Sound pickup device and program

Abstract

To separately collect a sound field radiated from a sound source in an area and a sound field coming from the outside of the area by using a distributed spherical microphone array.SOLUTION: A separation matrix calculation part 10 of a sound collection device 1 calculates a separation matrix in which position information and size are reflected and which is expressed by a spherical harmonic function on the basis of preset distributed spherical microphone array position information. The frequency analysis unit 11 obtains a time-frequency spectrum by performing Fourier transform on a sound collection signal from the distributed spherical microphone array M in a time domain. A spherical harmonic spectrum vector calculation part 12 calculates a spherical harmonic spectrum vector having a spherical harmonic spectrum of a sound field radiated from a sound source S in an area and a spherical harmonic spectrum of a sound field due to an interference wave coming from the outside of the area as elements on the basis of a separation matrix and a time frequency spectrum of a sound collection signal.SELECTED DRAWING: Figure 3

Description

[Technical Field]

[0001] The present invention relates to a sound collection device and a program that use a distributed spherical microphone array in which multiple spherical microphone arrays are distributed on a spherical surface to separate and collect sound fields radiated from an internal sound source and sound fields caused by interference waves arriving from outside. [Background technology]

[0002] Conventionally, methods have been proposed in which a microphone array is used to capture a sound field in the entire vicinity of the microphone array or in a specific range.

[0003] Higher-order Ambisonics technology is known as one method for capturing a sound field, and the best-known method of this technology is to capture the surrounding sound field centered on a rigid sphere using sound pressure acquired from a microphone array in which small microphone elements are placed on the surface of the rigid sphere (see, for example, non-patent document 1).

[0004] When capturing a sound field using the method described in Non-Patent Document 1, it is possible to separate the sound waves arriving from each direction, but if an interfering wave arrives from the same direction as the desired sound wave, it is not possible to distinguish (separate) and collect the sound.

[0005] Another method has been proposed, which uses a surrounding microphone array to capture a sound field (see, for example, Non-Patent Document 2). However, like the method in Non-Patent Document 1, this method does not take into consideration the ability to distinguish between desired sound waves and interference waves when they arrive simultaneously. [Prior art documents] [Non-patent literature]

[0006] [Non-Patent Document 1] B. Rafaely, “Analysis and design of spherical microphone arrays”, in IEEE Transactions on Speech and Audio Processing, vol.13, no.1, pp.135-143, Jan.2005 [Non-patent document 2] T. Okamoto, “Mode-matching-based sound field recording and synthesis with circular double-layer arrays”, Appl.Sci., vol.8, no.7,1048,July.2018. Summary of the Invention [Problem to be solved by the invention]

[0007] When a microphone array is used to capture sound waves emitted from a sound source located within the area surrounded by the microphone array, if there is a noise source or reflecting object outside that area, external interfering sounds will be mixed into the signals picked up by the individual microphone elements that make up the microphone array.

[0008] This has led to the problem that the sound field that the microphone array was originally intended to record cannot be accurately captured.

[0009] To solve this problem, it is conceivable to use directional microphones as microphone elements that make up a microphone array, or to form directivity using beamforming technology.

[0010] By using such directional microphones or beamforming technology, it is possible to block interfering sounds coming from outside the directional direction, but it is difficult to block interfering sounds coming from the same direction as the sound source being picked up.

[0011] Meanwhile, many methods have been proposed for removing interfering sounds that are mixed into microphone elements, using blind sound source separation technology, noise reduction technology, etc. However, while these methods generally achieve high performance when the statistical characteristics of the sound source itself or the propagation characteristics of the interfering sound are known, they often have drawbacks, such as impairing the sound quality of the sound being picked up, when the statistical characteristics of the sound source itself are unknown.

