Microphone device

Abstract

We provide a microphone device that allows adjustment of the sharpness of its directivity. [Solution] The microphone device 1 comprises two front microphones provided in the front row along a first direction, one rear microphone provided in the rear row behind the front row, a frequency conversion unit that converts the signals obtained from the two front microphones and the rear microphone into the frequency domain, a front extraction processing unit that extracts signals from frequency bins where the phase difference between at least one signal from the two front microphones and the signal from the rear microphone is positive, a front extraction processing unit that extracts signals from frequency bins where the phase difference between one signal from the two front microphones and the other signal falls within a predetermined range, an output spectrum calculation unit that obtains an output spectrum based on the extraction results of the front extraction processing unit and the front extraction processing unit, and an inverse frequency conversion unit that converts the output spectrum into a time domain signal, wherein the front extraction processing unit sets a predetermined range based on a parameter indicating the sharpness of the directivity.

Description

[Technical Field]

[0001] This invention relates to a microphone device. [Background technology]

[0002] Patent Document 1 discloses a sound acquisition device that calculates the ratio of the amplitude spectrum of the sound in the target area to the amplitude spectrum of the input signal of the microphone array for each frequency, then uses the sum of the amplitude spectrum ratios obtained by adding the amplitude spectrum ratios for each frequency and the sum of the coherence values ​​obtained by adding the coherence values ​​for each frequency to determine whether or not sound is present in the target area, and outputs the extracted sound from the target area if sound is present. [Prior art documents] [Patent Documents]

[0003] [Patent Document 1] Patent No. 6065030 [Overview of the Initiative] [Problems that the invention aims to solve]

[0004] However, in the invention described in Patent Document 1, the sharpness of the directivity depends on the arrangement of the microphone array, and it is not possible to adjust the sharpness of the directivity afterward.

[0005] This invention has been made in view of these circumstances, and aims to provide a microphone device that can adjust the sharpness of its directivity retrospectively. [Means for solving the problem]

[0006] To solve the above problems, the microphone device according to the present invention comprises, for example, two front microphones provided in a front row along a first direction, one rear microphone provided in a rear row along the first direction and behind the front row, a frequency conversion unit that converts the signals obtained by the two front microphones and the signals obtained by the rear microphone into the frequency domain, and a front extraction unit that extracts the signal of the frequency bin in which the phase difference between the signal obtained by at least one of the two front microphones and the signal obtained by the rear microphone is positive, from the signal converted into the frequency domain by the frequency conversion unit. The device comprises an output processing unit, a front extraction processing unit that extracts signals from the frequency bins of the signal converted to the frequency domain by the frequency conversion unit, wherein the phase difference between the signal obtained by one of the two front microphones and the signal obtained by the other is within a predetermined range, an output spectrum calculation unit that obtains an output spectrum based on the results extracted by the front extraction processing unit and the front extraction processing unit, an inverse frequency conversion unit that converts the output spectrum into a time-domain signal, and a receiving unit that receives parameters indicating the sharpness of directivity, wherein the front extraction processing unit is characterized by varying the predetermined range based on the parameters received by the receiving unit.

[0007] According to the microphone device of the present invention, when extracting the signal from a frequency bin whose phase difference between the signal obtained by one of the two front microphones and the signal obtained by the other falls within a predetermined range, the predetermined range is varied based on parameters received by the receiving unit. This allows for retrospective adjustment of the directivity sharpness. Furthermore, the adjustment of directivity sharpness can be performed at any stage after manufacturing, for example, by the user after shipment. The predetermined range may be any range where the phase difference is 0 or in its vicinity, or any range other than the front.

[0008] The system further includes a virtual microphone processing unit that calculates a virtual signal obtained by a virtual microphone virtually positioned at the foot of a perpendicular line drawn from the rear microphone to a virtual line segment connecting the two front microphones, based on the signals obtained by the two front microphones. The front extraction processing unit may extract the signal of the frequency bin where the phase difference between the virtual signal and the signal obtained by the rear microphone is positive. This allows the virtual signal obtained by a virtual microphone positioned in front of the rear microphone to be calculated from the two front microphones. Therefore, it is not necessary for either of the two front microphones to be positioned in front of the rear microphone, and the positions of the front and rear microphones can be designed flexibly.

[0009] The system further includes a synchronous subtraction unit that calculates the arrival time for sound reaching one of the virtual microphones and the rear microphone to reach the other based on the distance between the virtual microphone and the rear microphone, and calculates a synchronous subtraction signal by subtracting a delayed signal obtained by the rear microphone, which is delayed by the arrival time, from the virtual signal. The forward extraction processing unit extracts the signal of the frequency bin where the phase difference between the virtual signal and the signal obtained by the rear microphone is positive, and the output spectrum calculation unit may obtain the output spectrum based on the synchronous subtraction signal and the results extracted by the forward extraction processing unit and the front extraction processing unit. In this way, by using a synchronous subtraction signal, sound waves from the rear and from the sides can be suppressed, and the forward directivity can be further strengthened. The form using a synchronous subtraction signal is particularly effective when the phase difference is 0 (front) and an arbitrary range in its vicinity is defined as the predetermined range.

