JPH07240991A - Microphone device - Google Patents

Abstract

PURPOSE:To provide ultra sharp directivity without being influenced by dark noise or the like even with a small-sized constitution. CONSTITUTION:An adaptive microphone part 1 subtracts adaptive processing signals obtained by adaptively processing audio signals from a microphone M for reference input from the microphone M1 for main input and performs an adaptive processing so as to minimize the power of the subtracted output. An energy calculation part 6 segments the output signals of a sharp directional microphone M3 and the adaptive microphone part 1 by windows provided with a fixed length at all times and calculates the total sum of energy in the segmented section. A comparator 3 compares the energy of the output signals of the sharp directional microphone M3 and the adaptive microphone part 1 and supplies information for indicating which output signals are the ones of smaller energy to a changeover switch 4. A changeover switch part 4 outputs the output signals provided with the smaller energy in the segmented window section from an output terminal 5 corresponding to the information supplied from the comparator 3.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a microphone device capable of obtaining super directivity.

[0002]

2. Description of the Related Art Conventionally, unnecessary noise such as background sound is reduced,
In order to collect only the desired sound, a microphone having directivity such as unidirectionality is often used. In addition, an adaptive noise canceler (AN)
It is considered to apply C) to eliminate unnecessary noise and obtain only desired voice.

A microphone device to which this adaptive noise canceller is applied (hereinafter referred to as an adaptive processing microphone device) will be described with reference to FIG. The main input signal input through the main input terminal 51 is supplied to the adding circuit 53 via the delay circuit 52. In addition, the reference noise signal n ₁ input through the reference input terminal 54
Is supplied to the adder circuit 53 via the adaptive filter circuit 55 and subtracted from the signal from the delay circuit 52. The subtracted output of the adder circuit 53 is fed back to the adaptive filter circuit 55 and is also led to the output terminal 56.

The main input signal is a signal in which a desired signal S and a noise signal n ₀ which is uncorrelated with the desired signal S are added. On the other hand, the reference input noise signal n ₁ is uncorrelated with the desired signal S, but is correlated with the noise signal n ₀ .

The adaptive filter circuit 55 filters the reference input noise signal n ₁ and outputs a signal y that approximates the noise signal n ₀ . This signal y is subtracted from the main input signal in the adder circuit 53. Therefore, the signal derived from the output terminal 56 approximates to the desired signal S. The delay circuit 52 is inserted in order to compensate for the time delay required for the adaptive processing operation in the adaptive filter circuit 55, the propagation time of the filter, and other delays. Here, the main input signal supplied to the main input terminal 51 is picked up by the main input microphone 57. The reference input noise signal n ₀ supplied to the reference input terminal 54 is picked up by the reference input microphone 58.

In this way, the adaptive processing microphone device can obtain a sharp directivity, and can obtain a signal extremely close to the desired voice as an output, and has a good reception substantially equal to or higher than that of the superdirectional microphone. The sound quality can be realized as a small device instead of a large device.

[0007]

By the way, even if the adaptive processing microphone device as described above is designed so that the sound pressure-frequency characteristic becomes flat in a free sound field such as outdoors, the microphone is still used. When sound is picked up in a reverberant sound field (reverberant sound field or diffuse sound field) such as in a room, the sound is picked up as if the low-pitched sound such as surrounding voice or noise is expanded
It is known to be difficult to hear.

As described above, the sound field to be used is diffuse, the sound is increased, and the background noise level, which is the total noise of the system existing regardless of the presence or absence of the transmission signal, is equal to or higher than the level of the desired input voice. When the size becomes, the property of the adaptive processing microphone device is that the rejection rate of unnecessary noise is limited, and it becomes difficult to obtain good sound collection quality.

In particular, the adaptive processing microphone device may operate to amplify the background noise itself in an environment where background noise is incident. In addition, the directivity of the output may be significantly impaired due to the influence of background noise.

That is, depending on various situations in the real environment such as background noise, the adaptive processing microphone device interferes with its original operation, resulting in a loss of directivity or amplification of background noise. Sometimes.

