JP2000081900A - Sound absorbing method, and device and program recording medium therefor - Google Patents

Abstract

PROBLEM TO BE SOLVED: To suppress even such omnidirectional noise as an entire room is filled therewith. SOLUTION: Outputs of microphones 1, 2 are divided 4 into each frequency bands, and varied according to the positions of the microphones 1, 2. A difference between parameter values of each sound signal reaching the microphones is detected 3, and based on the detected difference, sound sources are separated by selecting a frequency component of each sound signal, and desired sound and undesired sound are discriminated 6 from each other from the difference between both frequency characteristics, and the undesired sound is suppressed on frequency base 7 and the output is synthesized with the sound source signal. The undesired signal is suppressed also on time base by paying attention to the difference in the periodicity of the waveforms between the desired and undesired sound.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a sound collecting method for separating at least one sound source from a plurality of sound sources using a plurality of microphones, an apparatus therefor, and a program recording medium.

[0002]

2. Description of the Related Art Conventionally, for example, in order to collect voice for a call or to operate a recognition process such as voice recognition or speaker recognition, the input signal does not contain noise and is collected in a clean state. It had to be sounded. However, when we actually pick up a sound or use a recognition process, it is generally difficult to pick up only the target sound clearly. Therefore, a method for improving the S / N of a picked-up signal and a technique for giving noise resistance to a recognition process have been developed. In particular, for non-stationary noise whose frequency characteristics are unknown or noise whose frequency characteristics are similar to the target voice (for example, when the noise is also a voice), the frequency characteristics of each sound source and the difference in the position of each sound source are determined. A method of improving the S / N ratio of a signal by calculating and utilizing the signal of each sound source has been developed (a sound source separating method, device and recording medium: Japanese Patent Application No. 09-2).
52212, Sep. 1996 Proceedings of the Acoustical Society of Japan 48
Pages 9 to 490 1-7-13 "Study of two sound source separation method focusing on information between channels"). Furthermore, in this conventional method, it is possible to separate and extract a plurality of sound sources only by selecting a frequency by dividing the frequency components of a signal into bands so fine that only the frequency components of one sound source are included.

[0003]

However, in the conventional method, since the difference in the spatial arrangement of the sound sources is used, it is difficult to suppress nondirectional noise that fills the entire room, for example. Therefore, in the present invention, a function of performing non-directional noise suppression is added to the conventional method, thereby suppressing both directional and non-directional noise.
N improves the effect of improvement.

[0004]

According to the present invention, a plurality of microphones provided apart from each other are used, and a difference between parameter values of respective acoustic signals reaching the microphones which varies depending on the positions of the plurality of microphones. , The frequency components of each acoustic signal are selected, and each sound source is separated and extracted. In addition, a difference in frequency characteristics between the target voice and noise is detected, and noise having no direction is suppressed on the frequency axis. Furthermore, paying attention to the difference in periodicity between the target sound and the noise waveform,
Unnecessary signals are suppressed even on the time axis.

The difference between the prior art of the present invention and the prior art is that in the prior art, only the difference in the spatial arrangement of each sound source was used or only the difference in the frequency characteristics between the target sound and noise was used. In the present invention, the S / N of the signal is further improved by combining both processes.

[0006]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS First, an embodiment of the present invention will be described. FIG. 1 shows an example of the functional configuration of this embodiment. Microphones 1 and 2 are arranged at intervals, for example, about 20 cm, and these microphones 1 and 2 collect acoustic signals from sound sources A and B, respectively, and convert them into electric signals. The output of the microphone 1 is called an L channel signal, and the output of the microphone 2 is called an R channel signal. The L-channel signal and the R-channel signal are output from the inter-channel time difference / level difference detection unit 3 in the band-specific inter-channel parameter value difference detection unit 3.
01 and supplied to the band division unit 4. In the band division unit 4, the L channel signal and the R channel signal are divided into a plurality of frequency band signals, respectively, and supplied to the band-by-band time difference / level difference detection unit 302 and the sound source determination signal selection unit 5.

[0007] In accordance with each detection output of the detection units 301 and 302, the sound source determination signal selection unit 5 selects one of the channel signals as an A component or a B component for each band.
With respect to the A component signal and the B component signal for each of the selected bands, the unnecessary component identification unit 6 detects a difference between the frequency characteristics of the target sound and the noise, and identifies the noise component. The identified noise component is attenuated by the unnecessary component attenuator 7. Finally, the A component signal and the B component signal for each of the selected bands are respectively synthesized by the sound source signal synthesizing unit 8 and separated and output as a sound source signal A and a sound source signal B.