[0010] The output spectrum calculation unit may obtain the output spectrum based on the virtual signal and the results extracted by the forward extraction processing unit and the frontal extraction processing unit. In this way, even without synchronous subtraction processing, a microphone device with forward directivity can be obtained.

[0011] The system further includes a synchronous subtraction unit for calculating a synchronous subtraction signal, wherein the triangle formed by the rear microphone and the two front microphones is a right-angled triangle, the rear microphone and the first front microphone, which is one of the two front microphones, are located on a line along a second direction perpendicular to the first direction, the front extraction processing unit extracts the signal of the frequency bin where the phase difference between the signal obtained by the first front microphone and the signal obtained by the rear microphone is positive, the synchronous subtraction unit calculates the arrival time from the sound that reaches one of the first front microphone and the rear microphone to the other based on the distance between the first front microphone and the rear microphone, and calculates the synchronous subtraction signal by subtracting a delayed signal obtained by the signal obtained by the first front microphone, which is the signal obtained by the rear microphone delayed by the arrival time, and the output spectrum calculation unit may obtain the output spectrum based on the synchronous subtraction signal and the results extracted by the front extraction processing unit and the front extraction processing unit. In this way, by using a synchronous subtraction signal, sound waves from the rear and sides can be suppressed, and the forward directivity can be further strengthened. Furthermore, by making the triangle formed by the rear microphone and the two front microphones a right triangle, the process of calculating the output signal of the virtual microphone becomes unnecessary, reducing the computational burden. Note that the configuration using a synchronous subtraction signal is particularly effective when the predetermined range is an arbitrary range in front and its vicinity.

[0012] The triangle formed by the rear microphone and the two front microphones is a right triangle. The rear microphone and the first front microphone, which is one of the two front microphones, are located on a line along a second direction orthogonal to the first direction. The front extraction processing unit extracts signals of frequency bins where the phase difference between the signal obtained by the first front microphone and the signal obtained by the rear microphone is positive. The output spectrum calculation unit may obtain the output spectrum based on the signal obtained by the first front microphone and the results extracted by the front extraction processing unit and the front extraction processing unit. Thus, even when the synchronous subtraction process is not performed, a microphone device having directivity in the forward direction can be obtained. Further, by making the triangle formed by the rear microphone and the two front microphones a right triangle, the process of calculating the output signal of the virtual microphone becomes unnecessary, and the computational processing burden can be reduced.

[0013] The front extraction processing unit is a frequency mask that extracts signals of frequency bins included in a range determined by a threshold value represented by an expression including the parameter, and uses a frequency mask in which the value smoothly decays as it approaches the threshold value to extract signals of frequency bins in which the phase difference between the two front microphones is included in the predetermined range. Thereby, even when the phase difference is near the threshold value, the value of the frequency mask does not alternate between 0 and 1, and musical noise can be reduced.

Advantages of the Invention

[0014] According to the present invention, the sharpness of the directivity can be adjusted retrospectively.

Brief Description of the Drawings

[0015] [Figure 1] It is a diagram showing an outline of an example of the microphone device 1. [Figure 2] It is a diagram for explaining an outline of a processing flow in the microphone device 1. [Figure 3] It is a block diagram showing an electrical schematic configuration of the microphone device 1. [Figure 4] This diagram schematically shows the positional relationship of the virtual microphones calculated based on the first front microphone 11 and the second front microphone 12. [Figure 5] This figure shows an example of a directional pattern after synchronous subtraction. [Figure 6] This diagram schematically shows an example of a frequency mask. [Figure 7] This figure shows an example of a directional pattern when forward extraction is performed using a frequency mask. [Figure 8] This diagram schematically shows an example of a frequency mask. [Figure 9] This is a schematic diagram illustrating the maximum angle θ. [Figure 10] This diagram schematically shows an example of a frequency mask. [Figure 11] This figure shows an example of a directional pattern when frontal extraction is performed using a frequency mask. [Figure 12] This figure shows an example of a simulation result of the directional pattern of sound picked up by microphone device 1. [Figure 13] This figure shows an example of a simulation result of the directional pattern of sound picked up by microphone device 1. [Figure 14] This figure shows an example of a simulation result of the directional pattern of sound picked up by microphone device 1. [Figure 15] This figure shows an example of a simulation result of the directional pattern of sound picked up by microphone device 1. [Figure 16] This figure shows an example of the microphone arrangement of microphone device 1A. [Figure 17] This figure shows an example of the microphone arrangement of microphone device 1B. [Figure 18] This is a schematic diagram illustrating the angles θ1 and θ2 when directing the beam in a direction other than directly in front. [Figure 19] This diagram schematically shows an example of a frequency mask when directing the beam in a direction other than directly in front. [Figure 20]This diagram illustrates the general processing flow of the microphone device 2. [Figure 21] This is a block diagram showing the schematic electrical configuration of microphone device 2. [Modes for carrying out the invention]

[0016] Hereinafter, embodiments of the microphone device according to the present invention will be described in detail with reference to the drawings. The microphone device according to the present invention is a device that can pick up only the target sound from a specific direction. <First Embodiment> Figure 1 is a schematic diagram of the microphone device 1. The microphone device 1 mainly comprises a first front microphone 11, a second front microphone 12, and a rear microphone 13. Of the target sound to be picked up by the microphone device 1, the sound waves from the front of the microphone device 1 are assumed to travel from the bottom to the top in the figure.