The present invention has been made in view of the above circumstances, and it is possible to reduce a large amount of the background noise or the like even in an environment where the background noise is incident or in a diffuse sound field or a reverberant sound field. It is also an object of the present invention to provide a microphone device having a high sound collection quality and being able to obtain a large reduction amount even for noise from directions other than the desired direction.

[0012]

A microphone device according to the present invention subtracts an adaptive processed signal obtained by adaptively processing a voice signal from a reference input microphone from a main input microphone and outputs the subtracted output power. Adaptive microphone means for performing adaptive processing so as to minimize, directional microphone means having a fixed directivity, level detecting means for detecting the level of the output signal of the adaptive microphone means and the directional microphone means, Comparing means for comparing the levels of the output signals of the adaptive microphone means and the directional microphone means detected by the level detecting means, and the adaptive having the smaller output signal level depending on the comparison result of the comparing means. Switching means for switching and outputting the output of the microphone means or the directional microphone means. To solve the above problem by.

In this case, the directional microphone means may have a fixed directional characteristic by the mixed output of the main input and the reference input.

The switching means may select and output the output of the adaptive microphone means or the directional microphone means according to the background noise detection output and the level detection comparison output.

The background noise detection output is detected by the background noise detection means. Background noise is noise having a random waveform that exists regardless of the presence or absence of a transmission signal, which is particularly noticeable in a reverberant sound field or a diffuse sound field. Background noise detection means, in the reverberation sound field or diffuse sound field, in the case of an omnidirectional microphone, in the case of a directional microphone, to detect the background noise using the property that the frequency characteristics change. There is. That is, in the case of an omnidirectional microphone, the frequency characteristic hardly changes from the free sound field even in a reverberant sound field or a diffuse sound field, but in the case of a directional microphone, in a reverberant sound field or a diffuse sound field. It takes advantage of the characteristic that the sound field sensitivity is high in the low range.

Further, the level detecting means cuts out each output signal corresponding to each directional characteristic of the adaptive microphone means and the directional microphone means at any time with a window of a constant length, and calculates the total energy in the cut-out section. You may do it. In this case, the windows used include various windows such as a directional window, a Hanning window, and a Hamming window. Then, the comparison means compares the total energy of the cutout sections, and the selection means responds to the comparison result of the comparison means.
One of the output signals having the smallest output signal level is selectively output to the cutout section.

[0017]

The comparing means compares the levels of the respective output signals of the adaptive microphone means and the directional microphone means detected by the level detecting means in the cut-out section of a fixed length, and the selecting means outputs the smaller one of the outputs depending on the result of the comparison. Since one of the output signals having the signal level is selected and output, the voice with the least noise from directions other than the desired direction can be output for each of the fixed length cutout sections. Further, even in an environment where background noise is incident, a large reduction amount can be obtained for the background noise, and
A large amount of reduction can be obtained for noise from directions other than the desired direction, and superdirectivity can be obtained with a very small configuration.

[0018]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of a microphone device according to the present invention will be described below with reference to the drawings. First, in the first embodiment, as shown in FIG. 1, a main input microphone M ₁ for collecting desired voice and a reference input microphone M ₂ for collecting voice in a direction to be removed as noise are provided. , The adaptive processed signal obtained by adaptively processing the audio signals from the microphone M ₃ having a fixed directivity and the reference input microphone M _{2 is} subtracted from the main input microphone M _1, and the power of the subtracted output An adaptive processing unit 1 that performs adaptive processing so as to minimize, and an energy calculation unit 6 that calculates the energy of the filter processing output of the adaptive processing unit 1 and the output energy of the microphone M ₃ having a fixed directivity. According to the result of the energy comparison by the comparator 3 that compares the energy calculated by the energy calculating unit 6, the output with the smaller energy is selected, and the output signal of that output signal is selected. Changeover switch unit 4 for outputting from the output terminal 5
And a microphone device including.

Here, the main input microphone M ₁ and the reference input microphone M ₂ have the same main axis. Further, the main axes of the microphone M ₃ having a fixed directivity are also the same as those main axes.

