JP2003299183A - Sound collecting apparatus and stereophonic computing method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To solve the problem of emphasis of a wind noise component caused by stereophonic arithmetic processing. <P>SOLUTION: Sound signals are supplied from a first nondirectional microphone 1R and a second nondirectional microphone 1L to the adding terminals of computing elements 3R, 3L via amplifiers 2R, 2L respectively; and are supplied to band dividing means 4R, 4L and 5R, 5L as well. In the dividing means 4R, 4L on this occasion, signals in band 1 corresponding to wind noises for example are extracted. These signals from the dividing means 4R-5L are added for a left channel and a right channel in adders 10R, 10L via delay apparatuses 6R, 6L and 7R, 7L, variable attenuators 8R, 8L and attenuators 9R, 9L respectively, and are supplied to the subtracting terminals of the computing elements 3L, 3R mutually. Moreover, a difference component of signals extracted by the dividing means 4R, 4L is outputted from a subtracter 13, and the level of its envelope is detected by a level detecting means 14 and supplied to a coefficient generation means 15. The attenuators 8R, 8L are controlled according to this control coefficient. <P>COPYRIGHT: (C)2004,JPO

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a sound pickup device and a stereo calculation method suitable for use in, for example, a digital still camera or a VTR integrated video camera device. Specifically, when outputting a directional stereo sound field signal to a voice signal picked up by a built-in microphone using an arithmetic circuit, for example, it is intended to reduce the occurrence of obstacles due to wind noise. is there.

[0002]

2. Description of the Related Art For example, in a digital still camera or a video camera integrated with a VTR, it is required to pick up an image and to collect sound by stereo. However, in such a device, since two microphones are provided in the vicinity of a small-sized housing, good stereo sound collection cannot be performed by itself. Therefore, conventionally, there has been proposed a technique for emphasizing the left and right directivity of an audio signal picked up by using a stereo operation circuit as shown in FIG. 5, for example.

That is, in FIG. 5, first and second omnidirectional microphones 51R and 5 are arranged at arbitrary intervals.
From 1L, the picked-up audio signals are supplied as a right channel input and a left channel input. These right channel and left channel audio signals are supplied to amplifiers 52R and 52L, respectively, to optimize the signal levels. The signals from these amplifiers 52R and 52L are supplied to the addition terminals of the arithmetic units 53R and 53L, respectively, and are also supplied to the delay units 54R and 54L which are generally formed by low-pass filters.

Therefore, from these delay devices 54R and 54L, a time delay corresponding to the time during which the sound wave is transmitted is provided between the first and second omnidirectional microphones 51R and 51L, respectively. The low frequency component signal of the audio signal thus extracted is extracted. Further, these low-frequency component signals of the audio signal subjected to the time delay are respectively attenuator 5
It is supplied to 5R and 55L and the signal level is optimized,
Mutually supplied to the subtraction terminals of the arithmetic units 53R and 53L.

That is, the signal from the right channel attenuator 55R is supplied to the left channel arithmetic unit 53L, and the signal from the left channel attenuator 55L is supplied to the right channel arithmetic unit 53R. As a result, the arithmetic units 53R, 5
In 3L, matrix processing is performed by subtracting the low-frequency components of the respective channels, and the stereo sound field is reproduced in the low-frequency band. That is, although the sound pickup by the microphones 51R and 51L is omnidirectional as shown in A of FIG.
It will be directed.

The audio signals from the arithmetic units 53R and 53L are supplied to equalizers 56R and 56L, which are respectively composed of filter circuits, and the frequency characteristics are adjusted, and the audio signals passed through the equalizers 56R and 56L are It is taken out from the output terminals 57R and 57L as a right channel output and a left channel output, respectively. These output audio signals are supplied to an audio signal circuit (not shown) for subsequent processing such as recording and transmission.