As a parameter for obtaining the time difference between channels for each band, for example, a phase difference when a signal is frequency-decomposed is used. In addition, as a parameter for obtaining a level difference between channels for each band, for example, a difference in power spectrum of each channel is used. As a parameter for obtaining the inter-channel time difference, for example, cross-correlation between channels is used. The band dividing method of the band dividing unit 4 is performed, for example, by performing a discrete Fourier transform to convert to a frequency domain signal, and then dividing the signal into each frequency band. The L channel signal and the R channel signal R are respectively converted into band signals L ( f1) to L (fn) and R (f1) to R (fn).

The sound source determination signal selection unit 5 uses the values detected by the band-based inter-channel parameter value difference detection unit 3 to output the band signals L (f1) to L (fn) and R (f1) to R ( f
The sound source signal determination unit 501 determines which of the corresponding items in (n) is to be selected. For example, when the sound source A is near the microphone 1 on the L side and the sound source B is near the microphone 2 on the R side, the time difference between the channels and the level difference between the channels in the ith band are positive values ( However, the level difference between channels and the time difference for each band are obtained by subtracting the value on the R side from the value on the L side.)
When, the sound source signal selection unit 502 attenuates R (fi).

The unnecessary component identifying section 6 identifies the frequency component of the target sound and the unnecessary sound frequency component of the signal that has been frequency-decomposed by the band dividing section by utilizing the difference in the frequency characteristics of each signal. Next, the unnecessary component attenuating unit 7 attenuates the frequency component identified as unnecessary. The sound source signal synthesizing unit 8 re-synthesizes the output frequency component into an audio signal, and performs, for example, inverse Fourier transform to return to a time waveform. By the above processing, even in an environment where a plurality of sounds are uttered simultaneously, the S / N of the signal can be improved by separating and extracting each signal. Detailed operation examples of the unnecessary component identifying unit 6 and the unnecessary component attenuating unit 7 will be described below.

The unnecessary component identifying section 6 and the unnecessary component attenuating section 7
It is desirable to place it in the latter stage as much as possible, since the processing amount is light. (For example, as shown in FIG. 1,
For example, between the sound source signal selector 5 and the sound source synthesizer 8. Because the unnecessary sound identification processing and the unnecessary sound attenuating processing are applied to only the components selected as the frequency components of each sound source instead of all the frequency components, the number of frequency components to be processed can be reduced and the weight is light. This is because it can be realized by processing. However, since the same function can be realized in other installation locations, the unnecessary component identification unit 6 and the unnecessary component attenuation unit 7
Is used between the band division unit 4 and the band-based inter-channel parameter value difference detection unit 3, or between the band-based inter-channel parameter value difference detection process and the sound source signal determination process, or the sound source signal judgment unit 501. Any number may be placed anywhere between the sound source signal selection unit 502 and the sound source signal selection unit 502 and the sound source synthesis unit 8.

In this embodiment, the inter-channel time difference / level difference between the L channel signal and the R channel signal is estimated by the inter channel time difference / level difference detection section 301. The inter-channel time difference / level difference of the channel signal may be known, and the value may be provided to the sound source determination signal selection unit 5 as data. next,
The unnecessary component identifying unit 6 and the unnecessary component attenuating unit 7 will be described.

Here, as an example, a method in the case where the unnecessary component identifying unit 6 and the unnecessary component attenuating unit 7 are placed between the band dividing unit 4 and the band-based inter-channel parameter value difference detecting unit 3 will be described. First, FIG. 2 shows a flow of the processing, and the description will be made according to the flow. In this embodiment, noise suppression is performed using the difference in statistical properties between the target sound and noise. Therefore, as an example, a signal to be suppressed will be described as a wideband signal. Here, it is necessary to determine a broadband component from frequency components having power equal to or higher than a certain value. First, the band division unit 4
(2-01). For each of the received frequency components, the power (Pow (k
k)), and a histogram is taken (2-02). The histogram counts the number of bands that fall within a certain range of power. FIG. 3 shows an example of a frequency characteristic when a signal does not include wideband noise, and FIG. 4A shows an example of a histogram thereof.
FIG. 4B shows the power in the power section. further,
FIG. 5 shows an example of a frequency characteristic of a signal including broadband noise.
FIG. 6 shows an example of the histogram.