The adaptive processing unit 1 is an adaptive filter circuit 8 to which the audio signal from the reference input microphone M ₂ is supplied.
And a subtraction circuit 7 for subtracting the output signal of the adaptive filter circuit 8 from the desired voice signal of the main input microphone M _1.
And have. The output signal of the subtraction circuit 7 is fed back to the adaptive filter circuit 8 and the changeover switch 4
And the energy calculation unit 6.

In the microphone device of the first embodiment, the main input microphone M ₁ is an omnidirectional microphone capable of picking up sound in all directions with substantially uniform sensitivity. Further, the reference input microphone M ₂ is a unidirectional microphone having sensitivity in the back direction which is an angle of 180 ° when the arrival direction of the desired audio signal is 0 °.

The adaptive filter circuit 8 is, as described later,
The reference input voice is controlled so as to approximate the voice as noise included in the main input voice. As a result, assuming that the desired voice in the voice picked up by the main input microphone M ₁ and the noise are uncorrelated, the subtraction circuit 7:
The noise signal picked up by the reference input microphone M ₂ is subtracted and removed from the audio signal from the main input microphone M _1, and only the desired signal is obtained from the subtraction circuit 7.

That is, in this configuration, the output signal of the main input microphone M ₁ is supplied as the main input, and the reference input microphone M ₂ is supplied as noise as the reference input.
It has a configuration to which an adaptive noise canceller supplied with the output signal of is applied. Hereinafter, the adaptive processing unit 1 to which this adaptive noise canceller is applied will be referred to as an adaptive microphone unit 1.

In this case, the main input voice from the main input microphone M ₁ is, as shown in FIG. 2, the desired voice signal S from the front direction of the arrow AR and the back voice direction which is considered to be uncorrelated with the desired voice signal S. It is the sum of the noise n ₀ which is a voice signal. On the other hand, the reference input voice signal n ₁ from the reference input microphone M ₂ is uncorrelated with the desired voice signal, but correlated with the noise n ₀ . The adaptive filter circuit 8 filters the reference input audio signal n ₁ and outputs the signal y, and the adaptive algorithm works to minimize the subtraction error e which is the output of the subtraction circuit 7.

Now, the desired input signal S, the noise signal n ₀ , the reference input voice signal n ₁ and the output signal y are statistically stationary. Assuming that the average value of these signals is 0, the residual output e is e = S + n ₀ −y (1). The (1) the expected value E [e ^2] Although the squared residuals output e of the desired input signal S is a noise signal n _0, The output signal y and because it is uncorrelated, E [e ² ] = become ^{E [S 2] + E [} (n 0 -y) 2 + 2E [S (n 0 -y)] = E [S 2] + E [(n 0 -y) 2] ··· (2) . Assuming that the adaptive filter circuit 8 converges,
The adaptive filter circuit 8 updates the adaptive filter coefficient so that E [e ² ] is minimized. At this time, since E [S ² ] is not affected, its minimum value E _min
[E ^{2] _{is, E min [e 2] =}} E [S 2] + E min [(n 0 -y) 2] a (3).

Here, minimizing E [e ² ] means minimizing E [(n ₀ -y) ² ]. Therefore, the filter output y is the best least-squares estimate of the noise signal n ₀ . When ^{_{E [(n 0 -y) 2}} ] is minimized, E [(e-S) 2] it is also minimized.
This is because e−S = n ₀ −y. That is, adjusting the filter to minimize the total output power is equivalent to the residual output e being the best least-squares estimate of the desired signal S.

The residual output e is generally the desired signal S with some noise remaining, but since the output noise is given by n ₀ -y, E [(n-y) ² ] should be minimized. Is equivalent to maximizing the output signal-to-noise ratio.