[0017] Hereinafter, the direction from the microphone device 1 toward the front (the direction from top to bottom in Figure 1) will be referred to as the front, and the opposite direction (the direction from bottom to top in Figure 1) will be referred to as the rear. Also, looking from the front to the rear, the direction toward the right will be referred to as the right, and the direction toward the left will be referred to as the left. The direction from right to left and from left to right will be referred to as the first direction, and the direction from front to rear and from rear to front will be referred to as the second direction. The first and second directions are perpendicular to each other. Note that the line connecting the first front microphone 11 and the second front microphone 12 is aligned in the left-right direction (first direction).

[0018] The first front microphone 11, the second front microphone 12, and the rear microphone 13 are omnidirectional microphones. The first front microphone 11, the second front microphone 12, and the rear microphone 13 have equivalent configurations, and for example, MEMS microphones can be used.

[0019] The first front microphone 11, the second front microphone 12, and the rear microphone 13 are housed, for example, inside the housing 20. The housing 20 is provided with appropriate sound holes at positions corresponding to each of the first front microphone 11, the second front microphone 12, and the rear microphone 13.

[0020] The first front microphone 11 and the second front microphone 12 are located in the front row of the microphone device 1. Here, the front row refers to the arrangement located at the front of the arrangement along the first direction. In this embodiment, the front row is adjacent to the frontmost surface of the housing 20, but the position of the front row is not limited to this. Also in this embodiment, the first front microphone 11 and the second front microphone 12 are located near the left and right ends of the front row. However, the first front microphone 11 and the second front microphone 12 may be located at any position in the front row.

[0021] The rear microphone 13 is located behind the first front microphone 11 and the second front microphone 12, i.e., in the rear row. Here, the rear row refers to the row located behind the front row in the arrangement along the first direction. In this embodiment, the rear microphone 13 is located midway between the first front microphone 11 and the second front microphone 12 in the lateral direction (left-right direction in Figure 1), but the lateral position of the rear microphone 13 is not limited to this.

[0022] The position of the rear microphone 13 relative to the first front microphone 11 and the second front microphone 12 can be arbitrarily set as long as it is between the first front microphone 11 and the second front microphone 12. Therefore, the position of the rear microphone 13 can be flexibly set considering the shape of the device in which the microphone device 1 is installed.

[0023] Figure 2 is a diagram illustrating the general processing flow in microphone device 1. Microphone device 1 primarily performs virtual microphone calculation processing S1, synchronous subtraction processing S2, frequency conversion processing S3, forward extraction processing S4, frontal extraction processing S5, output spectrum calculation processing S6, and inverse frequency conversion processing S7. Microphone device 1 processes in the following order: virtual microphone calculation processing S1, synchronous subtraction processing S2, frequency conversion processing S3, forward extraction processing S4, frontal extraction processing S5, output spectrum calculation processing S6, and inverse frequency conversion processing S7. On the connectors connecting each process S1 to S7, information in the time domain or frequency domain that is passed to the next process by the connector is indicated.

[0024] Figure 3 is a block diagram showing the schematic electrical configuration of the microphone device 1. The microphone device 1 includes a control unit 50 connected to a first front microphone 11, a second front microphone 12, and a rear microphone 13. The control unit 50 includes, as software resources, at least a virtual microphone processing unit 51, a synchronous subtraction unit 52, a frequency conversion unit 53, a front extraction processing unit 54, a front extraction processing unit 55, an output spectrum calculation unit 56, and an inverse frequency conversion unit 57, all powered by a CPU (Central Processing Unit) or other computing device and storage device.

[0025] The virtual microphone processing unit 51 is a functional unit that performs calculation processing S1 for the virtual microphone 14 based on signals obtained from the first front microphone 11 and the second front microphone 12.

[0026] Figure 4 schematically shows the positional relationship of the virtual microphone 14 calculated based on the first front microphone 11 and the second front microphone 12. The virtual microphone 14 is virtually positioned at the foot of the perpendicular line drawn from the rear microphone 13 to the virtual line segment connecting the first front microphone 11 and the second front microphone 12. This virtual line segment overlaps with the front row, and the virtual microphone 14 is virtually provided in the front row, similar to the first front microphone 11 and the second front microphone 12.