[0012] Therefore, the present invention has been made to solve the above-mentioned problems, and its purpose is to provide a sound collection device and program that can separate and collect the sound field radiated from a sound source within an area surrounded by the distributed spherical microphone array from the sound field caused by interference waves arriving from outside that area, by performing signal processing in the spatial frequency domain using a distributed spherical microphone array consisting of multiple spherical microphone arrays, even when the statistical characteristics of the sound source itself or the propagation characteristics of the interference sound are unknown. [Means for solving the problem]

[0013] In order to solve the above problem, the sound collection device of claim 1 is a sound collection device that receives a collected signal from a distributed spherical microphone array M in which a plurality of spherical microphone arrays are distributed and arranged on a predetermined spherical surface, and based on the collected signal, separates and collects a sound field radiated from a sound source S to be collected that exists within an area surrounded by the plurality of spherical microphone arrays and a sound field due to interference waves arriving from outside the area, and the sound collection device is configured such that a plurality of microphone elements are arranged on the surface of each of the plurality of spherical microphone arrays, and the positional information of the plurality of spherical microphone arrays that constitute the distributed spherical microphone array M, the positional information of the plurality of microphone elements arranged on the surface of the spherical microphone array for each of the plurality of spherical microphone arrays, and the size of the spherical microphone array are reflected, and The distributed spherical microphone array M is characterized by comprising a separation matrix calculation unit that calculates a separation matrix expressed in surface harmonic functions, a frequency analysis unit that inputs sound collection signals of all microphone elements of a plurality of spherical microphone arrays that constitute the distributed spherical microphone array M and calculates a time-frequency spectrum of the sound collection signals by performing a Fourier transform of the sound collection signals in the time domain, and a spherical harmonic spectrum vector calculation unit that calculates a spherical harmonic spectrum vector whose elements are the spherical harmonic spectrum of the sound field in the internal region of the distributed spherical microphone array M radiated from the sound source S and the spherical harmonic spectrum of the sound field due to the interference wave arriving from the external region of the distributed spherical microphone array M, based on the time-frequency spectrum of the sound collection signals calculated by the frequency analysis unit and the separation matrix calculated by the separation matrix calculation unit.

[0014] Furthermore, the sound collection device of claim 2 is characterized in that, in the sound collection device described in claim 1, each of the multiple spherical microphone arrays consists of a rigid spherical baffle, and the multiple microphone elements are arranged on the rigid spherical baffle.

[0015] A sound collection device according to claim 3 is a sound collection device that receives a collected signal from a distributed spherical microphone array M in which a plurality of spherical microphone arrays are distributed and arranged on a predetermined spherical surface, and based on the collected signal, separates and collects a sound field radiated from a sound source S to be collected that exists within an area surrounded by the plurality of spherical microphone arrays, and a sound field due to interference waves arriving from outside the area, and in which the distributed spherical microphone array M is composed of L spherical microphone arrays, and Q microphone elements are arranged on the surface of each of the L spherical microphone arrays, a separation matrix is ​​calculated based on position information of the L spherical microphone arrays that make up the distributed spherical microphone array M, position information of the Q microphone elements arranged on the surface of each of the L spherical microphone arrays, and a radius a of the sphere of the spherical microphone array. TIFF2026025419000002.tif10170, and a separation matrix calculation unit that calculates a time-frequency spectrum of the collected sound signal by inputting the collected sound signals of all the microphone elements of the L spherical microphone arrays that constitute the distributed spherical microphone array M and performing a Fourier transform on the collected sound signals in the time domain. TIFF2026025419000003.tif9170 (k is the wave number), α{n'm'}(k) is the spherical harmonic spectrum of the sound field radiated from the sound source S (n', m' are the order and magnitude of the spherical harmonic spectrum, respectively), β{n'm'}(k) is the spherical harmonic spectrum of the sound field caused by the interference wave, TIFF2026025419000004.tif9170 is expressed as a spherical harmonic spectrum vector having the spherical harmonic spectrum α{n′m′}(k) and the spherical harmonic spectrum β{n′m′}(k) as elements, using the following formula: a spherical harmonic spectrum vector calculation unit that calculates the spherical harmonic spectrum vector by the calculation of TIFF2026025419000005.tif9170, where l (El) is the number of the spherical microphone array (l=[1,L]), q is the number of the microphone element (q=[1,Q]), a coordinate system whose origin is the sound source S is called a global coordinate system, a coordinate system whose origin is the center of the spherical microphone array is called a local coordinate system, TIFF2026025419000006.tif9170 is the position vector of the global coordinate system in the l (el)-th spherical microphone array, TIFF2026025419000007.tif9170 is a position vector in a local coordinate system of the q-th microphone element installed on the surface of the l-th spherical microphone array, and i is an imaginary unit, and the separation matrix calculation unit calculates the following transfer matrix equation: TIFF2026025419000008.tif9170 is calculated, and the inverse matrix is ​​the separation matrix TIFF2026025419000009.tif10170 is calculated, and the transfer matrix calculation formula is TIFF2026025419000010.tif9170 with the following formula: Calculated using TIFF2026025419000011.tif36170 TIFF2026025419000012.tif8170 with the following formula: Calculated using TIFF2026025419000013.tif11170, j q is the q-th order spherical Bessel function, TIFF2026025419000014.tif8170 is the complex conjugate of the spherical harmonic function, TIFF2026025419000015.tif9170 is the gaunt coefficient, TIFF2026025419000016.tif8170 with the following formula: Calculated using TIFF2026025419000017.tif11170, hq (2) is a q-order second-order Hankel function, and in the transfer matrix calculation formula TIFF2026025419000018.tif9170 with the following formula: Calculated using TIFF2026025419000019.tif11170 TIFF2026025419000020.tif9170 with the following formula: Calculation is performed using TIFF2026025419000021.tif10170, and h'n is the derivative of the n-th order spherical Hankel function. TIFF2026025419000022.tif9170 with the following formula: Calculated using TIFF2026025419000023.tif11170 TIFF2026025419000024.tif9170 with the following formula: It is characterized in that it is calculated using TIFF2026025419000025.tif24170, and Y{n,m} is an n-th order, m-th order spherical harmonic function.