Next, a concrete example of the adaptive filter circuit 8 is shown in FIG.
Shown in. In this specific example, as shown in FIG. 3, an FIR filter type adaptive linear combiner 60 is used. The adaptive linear combiner 60 includes a plurality of delay circuits 61 ₁ , each having a delay time Z ⁻¹ of a unit sampling time.
61 ₂ , ... 61 _m (m is an integer), the reference input noise signal n ₁ and the weighting circuit 62 for multiplying the output signals of the delay circuits 61 ₁ , 61 ₂ , ... 61 _{m by} the weighting coefficient. ₀ , 62 ₁ , 6
2 ₂ , ... 62 _m and weighting circuits 62 ₀ , 62 ₁ , 62 ₂ ,
... Addition circuit 63 that adds the outputs of 62 _m . The output signal y of the adder circuit 63 is supplied to the subtractor circuit 7 of FIG. 1 via the output terminal 64.

Weighting circuits 62 ₀ , 62 ₁ , 62 ₂ , ... 6
The weighting coefficient supplied to 2 _m is the residual signal e supplied from the subtraction circuit 7 shown in FIG. 1 via the input terminal 66 in the LMS arithmetic circuit 65 composed of, for example, a microcomputer.
Or according to the reference input audio signal n ₁ . This L
The algorithm executed by the MS arithmetic circuit 65 is as follows.

Now, the input vector X _k at time k is
As shown in FIG. 3, X _k = [x _0k , x _1k , x _2k ... x _mk ] ^T (4), the output is y _k , and the weighting coefficient is w _jk (j = 0, 1, 2,
..M), the relationship between input and output is

[0031]

[Equation 1]

It becomes

Then, the weight vector W _{k at} time _k
Is defined as W _k = [w _0k , w _1k , w _2k ... w _mk ] ^T (6), the input-output relationship shown in the above equation (5) is y _k = X _k ^T · W _k is given by (7). If the desired response is d _k , the difference output (residual output) e _k becomes e _k = d _k −y _k = d _k −X _k ^T · W _k (8)

In the LMS (least mean square) method, the weight vector is updated by the formula W _{k + 1} = W _k +2 μ · e _k · X _k (9). Here, μ is a gain factor (step gain) that determines the speed and stability of adaptation.

In this way, the subtraction circuit 7 obtains an audio signal mainly consisting of the desired audio signal, from which noise, in this example, the audio signal from the back side is removed.

As described above, the adaptive filter circuit 8 operates so as to minimize the output power by updating the weight vector by the method as described above.

By the way, in order to reduce the noise in the main input by using the reference input by the adaptive processing as described above, it is necessary that the desired speech and the reference noise are uncorrelated. Therefore, conventionally, in this type of adaptive noise canceller, a desired voice is not picked up as a reference input,
Some measures have been taken, such as elaborating soundproofing on the reference input microphone, installing it as close as possible to the noise source, and keeping it away from the main input microphone. However, this makes the system large and difficult to move.

On the other hand, in the microphone device of the first embodiment, the desired voice and the noise are distinguished according to the direction of arrival of the voice. And the main input microphone M
₁ is omnidirectional so as to have a directional characteristic capable of picking up a voice from a desired voice arrival direction, and the reference input microphone M ₂ has no sensitivity in the desired voice arrival direction or has a low sensitivity. So that the desired voice in the voice picked up by the main input microphone M ₁ and the noise picked up by the reference input microphone M ₂ are uncorrelated. To do.

Next, the microphone M ₃ having a fixed directivity is a microphone having a fixed directivity relatively close to the super directivity. In the first place, what is required for a fixed directional microphone is a so-called sharp directional characteristic that has sufficient sensitivity in the forward direction (desired direction), but the lower the sensitivity in the other directions, the better. In this first embodiment,
As the microphone M ₃ having the fixed directivity, for example, a sharp directivity microphone such as a hyper cardioid type microphone or a super cardioid type microphone is used. Hereinafter, this microphone M _{3 is referred} to as a sharp directivity microphone M ₃ .

The output signals of the sharp directivity microphone M ₃ and the adaptive microphone section 1 are supplied to the energy calculation section 6 and the changeover switch section 4. Hereinafter, the energy calculation unit 6, the changeover switch unit 4, and the comparison unit 3 will be described.