In this way, a stereo operation circuit for enhancing the left and right directivity in the low frequency band of the audio signal is formed. In the high frequency band, the left and right directivity can be emphasized by a method such as devising the installation environment of the microphones 51R and 51L.
By combining these, the stereo sound field is reproduced over the entire frequency band. By using such a circuit, stereo sound can be collected even in a small device, for example.

[0008]

By the way, in the above-mentioned stereo operation circuit, the delay devices 54R and 54L have a time delay that depends on the interval between the first and second omnidirectional microphones 51R and 51L, and are of the normal type. The audio signals are output from the microphones 51R and 5R with a phase difference corresponding to this time delay.
A stereo operation process is performed by collecting sound by 1 L and subtracting the sound from each other. However, for example, since the wind noise component is picked up by the microphones 51R and 51L in a random phase, the noise component may be emphasized on the contrary.

That is, the wind noise is generated when the wind pressure is converted into an electric signal as it is, and is not propagated like a sound wave. Therefore, the microphone 5 is used.
It occurs in a random phase irrelevant to the time delay depending on the interval of 1R and 51L. Therefore, in the above-described stereo operation circuit, for example, when the inputs of the operation units 53R and 53L are just in opposite phases and the signals of the respective channels are added, the noise level is deteriorated by 6 dB. Become.

Further, since such a band of wind noise is a band in which the gain is normally increased by the equalizers 56R and 56L in the subsequent stage, the wind noise is easily emphasized as compared with the voice signal. Therefore, emphasizing wind noise in the above-mentioned stereo operation circuit may cause a serious problem.
In addition, due to the miniaturization of products in recent years, the microphone 51R,
The interval of 51L is further reduced, and equalizers 56R, 5R
Since the gain at 6L tends to increase, the above-mentioned problem becomes more remarkable.

This application has been made in view of such a point, and the problem to be solved is that in the conventional apparatus, the wind noise generated in the random phase is emphasized by the noise component in the stereo operation circuit. Such emphasis on wind noise may cause extremely troubles in relation to the circuit in the subsequent stage, and the gap between microphones has been further narrowed due to miniaturization of products in recent years. For this reason, this tendency is becoming more prominent.

[0012]

Therefore, in the present invention, for example, the occurrence of wind noise is detected and the parameters of the stereo operation processing are automatically controlled.
This eliminates the possibility that the wind noise component will be emphasized due to the stereo calculation processing even when wind noise occurs, and in particular, even in microphones that are provided close to each other in a small device, good stereo output is always obtained. Can enable sound.

[0013]

BEST MODE FOR CARRYING OUT THE INVENTION That is, the sound collecting device of the present invention has first and second omnidirectional microphones arranged at arbitrary intervals, and a first voice from the first omnidirectional microphone. A sound collecting device provided with a stereo operation circuit for outputting a directional stereo sound field signal using a signal and a second audio signal from a second omnidirectional microphone, comprising:
And a band dividing unit for dividing the voice band of the second voice signal into a plurality of bands, and the voice signals of the respective bands divided by the band dividing unit according to the intervals between the first and second omnidirectional microphones, respectively. By means of delaying means for delaying the time and a signal attenuating means for setting the level of the audio signal of each band respectively, and an adding means for adding each of the first and second audio signals and a signal from the adding means to each other. A first band signal and a second band signal from the first band signal of the respective band band signals divided by the band band dividing unit. The level of the signal attenuating means in the predetermined band is controlled by the control signal according to the level detected by the difference component from the signal in the predetermined band.

Further, the sound collecting device of the present invention has first and second omnidirectional microphones arranged at arbitrary intervals, and the first voice signal from the first omnidirectional microphone and the first voice signal. A sound collection device provided with a stereo operation circuit for outputting a directional stereo sound field signal using a second sound signal from an omnidirectional microphone of No. 2, and
Band dividing means for dividing the voice band of the voice signal into a plurality of bands, and the voice signal of each band divided by the band dividing means is time-delayed according to the interval between the first and second omnidirectional microphones. Calculating means for mutually subtracting from the first and second audio signals via delay means for delaying and signal attenuating means for setting the level of the audio signal of each band respectively, and the audio of each divided band from the operating means And adding means for adding signals to each of the first and second audio signals via the gain control means, respectively, and adding the signals from the first audio signal of the audio signals of the respective bands divided by the band dividing means. The signal attenuating means and the gain controlling means in the predetermined band are controlled by a control signal corresponding to the level detected by detecting the difference component between the signal in the predetermined band and the signal in the predetermined band from the second audio signal.