[0016] Furthermore, a program according to claim 4 causes a computer to function as the sound collection device according to any one of claims 1 to 3. [Effects of the Invention]

[0017] As described above, according to the present invention, by performing signal processing in the spatial frequency domain using a distributed spherical microphone array consisting of multiple spherical microphone arrays, it is possible to separate and collect the sound field radiated from the sound source within the area surrounded by the distributed spherical microphone array from the sound field due to the interference wave arriving from outside that area, even if the statistical characteristics of the sound source itself or the propagation characteristics of the interference sound are unknown. [Brief explanation of the drawings]

[0018] [Figure 1] 1 is a diagram showing an example of the arrangement of a distributed spherical microphone array M used in the sound field capturing method of the present invention. [Figure 2] 1 is a diagram showing an example of the arrangement of a sound source S and spherical microphone arrays r1, r2, and r3 assumed in the sound field capture method of the present invention, as well as a sound field radiated from the sound source S within the region and a sound field due to interference waves arriving from outside the region. [Figure 3] 1 is a block diagram showing an example of the configuration of a sound collection device according to an embodiment of the present invention. [Figure 4] 10 is a flowchart illustrating an example of processing performed by a sound collection device according to an embodiment of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0019] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings. [Sound field capturing method of the present invention] First, an outline of the sound field capturing method of the present invention will be described. Figure 1 is a diagram showing an example of the arrangement of a distributed spherical microphone array M used in the sound field capturing method of the present invention.

[0020] This distributed spherical microphone array M is composed of L spherical microphone arrays r1,...,rL that are distributed and arranged on a specific spherical surface. L is an integer greater than or equal to 2. Let l be the number of the spherical microphone array, and the l-th spherical microphone array be denoted as rl. l = [1,L].

[0021] A spherical microphone array rl is a microphone array in which, for example, Q microphone elements rl^1,...,rl^Q are arranged on a rigid spherical baffle, and the radius of the sphere of the spherical microphone array rl (for example, the radius of the rigid spherical baffle) is a. Q is an integer greater than or equal to 2. Let q be the microphone element number, and the qth microphone element arranged in the lth spherical microphone array rl is denoted as rl^q. q = [1,Q].

[0022] Figure 2 shows an example of the arrangement of the sound source S and spherical microphone arrays r1, r2, and r3 assumed in the sound field capture method of the present invention, as well as the sound field radiated from the sound source S within the region and the sound field due to interference waves arriving from outside the region.

[0023] The sound source S to be picked up is assumed to be located at the origin of the coordinate system. The coordinate system with the position of the sound source S to be picked up as the origin will be called the global coordinate system, in contrast to the local coordinate systems centered on the individual spherical microphone arrays r1, r2, and r3, which will be described later.

[0024] In the global coordinate system, TIFF2026025419000026.tif9170 is the position vector of the l (El)th spherical microphone array rl. In this figure, the position vectors of the 1st, 2nd, and 3rd spherical microphone arrays r1, r2, and r3 are TIFF2026025419000027.tif9170 is shown.