Energy calculation unit 6 which is a level detection unit
Outputs the output signals of the sharp directivity microphone M ₃ and the adaptive microphone unit 1 at any time with a window having a fixed length, and calculates the total energy in the cut-out section. This total energy is the sum of all squared values of each data sample in the cutout section. As the window used in this case, various windows such as a rectangular window, a Hanning window and a Hamming window can be used. The window length is, for example, about 42 mse.
It is a very short time like c.

In the comparator 3, the sharp directivity microphone M ₃ calculated by the energy calculation unit 6 in the cutout window section having a constant length is used.
And the power of the output signal of the adaptive microphone unit 1 are compared with each other, and information indicating which output signal has the smaller energy is supplied to the changeover switch 4.

The changeover switch section 4 responds to the above information supplied from the comparator 3 within the cutout window section,
The output signal having the smaller energy is output from the output terminal 5.

Here, the changeover switch section 4 selects the output signal having a small energy because the output signal having a small energy does not contain a noise component. As the noise signal, in addition to the background noise, there is background noise that is noticeable in a diffuse sound field or a reverberant sound field regardless of the presence or absence of a transmission signal. This background noise has a random waveform,
For example, there is no correlation when the case where sound is picked up in two directions is considered. Therefore, if the adaptive microphone unit 1 alone attempts to reduce the overall noise, the subtraction circuit 7 adds background noise having a random waveform. For this reason,
In the first embodiment, the energy levels of the output signals are compared, and the output signal with the smaller energy level (the output signal of the adaptive microphone unit 1 or the microphone M ₃ ) is output.

Therefore, the microphone device of the first embodiment can output the sound with the noise and background noise from the directions other than the desired direction most reduced for each of the above-mentioned cut-out sections of a certain length. Therefore, the first embodiment has a small structure, but is sharp even when the directivity of the output of the adaptive microphone unit 1 is significantly impaired or the background noise is amplified due to the influence of background noise. A directional microphone can temporarily output a signal having a sharp directivity, and superdirectivity can be obtained.

Next, the second embodiment will be described with reference to FIG. The second embodiment has an adaptive microphone unit 1 similar to the configuration of the first embodiment shown in FIG.
The energy calculation unit 6, the comparator 3, and the changeover switch 4 are provided, but the sharp directivity microphone M ₃ shown in FIG. 1 is not provided. A directional microphone unit 9 is provided instead of the sharp directional microphone M ₃ .

This directional microphone unit 9 is the same as the _one that the main input from the main input microphone M ₁ and the reference input from the reference input microphone M ₂ are mixed to output from the sharp directional microphone M _3. I am getting such an output signal. That is, the subtraction circuit 11 subtracts the reference input amplified by the amplifier 10 from the main input.

The reference input picked up by the reference input microphone M ₂ is the angle 1 with respect to the arrival direction of the desired audio signal.
It is a noise signal from the back direction of 80 °. By subtracting a value obtained by multiplying the noise signal by -1.5 in the amplifier 10 from the desired voice signal which is the main input, an output signal having a polar pattern as shown in FIG. 5 can be obtained. The polar pattern as shown in FIG. 5 represents sharp directivity, and the directional microphone unit 9 of the second embodiment produces an output equivalent to that of the sharp directional microphone ₃ used in the first embodiment. You know you can get it.

The output signals of the directional microphone unit 9 and the adaptive microphone unit 1 are supplied to the energy calculation unit 6 and the changeover switch 4.

The energy calculation unit 6 cuts out the output signals of the directional microphone unit 9 and the adaptive microphone unit 1 at any time using a window having a fixed length, and calculates the total energy in the cut-out section. . This total energy is the sum of all squared values of each data sample in the cutout section. As the window used in this case, a rectangular window, as in the first embodiment,
Various windows such as a Hanning window and a Hamming window can be used.

The comparator 3 compares the powers of the output signals of the directional microphone unit 9 and the adaptive microphone unit 1 calculated by the energy calculation unit 6 in the cut-out window section of a constant length, and the output signal of the smaller energy is compared. The information indicating which is which is supplied to the changeover switch 4.

The change-over switch section 4 responds to the above information supplied from the comparator 3 within the cutout window section,
The output signal having the smaller energy is output from the output terminal 5.