Further, according to the stereo operation method of the present invention, the directional stereo sound is generated by using the first and second audio signals picked up by the first and second omnidirectional microphones arranged at arbitrary intervals. A stereo operation method for outputting a field signal, wherein the audio bands of the first and second audio signals are each divided into a plurality of bands, and the divided audio signals of the respective bands are respectively first and second omnidirectional. The time delay is delayed according to the interval of the characteristic microphones, the level of the audio signal in each band is set for each, and the first and second audio signals are added, and the added signals are first and second mutually. A level obtained by detecting a difference component between a signal of a predetermined band from the first audio signal and a signal of a predetermined band from the second audio signal among the divided audio signals of the respective bands which are subtracted from the second audio signal. Control signal according to the specified band Those obtained by controlling the level of the definitive sound signal.

Further, the stereo operation method of the present invention uses the first and second voice signals picked up by the first and second omnidirectional microphones arranged at arbitrary intervals to produce a directional stereo. A stereo operation method for outputting a sound field signal, wherein the audio bands of the first and second audio signals are each divided into a plurality of bands, and the audio signals of the respective divided bands are respectively divided into first and second audio signals. While delaying the time according to the interval between the directional microphones, setting the level of the audio signal of each band to each, subtracting from the first and second audio signals mutually, and subtracting the audio signal of each band The gain control is performed to add each of the first and second audio signals, and the divided audio signals of the respective bands are divided into a predetermined band signal from the first audio signal and a predetermined band from the second audio signal. The difference component from the signal Those obtained by controlling the level and gain of the audio signal in a predetermined band by the control signal in accordance with Le.

The present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing a configuration of an embodiment of a stereo operation circuit applied to a sound collecting device and a stereo operation method of the present invention.

In FIG. 1, the sound signals picked up from the first and second omnidirectional microphones 1R and 1L arranged at arbitrary intervals are supplied as a right channel input and a left channel input, respectively. These right channel and left channel audio signals are amplified by amplifiers 2R and 2R, respectively.
It is supplied to L to optimize the signal level. The signals from these amplifiers 2R and 2L are respectively supplied to the computing unit 3
It is supplied to the R and 3L addition terminals and is also supplied to the band dividing means 4R, 4L and 5R and 5L.

In the band dividing means 4R, 4L, a signal of band 1 corresponding to wind noise, for example, is extracted. That is, as for the frequency characteristic of the wind noise signal in a general video camera device, as shown in FIG. 2, for example, the level increases with the 1 / F characteristic (F is the frequency) as the frequency decreases from about 1 kHz. , The level of the extremely low frequency decreases due to the characteristics of the microphone unit used and the coupling capacitor of the analog circuit.
The characteristic has a peak near z.

Therefore, the band dividing means 4R, 4L and 5R, 5L described above perform band division with the frequency characteristics shown in FIG. 3, for example. That is, A of FIG. 3 is a case of band division by a bandpass filter, for example, a band 1 signal is extracted by the band division means 4R, 4L composed of a bandpass filter, and a band 2 signal is extracted by the band division means 5R, 5L. To be done. Here, the band 1 is about 2 above.
This corresponds to the band 1 that matches the frequency characteristic of the wind noise signal having a peak near 00 Hz.