[0027] The virtual microphone processing unit 51 processes the signal x picked up by the first front microphone 11. L (t) and the signal x picked up by the second front microphone 12 R By linearly interpolating (t) and , the virtual signal x picked up by the virtual microphone 14 is obtained. V Calculate (t). Let d be the distance between the first front microphone 11 and the second front microphone 12, d1 be the distance between the first front microphone 11 and the virtual microphone 14, and d2 be the distance between the second front microphone 12 and the virtual microphone 14. Then the virtual signal x picked up by the virtual microphone 14 is calculated. V (t) is calculated by the following equation (1).

number

[0028] Returning to the explanation of Figure 3, the synchronous subtraction unit 52 subtracts the output signal x of the rear microphone 13. B (t) and the virtual signal x of the virtual microphone 14 V This is a functional unit that performs the synchronous subtraction process S2 of (t). As shown in Figure 4, the rear microphone 13 is located behind the virtual microphone 14, so there is a difference in arrival time. The arrival time τ(s) for sound that has reached one of the virtual microphone 14 or the rear microphone 13 to reach the other (the difference in arrival time between the rear microphone 13 and the virtual microphone 14) is calculated using the following equation (2), where c (m / s) is the speed of sound and h (m) is the distance between the rear microphone 13 and the virtual microphone 14.

number

[0029] Furthermore, the synchronous subtraction unit 52 uses the virtual signal x obtained by the virtual microphone 14. V (t) From this, the signal x obtained by the rear microphone 13 B The synchronous subtracted signal y(t), obtained by subtracting the delayed signal obtained by delaying (t) by the arrival time τ, is calculated using the following equation (3). [Number] ···(3)

[0030] FIG. 5 is a diagram showing an example of a directivity pattern by the synchronous subtraction process S2. By the synchronous subtraction process S2, the sound wave from the rear side is delayed until it reaches the position of the virtual microphone 14, and the phase is adjusted and subtracted from the virtual signal x V (t) of the virtual microphone 14, so that the sound wave from the rear can be suppressed and the directivity with respect to the front can be emphasized.

[0031] Returning to the description of FIG. 3. The frequency conversion unit 53 is a functional unit that performs the frequency conversion process S3 of the signal. The frequency conversion unit 53 uses the fast Fourier transform (FFT) or the like to perform the power spectrum X L (f) of the output signal x L (t) of the first front microphone 11, the power spectrum X R (f) of the output signal x R (t) of the second front microphone 12, the power spectrum X B (f) of the output signal x B (t) of the rear microphone 13, the power spectrum X V (f) of the virtual signal x V (t) of the virtual microphone 14, and the power spectrum Y(f) of the synchronous subtraction signal y(t) obtained by the synchronous subtraction process S2 are obtained by the following equations (4) to (8), respectively. [Number]

[0032] The front extraction processing unit 54 is a functional unit that performs the front extraction process S4 of extracting the signal of the frequency bin in which the phase difference between the signals obtained by the first front microphone 11 and the second front microphone 12 and the signal obtained by the rear microphone 13 is positive.

[0033] The forward extraction processing unit 54 processes the power spectrum X of the virtual signal virtually obtained by the virtual microphone 14, for example, as the signal obtained by the first forward microphone 11 and the second forward microphone 12. V (f) is adopted. Power spectrum X of the virtual signal V (f) Power spectrum X of the signal obtained by the rear microphone 13 B The phase difference φ1(f) of (f) is calculated by the following equation (9).

number

[0034] Figure 6 schematically shows an example of a frequency mask applied to the phase difference φ1(f) in the microphone device 1. The frequency mask G1(f) is expressed by, for example, the following equation (10). This allows only the signals of the frequency bins where the phase difference φ1(f) is positive to be retained.

number

[0035] Figure 7 shows an example of the directional pattern when forward extraction processing S4 is performed on the microphone device 1 using the frequency mask G1(f). As shown in the figure, sound waves in the upper region of the figure, which represents the rear, disappear, while sound waves in the lower region of the figure, which represents the front, are maintained. In other words, at the stage when forward extraction processing S4 is completed, the microphone device 1 has directionality only in the forward direction.

[0036] Returning to the explanation of Figure 3, the front extraction processing unit 55 analyzes the power spectrum X of the signal obtained by the first front microphone 11. L (f) and the power spectrum X of the signal obtained by the second front microphone 12 R This is a functional unit that performs a frontal extraction process S5 to extract signals from frequency bins whose phase difference with (f) falls within a predetermined range.

[0037] The front extraction processing unit 55 analyzes the power spectrum X of the signal obtained by the first front microphone 11. L (f) and the power spectrum X of the signal obtained by the second front microphone 12 R The phase difference φ2(f) with (f) is calculated using the following equation (11).

number

[0038] The front-facing signal extraction unit 55 extracts signals whose phase difference φ2(f) is within a predetermined range using a threshold T(f). In this embodiment, the predetermined range includes the case where the phase difference φ2(f) is 0, and is any range in the vicinity of the case where the phase difference φ2(f) is 0. Since sound waves from the front have a phase difference φ2(f) of 0, sound waves from the front can be extracted by extracting signals with a phase difference φ2(f) in the vicinity of 0.