[0025] In addition, the coordinate system centered on each of the spherical microphone arrays r1, r2, and r3 (the coordinate system in which the center of each of the spherical microphone arrays r1, r2, and r3 is located at the origin) is defined as a local coordinate system. TIFF2026025419000028.tif9170 is a position vector expressed in a local coordinate system centered on the l (El)-th spherical microphone array rl. In this figure, the position vectors expressed in a local coordinate system centered on the 1st, 2nd, and 3rd spherical microphone arrays r1, r2, and r3 are TIFF2026025419000029.tif9170 is shown.

[0026] α{n'm'}(k) is the spherical harmonic spectrum (spherical harmonic function expansion coefficients) of the sound field radiated from the target sound source S located inside the area surrounded by the spherical microphone arrays r1,...,rL (spherical microphone arrays r1,r2,r3 in this figure) (hereinafter referred to as the "spherical harmonic spectrum of the target sound field"). β{n'm'}(k) is the spherical harmonic spectrum of the sound field due to interference waves arriving from outside the area surrounded by the spherical microphone arrays r1,...,rL (spherical microphone arrays r1,r2,r3 in this figure) (hereinafter referred to as the "spherical harmonic spectrum of the interference wave sound field").

[0027] n' and m' are the order and degree of the spherical harmonic spectrum, respectively. k is the wave number, which is obtained by dividing the angular frequency ω by the sound speed c (k=ω / c).

[0028] Here, the spherical harmonic spectrum α{n'm'}(k) of the sound field to be collected represents the sound pressure distribution of the sound field radiated from the sound source S to be collected that exists within the region (surrounded by the spherical microphone arrays r1,...,rL that make up the distributed spherical microphone array M) at wave number k (angular frequency ω). Also, the spherical harmonic spectrum β{n'm'}(k) of the interference wave sound field represents the sound pressure distribution of the sound field due to interference waves that arrive from outside that region at wave number k (angular frequency ω). Acquiring these independently is equivalent to separately collecting (capturing) the two sound fields.

[0029] Wave number k, position vector in the global coordinate system Sound field emitted from the sound source S to be recorded in TIFF2026025419000030.tif9170 TIFF2026025419000031.tif11170 is represented as follows:

number

[0030] hn' is the n'th order spherical Hankel function, Y{n'm'} is the n'th order m'th order spherical harmonic function, and Σn'Σm' is the infinite sum in n' and m'.

[0031] Here, by converting the global coordinate system shown in equation (1) above into a local coordinate system centered on the l (el)th spherical microphone array rl using the addition theorem of spherical wave functions, the sound field of equation (1) above can be transformed as follows:

number

[0032] jn is the nth order spherical Bessel function. TIFF2026025419000034.tif8170 is

number

[0033] TIFF2026025419000036.tif8170 is the complex conjugate of the spherical harmonic function, TIFF2026025419000037.tif9170 is the gaunt coefficient.

[0034] On the other hand, the wave number k and the position vector in the global coordinate system Sound field due to interference waves coming from outside the area surrounded by the spherical microphone array r1,...,rL in TIFF2026025419000038.tif9170 TIFF2026025419000039.tif10170 is represented as follows:

number

[0035] Here, by converting the global coordinate system shown in equation (4) above into a local coordinate system centered on the l (el)th spherical microphone array rl using the addition theorem of spherical wave functions, the sound field of equation (4) above can be transformed as follows:

number

[0036] Conversion Factor TIFF2026025419000042.tif8170 is

number

[0037] Therefore, when viewed in a local coordinate system centered on the l-th spherical microphone array rl, the sound field (summed sound field) obtained by adding these (the sound field radiated from the sound source S within the region and the sound field due to the interference wave arriving from outside the region) is TIFF2026025419000044.tif10170 is as follows.

number

[0038] The sum sound field shown in the above equation (7) is a spherical harmonic domain expression when viewed in a local coordinate system centered on the l (el)th spherical microphone array rl.