Alternatively, FIG. 3B shows a case of band division by a low-pass filter. For example, a band 1 signal is extracted by the band-dividing means 4R, 4L composed of a low-pass filter, and the band-dividing means is output from the band-dividing means 5R, 5L. Four
The band 2 signal is extracted by subtracting the outputs of R and 4L. Also in this case, the band 1 is about 20
This corresponds to the band 1 which is matched with the frequency characteristics of the wind noise signal having a peak near 0 Hz. Band division is performed in this manner.

These band dividing means 4R, 4L and 5
The signals of the respective bands divided by R and 5L are, for example, according to the intervals of the microphones 1R and 1L, and delay devices 6R and 6L having a delay time corresponding to the time during which sound waves are transmitted between them.
And 7R and 7L. Further, these time-delayed audio signals are output to the variable attenuators 8R and 8L, respectively.
And the attenuators 9R and 9L so as to have arbitrary signal levels, and then supplied to the adders 10R and 10L, respectively, and added to the left and right channels.

Then, the signals from these adders 10R and 10L are mutually supplied to the subtraction terminals of the arithmetic units 3R and 3L. That is, the signal from the adder 10R for the right channel is supplied to the calculator 3L for the left channel, and the signal from the adder 10L for the left channel is supplied to the calculator 3R for the right channel.
Is supplied to. As a result, the calculators 3R and 3L perform matrix processing by subtracting the low frequency components of the channels from each other, and reproduce the stereo sound field in the low frequency band.

Further, the audio signals from these arithmetic units 3R and 3L are supplied to equalizers 11R and 11L, which are respectively composed of filter circuits, to adjust the frequency characteristics,
The audio signals passed through the equalizers 11R and 11L are taken out from the output terminals 12R and 12L as right channel output and left channel output, respectively. These output audio signals are supplied to an audio signal circuit (not shown) for subsequent processing such as recording and transmission.

In this circuit, the signal of band 1 extracted by the band dividing means 4R, 4L is further subtracted by the subtractor 13.
And the difference component of the signal of band 1 of the audio signals of the right channel and the left channel is output. The difference component is supplied to the level detecting means 14 to extract the envelope level thereof, and the information of the envelope level is supplied to the coefficient generating means 15 to generate the control coefficient corresponding to the envelope level. Then, the attenuation rates of the variable attenuators 8R and 8L described above are controlled according to this control coefficient.

That is, in this circuit, most of the output signal of the subtracter 13 which is the difference component of the right-channel and left-channel audio signals of the band 1 extracted by the band dividing means 4R, 4L has the non-correlation component. The wind noise signal is extracted and the level of wind noise is detected by extracting the envelope level. Then, for example, control is performed to increase the attenuation rate of the attenuators 8R and 8L when the level of wind noise is high, and to decrease the attenuation rate of the attenuators 8R and 8L when the level of wind noise is low.

Thus, in this circuit, a stereo operation circuit for emphasizing the left and right directivity of the audio signal is formed, and the attenuator 8 is used when the level of wind noise is high, for example.
The attenuation rates of R and 8L are increased, and the amount of signals subtracted by the arithmetic units 3R and 3L is decreased. With this, for example, when the level of wind noise is large, the stereo effect formed by the stereo operation circuit is weakened or stopped,
It is possible to solve the problem that wind noise is emphasized in the stereo operation circuit described in the above problem.

Therefore, in this embodiment, for example, by detecting the occurrence of wind noise and automatically controlling the parameters of the stereo operation processing, for example, even when the wind noise occurs, the wind noise component is generated due to the stereo operation processing. It is possible to eliminate the fear of being emphasized, and always enable good stereo sound collection even with a microphone provided closely in a small device.

As a result, in the conventional apparatus, there is a risk that the noise component of the wind noise generated in the random phase will be emphasized in the stereo operation circuit. There is a possibility that an extremely trouble will occur due to the relationship between the microphones and the microphones, and the distance between the microphones is further narrowed due to the miniaturization of products in recent years. According to the calculation method, these problems can be easily solved.