When no broadband noise is included, FIG.
As shown in FIG. 4, components other than the harmonic structure have only a small power. Therefore, even in the histogram of FIG.
It is limited to a section where the power is very small (section 1 in FIG. 4A). However, when broadband noise is included, FIG.
As shown in (1), there are many bands having relatively large powers other than the harmonic components of the signal. Therefore, FIG.
Also in the histogram of, the frequency of a section having a relatively large power is prominent among sections smaller than the power of the harmonic structure, such as a hatched portion (section 3).

Therefore, if the number of bands falling within a predetermined range (power section) (section 3 in FIG. 6) exceeds a predetermined value (threshold) x (2-03), unnecessary components are included. It is determined that it has been performed (2-04), and an unnecessary component attenuation process is started (2-05). In the unnecessary component attenuating step (2-05), the band having the power of the section where the histogram is prominent (section 3 in FIG. 6) is attenuated. As a method of attenuating, for example, a predetermined value α (0 <0 <
= Α <1) is multiplied by the frequency component of the band to be attenuated.
Here, the value of α may be the same value for all bands or may be changed for each band. Further, the band to be attenuated may be not only the band included in the section where the histogram protrudes, but also, for example, all the bands having power equal to or less than the section where the histogram protrudes. According to the above method, the presence / absence of broadband noise can be determined by the unnecessary component identification unit 6 and the wideband noise component can be attenuated by the unnecessary component attenuator 7.

Here, it should be noted that the unnecessary component identifying unit 6 and the unnecessary component attenuating unit 7 are provided between the band dividing unit 4 and the band-specific inter-channel parameter value difference detecting unit 3 or the band-specific inter-channel parameter value difference. Where between the detection unit 3 and the sound source signal judgment unit 501, between the sound source signal judgment unit 501 and the sound source signal selection unit 502, or between the sound source signal selection unit 502 and the sound source synthesis unit 8 , You can put any number. For example, comparison is made between the case where the signal is placed between the sound source signal selection unit 502 and the sound source synthesis unit 8 and the case where the signal is placed between the band division unit 4 and the band-based inter-channel parameter value difference detection unit 3. As an advantage of the former, there is an advantage that the number of bands to be identified is reduced because only those that have already been selected as frequency components of each sound source are identified, and processing is lightened.

Next, another embodiment will be described. In this embodiment, the target sound is distinguished from noise using the harmonic structure of the target sound, and noise suppression is performed. Therefore, as an example, the target of the noise is periodic noise or quasi-periodic noise (for example, voice or animal cry). In this case, the sound source signal determination unit 5
The harmonic structure of the sound source signal is estimated from the components determined in step 01 (which mainly include the frequency components of each sound source), and those having large power in other frequency components are determined as noise. FIG. 7 shows the processing procedure. Hereinafter, description will be given according to this procedure.

Here, a case where the unnecessary component identifying section 6 and the unnecessary component attenuating section 7 are placed between the sound source signal selecting section 502 and the sound source signal combining section 8 will be described. First, the sound source signal selection unit 50
2 is received (7-01).
With respect to the received frequency components, a predetermined number of bands (LL (f)) having larger powers are taken out in the descending order (7-02). Next, a fundamental frequency is selected from these bands. That is, since the fundamental frequency of human voice exists within approximately 100 Hz to 250 Hz,
Of the bands selected in step 7-02, for example, 100H
Only those that fall within 250Hz from z are selected (7-0
3). Next, the band (LLs) selected in step 7-03
The following processing is performed to estimate the fundamental frequency from (f)). That is, for each of the selected bands (LLs (f)), the harmonic power is added. The added power is set to SumLLs (f) (7-04). LL at which the sum SumLLs (f) becomes the largest
It is determined that s (f) is the fundamental frequency (7-05).