[0039] Figure 8 schematically shows an example of a frequency mask applied to the phase difference φ2(f) between the left and right microphones. If the distance between the first front microphone 11 and the second front microphone 12 is d (m), the speed of sound is c (m / s), and the maximum angle between the source of the target sound and the direction of propagation of the sound wave from the front is θ (rad), then the threshold T(f) of the phase difference φ2(f) is calculated by the following equation (12). In other words, the threshold T(f) is expressed by an equation that includes the maximum angle θ, a parameter indicating the sharpness of directivity, and the predetermined range from which the signal is extracted is changed by the maximum angle θ. Note that the maximum angle θ is the angle between the direction in which a sound enters the first front microphone 11 and the second front microphone 12, and a line virtually drawn forward from the first front microphone 11 and the second front microphone 12, as shown in Figure 9. Furthermore, in Figure 9, the maximum angle θ is considered positive when it is pointing to the left from the virtual line.

number

[0040] Parameters indicating the sharpness of directivity, such as the maximum angle θ, are received by the reception unit 21 (see Figure 3) via an input device 25 (see Figure 3) that is either present in the microphone device 1 or appropriately connected to the microphone device 1. The reception unit 21 can receive input at any stage after the manufacturing of the microphone device 1. The reception unit 21 outputs information indicating the acquired maximum angle θ to the control unit 50, and the front extraction processing unit 55 uses the acquired maximum angle θ to determine the threshold T(f). With this configuration, the sharpness of the directivity of the microphone device 1 can be easily changed.

[0041] The front extraction processing unit 55 then uses the frequency mask G2(f) shown in equation (13) to extract signals from frequency bins in which the phase difference φ2(f) of the left and right microphones falls within a predetermined range. The frequency mask G2(f) is "1" only in the shaded area in Figure 8 (i.e., when the phase difference φ2(f) of the left and right microphones is from -T(f) to T(f)).

number

[0042] In addition, with the frequency mask G2(f) shown in equation (13), musical noise may occur when the phase difference φ2(f) is near the threshold ±T(f), as the value of the frequency mask G2(f) fluctuates between 0 and 1. Therefore, instead of the binary mask of 0 and 1 shown in equation (13), the frequency mask G2(f) may be a frequency mask that smoothly attenuates as it approaches the threshold. In this case, the frequency mask G2(f) used by the front extraction processing unit 55 will be the one shown in equation (14) instead of equation (13).

number

[0043] Figure 10 schematically shows an example of the frequency mask G2(f) represented by equation (13) and the frequency mask G2(f) represented by equation (14). In Figure 10, graph M1 is a graph showing the frequency mask G2(f) represented by equation (13), and graph M2 is a graph showing the frequency mask G2(f) represented by equation (14).

[0044] The frequency mask G2(f) expressed in equation (14) decays smoothly as it approaches the threshold. Therefore, even when the phase difference φ2(f) is near the threshold ±T(f), the value of the frequency mask G2(f) does not fluctuate between 0 and 1, thus reducing musical noise.

[0045] Figure 11 shows an example of a directional pattern after frontal extraction processing S5 using frequency mask G2(f). As shown in the figure, the microphone device 1 can be seen to be able to extract signals in a narrow range including zero phase difference in the front and rear directions by frontal extraction processing S5.

[0046] Returning to the explanation of Figure 3, the output spectrum calculation unit 56 is a functional unit that performs the output spectrum calculation process S6. The output spectrum calculation unit 56 calculates the output spectrum Z(f) by integrating the power spectrum Y(f) of the synchronous subtraction signal y(t) obtained in the synchronous subtraction process S2 and the frequency conversion process S3, the frequency mask G1(f) obtained in the forward extraction process S4, and the frequency mask G2(f) obtained in the frontal extraction process S5. The output spectrum Z(f) is expressed by the following equation (15).

number

[0047] The inverse frequency conversion unit 57 is a functional unit that performs an inverse frequency conversion process S7 to convert the power spectrum into a time-domain signal. The inverse frequency conversion unit 57 obtains a time signal z(t) by inverse frequency conversion of the output spectrum Z(f). The time signal z(t) is expressed by the following equation (16).

number

[0048] Figures 12-15 show examples of simulator results for the directional pattern of sound picked up by microphone device 1. Figure 12 shows an example where the maximum angle θ, which indicates the threshold T(f) range, is 10 degrees; Figure 13 shows an example where the maximum angle θ is 20 degrees; Figure 14 shows an example where the maximum angle θ is 30 degrees; and Figure 15 shows an example where the maximum angle θ is 45 degrees. Figures 12-15 show the results of measurements taken with the first front microphone 11, the second front microphone 12, and the rear microphone 13 arranged in an equilateral triangle with sides of 21 mm (see Figure 1). In all of Figures 12-15, a signal with strong directivity towards the front is output. Also, as the maximum angle θ increases, the strength of the directivity decreases (the width of the signal extraction range increases). The differences in the directional patterns shown in Figures 12-15 can be recognized by listening to and comparing the sounds actually picked up by microphone device 1.