That is, according to the sound collecting device of the present invention,
And a band dividing unit for dividing the voice band of the second voice signal into a plurality of bands, and the voice signals of the respective bands divided by the band dividing unit according to the intervals between the first and second omnidirectional microphones, respectively. By means of delaying means for delaying the time and a signal attenuating means for setting the level of the audio signal of each band respectively, and an adding means for adding each of the first and second audio signals and a signal from the adding means to each other A first band signal and a second band signal from the first band signal of the respective band band signals divided by the band band dividing unit. By controlling the level of the signal attenuating means in the predetermined band with a control signal corresponding to the level detected by detecting the difference component from the signal in the predetermined band, the wind noise component is generated for stereo calculation processing even when wind noise occurs. To eliminate the fear of being adjusted, in particular at the microphone that is provided close to each other in small devices, in which always allow for good stereo sound pickup.

Further, according to the stereo operation method of the present invention, the audio bands of the first and second audio signals are respectively divided into a plurality of bands, and the audio signals of the respective divided bands are respectively divided into the first and second audio signals. Of the omnidirectional microphones, the levels of the audio signals in the respective bands are set respectively, the first and second audio signals are added, and the added signals are mutually added to the second. Subtracting from the first and second audio signals, the difference component between the signal of the predetermined band from the first audio signal and the signal of the predetermined band from the second audio signal among the divided audio signals of each band is calculated. By controlling the level of the audio signal in the predetermined band with the control signal according to the detected level, it eliminates the possibility that the wind noise component is emphasized due to the stereo calculation processing even when wind noise occurs, For small devices Even a microphone provided close you are, in which always allow for good stereo sound pickup.

Further, another embodiment of the stereo operation circuit applied to the sound collecting device and the stereo operation method of the present invention will be described with reference to FIG. In the following description, parts corresponding to those in the above-described embodiment of FIG. 1 will be assigned the same reference numerals and overlapping description will be omitted.

In FIG. 4, in this embodiment, the audio signals of the right channel and the left channel from, for example, the amplifiers 2R and 2L are converted into band dividing means 4R, 4L and 5R, 5L.
Supplied only to. And these band dividing means 4R,
The signals from 4L, 5R, and 5L are respectively calculated by the arithmetic unit 30.
It is supplied to the addition terminals of R, 30L and 31R, 31L. Further, variable attenuators 8R and 8L and attenuators 9R and 9L
Signals from the two calculators 30R, 30L and 31
It is supplied to the subtraction terminals of R and 31L.

That is, the signal from the variable attenuator 8L is supplied to the arithmetic unit 30R to which the signal from the band dividing means 4R is supplied, and the variable attenuator is supplied to the arithmetic unit 30L to which the signal from the band dividing means 4L is supplied. The signal from 8R is supplied. Further, the signal from the attenuator 9L is supplied to the arithmetic unit 31R to which the signal from the band dividing unit 5R is supplied, and the arithmetic unit 31L to which the signal from the band dividing unit 5L is supplied is from the variable attenuator 9R. Signal is supplied.

Further, the audio signals from the arithmetic units 30R, 30L and 31R, 31L are supplied to the variable gain units 32R, 32L and the gain units 33R, 33L, respectively, and have arbitrary signal levels. The signals from the variable gain units 32R and 32L and the gain units 33R and 33L are supplied to the adders 34R and 34L, respectively, and are added to the left and right channels, and the added audio signals are output from the output terminals 12R and 12L, respectively. It is taken out as the right channel output and the left channel output.

Further, the difference component of the signals of the band 1 of the audio signals of the right channel and the left channel from the above subtractor 13 is supplied to the level detecting means 14, and the information of the envelope level thereof is supplied to the coefficient generating means 15. Then, a control coefficient corresponding to the envelope level is generated. Then, according to this control coefficient, the attenuation rates of the variable attenuators 8R and 8L and the gains of the variable gain units 32R and 32L are controlled.