[0039] When scattering by a spherical microphone array rl is taken into account, the sound pressure on the surface of the spherical microphone array is TIFF2026025419000046.tif11170, that is, the sound pressure that can be observed at the qth microphone element rl^q arranged in the lth spherical microphone array rl can be expressed as follows:

number

[0040] However, for the sake of simplicity, the following definitions are used.

number

[0041] If the infinite sum of the sound pressure shown in equation (8) (the sound pressure observable by the q-th microphone element rl^q arranged in the l-th spherical microphone array rl) is truncated to a finite sum and expressed as a matrix, the result is as follows:

number

[0042] However, the following applies.

number

number

number

[0043] In addition, the spherical harmonic spectrum vector, whose elements are the spherical harmonic spectrum α{n'm'}(k) of the sound field to be picked up and the spherical harmonic spectrum β{n'm'}(k) of the interfering wave sound field, is TIFF2026025419000055.tif9170 is represented as follows:

number

[0044] Since Q microphone elements rl^1,...,rl^Q are arranged on the surface of the l (L)th spherical microphone array rl, the simultaneous equation for these is as follows:

number

number

[0045] Finally, for a distributed spherical microphone array M consisting of L spherical microphone arrays r1, . . . , rL, the above equations can be simultaneously formulated as follows:

number

[0046] Therefore, the time-frequency spectrum of the sound signal picked up by the distributed spherical microphone array M is TIFF2026025419000061.tif9170 and transfer matrix By performing signal processing in the spatial frequency domain using a separation matrix, which is the inverse matrix of TIFF2026025419000062.tif9170, and solving the simultaneous equations of the above equation (17), a spherical harmonic spectrum vector is obtained, which has the spherical harmonic spectrum α{n'm'}(k) of the sound field to be picked up and the spherical harmonic spectrum β{n'm'}(k) of the interference wave sound field as elements. TIFF2026025419000063.tif9170 using the following formula This can be obtained from TIFF2026025419000064.tif9170. TIFF2026025419000065.tif10170 is the separation matrix.

[0047] Furthermore, the transfer matrix If TIFF2026025419000066.tif9170 is an irregular matrix, the inverse matrix is ​​the separation matrix TIFF2026025419000067.tif10170 cannot be calculated. Therefore, the pseudo-inverse matrix TIFF2026025419000068.tif9170 is the separation matrix Replaces TIFF2026025419000069.tif10170.

[0048] In other words, it is possible to separately record the spherical harmonic spectrum α{n'm'}(k) of the sound field radiated from the sound source S to be recorded that is located within the area surrounded by the spherical microphone arrays r1,...,rL, and the spherical harmonic spectrum β{n'm'}(k) of the sound field due to interference waves arriving from outside the area surrounded by the spherical microphone arrays r1,...,rL, i.e., it is possible to individually obtain the sound pressure distributions of these sound fields.

[0049] [Sound recording device] Next, a description will be given of a sound collection device according to an embodiment of the present invention. Fig. 3 is a block diagram showing an example of the configuration of a sound collection device according to an embodiment of the present invention, and Fig. 4 is a flowchart showing an example of processing by the sound collection device according to an embodiment of the present invention.

[0050] The sound collection device 1 is a device that inputs a collection signal from a distributed spherical microphone array M, and collects sound by separating the sound field radiated from a sound source S within the area and the sound field due to interference waves arriving from outside the area based on a separation matrix calculated based on the position information and size of the distributed spherical microphone array M (distributed spherical microphone array position information) and the collection signal.

[0051] The sound collection device 1 includes a separation matrix calculation unit 10, a frequency analysis unit 11, and a spherical harmonic spectrum vector calculation unit 12.

[0052] The separation matrix calculation unit 10 receives preset distributed spherical microphone array position information. The distributed spherical microphone array position information includes position information of L spherical microphone arrays r1,...,rL (r1) constituting the distributed spherical microphone array M in the global coordinate system, position information of Q microphone elements r1,...,r1Q installed on the surface of each of the L spherical microphone arrays r1,...,rL (r1) in the local coordinate system, and the size (radius a) of each of the spherical microphone arrays r1,...,rL (r1).

[0053] The separation matrix calculation unit 10 calculates the following transfer matrix formula based on the distributed spherical microphone array position information: TIFF2026025419000070.tif9170 reflects the position information of the spherical microphone array rl, the position information of the microphone elements rl^q, and the size of the spherical microphone array rl, and is a transfer matrix expressed as a spherical harmonic function. TIFF2026025419000071.tif9170 is calculated, and the inverse matrix is ​​the separation matrix TIFF2026025419000072.tif10170 is calculated (step S401), and this is output to the spherical harmonic spectrum vector calculation unit 12. This transfer matrix calculation formula is a part of the above formula (17).