[0049] According to this embodiment, a microphone device 1 with strong forward directivity can be obtained, and the strength of the directivity can be adjusted retrospectively by accepting input of a maximum angle θ, which is a parameter indicating the sharpness of the directivity, and using the accepted maximum angle θ to change the range in which the microphone device 1 extracts sound.

[0050] Furthermore, according to this embodiment, a microphone device 1 that selectively picks up sound from the front using three omnidirectional microphones can be realized, making it possible to create an inexpensive and compact microphone device 1.

[0051] Furthermore, according to this embodiment, since the output spectrum Z(f) is obtained using the power spectrum Y(f) of the synchronous subtraction signal y(t), even if there are any remaining sounds after processing using frequency masks G1(f) and G2(f), these remaining sounds can be eliminated with the power spectrum Y(f). Therefore, unwanted sounds are prevented from being extracted by the microphone device 1, and the directivity of the microphone device 1 can be further enhanced.

[0052] In this embodiment, the threshold T(f) was varied by the maximum angle θ as shown in equation (12), but the form of the threshold is not limited to this. For example, the threshold may be varied by an exponent α, which is a parameter indicating the strength of directivity. For example, as shown in equation (17), the threshold T(f) is varied by the exponent α, which is a parameter indicating the strength of directivity, and the sampling frequency f s It may also be calculated using the following method. In this case, the front extraction processing unit 55 receives the input of the exponent α via the reception unit 21, allowing the parameter to be set to any size. For example, increasing the exponent α increases the threshold T(f), i.e., widens the sound collection range, while decreasing the exponent α decreases the threshold T(f), i.e., strengthens the directivity.

number

[0053] Furthermore, in this embodiment, the triangle formed by the first front microphone 11, the second front microphone 12, and the rear microphone 13 can be of any shape, but the triangle formed by the first front microphone 11, the second front microphone 12, and the rear microphone 13 may be an isosceles triangle, a right triangle, or the like.

[0054] Figure 16 shows an example of a microphone arrangement for microphone device 1A, where the triangle formed by the first front microphone 11, the second front microphone 12, and the rear microphone 13 is an isosceles triangle. In this case, d1=d2=d / 2, and equation (1) becomes equation (18) below.

number

[0055] Figure 17 shows an example of a microphone arrangement for microphone device 1B, where the triangle formed by the first front microphone 11, the second front microphone 12, and the rear microphone 13 is a right-angled triangle. In this case, d1=0 and d2=d, so x V (t) = x L (t) is obtained, and the process of calculating the output signal of the virtual microphone 14 is unnecessary. In other words, by making the triangle formed by the first front microphone 11, the second front microphone 12, and the rear microphone 13 a right-angled triangle, the computational burden can be reduced.

[0056] Furthermore, if the triangle formed by the rear microphone 13, the first front microphone 11, and the second front microphone 12 is a right-angled triangle, the forward extraction processing unit 54 extracts the signal of the frequency bin where the phase difference between the signal obtained by the first front microphone 11 or the second front microphone 12 and the signal obtained by the rear microphone 13 is positive, as forward extraction processing S4. For example, if the triangle formed by the rear microphone 13, the first front microphone 11, and the second front microphone 12 is a right-angled triangle as shown in Figure 17, the first front microphone 11 (corresponding to the first front microphone of the present invention) and the rear microphone 13 are located on a line along the second direction. Therefore, in equation (9), the forward extraction processing unit 54 extracts the power spectrum X of the virtual signal. V Instead of (f), the power spectrum X of the signal obtained by the first front microphone 11 located in front of the rear microphone 13. L (f) is used. Also, if the triangle formed by the rear microphone 13, the first front microphone 11, and the second front microphone 12 is a right triangle, and the second front microphone 12 is located in front of the rear microphone 13, the front extraction processing unit 54 calculates the power spectrum X of the virtual signal in equation (9). V Instead of (f), the power spectrum X of the signal obtained by the second front microphone 12. R Use (f).

[0057] Furthermore, in this embodiment, the front extraction processing unit 55 uses the threshold T(f) shown in equation (12) to extract signals when the phase difference φ2(f) is 0 and in an arbitrary range in its vicinity (i.e., a predetermined range). However, this predetermined range does not necessarily include the case where the phase difference φ2(f) is 0. In other words, the front extraction processing unit 55 may extract sounds from any range other than the front as the predetermined range. Below, we will describe an example of how the front extraction processing unit 55 extracts sounds from the diagonally left front, and how it extracts sounds from a predetermined range other than the front.

[0058] First, the front extraction processing unit 55 analyzes the power spectrum X of the signal obtained by the first front microphone 11. L (f) and the power spectrum X of the signal obtained by the second front microphone 12 R The phase difference φ2(f) with (f) is calculated using equation (11).

[0059] The front extraction processing unit 55 extracts signals whose phase difference φ2(f) is within a predetermined range using thresholds T1(f) and T2(f). The predetermined range is an arbitrary range indicated by two angles, and in this case, it is the range between angles θ1 and θ2 as shown in Figure 18. Angles θ1 and θ2 are the angles between the direction in which a sound enters the first front microphone 11 and the second front microphone 12, and a virtual line virtually drawn forward from the first front microphone 11 and the second front microphone 12. Also, angles θ1 and θ2 are considered positive when they are pointing to the left from the virtual line in Figure 18.