As a result, the computing units 30R, 30L and 31
In R and 31L, the matrix processing is performed by subtracting the signal components of the respective channels, and the stereo sound field is reproduced. Further, the level of wind noise is detected from the output signal of the subtractor 13, and when the level of wind noise is high, the attenuation rates of the variable attenuators 8R and 8L are increased, and when the level of wind noise is low, the variable attenuator is reduced. Control is performed to reduce the attenuation rate of 8R and 8L.

At the same time, in this circuit, variable gain units 32R and 32R are used instead of the equalizer of the conventional signal output section.
L and gain devices 33R and 33L are provided so that desired frequency characteristics can be obtained by these circuits, and a variable gain device to which a band 1 signal containing a lot of wind noise components is supplied is supplied. For example, the gains of 32R and 32L are reduced when the level of wind noise is high, and are increased when the level of wind noise is low.

Thus, in this circuit, a stereo operation circuit for emphasizing the left and right directivity of the audio signal is formed.
Further, for example, when the level of wind noise is large, the attenuator 8
The attenuation rates of R and 8L are increased, the amount of signals subtracted by the arithmetic units 3R and 3L is reduced, and the variable gain unit 3
The gains of 2R and 32L are reduced to reduce the band 1 signals of the right channel and left channel audio signals.

Thus, in this circuit, for example, when the level of wind noise is high, the stereo effect formed by the stereo operation circuit is weakened or stopped, and the problem that wind noise is emphasized in the stereo operation circuit is solved. At the same time, for example, when the level of wind noise is high, the signal in band 1 containing many wind noise components is reduced, and the influence of wind noise is further reduced.

Therefore, in this embodiment, by detecting the occurrence of wind noise and automatically controlling the parameters of the stereo operation processing, for example, even when the wind noise is generated, the wind noise component is generated due to the stereo operation processing. Eliminating the fear of being emphasized, and further, for example, when wind noise occurs, the signal in the band containing a lot of wind noise components is reduced. Can enable sound.

As a result, in the conventional apparatus, there is a risk that the noise component will be emphasized in the stereo operation circuit in the wind noise generated in a random phase. There is a possibility that an extremely trouble will occur due to the relationship between the microphones and the microphones, and the distance between the microphones is further narrowed due to the miniaturization of products in recent years. According to the calculation method, these problems can be easily solved.

That is, according to the sound collecting device of the present invention, the directional sound is obtained by using the first and second voice signals picked up by the first and second omnidirectional microphones arranged at arbitrary intervals. A stereo operation method for outputting a stereo sound field signal, wherein the audio bands of the first and second audio signals are each divided into a plurality of bands, and the audio signals of the respective divided bands are respectively divided into first and second audio signals. The time delay is delayed according to the interval of the omnidirectional microphones, the levels of the audio signals in the respective bands are set respectively, the first and second audio signals are added, and the added signals are mutually first And a difference component between a signal of a predetermined band from the first audio signal and a signal of a predetermined band from the second audio signal among the divided audio signals of each band, which is subtracted from the second audio signal. A predetermined band with a control signal according to the level By controlling the level of the definitive sound signals, to eliminate for example the risk of wind noise components for stereo processing even when wind noise generated from being emphasized,
Further, for example, when wind noise occurs, the signal in the band containing a lot of wind noise components is reduced, and it is possible to always achieve good stereo sound collection even with a microphone provided closely in a small device. Is.

Further, according to the stereo operation method of the present invention, the directional pattern is obtained by using the first and second audio signals picked up by the first and second omnidirectional microphones arranged at arbitrary intervals. Is a stereo operation method for outputting a stereo sound field signal, wherein the audio bands of the first and second audio signals are each divided into a plurality of bands, and the divided audio signals of the respective bands are divided into first and second audio signals. Of the omni-directional microphones, time delays are set, the levels of the audio signals in the respective bands are set respectively, and the subtracted sounds are mutually subtracted from the first and second audio signals. Gain control the signals respectively and add them for each of the first and second audio signals,
Predetermined by the control signal according to the level detected by detecting the difference component between the signal of the predetermined band from the first audio signal and the signal of the predetermined band from the second audio signal among the divided audio signals of each band By controlling the level and the gain of the audio signal in the band, for example, even when the wind noise occurs, the fear that the wind noise component is emphasized due to the stereo calculation processing is eliminated, and further, when the wind noise occurs, the wind noise component is reduced. It is possible to reduce the signal of a band included in a large amount and always enable good stereo sound collection even with a microphone provided closely in a small device.