[0054] Here, in the transfer matrix calculation formula TIFF2026025419000073.tif9170 is expressed by the above formula (13) and calculated by the above formulas (3) and (6). TIFF2026025419000074.tif9170 is expressed by the above formula (12) and calculated by the above formula (9). TIFF2026025419000075.tif9170 is expressed by the above formula (16) and calculated by the above formula (11).

[0055] The frequency analysis unit 11 receives as input a time-domain sound collection signal from the distributed spherical microphone array M (step S402). The sound collection signal of the distributed spherical microphone array M is composed of sound collection signals of L spherical microphone arrays r1,...,spherical microphone array rL, and each of the sound collection signals of the spherical microphone arrays r1,...,rL (r1(l)) is composed of sound collection signals of M microphone elements r1(l)^1,...,sound collection signals of microphone elements r1(l)^Q.

[0056] Here, when the spherical microphone array rl (L) is expressed as a unit, the frequency analysis unit 11 inputs the collected signals of L spherical microphone arrays r1, ..., spherical microphone array rL that make up the distributed spherical microphone array M. Furthermore, when the microphone element rl (L)^q is expressed as a unit, the frequency analysis unit 11 inputs the collected signals of all microphone elements rl (L)^q for the L spherical microphone arrays r1, ..., rL that make up the distributed spherical microphone array M.

[0057] The frequency analysis unit 11 performs frequency analysis (Fourier transform in the time domain) on the sound pickup signal of the distributed spherical microphone array M, and obtains the time-frequency spectrum of the sound pickup signal of the distributed spherical microphone array M. TIFF2026025419000076.tif9170 is calculated (step S403) and output to the spherical harmonic spectrum vector calculation unit 12. This time-frequency spectrum is shown on the left side of the above-mentioned equation (17).

[0058] Here, when the spherical microphone array r1 is expressed as a unit, the frequency analysis unit 11 performs a Fourier transform in the time domain on each of the sound pickup signals of the L spherical microphone arrays r1, ..., and the sound pickup signals of the spherical microphone array rL that make up the distributed spherical microphone array M, thereby obtaining the time-frequency spectrum of the sound pickup signal of the spherical microphone array r1, ..., the time-frequency spectrum of the sound pickup signal of the spherical microphone array rL (the time-frequency spectrum of the sound pickup signal of the spherical microphone array r1). TIFF2026025419000077.tif10170 is obtained. This time-frequency spectrum is shown on the left side of the above-mentioned equation (15).

[0059] Furthermore, when expressed in units of microphone element r1^q, the frequency analysis unit 11 performs a Fourier transform in the time domain on each of the collected signals of all microphone elements r1^q for L spherical microphone arrays r1,...,rL that make up the distributed spherical microphone array M, thereby obtaining the time-frequency spectrum of the collected signal of the first microphone element r1^1 installed on the surface of the first spherical microphone array r1,..., the time-frequency spectrum of the collected signal of the Qth microphone element r1^Q installed on the surface of the Lth spherical microphone array rL (the time-frequency spectrum of the collected signal of microphone element r1^q). TIFF2026025419000078.tif11170 is obtained. This time-frequency spectrum is shown in the elements of the matrix on the right side of equation (15) and equation (10).

[0060] The spherical harmonic spectrum vector calculation unit 12 receives the separation matrix from the separation matrix calculation unit 10. TIFF2026025419000079.tif10170 is input, and the time-frequency spectrum of the picked-up signal is output from the frequency analysis unit 11. Enter TIFF2026025419000080.tif9170.

[0061] Based on these data, the spherical harmonic spectrum vector calculation unit 12 calculates the separation matrix by the calculation of the above-mentioned equation (17). TIFF2026025419000081.tif10170 and the time-frequency spectrum of the picked-up signal TIFF2026025419000082.tif9170, the spherical harmonic spectrum vector is obtained by multiplying the spherical harmonic spectrum α{n'm'}(k) of the sound field to be picked up and the spherical harmonic spectrum β{n'm'}(k) of the interference wave sound field, as shown in the formula (14). TIFF2026025419000083.tif9170 is calculated in the spatial frequency domain (step S404) and output.

[0062] By performing a spherical harmonic inverse expansion (equivalent to an inverse Fourier transform in the spatial domain) of the spherical harmonic spectrum α{n'm'}(k) of the sound field to be picked up and the spherical harmonic spectrum β{n'm'}(k) of the interference wave sound field, which constitute the spherical harmonic spectrum vector calculated in this way, the sound field radiated from the sound source S and the sound field due to the incoming interference wave can be obtained.