Next, it is determined whether or not a band mm having a large power exists in the components other than the determined fundamental frequency and its harmonics (7-06). In this determination, for example, if there is a band having power exceeding a predetermined value x in a band other than the fundamental frequency and its harmonic components, it is determined that the band is an unnecessary component. Is determined. If it is determined that there is an unnecessary component, the unnecessary component attenuator 7 attenuates the power of the band mm determined to be unnecessary (7-07). The method of this attenuation is, for example,
A certain value α (0 <= α <1) is multiplied by the power in the band mm. Here, as a precaution similar to the case of the embodiment shown in FIG. 2, the unnecessary component identifying unit 6 and the unnecessary component attenuating unit 7 Or, a band-based inter-channel parameter value difference detection unit 3
Between the sound source signal determination unit 501 and the sound source signal determination unit 501 and the sound source signal selection unit 502, or
Any number may be placed anywhere between the sound source signal selection unit 502 and the sound source synthesis unit 8. However, if it is placed only between the band division unit 4 and the band-based inter-channel parameter value difference detection unit 3 or between the band-based inter-channel parameter value difference detection unit 3 and the sound source signal determination unit 501, Before the components of the sound source are determined, there is a possibility that the harmonic structure estimation accuracy of each sound source is deteriorated.

Finally, another embodiment will be described. Here, the unnecessary sound is identified in the time waveform region, and the identified signal is suppressed. FIG. 8 shows the processing procedure. Here, a case where the unnecessary sound discriminating unit and the unnecessary sound attenuating unit are placed between the plurality of microphones 1 and 2 and the band dividing unit 4 will be described. First, a signal is read for a fixed section length (8-01). In FIG. 8, as an example, a case will be described in which a frame length is read for the Fourier transform. Next, it is determined whether or not the read signal is a voice section. The determination is made, for example, by calculating the autocorrelation function of the signal and obtaining the peak value (8-
02). The determination as to whether or not a peak exists, for example, when the difference between the maximum value of the autocorrelation function and the second largest value exceeds a predetermined threshold, the peak exists,
If not, the peak is not present (8-03). If there is a peak, the signal section is determined to be a voice section, and is sent to the next band division section 4 (8
-04). If no peak is detected, the signal section is determined to be an unnecessary sound section, sent to the unnecessary sound attenuating section, and the signal amplitude is attenuated by a predetermined magnitude (8-0).
5). After being attenuated, it is sent to the band dividing unit 4 which is the next processing. Further, the unnecessary sound discriminating unit and the unnecessary sound attenuating unit may be provided at the subsequent stage of the sound source synthesizing unit 8 or both.

The signal whose noise has been suppressed by the method described above can be used not only for the purpose of direct listening by a human, such as for speech, but also for the input signal of a voice recognition device or a speaker recognition device, for example. . According to the present invention, it is possible to improve the S / N of a signal and obtain a clear signal. Not only the case where each of the above-described units is configured by hardware, but also each function can be executed by reading and decoding and executing a program by a computer. Further, the technique shown in FIG. 2 or FIG. 7 and the technique shown in FIG. 8 can be used together.

[0022]

As described above, according to the present invention, the unnecessary sound discriminating section discriminates unnecessary voice and noise components with respect to the target signal, and the unnecessary sound component is attenuated by the unnecessary sound attenuating section. By doing so, it is also possible to suppress nondirectional noise that fills the entire room, improve the S / N of the signal, and obtain a clear signal.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a functional configuration of an embodiment of the present invention.

FIG. 2 is an unnecessary component identifying unit 6 and an unnecessary component attenuating unit 7 in FIG.
4 is a flowchart showing an example of the processing procedure.

FIG. 3 is a diagram illustrating an example of frequency characteristics when there is no broadband noise.

FIG. 4A is a diagram illustrating an example of a histogram when there is no broadband noise, and FIG. 4B is a diagram illustrating an example of a value in each power section.

FIG. 5 is a diagram illustrating an example of frequency characteristics when there is wideband noise.

FIG. 6 is a diagram showing an example of a histogram when there is broadband noise.

FIG. 7 is a flowchart showing another example of the processing procedure of the unnecessary component identifying unit 6 and the unnecessary component attenuating unit 7;

FIG. 8 is a flowchart showing still another example of the processing procedure of the unnecessary sound discriminating unit 6 and the unnecessary sound attenuating unit 7.

 ────────────────────────────────────────────────── ─── Continuing on the front page (72) Inventor Hiroyuki Matsui 3-19-2 Nishi-Shinjuku, Shinjuku-ku, Tokyo F-term in Nippon Telegraph and Telephone Corporation (reference) 5D015 CC14 DD02 EE04

Claims (12)

[Claims] 1. A sound collection method for separating at least one sound source from a plurality of sound sources by using a plurality of microphones provided apart from each other, wherein each output channel signal of each of the microphones is divided into a plurality of frequency bands. And, for each of the same bands of the output channel signals divided in the above-mentioned band division process, the value of the parameter of the acoustic signal reaching the microphone that changes due to the positions of the plurality of microphones. A difference, a band-by-band parameter value difference detecting step of detecting as a band-by-band parameter value difference, and based on the band-by-band parameter value difference of each band, any of the band-divided output channel signals. A sound source signal determination process of determining whether to select, and, based on the determination result, each of the band-divided output channel signals, A sound source signal selecting step of selecting at least one signal input from one sound source; and a sound source synthesizing step of combining a plurality of band signals selected as signals from the same sound source in the sound source signal selecting step as a sound source signal. In the method, between the band dividing process and the band-based inter-channel parameter value difference detecting process, between the band-based inter-channel parameter value difference detecting process and the sound source signal determining process, the sound source signal determining process and the sound source Selecting at least one of the frequency components of each output channel signal divided in the band dividing step during at least one of the sound source signal selecting step and the sound source synthesizing step; An unnecessary component identification process for identifying unnecessary frequency components other than the signal from the sound source to be performed, and an unnecessary component for attenuating the frequency components identified in the unnecessary component identification process. Sound collecting method characterized by having a decay process. 2. The method according to claim 1, wherein the unnecessary component identifying step includes identifying unnecessary frequency components by focusing on a statistical property of a signal from a sound source to be selected and an unnecessary signal. Characteristic sound collection method. 3. The method according to claim 1, wherein said unnecessary component identifying step identifies a frequency component of an unnecessary signal by focusing on a harmonic structure of a signal of a sound source to be selected. Sound method. 4. A sound collecting method for separating at least one sound source from a plurality of sound sources by using a plurality of microphones provided apart from each other, wherein each output channel signal of each of the microphones is divided into a plurality of frequency bands. And, for each of the same bands of the output channel signals divided in the above-mentioned band division process, the value of the parameter of the acoustic signal reaching the microphone that changes due to the positions of the plurality of microphones. Detecting a difference as an inter-channel parameter value difference for each band; and a step of detecting an inter-channel parameter value difference for each band, based on the inter-band parameter value difference for each band of the respective bands. A sound source signal selecting step of selecting at least one signal input from the sound source; and a signal from the same sound source in the sound source signal selecting step. In the method having a sound source synthesizing process of synthesizing a plurality of band signals selected as a sound source signal, an unnecessary sound identifying process for identifying an unnecessary signal by focusing on a signal level; A sound attenuating step of attenuating the extracted signal between the plurality of microphones and the band dividing step or after the sound source synthesizing step. 5. A sound pickup device for separating at least one sound source from a plurality of sound sources by using a plurality of microphones provided apart from each other, wherein each output channel signal of each of the microphones is divided into a plurality of frequency bands. For each of the same bands of each of the output channel signals divided by the band dividing means, the value of the parameter of the acoustic signal reaching the microphone that changes due to the positions of the plurality of microphones. A difference between band-based channel parameter value difference detecting means for detecting a difference as a band-based channel-to-channel parameter value difference, and any of the band-divided output channel signals based on the band-based inter-channel parameter value difference of each band. Sound source signal determining means for determining whether or not to select, based on the determination result, the same from each of the band-divided output channel signals. Sound source signal selecting means for selecting at least one signal input from a sound source; and sound source synthesizing means for synthesizing, as a sound source signal, a plurality of band signals selected as signals from the same sound source by the sound source signal selecting means. In the apparatus, between the band dividing means and the band-specific inter-channel parameter value difference detecting means, between the band-specific inter-channel parameter value difference detecting means and the sound source signal judging means, the sound source signal judging means and the sound source Between the signal selecting means and at least one between the sound source signal selecting means and the sound source synthesizing means, the frequency component of each output channel signal divided by the band dividing means should be selected. Unnecessary component identification means for identifying unnecessary frequency components other than the signal from the sound source; and unnecessary component reduction for attenuating the frequency components identified by the unnecessary component identification means. Sound collecting device, characterized in that it comprises a means. 6. An apparatus according to claim 5, wherein said unnecessary component identifying means is means for creating a histogram with respect to the power of each band for each of the band-divided signals, and wherein the frequency of a predetermined power section is a predetermined value. A sound pickup device comprising means for identifying an unnecessary component based on the above. 7. The apparatus according to claim 5, wherein said unnecessary component identifying means selects a high power component LLs (f) in a predetermined frequency band from the frequency components selected by said sound source signal selecting means. Means, means for adding the powers of the respective harmonic components of the selected component LLs (f), and harmonic components other than the fundamental wave component LLs (f) having the maximum added power to the unnecessary signal. A sound collection device comprising: means for identifying a frequency component. 8. A sound pickup device for separating at least one sound source from a plurality of sound sources by using a plurality of microphones provided apart from each other, wherein each output channel signal of each of the microphones is divided into a plurality of frequency bands. For each of the same bands of each of the output channel signals divided by the band dividing means, the value of the parameter of the acoustic signal reaching the microphone that changes due to the positions of the plurality of microphones. The band-to-band parameter value difference detecting means for detecting the difference as a band-to-band parameter value difference; and the same from each of the band-divided output channel signals based on the band-to-band parameter value difference of each band. Sound source signal selecting means for selecting at least one signal input from a sound source; A sound source synthesizing means for synthesizing a plurality of band signals selected as a sound source signal, calculating a peak value of an autocorrelation function of the signal, and generating an unnecessary signal depending on whether or not the peak value exists. Unnecessary sound identification means for identifying, and unnecessary sound attenuation means for attenuating a signal identified as unnecessary by the unnecessary sound identification means, between the plurality of microphones and the band division means, or at a subsequent stage of the sound source synthesis means A sound pickup device, which is provided in a sound pickup device. 9. A recording medium recording a program for separating at least one sound source from a plurality of sound sources using a plurality of microphones provided apart from each other, wherein each output channel signal of each of the microphones is stored in a plurality. And a parameter of an acoustic signal reaching the microphone, which varies depending on the positions of the plurality of microphones, for each of the same bands of the output channel signals divided by the above-described band division processing. , A band-based inter-channel parameter value difference detection process for detecting the band-based inter-channel parameter value difference, and the band-divided output channel signal based on the band-based inter-channel parameter value difference of each band. Sound source signal determination processing for determining which of the above is to be selected, and each of the band-divided channels based on the determination result. Sound source signal selection processing for selecting at least one signal input from the same sound source from the channel signal, and sound source synthesis processing for synthesizing, as a sound source signal, a plurality of band signals selected as signals from the same sound source in the sound source signal selection processing Between the band division processing and the band-based inter-channel parameter value difference detection processing, between the band-based inter-channel parameter value difference detection processing and the sound source signal determination processing, and the sound source signal determination processing During at least one of the sound source signal selection process and the sound source signal selection process and the sound source synthesis process, a selection is made of frequency components of each output channel signal divided by the band division process. Unnecessary component identification processing for identifying unnecessary frequency components other than the signal from the sound source to be reduced, and unnecessary component reduction for attenuating the frequency components identified in the unnecessary component identification processing. Computer readable recording medium and a process characterized by having the above-mentioned program. 10. The recording medium according to claim 9, wherein the unnecessary component identification process includes a process of creating a histogram with respect to the power of each band for each of the band-divided signals; A process for identifying an unnecessary component based on whether the value is equal to or greater than a value. 11. The recording medium according to claim 9, wherein said unnecessary component identifying process selects a component LLs (f) having a large power in a predetermined frequency band from frequency components selected by said sound source signal selecting process. , A process of adding the power of each harmonic component of the selected component LLs (f), and a process of removing harmonic components other than the fundamental component LLs (f) whose added power is maximum. And a process for identifying the frequency component. 12. A recording medium recording a program for separating at least one sound source from a plurality of sound sources by using a plurality of microphones provided apart from each other, wherein each output channel signal of each of the microphones is stored in a plurality. And a parameter of an acoustic signal reaching the microphone, which varies depending on the positions of the plurality of microphones, for each of the same bands of the output channel signals divided by the above-described band division processing. , A band-based inter-channel parameter value difference detection process for detecting the band-based inter-channel parameter value difference, and the band-divided output channel signal based on the band-based inter-channel parameter value difference of each band. A sound source signal selection process for selecting at least one signal input from the same sound source; The above program has a sound source synthesizing process for synthesizing a plurality of band signals selected as signals from the same sound source as a sound source signal, and calculates a peak value of an autocorrelation function of the signal, and the peak value exists. An unnecessary sound discriminating process for discriminating an unnecessary signal depending on whether the signal is unnecessary, and an unnecessary sound attenuating process for attenuating a signal identified as unnecessary in the unnecessary sound discriminating process, between the plurality of microphones and the band dividing process. Or a computer-readable recording medium that the program has after the sound source synthesis processing.