessive volume of treble.

As a solution to this problem, it is possible to compensate the frequency characteristics such as boosting the low frequency range of the audio signal and attenuating the high frequency range of the audio signal on the assumption of the noise spectrum, but when the noise spectrum changes. Is not an adequate solution to.

Another problem in the conventional apparatus is that the gain of the audio signal with respect to the noise level is determined by representing the level of the audio signal by setting the volume, so that the difference in the recording level of the music source or the phrase of the music is different. It is inevitable that the perceived sound volume is excessive or deficient due to changes in the level.

Another problem of the conventional sound reproducing apparatus is that the characteristics of each listener are not taken into consideration. In other words, not all individual hearing characteristics are the same,
In particular, it is well known that the older the elderly, the lower the sensitivity in the high frequency range.

As a solution to this problem, it is conceivable to provide a sound quality controller such as a sound quality controller that is entrusted to the adjustment of the listener to adjust the volume of the high frequency range. Most of them are phenomena in which the minimum audible value rises, and the sensitivity when the level of the reproduced sound is sufficiently high does not decrease so much. In any case, the maximum audible limit does not change so much, and the problem that simply increasing the volume is unnecessarily too loud for a sound that is originally loud is not a sufficient solution.

Further, since the auditory characteristics of individuals are not uniform, there is a problem that appropriate adjustment can be performed according to the individual characteristics.

The present invention has been made to solve the above-mentioned problems, and appropriately reduces the perceived sound volume due to ambient noise during music reproduction and properly compensates the auditory characteristics of the listener, and thus the entire frequency band of the audio signal. An object is to obtain an automatic volume control device that maintains an appropriate volume over a period of time.

[0014]

SUMMARY OF THE INVENTION An automatic volume control device according to the invention of claim 1 is an audio signal analyzing means for dividing an audio signal into a plurality of frequency band components and detecting respective levels.
Ambient sound analysis means for dividing the noise into a plurality of frequency band components corresponding to this and detecting the respective levels, and hearing characteristic holding means for holding data corresponding to the minimum audible value of the listener for each of the divided frequency bands. And gain calculating means for calculating the gain to be given to the voice signal for each of the divided frequency bands,
Filter means for roughly giving the frequency-gain characteristic to the audio signal.

According to the second aspect of the present invention, the automatic volume control device comprises a voice signal analyzing means for dividing the voice signal into a plurality of frequency band components and determining the level of the voice signal component falling within each frequency band, a microphone, and a microphone. An adaptive filter unit configured to receive the obtained ambient sound signal and the voice signal as input and subtract a component having a high correlation with the voice signal from the ambient sound signal and output the adaptive ambient sound signal and a plurality of ambient sound signals provided as the output of the adaptive filter unit. Ambient sound signal analysis means for obtaining the level of the ambient sound signal component that is divided into frequency band components and that enters each frequency band, auditory characteristic holding means for holding the auditory characteristics of the listener, output of the audio signal analysis means, ambient sound signal The gain to be given to each audio signal component in the plurality of frequency bands is calculated from the output of the analysis means and the auditory characteristic held in the auditory characteristic holding means. A resulting calculation unit, those comprising a filter means for schematic realizing a gain frequency characteristic given as the output of the gain calculating means.

According to a third aspect of the present invention, there is provided an automatic volume control device, wherein the audio signal is divided into a plurality of frequency band components, the audio signal analyzing means for detecting respective levels, and the noise is divided into a plurality of frequency bands corresponding thereto. Ambient sound analysis means for dividing each component into each level and detecting each level, auditory characteristic holding means for holding data corresponding to the minimum audible value of the listener for each divided frequency band, and each divided frequency band The audio characteristic is provided with a gain calculating means for calculating a gain to be given to the audio signal and a filter means for roughly giving the frequency-gain characteristic to the audio signal. A test signal generating means for generating a test signal having a central power in a frequency band which is symmetrical to the characteristic change. Than it is.