[0063] As described above, according to the sound collection device 1 of the embodiment of the present invention, the separation matrix calculation unit 10 calculates a separation matrix expressed in spherical harmonic functions based on the preset distributed spherical microphone array position information, which reflects the position information of all spherical microphone arrays rl (L) and all microphone elements rl (L)^q, as well as the size of the spherical microphone array rl (L).

[0064] The frequency analysis unit 11 receives the collected sound signal from the distributed spherical microphone array M and performs a Fourier transform on the collected sound signal in the time domain to obtain the time-frequency spectrum of the collected sound signal.

[0065] Based on the separation matrix and the time-frequency spectrum of the picked-up signal, the spherical harmonic spectrum vector calculation unit 12 calculates a spherical harmonic spectrum vector whose elements are the spherical harmonic spectrum of the sound field in the internal region of the distributed spherical microphone array M radiated from the sound source S and the spherical harmonic spectrum of the sound field due to interference waves arriving from the external region of the distributed spherical microphone array M, by calculating the equation (17).

[0066] As a result, by performing signal processing in the spatial frequency domain using a distributed spherical microphone array M consisting of L spherical microphone arrays r1,...,rL, it is possible to separate and collect the sound field radiated from the sound source S within the area surrounded by the distributed spherical microphone array M from the sound field due to the interference wave arriving from outside the area surrounded by the distributed spherical microphone array M, even if the statistical characteristics of the sound source S itself or the propagation characteristics of the interference sound are unknown.

[0067] Although the present invention has been described above with reference to the embodiments, the present invention is not limited to the above-described embodiments and can be modified in various ways without departing from the technical concept thereof.

[0068] For example, in the sound collection device 1 shown in Fig. 3, the spherical harmonic spectrum vector calculation unit 12 calculates a spherical harmonic spectrum vector having the spherical harmonic spectrum α{n'm'}(k) of the sound field to be collected and the spherical harmonic spectrum β{n'm'}(k) of the interference wave sound field as elements by solving the simultaneous equations of the above-mentioned formula (17). On the other hand, the spherical harmonic spectrum vector calculation unit 12 may calculate only the spherical harmonic spectrum α{n'm'}(k) of the sound field to be collected, or may calculate only the spherical harmonic spectrum β{n'm'}(k) of the interference wave sound field.

[0069] A typical computer can be used as the hardware configuration of the sound collection device 1 according to the embodiment of the present invention. The sound collection device 1 is configured by a computer equipped with a CPU, a volatile storage medium such as RAM, a non-volatile storage medium such as ROM, an interface, etc.

[0070] The functions of the separation matrix calculation unit 10, frequency analysis unit 11, and spherical harmonic spectrum vector calculation unit 12 provided in the sound collection device 1 are realized by causing a CPU to execute a program describing these functions. These programs are stored in the storage medium and are read and executed by the CPU. Furthermore, these programs can be stored in a storage medium such as a magnetic disk (hard disk, etc.), an optical disk (CD-ROM, DVD, etc.), or a semiconductor memory and distributed, or can be transmitted and received via a network. [Explanation of symbols]

[0071] 1. Sound pickup device 10 Separation matrix calculation unit 11 Frequency analysis section 12 Spherical harmonic spectrum vector calculation unit M-Distributed Spherical Microphone Array rl lth spherical microphone array The qth microphone element arranged in the rl^ql (el)th spherical microphone array α{n'm'}(k) is the spherical harmonic spectrum of the target sound field. β{n'm'}(k) Spherical harmonic spectrum of the interfering sound field S sound source a) Radius of the sphere of the spherical microphone array rl k wavenumber n' is the order of the spherical harmonic spectrum of the sound field to be estimated m' is the order of the spherical harmonic spectrum of the sound field to be estimated n is the order of the spherical harmonics of the sound field captured by the individual microphone arrays m is the order of the spherical harmonics of the sound field captured by the individual microphone arrays