In the embodiment of FIG. 4, variable gain units 32R and 3R are used instead of the equalizer of the conventional signal output unit.
2L and gain devices 33R and 33L are provided. For this reason, in the conventional equalizer, a high-order filter is usually used, which may deteriorate the phase characteristics and the SN ratio. However, in the configuration using the gain control function as described above, such characteristics are There is no fear of deterioration.
Further, for example, by making it possible to change the gain of each band independently, it is possible to have a function as a level equalizer, and it is possible to realize an optimum frequency characteristic.

Further, in the above-described embodiment shown in FIGS. 1 and 4,
Although the mode in which the band division is divided into two has been described, the number of band divisions is not limited to this, and for example, the wind noise band is further finely divided, and the attenuator and the gainer are controlled separately for each band. You may do it. By doing so, it is possible to perform a reduction process that is more comfortable in hearing.

Further, the applicant of the present application has previously proposed Japanese Patent No. 3186411 as a sound collecting device in which wind noise is reduced. In this patented invention, a wind noise component is extracted by adaptive processing, and The wind noise is reduced by subtracting the extracted wind noise component from the original signal, and the assumption is completely different from the wind noise reduced by controlling the stereo effect in the present invention. is there.

The present invention is not limited to the above-described embodiments, and various modifications can be made without departing from the spirit of the present invention.

[0049]

Therefore, according to the inventions of claims 1 and 3, by detecting the occurrence of wind noise and automatically controlling the parameters of the stereo operation processing, for example, even when wind noise occurs. By eliminating the fear that the wind noise component is emphasized due to the stereo operation processing, it is possible to always achieve good stereo sound collection even with a microphone provided in close proximity, especially in a small device. .

Further, according to the inventions of claim 2 and claim 4, for example, by detecting the occurrence of wind noise and automatically controlling the parameters of the stereo operation processing,
For example, when the wind noise is generated, the fear that the wind noise component is emphasized due to the stereo calculation processing is eliminated. Further, for example, when the wind noise is generated, the signal in the band including a large amount of the wind noise component is reduced, and the size is particularly small. Even with microphones provided close to each other in the device, good stereo sound pickup can be always performed.

As a result, in the conventional apparatus, there is a risk that the noise component of wind noise generated at random phases will be emphasized in the stereo operation circuit. There is a possibility that an extremely trouble will occur due to the relationship between the microphones and the microphones, and the distance between the microphones is further narrowed due to the miniaturization of products in recent years. According to the calculation method, these problems can be easily solved.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of an embodiment of a stereo operation circuit applied to a sound collecting device and a stereo operation method of the present invention.

FIG. 2 is a diagram for explaining the explanation.

FIG. 3 is a diagram for explaining the explanation.

FIG. 4 is a block diagram showing a configuration of another embodiment of a stereo operation circuit applied to the sound collecting device and the stereo operation method of the present invention.

FIG. 5 is a block diagram showing a configuration of a conventional stereo operation circuit.

FIG. 6 is a diagram for explaining the explanation.