[0017]

According to the automatic volume control device of the first aspect of the present invention, the reduction of the perceptual volume and the hearing ability of the listener due to the masking of the audio signal by noise can be obtained for a plurality of frequency bands. It is possible to calculate an appropriate correction amount (gain). In addition, by constructing a filter that gives the gain-frequency characteristic as the correction amount obtained in this way, and letting the audio signal pass through this, the audible volume can be appropriately maintained over the entire frequency range of the audio signal regardless of the type of noise. You can

In the automatic volume control device according to the second aspect of the present invention, the adaptive filter means removes the component having a high correlation with the sound signal from the input of the ambient sound signal analysis means, thereby eliminating the influence of the reproduced sound and improving the accuracy. Since good noise level estimation is performed, it is possible to accurately compensate for the reduction in the audible volume due to noise masking.

In the automatic volume control device according to the invention of claim 3, the auditory characteristic held in the auditory characteristic holding means, or
Since the listener is able to change one of the plurality of retained auditory characteristics, the volume control operation can be adjusted to better suit the characteristics of each listener. Furthermore, a test signal generating means for generating a test signal having a central power in a frequency band symmetrical to the characteristic change is provided, the listener adjusts the reproduction level of this signal sound, and the auditory characteristic of the listener is estimated from this adjustment amount. By doing so, the auditory characteristics of the listener can be accurately and easily measured.

[0020]

EXAMPLES Example 1. FIG. 1 is a block diagram showing an automatic volume control device according to an embodiment of the invention of claim 1. In the figure, 1 is an input terminal, 2 is an output terminal, 3 is an A / D converter, 4 is an FIR filter, 5 is a D / A converter, 6 is a fast Fourier transformer, 7 is a level detector, and 8 is a gain. Calculation means, 9 is an interpolator, 10 is a filter coefficient calculation means, 11 is a microphone, 12 is an amplifier, 13 is an A / D converter, 1
Reference numeral 4 is a fast Fourier transformer, 15 is a level detector, 18 is an A / D converter, 20 is a hearing characteristic holding means, and 21 is a comparator. Further, 100 is an audio signal analyzing means, 101 is an ambient sound analyzing means, and 102 is a filter means.

The input voice signal is converted into a digital signal by the A / D converter 18, and then converted by the fast Fourier transformer 6 into a set of data giving the level of the signal component at each center frequency at equal frequency intervals. . From this, it is possible to obtain the signal component level in the analysis period for each frequency band by dividing 1/2 of the sampling frequency into n equal to the number n of sample data given as an input. This fast Fourier transform method is well known and can be executed by a digital signal processor (hereinafter referred to as DSP) which has already been put into practical use.

Since the masking for a voice signal component (spectrum) of a certain frequency is determined by the level of noise falling within the critical bandwidth centered on that frequency, the analysis frequency band of the voice signal and the noise is about the critical bandwidth. It is desirable to subdivide the data into smaller parts, but since the critical bandwidth is narrow in the low frequency range and wide in the high frequency range, equidistant data obtained by the fast Fourier transform should have an appropriate bandwidth in the low frequency range. And the frequency interval of data becomes too close in the high frequency range.

The level detector 7 is a fast Fourier transformer 6
The output of is divided into appropriate frequency bands, and the level of the audio signal contained in each frequency band is given, and the processing for obtaining the effective value of all the data within the set frequency band is performed. For the frequency band division, for example, the audio frequency band (20 to 20 kHz) is divided into about 10 octaves to obtain a sufficient audible performance. In addition, the general noise spectrum has a high level in the low frequency range and a lower level in the high frequency range (normal household noise decreases by -6 dB / oct, and vehicle interior noise decreases by about -10 to -12 dB). )It is known. For this reason, it is practically possible to narrow the detection range of the high frequency range of the voice signal level and the ambient sound level to reduce the processing amount in the fast Fourier transformer.

The fast Fourier transformer 6 and the level detector 7 form the audio signal analyzing means 100. As another means, a plurality of band-pass filters corresponding to a required frequency band and their outputs are respectively detected. The smoothing level detector can be replaced with this.

Next, the ambient sound signal obtained from the microphone 11 is amplified by the amplifier 12, converted into a digital signal by the A / D converter 13, and converted into level data at equal frequency intervals by the fast Fourier transformer 14. . The level detector 15 is
The output of the fast Fourier transformer 14 is divided into appropriate frequency bands, and the level of the ambient sound signal contained in each frequency band is given.

The fast Fourier transformer 14 and the level detector 15 form the ambient sound signal analyzing means 101. As another means, a plurality of band filters corresponding to a required frequency band and their outputs are respectively provided. The level detector for detecting and smoothing can be replaced with this.

On the other hand, the hearing characteristic holding means 20 holds an average hearing characteristic for each age of a person. Specifically, FIG.
As illustrated in FIG. 5, the hearing ability (minimum audible value) in the same frequency band as the frequency band division of the ambient sound signal is held as data. Which of the several characteristics to retain is used depends on the listener's choice.

The comparator 21 compares the output data of the ambient sound analyzing means 101 and the output of the hearing characteristic holding means 20 for each divided frequency band, and outputs the larger one to the gain calculating means 8. . The gain calculating means 8 determines the gain for each frequency band.

Here, the concept of gain calculation will be described. First, a description will be given of the reduction of the audible volume of a voice signal caused by masking due to ambient noise. The decrease in the audible volume of the audio signal caused by masking due to ambient noise is considered to be a phenomenon in which the origin moves to the masking level (unit sone) at the sensory level in units of sone (hereinafter referred to as "audible volume"). That is, when S _m is the audible sound volume with masking, S is the audible sound volume without masking, and S _th is the audible sound volume corresponding to the masking level, the following relationships are established. _{S m = S-S th =} K (I a -I th a) - (1) where the strength of the I sound (in W / m ^2), I _th is the intensity of the masking level equivalent sound K and a are constants depending on the frequency.

The relationship between the sound pressure level P (unit dBspl) and the sound intensity I is as follows in the case of a plane wave. P = 10 · log (I) +150 −− (2)

FIG. 7 is a diagram showing the relationship between the sound pressure level and the sound intensity. A curve 200 shows the case where there is ambient noise, and the masking level 201 due to ambient noise is shown at 201.
Curve 202 is the case where there is no ambient noise. In order to obtain the same perceived sound volume when there is ambient noise than when there is no ambient noise, it is necessary to increase the sound pressure level of the audio signal so as to map from the curve 202 to the curve 200, as indicated by the arrow in the figure. I understand that there is.

The gain calculating means 8 calculates the gain to be given in order to make the perceived volume of the voice signal equal to that in the case where there is no ambient noise for each frequency band analyzed by the voice signal analyzing means based on the relationship of FIG. Is. Specifically, first, the volume P ₀ of the reproduced sound when the volume control is not applied is estimated from the output of the audio signal analysis unit 100, and the masking level P _{t is} calculated from the ambient sound level obtained from the output of the ambient sound analysis unit 101. Is estimated and the loudness P _{1 of the} reproduced sound to be given in order to compensate for the reduction in the auditory volume due to masking.
Is calculated by the following formula. This formula is derived from the previous formula. _{P 1 = P 0 +10 · log} {1 + 10 (pt-pO) · a / 10} / a - (3) gain to give here a speech signal, becomes the P ₁ -P ₀ equivalent. The value of a is about 0.3 for a 1 kHz pure tone.
Although it takes a different value for each frequency, it is calculated as 0.3 except for the low frequency range, and there is no practical problem. For the low frequency range, for example, set the audio frequency band to 1 for each octave.
In the case of 0 division, it is possible to calculate with a value of a for each of 3 bands including 100 Hz or less, and a for 0.3 for the other 7 bands. The calculation of the equation (3) can be performed by an approximation calculation by a DSP or a microcomputer, or by a look-up process and an interpolation process using a table of the relationship of P _O , P _t and P ₁ .

Next, it is known that, in many cases, the volume decrease due to the auditory characteristic (hearing loss) exhibits the same characteristic as the increase of the minimum audible value due to the masking of ambient noise as shown in FIG. In other words, it is known that the hearing loss caused by the sensory system when the auditory formation is divided into the transmissive system and the sensory system is of this type, and the hearing loss in the high range due to aging of humans also occurs. Many will be this type.

Therefore, the actual decrease in volume (including hearing loss due to aging) is the largest of the minimum audible value increased by the masking of ambient noise and the minimum audible value given as the output of the hearing characteristic holding means 20. It can be obtained by applying the equation (2) and the equation (1) with P _th being the other. Further, the gain to be given to the audio signal can be obtained by applying this P _th to the equation (3). As described above, the gain calculating means 8 performs this calculation for each frequency band and outputs the gain to be given to the audio signal.

Next interpolator 9, filter coefficient calculation means 1
The 0, FIR filter 4 forms a filter means 102 for providing this gain-frequency characteristic to the audio signal.

Here, it is known that the coefficient of the FIR filter becomes almost the impulse response of the desired filter.
It is also well known that the frequency response and impulse response of an arbitrary filter can be mutually transformed by Fourier transform.
From this relationship, an arbitrary frequency response is subjected to inverse Fourier transform to obtain an impulse response and the impulse response is used as an FIR filter coefficient, whereby a characteristic approximate to a desired frequency response can be obtained.

Here, the filter coefficient calculating means 10 inverse Fourier transforms the output of the interpolator 9 to obtain FIR filter coefficients, and sends the coefficients to the FIR filter 4. At this time, the input to be given to the filter coefficient calculation means 10 is 1 / the sampling frequency of the voice signal when the number of FIR filter stages is n.
It is a gain at each frequency obtained by dividing 2 into n equal parts. Interpolator 9
Is converted into a gain at each frequency obtained by the gain calculating means 8 according to the auditory characteristics for each bandwidth having substantially the same ratio, that is, 1/2 of the sampling frequency of the voice signal is divided into n equal parts. Do the operation. Specific numerical processing can be executed by a DSP, a microcomputer, or the like.

Example 2. FIG. 2 is a block diagram showing an automatic volume control device according to another embodiment of the invention of claim 1.
In the figure, 1 is an input terminal, 2 is an output terminal, and 3 is an A / D.
Converter, 4 FIR filter, 5 D / A converter, 6 fast Fourier transformer, 7 level detector, 8 gain calculating means, 9 interpolator, 10 filter coefficient calculating means, 11
Is a microphone, 12 is an amplifier, 13 is an A / D converter,
14 is a fast Fourier transformer, 15 is a level detector, 16
Is a comparator, 18 is an A / D converter, 20 is a hearing characteristic holding means, and 21 is a comparator.

In contrast to the first embodiment, this embodiment is based on the assumption that the reproduction means of the audio signal is an earphone, for example, and the reproduced sound hardly enters the ambient sound collecting microphone. The case where the sound reproducing means is a speaker or the like and the reproduced sound enters the sound collecting microphone at the same time as the ambient sound.

Here, the blocks having the same numbers as in the first embodiment perform the same operations. That is, the fast Fourier transformer 6
Outputs the audio signal component level at equal frequency intervals, and the fast Fourier transformer 14 outputs the ambient sound signal component level at equal frequency intervals as the fast Fourier transformer 6. The comparator 16 includes a fast Fourier transformer 6 and a fast Fourier transformer 1
The same frequency component levels of 4 are compared, and the output level of the fast Fourier transformer 14 is
If the audio signal component is sufficiently higher than the level expected to occur when the audio signal component is reproduced by a speaker or the like and is picked up by a microphone from the output level of, the corresponding output of the fast Fourier transformer 14 is directly output, and is not In this case, the previously output value is output again (holding the previous value),
One of the following processes is performed: outputting a predetermined value for each band, outputting the same value as the smaller one of the levels of adjacent bands, and outputting a value that is smaller by a predetermined level than the previously output value. By this processing, it is possible to avoid erroneously recognizing the reproduced sound from the speaker or the like as noise and performing the subsequent volume adjustment processing.

Since the fast Fourier transformers 6 and 14 subdivide the frequency band, the probability that a large audio signal level exists in a specific band is low. On the other hand, noise often has a continuous frequency spectrum, and often has a certain level in all divided bands, so noise can be detected reliably from a probabilistic perspective. In addition, the noise level cannot be detected in some bands, and even if the above-mentioned measures are taken, the possibility of obtaining the correct noise level as a whole increases.

In the above description, it is assumed that the comparator 16 compares the outputs of the fast Fourier transformer 6 and the fast Fourier transformer 14, but when the level of the ambient sound is roughly known in advance, the sound signal level is sufficiently low. May be detected, that is, the output of the fast Fourier transformer 6 is compared with a predetermined value for each band, and when the output of the fast Fourier transformer 6 is small, the output of the fast Fourier transformer 14 is output as it is. ， Otherwise, it may be possible to perform the process (1) of the above description.

In the second embodiment, the fast Fourier transformer 6 is used in order to eliminate the influence of the reproduced voice signal entering the microphone.
The output of the fast Fourier transformer 14 and the output of the fast Fourier transformer 14 are compared by a comparator 16. The voice signal entering the microphone propagates through the space from the electroacoustic converting means such as a speaker to reach the microphone, and thus has a time delay corresponding to the distance. It will be. In order to avoid the inconvenience of erroneously recognizing a voice signal as ambient sound, it is necessary to compare the levels of the voice signals included in the outputs of the fast Fourier transformer 6 and the fast Fourier transformer 14 at the same timing. Therefore, when the distance between the microphone and the speaker is large, it is necessary to give a considerable time delay to the path on the audio signal level detection side.

Further, when the level of the voice signal and the ambient sound is detected by the fast Fourier transform, it takes a certain amount of time to obtain the result because the data block of a certain time related to the frequency resolution is analyzed. Control is delayed by this amount of time. The volume control is performed by the audio signal level and the ambient noise level as described above, but the delay in following the ambient noise level is not a problem. However, if the control for decreasing the gain is delayed when the audio signal level is rapidly increased, the audio is temporarily reproduced at an unnecessarily high level, which causes great discomfort during listening.
In order to avoid this, it is appropriate to provide a delay corresponding to this detection delay time before the audio signal path and the filter means 102. However, since the delay time in the filter means 102 can be subtracted from this, if the delay in the FIR filter 4 is large, it may not be necessary to provide another delay.

Example 3. FIG. 3 is a block diagram showing an automatic volume control device according to an embodiment of the present invention. In the figure, 1 is an input terminal, 2 is an output terminal, 3 is an A / D converter, 4 is an FIR filter, 5 is a D / A converter, 6 is a fast Fourier transformer, 7 is a level detector, and 8 is a gain. Calculation means, 9 is an interpolator, 10 is a filter coefficient calculation means, 11 is a microphone, 12 is an amplifier, 13 is an A / D converter, 1
4 is a fast Fourier transformer, 15 is a level detector, 16 is a comparator, 17 is an adaptive filter means, 18 and 19 are A
An A / D converter, 20 is a hearing characteristic holding means, and 21 is a comparator. Further, 100 is an audio signal analyzing means, 101 is an ambient sound analyzing means, and 102 is a filter means.

Here, blocks having the same numbers as in the first and second embodiments perform the same operation. Similar to the second embodiment, this embodiment also corresponds to the case where the sound reproducing means is a speaker or the like and the reproduced sound enters the sound collecting microphone at the same time as the ambient sound. The adaptive filter means 17 includes the microphone 11.
The ambient sound signal obtained by amplifying and A / D converting the output of is used as one input, and a voice signal having a correlation as high as possible with a voice signal reproduced from a speaker or the like is input to the other input (hereinafter referred to as a reference voice signal; Then, the output of the D / A converter 5 is A / D
It is used after being converted into a digital signal by the converter 19), and the operation of removing only the ambient sound by removing the component correlated with the reference audio signal, that is, the reproduced sound component from the speaker from the ambient sound signal. To do.

FIG. 4 is a diagram for explaining the operation of the automatic volume control device according to the third embodiment and shows the internal processing of the adaptive filter means 17. In the figure, 30 is an ambient sound signal input terminal, 31
Is an audio signal input terminal, 32 is an output terminal, 33 is a variable coefficient FIR filter process, 34 is a subtraction process, and 35 is a coefficient update process. Adaptive filters of this construction are well known.
The subtraction process 34 is for subtracting the reference voice signal passed through the variable coefficient FIR filter process 33 from the ambient sound signal,
The coefficient update processing 35 updates the coefficient of the variable coefficient FIR filter 33 so as to minimize the output of the subtraction processing 34 (hereinafter referred to as an error signal). Due to this coefficient update, the variable coefficient FIR filter processing 33 results in a system through which the audio signal obtained from the microphone passes after the reference audio signal extraction point, that is, an electroacoustic converter such as a D / A converter, a power amplifier, a speaker, or a voice. It simulates the transfer characteristics of the playback space and microphone. In other words, the variable coefficient FIR
The filtering process transforms the reference voice signal so that it becomes equivalent to the voice signal component contained in the ambient sound signal by convolving the reference voice signal with a characteristic that simulates the transfer characteristic of the system from the reproduction of the voice signal to the sound collection. It is a thing.

The signal processing in the above adaptive filter means can be executed by a DSP or the like which has already been put into practical use.

Example 4. FIG. 5 is a block diagram showing an automatic volume control device according to an embodiment of the present invention. In the figure, 1 is an input terminal, 2 is an output terminal, 3 is an A / D converter, 4 is an FIR filter, 5 is a D / A converter, 6 is a fast Fourier transformer, 7 is a level detector, and 8 is a gain. Calculation means, 9 is an interpolator, 10 is a filter coefficient calculation means, 11 is a microphone, 12 is an amplifier, 13 is an A / D converter, 1
4 is a fast Fourier transformer, 15 is a level detector, 16 is a comparator, 18 is an A / D converter, 20 is an auditory characteristic holding means, 21 is a comparator, 22 is adjustment control means, 23 is an operation key, 24 Is a display, 25 is a signal generating means, and 26 is a signal switching means. Further, 100 is a voice signal analyzing means, 1
Reference numeral 01 is an ambient sound analyzing means, and 102 is a filter means.

Here, the blocks having the same numbers as in the first, second and third embodiments perform the same operation. The present embodiment has a configuration in which the characteristics held by the hearing characteristic holding means 20 are adjusted according to the hearing characteristics of the listener. The adjustment control means 22 is, for example, a microcomputer, and the auditory characteristic holding means 20.
Can be a specific data area configured on the memory of this microcomputer. Operation key 23
The display 24 is a means for the adjustment control device 20 to receive the reaction of the listener and to display a status related to the adjustment. Usually, it can also be used for another purpose such as a volume adjustment operation and a sound signal level display. . The signal generating means 25 is configured by using, for example, a partial function of the DSP, and changes the center frequency and level of the output signal in response to a command from the adjustment control device 20. The signal switching means 26 switches the signal to be given to the D / A converter 5 between the output of the FIR filter 4 and the output of the signal generating means 25 according to a command from the adjustment control device 20.

In the automatic sound volume adjusting device configured as described above, adjustment for adjusting the device operation to the auditory characteristics of the listener can be performed by the following procedure. 1. Adjustment control means 20 in response to a listener operation on the operation keys 23
Starts the following adjustment operation. 2. The start of the adjusting operation is displayed on the display 24. The signal generating means 25 generates a test signal having a center frequency of 25 Hz. 3. The signal switching means 26 is switched to give the output of the signal generating means 25 to the D / A converter 5. 4. The test signal is barely heard by the listener from the operation key 23, and the display 24 is instructed to perform the adjustment end operation. 5. In response to the volume adjustment operation on the operation key 23 by the listener,
The output signal level of the signal generating means 25 is adjusted. In addition, the adjustment end operation for the operation key 23 by the listener is received and 25 Hz
After the adjustment for the band is completed, the reproduced sound pressure level is known from the output level of the signal generator 25 at that time. Since this level is almost the minimum audible value of the listener, this is recorded as data in the 25 Hz band of the listener-specific characteristics of the auditory characteristics holding means 20. About the band of 6.50 Hz or more 4.5. The above operation is repeated to record the minimum audible value data in the 50 Hz band in the auditory characteristic holding means 20. 7. Repeatedly adjust from 100 Hz to 12.8 kHz band. 8. When the adjustment of the entire region is completed, the adjustment operation is completed, and the signal switching means 26 is switched to return to the normal operation of giving the output of the FIR filter 4 to the D / A converter 5 and reproducing the audio signal.

As described above, in the fourth embodiment, since the auditory characteristics to be compensated are measured and operated for each listener, it is possible to perform a more appropriate automatic volume control operation.

In the above description, the FIR filter 4,
The fast Fourier transformer 6, the level detector 7, the gain calculating means 8, the interpolator 9, and the filter coefficient calculating means 10, 14 are independent of the fast Fourier transformer 14, the detector 15, the comparator 16, and the adaptive filter means 17, for convenience. However, these can be processes executed sequentially in the DSP. Further, even when all the processes cannot be executed by a single DSP due to the limitation of the processing capacity, the distributed processing can be executed by a plurality of DSPs or microcomputers.

Further, in this case, the A / D converters 13, 18 and 19 can be shared by using the same element in a time division manner.

[0055]

Since the present invention is constructed as described above, it has the following effects.

At the time of music reproduction, the reduction of the perceptual sound volume due to the ambient noise is appropriately perceived in accordance with the level and spectrum of the noise and the level and spectrum of the voice signal, and the auditory characteristics of the listener and the level and spectrum of the music signal. It is possible to provide an unprecedented automatic volume control function of compensating and maintaining an appropriate volume over the entire frequency band of the audio signal.

Further, since the means for easily measuring the auditory characteristics of the listener is provided, it is possible to provide the automatic volume control function which accurately compensates for the auditory characteristics that vary widely among individuals and enables comfortable music listening. .

Further, when the ambient noise is detected, it is possible to eliminate the inconvenience of erroneously recognizing the reproduced sound of the audio signal as noise, thereby ensuring the operation of the automatic volume control function.

Also, the influence of the reproduced voice signal is surely eliminated by matching the timing of the voice signal referred to for eliminating the influence of the reproduced voice signal when the ambient noise is detected and the voice signal component included in the ambient sound signal. As a result, the operation of the automatic volume control function can be ensured.

Further, by using the adaptive filter means for noise detection to improve the detection accuracy of ambient noise, the operation of the automatic volume control function can be ensured.

Further, by relatively compensating for the delay of the volume control with respect to the audio signal, a temporary and unnecessary increase in volume is suppressed and secondary effects such as discomfort caused by the operation of the automatic volume control function are suppressed. it can.

[Brief description of drawings]

FIG. 1 is a block diagram showing an automatic volume control device according to an embodiment of the present invention.

FIG. 2 is a block diagram showing an automatic volume control device according to another embodiment of the invention of claim 1;

FIG. 3 is a block diagram showing an automatic volume control device according to an embodiment of the present invention.

FIG. 4 is an operation explanatory diagram of the automatic volume control device according to the third embodiment.

FIG. 5 is a block diagram showing an automatic volume control device according to an embodiment of the invention of claim 3;

FIG. 6 is a diagram for explaining auditory characteristics in the present invention.

FIG. 7 is a diagram showing a relationship between sound pressure level and sound intensity in the present invention.

FIG. 8 is a block diagram of a conventional automatic volume control device.

FIG. 9 is an operation explanatory diagram of a conventional automatic volume control device.

[Explanation of symbols]

 1 Input Terminal 2 Output Terminal 3, 13, 18, 19 A / D Converter 4 FIR Filter 5 D / A Converter 6, 14 Fast Fourier Transform 7, 15 Level Detector 8 Gain Calculator 9 Interpolator 10 Filter Coefficient Calculation means 11 Microphone 12 Amplifier 16, 21 Comparator 17 Adaptive filter means 20 Auditory characteristic holding means 100 Voice signal analysis means 101 Ambient sound analysis means 102 Filter means

─────────────────────────────────────────────────── ───

[Procedure amendment]

[Submission date] January 11, 1994

[Procedure Amendment 1]

[Document name to be amended] Statement

[Name of item to be corrected] 0029

[Correction method] Change

[Correction content]

Here, the concept of gain calculation will be described. First, a description will be given of the reduction of the audible volume of a voice signal caused by masking due to ambient noise. The decrease in the audible volume of the audio signal caused by masking due to ambient noise is considered to be a phenomenon in which the origin moves to the masking level (unit sone) at the sensory level in units of sone (hereinafter referred to as "audible volume"). That is, when S _m is the audible sound volume with masking, S is the audible sound volume without masking, and S _th is the audible sound volume corresponding to the masking level, the following relationships are established. _{S m = S-S th =} K (I a -I th a) - (1) where the strength of the I sound (in W / m ^2), I _th is the intensity of the masking level equivalent sound K and a are constants depending on the frequency.

[Procedure Amendment 2]

[Document name to be amended] Statement

[Name of item to be corrected] 0030

[Correction method] Change

[Correction content]

The relationship between the sound pressure level P (unit dBspl) and the sound intensity I is as follows in the case of a plane wave. P = 10 · log (I) +120 −− (2)

[Procedure 3]

[Document name to be corrected] Drawing

[Name of item to be corrected] Figure 2

[Correction method] Change

[Correction content]

[Fig. 2]

[Procedure amendment 4]

[Document name to be corrected] Drawing

[Name of item to be corrected] Figure 3

[Correction method] Change

[Correction content]

[Figure 3]

─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁵ Identification code Internal reference number FI Technical indication location H03H 17/02 D 7037-5J K 7037-5J 17/06 A 7037-5J 21/00 7037-5J

Claims (3)

[Claims] 1. A voice signal analyzing means for dividing a voice signal into components of a plurality of frequency bands to obtain a level of a voice signal component in each frequency band, a microphone, and an ambient sound signal obtained from the microphone are divided into a plurality of frequency bands. The ambient sound signal analyzing means for obtaining the level of the ambient sound signal component that is divided into each frequency band and the level of the ambient sound signal component, the auditory characteristic holding means for retaining the auditory characteristics of the listener, the output of the voice signal analyzing means, and the ambient sound signal analyzing means. Gain calculation means for calculating a gain to be given to each audio signal component in the plurality of frequency bands from the output and auditory characteristics held by the auditory characteristics holding means, and gain frequency characteristics given as an output of the gain calculating means An automatic volume control device, comprising: 2. An audio signal analyzing means for dividing an audio signal into a plurality of frequency band components to obtain the level of an audio signal component in each frequency band, a microphone, and an ambient sound signal and an audio signal obtained from the microphone as inputs. An adaptive filter means configured to subtract a component having a high correlation with an audio signal from an ambient sound signal and outputting the same, and an ambient sound signal provided as an output of the adaptive filter means is divided into a plurality of frequency band components and enters each frequency band. Ambient sound signal analyzing means for obtaining the level of the ambient sound signal component, auditory characteristic holding means for holding the auditory characteristics of the listener, output of the audio signal analyzing means, output of the ambient sound signal analyzing means and auditory characteristic holding means A gain calculating means for calculating a gain to be given to each voice signal component in the plurality of frequency bands from the auditory characteristics, and an output of the gain calculating means. And a filter means for substantially realizing the gain-frequency characteristic given by the automatic volume control device. 3. A voice signal analyzing means for dividing a voice signal into components of a plurality of frequency bands to obtain a level of a voice signal component falling in each frequency band, a microphone, and an ambient sound signal obtained from the microphone are divided into a plurality of frequency bands. The ambient sound signal analyzing means for obtaining the level of the ambient sound signal component that is divided into each frequency band and the level of the ambient sound signal component, the auditory characteristic holding means for retaining the auditory characteristics of the listener, the output of the voice signal analyzing means, and the ambient sound signal analyzing means. Gain calculation means for calculating a gain to be given to each audio signal component in the plurality of frequency bands from the output and auditory characteristics held by the auditory characteristics holding means, and gain frequency characteristics given as an output of the gain calculating means Filter means and a test signal generating means, and the auditory characteristic held by the auditory characteristic holding means or a plurality of held auditory characteristics One of the above is configured to be changeable by a listener, and the test signal generating means is configured to generate a test signal having a central power in a frequency band symmetrical to the characteristic change, and automatic volume control. apparatus.