Claims

1. A sound collection device that receives a sound collection signal from a distributed spherical microphone array M in which a plurality of spherical microphone arrays are distributed and arranged on a predetermined spherical surface, and collects a sound field radiated from a sound source S to be collected that exists within an area surrounded by the plurality of spherical microphone arrays, and a sound field due to interference waves arriving from outside the area, based on the sound collection signal, It is assumed that a plurality of microphone elements are arranged on the surface of each of the plurality of spherical microphone arrays, a separation matrix calculation unit that calculates a separation matrix expressed as a spherical harmonic function, which reflects positional information of the plurality of spherical microphone arrays that constitute the distributed spherical microphone array M, positional information of the plurality of microphone elements arranged on the surface of the spherical microphone array for each of the plurality of spherical microphone arrays, and the size of the spherical microphone array; a frequency analysis unit that receives input of pickup signals from all microphone elements of a plurality of spherical microphone arrays that constitute the distributed spherical microphone array M, and performs a Fourier transform of the pickup signals in the time domain to obtain a time-frequency spectrum of the pickup signals; a spherical harmonic spectrum vector calculation unit that calculates a spherical harmonic spectrum vector having elements of the spherical harmonic spectrum of the sound field in the internal region of the distributed spherical microphone array M radiated from the sound source S and the spherical harmonic spectrum of the sound field due to the interference wave arriving from the external region of the distributed spherical microphone array M, based on the time-frequency spectrum of the picked-up signal obtained by the frequency analysis unit and the separation matrix calculated by the separation matrix calculation unit; and A sound collection device comprising:

2. The sound collection device according to claim 1, A sound collection device characterized in that each of the plurality of spherical microphone arrays is made up of a rigid spherical baffle, and the plurality of microphone elements are arranged on the rigid spherical baffle.

3. A sound collection device that receives a sound collection signal from a distributed spherical microphone array M in which a plurality of spherical microphone arrays are distributed and arranged on a predetermined spherical surface, and collects a sound field radiated from a sound source S to be collected that exists within an area surrounded by the plurality of spherical microphone arrays, and a sound field due to interference waves arriving from outside the area, based on the sound collection signal, Assuming that the distributed spherical microphone array M is composed of L spherical microphone arrays, and that Q microphone elements are arranged on the surface of each of the L spherical microphone arrays, A separation matrix is ​​calculated based on the position information of the L spherical microphone arrays constituting the distributed spherical microphone array M, the position information of the Q microphone elements arranged on the surface of the spherical microphone array for each of the L spherical microphone arrays, and the radius a of the sphere of the spherical microphone array. a separation matrix calculation unit that calculates The sound pickup signals of all the microphone elements of the L spherical microphone arrays constituting the distributed spherical microphone array M are input, and the sound pickup signals are Fourier-transformed in the time domain to obtain a time-frequency spectrum of the sound pickup signals. (k is the wave number), α{n'm'}(k) is the spherical harmonic spectrum of the sound field radiated from the sound source S (n', m' are the order and magnitude of the spherical harmonic spectrum, respectively), β{n'm'}(k) is the spherical harmonic spectrum of the sound field due to the interference wave, is defined as a spherical harmonic spectrum vector having the spherical harmonic spectrum α{n'm'}(k) and the spherical harmonic spectrum β{n'm'}(k) as elements, The following formula: a spherical harmonic spectral vector calculation unit that calculates the spherical harmonic spectral vector by the calculation of l (L) is the number of the spherical microphone array (l = [1, L]), q is the number of the microphone element (q = [1, Q]), a coordinate system in which the sound source S is located at the origin is a global coordinate system, and a coordinate system in which the center of the spherical microphone array is located at the origin is a local coordinate system. is the position vector of the global coordinate system in the l-th spherical microphone array, is the position vector in the local coordinate system of the qth microphone element installed on the surface of the lth spherical microphone array, and i is the imaginary unit. The separation matrix calculation unit The transfer matrix formula is: The inverse matrix of this is the separation matrix shall be calculated, In the transfer matrix equation to the following formula: Calculate by to the following formula: Calculation is performed by j q is the qth order spherical Bessel function, is the complex conjugate of the spherical harmonic function, is the gaunt coefficient, to the following formula: Calculation is performed by h q (2) is the q-order Hankel function of the second kind, In the transfer matrix equation to the following formula: Calculate by to the following formula: where h'n is the derivative of the n-th order spherical Hankel function, In the transfer matrix equation to the following formula: Calculate by to the following formula: and Y{n, m} is an n-th order, m-th order spherical harmonic function.

4. A program for causing a computer to function as the sound collection device according to any one of claims 1 to 3.

Images