The effect of the second embodiment when the background noise is incident will be described below with reference to FIG. This Figure 6
In the situation where the female voice comes from the desired direction, while another female voice and male voice come from the side and the back as noise component voices, respectively, the background noise of the same level as each voice is further added. It is the figure which has shown the effect. A on the horizontal axis indicates the case of only the adaptive microphone unit 1, and B indicates the effect of the second embodiment. The triangular points connected by the straight line S _N indicate the noise reduction amount from directions other than the desired voice arrival direction. The black dots connected by the straight line S _B indicate the amount of background noise reduction. Regarding the reduction of noise sound, a considerable reduction amount can be obtained only with the adaptive microphone unit 1, but the reduction of background noise is increased and is rather increased. In contrast to this, in the second embodiment, the background noise can be remarkably reduced, and the noise voice reduction amount is also higher than that of the adaptive microphone unit 1 alone.

Therefore, the microphone device of the second embodiment can output the voice with the noise and background noise from the directions other than the desired direction most reduced for each cut-out section of the constant length. Therefore, although the second embodiment has a smaller size, even if the directivity of the output of the adaptive microphone unit 1 is significantly impaired or the background noise is amplified due to the effects of background noise, A signal having sharp directivity can be temporarily output from the directivity micron unit 9, and superdirectivity can be obtained.

Next, a third embodiment will be described with reference to FIG. The third embodiment has an adaptive microphone unit 1 similar to the configuration of the second embodiment shown in FIG.
The directional microphone unit 9, the energy calculating unit 6, the comparator 3, and the changeover switch 4 are provided, and further, a comparator 12 that plays a role of detecting a background noise level is provided.

The comparator 12 is a microphone M for main input.
The main input from ₁ and the reference input from the reference input microphone M ₂ are compared, and the level of background noise is detected from the comparison result. That is, the comparator 12 is background noise detection means. Background noise is noise having a random waveform that exists regardless of the presence or absence of a transmission signal, which is particularly noticeable in a reverberant sound field or a diffuse sound field. The frequency characteristic of the main input microphone M ₁ which is an omnidirectional microphone in the reverberant sound field or the diffuse sound field is almost the same as that in the case of the free sound field, but the frequency characteristic of the reference input microphone M ₂ which is a directional microphone. The frequency characteristic in the reverberant sound field or the diffuse sound field is high in the low frequency range. By utilizing such a property, the comparator 12 detects the background noise. Specifically, when the comparator 12 detects that the level of the output signal supplied from the reference input microphone M ₂ is high, the comparator 12 informs that the changeover switch 4 is in an environment with a lot of background noise. To supply.

Then, the changeover switch 4 which has received the high level information of the background noise adjusts the changeover to the selection of the output signal from the directional microphone section 9.

On the other hand, when the comparator 12 does not detect a high level background noise, the changeover switch 4 selectively outputs the output signal from the adaptive microphone section 1.

Here, the energy calculation unit 6, the comparator 3,
The description of the operation of the changeover switch 4 is omitted.
The configuration around the changeover switch 4 may be a specific example as shown in FIG. That is, in FIG. 8, the state detection unit 24 is caused to judge the detection signal from the comparator 12 and the detection signal from the comparator 3 to control the switching of the changeover switch 4. However, this state detector 2
4 performs an operation of giving priority to the comparison result of the comparator 12 even when the output energy of the adaptive microphone unit 1 is smaller than that of the directional microphone unit 9 when the background noise is large.

A concrete example of the inside of the changeover switch 4 used in the third embodiment is shown in FIG. According to this FIG. 9, when the background noise level is high, the comparator 1
Since the switch SW ₂ controlled by the detection result of ₂ can be prioritized, the output of the directional microphone unit 9 can be output from the output terminal 5.

Further, the changeover switch 4 may be a concrete example as shown in FIG. Also in this case, the switch S
Priority is given to the switching of W ₂ .