[0060] Figure 19 schematically shows an example of a frequency mask applied to the phase difference φ2(f) between the left and right microphones. If the distance between the first front microphone 11 and the second front microphone 12 is d (m), the speed of sound is c (m / s), and the predetermined range of the target sound to be extracted is angles θ1 (rad) and θ2 (rad), then the thresholds T1(f) and T2(f) of the phase difference φ2(f) are calculated by the following equations (19) and (20). In other words, in the thresholds T1(f) and T2(f), the parameter indicating the sharpness of directivity is the difference with angles θ1 and θ2, and the predetermined range from which the signal is extracted is changed by angles θ1 and θ2.

number

[0061] The front extraction processing unit 55 then uses the frequency mask G2(f) shown in equation (21) to extract signals from frequency bins in which the phase difference φ2(f) between the left and right microphones is included within a predetermined range. The frequency mask G2(f) is shown as "1" only within the shaded area in Figure 19. This makes it possible to extract sound waves from any predetermined range other than the front.

number

[0062] The angles θ1 and θ2, which are parameters indicating the sharpness of the directivity, are received by the reception unit 21 (see Figure 3) via the input device 25 (see Figure 3), and the frontal extraction processing unit 55 uses the acquired angles θ1 and θ2 to determine thresholds T1(f) and T2(f). This makes it easy to change the predetermined range and the sharpness of the directivity.

[0063] <Second Embodiment> In the first embodiment of the present invention, the microphone device 1 performs synchronous subtraction processing S2 using a synchronous subtraction unit 52, but the synchronous subtraction unit 52 is not essential. The microphone device 2 according to the second embodiment differs from the microphone device 1 of the first embodiment in that it does not perform synchronous subtraction processing S2. The microphone device 2 will be described below. Parts identical to those in the first embodiment are denoted by the same reference numerals and their descriptions are omitted.

[0064] Figure 20 is a diagram illustrating the schematic processing flow of the microphone device 2 according to the second embodiment. The microphone device 2 mainly performs virtual microphone calculation processing S1, frequency conversion processing S3, forward extraction processing S4, frontal extraction processing S5, output spectrum calculation processing S6A, and inverse frequency conversion processing S7. The microphone device 2 processes in the following order: virtual microphone calculation processing S1, frequency conversion processing S3, forward extraction processing S4, frontal extraction processing S5, output spectrum calculation processing S6A, and inverse frequency conversion processing S7.

[0065] Figure 21 is a block diagram showing the schematic electrical configuration of the microphone device 2. The microphone device 2 includes a control unit 50A connected to a first front microphone 11, a second front microphone 12, and a rear microphone 13. The control unit 50A includes, as software resources, at least a virtual microphone processing unit 51, a frequency conversion unit 53, a front extraction processing unit 54, a front extraction processing unit 55, an output spectrum calculation unit 56A, and an inverse frequency conversion unit 57, all powered by a computing device such as a CPU (Central Processing Unit) and a storage device.

[0066] The output spectrum calculation unit 56A is a functional unit that performs the output spectrum calculation process S6A. The output spectrum calculation unit 56A performs the frequency conversion process S3 and calculates the power spectrum X of the virtual signal obtained by the virtual microphone 14. V The output spectrum Z(f) is calculated by integrating (f), the frequency mask G1(f) obtained in the forward extraction process S4, and the frequency mask G2(f) obtained in the frontal extraction process S5. The output spectrum Z(f) is expressed by the following equation (22).

number

[0067] Then, the inverse frequency conversion unit 57 performs an inverse frequency conversion process S7 to convert the output spectrum Z(f) obtained by the output spectrum calculation process S6A into a time-domain signal to obtain a time signal z(t).

[0068] This configuration also allows for obtaining a microphone device 2 with strong forward directivity, and the strength of the directivity can be adjusted retrospectively by changing parameters such as the maximum angle θ.

[0069] In this embodiment, the frequency mask G1(f) obtained in the forward extraction process S4 is used to eliminate unwanted sounds from the rear. However, since there is a possibility that some sounds may remain, it is desirable to generate a synchronous subtraction signal y(t) and obtain the output spectrum Z(f) using formula (15), as in the microphone device 1 of the first embodiment.

[0070] Although embodiments of this invention have been described in detail above with reference to the drawings, the specific configuration is not limited to these embodiments, and design modifications and the like are also included within the scope of the gist of this invention. Those skilled in the art can modify, add, or change each element of the embodiments as appropriate.

[0071] Furthermore, in this invention, "approximately" is a concept that includes not only cases where things are strictly identical, but also errors or modifications that do not result in a loss of identity. For example, "orthogonal" is not limited to cases where things are strictly orthogonal, but also includes cases where there is an error of a few degrees from orthogonality. Also, for example, when simply using terms like "orthogonal" or "parallel," it includes not only cases where things are strictly orthogonal or parallel, but also cases where things are approximately orthogonal or approximately parallel. Furthermore, in this invention, "neighborhood" means including a certain range (which can be arbitrarily defined) of area near a reference position. For example, when referring to the neighborhood of A, it means a certain range of area near A, which may or may not include A. [Explanation of symbols]

[0072] 1, 1A, 1B, 2: Microphone equipment 11: First front microphone 12: Second front microphone 13: Rear microphone 14: Virtual Microphone 20: Cabinet 21: Reception Department 25: Input device 50, 50A: Control Unit 51: Virtual Microphone Processing Unit 52: Synchronous subtraction unit 53: Frequency conversion section 54: Forward extraction processing unit 55: Frontal extraction processing unit 56, 56A: Multiplication part 57: Inverse frequency conversion section

Claims

1. Two front microphones are positioned in the front row along the first direction, A rear microphone is provided in the rear row, which is positioned along the first direction and behind the front row, A frequency conversion unit that converts the signals obtained by the two front microphones and the signals obtained by the rear microphone into the frequency domain, A forward extraction processing unit extracts signals from the frequency domain of the signal converted to the frequency domain by the frequency conversion unit, where the phase difference between the signal obtained by at least one of the two forward microphones and the signal obtained by the rear microphone is positive. A front extraction processing unit extracts signals from the frequency domain of the signal converted to the frequency domain by the frequency conversion unit, wherein the phase difference between the signal obtained by one of the two front microphones and the signal obtained by the other microphone is within a predetermined range. An output spectrum calculation unit that obtains an output spectrum based on the results extracted by the forward extraction processing unit and the forward extraction processing unit, An inverse frequency converter that converts the output spectrum into a time-domain signal, A reception unit that accepts parameters indicating the sharpness of directivity, Equipped with, The front extraction processing unit causes the predetermined range to differ based on the parameters received by the receiving unit. A microphone device characterized by the following features.

2. The system further includes a virtual microphone processing unit that calculates a virtual signal obtained by a virtual microphone virtually positioned at the foot of a perpendicular line drawn from the rear microphone to the virtual line segment connecting the two front microphones, based on the signals obtained by the two front microphones. The forward extraction processing unit extracts signals from frequency bins where the phase difference between the virtual signal and the signal obtained by the rear microphone is positive. The microphone device according to feature 1.

3. The system further includes a synchronization subtraction unit that calculates the arrival time for sound reaching one of the virtual microphones and the rear microphone to reach the other, based on the distance between the virtual microphone and the rear microphone, and calculates a synchronization subtraction signal by subtracting a delayed signal obtained by the rear microphone, which is delayed by the arrival time, from the virtual signal. The forward extraction processing unit extracts the signal of the frequency bin where the phase difference between the virtual signal and the signal obtained by the rear microphone is positive. The output spectrum calculation unit obtains the output spectrum based on the synchronous subtraction signal and the results extracted by the forward extraction processing unit and the frontal extraction processing unit. The microphone device according to feature 2.

4. The output spectrum calculation unit obtains the output spectrum based on the virtual signal and the results extracted by the forward extraction processing unit and the forward extraction processing unit. The microphone device according to feature 2.

5. It further includes a synchronous subtraction unit that calculates a synchronous subtraction signal, The triangle formed by the rear microphone and the two front microphones is a right-angled triangle. The rear microphone and the first front microphone, which is one of the two front microphones, are located on a line along a second direction perpendicular to the first direction. The forward extraction processing unit extracts the signal of the frequency bin where the phase difference between the signal obtained by the first forward microphone and the signal obtained by the rear microphone is positive. The synchronous subtraction unit calculates the arrival time for sound that has reached one of the first front microphone and the rear microphone to reach the other, based on the distance between the first front microphone and the rear microphone, and calculates the synchronous subtraction signal by subtracting a delayed signal obtained by delaying the signal obtained by the rear microphone by the arrival time from the signal obtained by the first front microphone. The output spectrum calculation unit obtains the output spectrum based on the synchronous subtraction signal and the results extracted by the forward extraction processing unit and the frontal extraction processing unit. The microphone device according to feature 1.

6. The triangle formed by the rear microphone and the two front microphones is a right-angled triangle. The rear microphone and the first front microphone, which is one of the two front microphones, are located on a line along a second direction perpendicular to the first direction. The forward extraction processing unit extracts the signal of the frequency bin where the phase difference between the signal obtained by the first forward microphone and the signal obtained by the rear microphone is positive. The output spectrum calculation unit obtains the output spectrum based on the signal obtained by the first front microphone and the results extracted by the front extraction processing unit and the front extraction processing unit. The microphone device according to feature 1.

7. The front extraction processing unit extracts signals from frequency bins whose phase difference between the two front microphones falls within a predetermined range, using a frequency mask that extracts signals from frequency bins whose phase difference falls within a predetermined range, wherein the value of the frequency mask smoothly decays as it approaches the threshold. A microphone device according to any one of claims 1 to 6.

Images