[Explanation of symbols]

1R, 1L ... First and second omnidirectional microphones, 2
R, 2L ... Amplifier, 3R, 3L ... Operator, 4R, 4L and 5R, 5L ... Band dividing means, 6R, 6L and 7R, 7
L ... Delay device, 8R, 8L ... Variable attenuator, 9R, 9L ... Attenuator, 10R, 10L ... Adder, 11R, 11L ... Equalizer, 12R, 12L ... Output terminal, 13 ... Subtractor, 1
4 ... Level detecting means, 15 ... Coefficient generating means, 30R, 3
0L and 31R, 31L ... arithmetic unit, 32R, 32L ... variable gain unit, 33R, 33L ... gain unit, 34R, 34L ...
Adder

Claims (4)

[Claims] 1. A first voice signal from the first omnidirectional microphone and first and second omnidirectional microphones arranged at arbitrary intervals, and the second omnidirectional microphone. Is a sound collecting device provided with a stereo operation circuit for outputting a directional stereo sound field signal by using the second sound signal from the first sound signal and a plurality of sound bands of the first and second sound signals, respectively. Band dividing means for dividing the band into a plurality of bands, delay means for delaying the audio signal of each band divided by the band dividing means according to the interval between the first and second omnidirectional microphones, and An adding unit that adds each of the first and second audio signals via a signal attenuating unit that sets the level of the audio signal to each other, and the signals from the adding unit to the first and second audio signals. Subtract from A signal of a predetermined band from the first audio signal and a signal of a predetermined band from the second audio signal of the audio signals of each band divided by the band dividing unit. The sound collecting device is characterized in that the level of the signal attenuating means in the predetermined band is controlled by a control signal according to the level obtained by detecting the difference component of. 2. A first audio signal from the first omnidirectional microphone and a second omnidirectional microphone, which have first and second omnidirectional microphones arranged at arbitrary intervals. Is a sound collecting device provided with a stereo operation circuit for outputting a directional stereo sound field signal by using the second sound signal from the first sound signal and a plurality of sound bands of the first and second sound signals, respectively. Band dividing means for dividing the band into a plurality of bands, delay means for delaying the audio signal of each band divided by the band dividing means according to the interval between the first and second omnidirectional microphones, and Arithmetic means for mutually subtracting from the first and second audio signals via signal attenuating means for setting the level of the audio signal respectively; and gains for the audio signals of the respective divided bands from the arithmetic means. A predetermined addition from the first audio signal of the audio signals of the respective bands divided by the band dividing means, and an adding means for adding each of the first and second audio signals via the control means. The signal attenuating means and the gain controlling means in the predetermined band are controlled by a control signal corresponding to a level obtained by detecting a difference component between a signal in the band and a signal in the predetermined band from the second audio signal. Sound collection device. 3. A stereo operation for outputting a directional stereo sound field signal using first and second audio signals picked up by first and second omnidirectional microphones arranged at arbitrary intervals. A method, wherein the audio bands of the first and second audio signals are each divided into a plurality of bands, and the audio signals of each of the divided bands are respectively separated by the interval between the first and second omnidirectional microphones. According to the above, the level of the audio signal of each band is set to each, and added for each of the first and second audio signals, and the added signals are mutually added to the first and second audio signals. Of the divided audio signals of the respective bands, and detects a difference component between the signal of the predetermined band from the first audio signal and the signal of the predetermined band from the second audio signal. With a control signal according to the level Stereo calculation method characterized by controlling the level of the audio signal in a predetermined band. 4. A stereo operation for outputting a directional stereo sound field signal using first and second audio signals picked up by first and second omnidirectional microphones arranged at arbitrary intervals. A method, wherein the audio bands of the first and second audio signals are each divided into a plurality of bands, and the audio signals of each of the divided bands are respectively separated by the interval between the first and second omnidirectional microphones. According to the above, the level of the audio signal in each band is set to each and subtracted from the first and second audio signals mutually, and the subtracted audio signals in each band are respectively gain controlled. And adding for each of the first and second audio signals, a predetermined band signal from the first audio signal and a predetermined signal from the second audio signal among the divided audio signals of the respective bands. Detects the difference component with the band signal Stereo calculation method characterized by controlling the level and the gain of the audio signal in the predetermined band by a control signal corresponding to the level.