As described above, in the third embodiment, in the case where the directional characteristic of the output of the adaptive microphone unit 1 is remarkably impaired or the background noise is amplified due to the influence of background noise, etc., even though the structure is small. But the directional Micron part 9
Can temporarily output a signal having sharp directivity, and superdirectivity can be obtained.

The microphone device according to the present invention is
The present invention is not limited to the first to third embodiments described above. For example, the comparator may compare not only the total energy of the output signals but also the amplitudes, peak values, etc. of the output signals. The level detecting means may detect the level of each output signal instantaneously.

The microphone device described above is not only used as a sound pickup microphone device of a camera-integrated VTR, but also applicable to all microphone devices such as a commercial video camera and a measurement microphone device.

[0065]

In the microphone device according to the present invention, the adaptive processed signal obtained by adaptively processing the audio signal from the reference input microphone is subtracted from the main input microphone to minimize the power of the subtracted output. Adaptive microphone means for performing adaptive processing, directional microphone means having a fixed directivity, level detecting means for detecting the levels of the output signals of the adaptive microphone means and the directional microphone means, and the level detection Comparing means for comparing the levels of the output signals of the adaptive microphone means and the directional microphone means detected by the means, and the adaptive microphone means having the smaller output signal level according to the result of the comparison by the comparing means or the above. Since it has switching means for switching and outputting the output of the directional microphone means, The most reduced audio noise from directions other than the direction can be outputted for each cut section of the fixed length. Further, even in an environment where background noise is incident, a large reduction amount can be obtained for the background noise, and a large reduction amount can be obtained for noise from directions other than the desired direction. Therefore, superdirectivity can be obtained with a very small configuration.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of a microphone device according to a first embodiment of the present invention.

FIG. 2 is a diagram showing directivities of a main input microphone and a reference input microphone used in an adaptive microphone unit of the microphone device shown in FIG.

FIG. 3 is a block diagram of an adaptive filter circuit of an adaptive microphone unit.

FIG. 4 is a block diagram of a microphone device according to a second embodiment of the present invention.

5 is a diagram showing the directivity of a directional microphone unit 9 of the microphone device of the second embodiment shown in FIG.

FIG. 6 is a diagram for explaining the effect of the second embodiment.

FIG. 7 is a block diagram of a microphone device according to a third embodiment of the present invention.

FIG. 8 is a block diagram showing a specific example around a changeover switch of the microphone device of the third embodiment shown in FIG.

FIG. 9 is a block diagram showing a specific example of a changeover switch.

FIG. 10 is a block diagram showing a specific example of a switching style.

FIG. 11 is a block diagram showing a configuration of a conventional adaptive microphone device.

[Explanation of symbols]

 1 Adaptive Microphone Section 3 Comparator 4 Changeover Switch 6 Energy Calculation Section 7 Subtraction Circuit 8 Adaptive Filter Circuit 9 Directional Microphone Section

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Ikue Akiba 6-735 Kitashinagawa, Shinagawa-ku, Tokyo Sony Corporation

Claims (4)

[Claims] 1. An adaptive microphone that subtracts an adaptive processed signal obtained by adaptively processing a voice signal from a reference input microphone from a main input microphone and performs adaptive processing so as to minimize the power of the subtracted output. Means, a directional microphone means having a fixed directivity, a level detecting means for detecting the levels of the output signals of the adaptive microphone means and the directional microphone means, and the adaptive microphone means detected by the level detecting means. And comparing means for comparing the levels of the output signals of the directional microphone means, and switching between the outputs of the adaptive microphone means or the directional microphone means having the smaller output signal level depending on the comparison result of the comparing means. A microphone device having a switching means for outputting the output. 2. The microphone device according to claim 1, wherein the directional microphone means has a fixed directional characteristic due to a mixed output of a main input and a reference input. 3. The switching means selects and outputs the output of the adaptive microphone means or the directional microphone means according to the background noise detection output and the level detection comparison output. 1. The microphone device according to 1. 4. The level detecting means cuts out each output signal corresponding to each directional characteristic of the adaptive microphone means and the directional microphone means at any time with a window of a constant length, and calculates the total energy in the cut-out section. The microphone device according to claim 1, wherein: