Patent Translate Powered by EPO and Google Notice This translation is machine-generated. It cannot be guaranteed that it is intelligible, accurate, complete, reliable or fit for specific purposes. Critical decisions, such as commercially relevant or financial decisions, should not be based on machine-translation output. DESCRIPTION JP2013110495 Abstract: A local reproduction device capable of setting a reproduction region at an arbitrary place, and a filter coefficient generation device generating a digital filter coefficient used in the local reproduction device are provided. SOLUTION: A filter coefficient H (ω) is an element with the position of the speaker and the total number N, the position of the control point and the total number M, and the control amount C specifying the value of the complex amplitude at the control point. Complex matrix H of N rows by 1 column, M matrix of G rows by N columns of elements having transfer function G (ω) from speaker n to control point m, complex conjugate transpose of the complex matrix G Assuming that G is a complex matrix of M rows × 1 column having C as an element, the filter coefficients are calculated by solving a linear equation of GH = C for the complex matrix H, and in the case of M = N, In the case of H = GC, M> N, H = (GG) GC, and in the case of M <N, H = G (GG) C is obtained (G is a complex conjugate transpose matrix of G). [Selected figure] Figure 1 Local reproduction device and filter coefficient generation device, and methods and programs therefor [0001] The present invention relates to a local reproduction device and a filter coefficient generation device capable of transmitting sound only to people in a specific place, and methods and programs therefor. [0002] 11-04-2019 1 When an acoustic signal is reproduced using a speaker, it is possible to listen to the reproduced sound from almost all directions with respect to the speaker although there is an influence of the directional characteristic of the speaker. Therefore, when aiming at construction of a local sound reproduction system which reproduces a sound only in a specific area (area), it is necessary to devise a loud speaker such as a speaker and a reproduction system. [0003] If it is possible to transmit the sound only to people in a specific place, the reproduced sound will not be a noise to people other than the listener when performing communication by loud voice. In addition, since communication contents do not leak around, privacy can be protected. [0004] As a reproduction method devised to confine sound in a specific area, the multipoint control method disclosed in Non-Patent Document 1 is known. The multipoint control method is a method of controlling emitted sound using a plurality of speakers and digital filters so as to obtain desired sound pressure characteristics at a plurality of control points arranged in space. Assuming that the transfer function from a certain speaker n to a certain control point m is Gmn (ω), a filter coefficient Hn (ω) for determining the radiation sound of each speaker is obtained so as to satisfy equation (1). [0005] [0006] The vector on the right side of Equation (1) specifies the value of the complex amplitude corresponding to the desired sound pressure characteristic at each control point. [0007] 11-04-2019 2 An exemplary functional configuration of the local reproduction device 900 based on this idea is shown in FIG. The local reproduction device 900 includes speaker arrays 91 and 92, and digital filters 93 and 94 connected to the respective speakers. The reproduction control point a is disposed at the center (origin) where the speaker arrays 91 and 92 are disposed to face each other on the sound emission side of the speaker. The suppression control points c are disposed at a plurality of locations spaced apart on the circumference of a semicircle of radius R centered on the reproduction control point a. A plurality of attenuation control points b are respectively disposed outside from positions of R and -R on the y axis orthogonal to the reproduction control point a, and are arranged in a direction away from the origin from the position of R in the x axis. [0008] Then, a control amount specifying the complex amplitude value of each control point is given as shown in equation (2). [0009] [0010] Here, L <−s> b * represents the distance from the origin to the attenuation control point b *, and s is an arbitrary constant. By arranging the damping control point b * outside the reproduction area inside the radius R, the attenuation characteristics of the sound pressure outside the reproduction area can be sharply attenuated. [0011] Hiroaki Ito, Kenichi Furuya, Yoichi Haneda, "Attenuation Control of Sound Waves by Multipoint Control for Area Reproduction" Proceedings of the 2011 IEICE General Conference, p. 174, 2011. 11-04-2019 3 [0012] However, in the conventional method, a speaker array in which a plurality of speakers are linearly arranged is made to face each other, and the middle point (center) is designated as the reproduction area, so the directivity characteristics of the sound from each speaker is It is controlled concentrically around the middle point. Therefore, there is a problem that the reproduction area can not be set at an arbitrary place. [0013] The present invention has been made in view of this problem, and it is an object of the present invention to provide a local reproduction device and a filter coefficient generation device capable of setting a reproduction region at an arbitrary place, and methods and programs thereof. Do. [0014] The filter coefficient generation device according to the present invention receives as input the position of the speaker and the total number N thereof, the position of the control point and the total number M thereof, and the control amount Cm specifying the value of the complex amplitude at the control point. A complex matrix H of N rows × 1 column having Hn (ω) as an element, a complex matrix G of M rows × N columns having transfer function Gmn (ω) from the speaker n to the control point m as an element, its complex matrix G The filter coefficient is calculated by solving the following linear equation with respect to the complex matrix H, where G <†> is a complex conjugate transpose matrix of and M is a complex matrix of M rows × 1 column having a control amount Cm as an element A filter coefficient generator, [0015] [0016] According to the magnitude relationship between the total number N of speakers and the total number M of control points, the complex matrix H can be expressed by [0017] 11-04-2019 4 [0018] Ask for. [0019] Further, the local reproduction device of the present invention comprises two or more speakers and a signal supply unit. Two or more speakers are disposed at arbitrary positions. Then, three or more suppression control points provided at different positions from each speaker where the value of the complex amplitude at an arbitrary position is 0 are different, and the complex amplitude at a position between each speaker and the suppression control point A reproduction control point having a value of 1 is provided, and the filter coefficient of the digital filter in the signal supply unit is a value generated by the above-described filter coefficient generation device. [0020] According to the filter coefficient generation device of the present invention, the filter coefficient capable of setting the reproduction region in any place by the speaker arbitrarily arranged based on the magnitude relation between the total number N of the speakers and the total number M of the control points Can be generated. Further, the local reproduction apparatus of the present invention uses the filter coefficients obtained by the filter coefficient generation apparatus of the present invention to perform reproduction control at any position between two or more speakers and three or more suppression control points. You can set points. [0021] 11-04-2019 5 The figure which shows the function structural example of the filter coefficient determination apparatus 100 of this invention. FIG. 7 is a diagram showing an operation flow of the filter coefficient determination device 100. FIG. 2 is a diagram showing an example of a functional configuration of a local reproduction device 200 of the present invention. FIG. 2 is a diagram showing an example of a functional configuration of a local reproduction device 300 of the present invention. FIG. 7 shows an example of a functional configuration of a local reproduction device 400 according to the present invention. The figure which shows the conditions of simulation. The figure which shows a simulation result. FIG. 18 is a diagram showing an example of a functional configuration of a local reproduction device 900 disclosed in Non-Patent Document 1. [0022] Hereinafter, embodiments of the present invention will be described with reference to the drawings. The same reference numerals are assigned to the same components in the drawings, and the description will not be repeated. [0023] [Filter Coefficient Generating Device] FIG. 1 shows a functional configuration example of the filter coefficient generating device 100 of the present invention. The filter coefficient generation device 100 generates filter coefficients of digital filters included in a local reproduction device described later, and includes the positions of the speakers and the total number N thereof, the positions of the control points and the total number M thereof, and the respective control points. A filter coefficient corresponding to the total number N of speakers and the total number M of control points is automatically input with a control amount Cm specifying the value of the complex amplitude and the presence or absence of a constraint condition for constraining the 11-04-2019 6 value of the complex amplitude at the control point New in that it generates. [0024] The filter coefficient generation device 100 includes a transfer function calculation unit 10 and a filter coefficient calculation unit 20. Transfer function calculation unit 10 receives speaker position Xn = [xn, yn, zn] and control point position Xm = [xm, ym, zm] as input, and provides a transfer function Gmn (ω from the speaker to each control point ) Is obtained by the following equation. [0025] [0026] Here, ‖Xm−Xn‖ represents the Euclidean distance from the speaker n to the control point m. c is the speed of sound. ω is an angular frequency, and the relationship with frequency f is expressed as ω = 2πf. [0027] The filter coefficient calculation unit 20 has the transfer function Gmn (ω), the total number N of speakers of the local reproduction device, the total number M of control points, the control amount Cm specifying the value of the complex amplitude at the control points, and the constraint condition The presence or absence of is input. The constraint condition is a condition in which the observed sound pressure (complex amplitude) at the control point is always the control amount, and the constraint condition is that there is a point where you want to reproduce the reproduced sound and a point where you want to suppress the sound pressure. Set with. [0028] 11-04-2019 7 The operation of the filter coefficient calculation unit 20 will be described in the case where there is no constraint condition. In the filter coefficient calculation unit 20, the position of the speaker and its total number N, the position of the control point and its total number M, the control amount Cm specifying the value of the complex amplitude at the control point, and the value of the complex amplitude at the control point A filter coefficient Hn (ω) given to the digital filter is calculated and output with the presence or absence of a constraint condition to be constrained as an input. The control amount Cm is determined as follows. [0029] [0030] When there is no constraint condition (NO in step S21), an element of complex matrix H of N rows × 1 column having filter coefficient Hn (ω), and transfer function Gmn (ω) from speaker n to control point m are elements. Assuming that a complex matrix G of M rows × N columns and a complex matrix of M rows × 1 columns having the control amount Cm as elements, filter coefficients Hn are obtained by solving the linear equation of equation (5) for the complex matrix H Calculate (ω). [0031] [0032] Therefore, the complex matrix H is calculated by the following equation based on the magnitude relationship between the total number N of speakers and the total number M of control points. [0033] [0034] The operation will be described with reference to the operation flow of the filter coefficient determination device 100 shown in FIG. The transfer function calculation unit 10 receives the position Xn of the speaker Xn = [xn, yn, zn] and the position Xm of the control point Xm = [xm, ym, zm] as input from the position 11-04-2019 8 information of each speaker n and the control point m The distance ‖Xm-Xn‖ between the speaker and the control point is calculated (step S11). Then, a transfer function Gmn (ω) from the speaker to each control point is obtained by equation (3) (step S12). [0035] When there is no constraint condition (NO in step S21), the filter coefficient calculation unit 20 calculates the filter coefficient Hn (ω) when the total number N of speakers and the total number M of attenuation control points are equal (YES in step S22). Is calculated as Hn (ω) = G <−1> C (Equation (6-1)) (step S23). [0036] When the total number M of attenuation control points is larger than the total number N of speakers (YES in step S24), the filter coefficient Hn (ω) is obtained as an approximate solution by the least square method. Let ε (Eq. (7)) be the square of the error between the right and left sides of the above equation (5). [0037] [0038] Find H such that the square ε of this error is minimized. [0039] [0040] Equation (8) can be solved by setting ∂ε / ∂H = 0 (Equation (6-2) step S25). 11-04-2019 9 [0041] [0042] If the total number M of control points is smaller than the total number N of speakers (NO in step S24), the filter coefficient Hn (ω) is obtained as a solution that minimizes its own nom ‖H‖. When the solution of equation (4) described above is expressed as H = G <-> C, G <-> is defined as a general inverse matrix of G. When nomul ‖ H ‖ is minimum, G satisfies the following equation (step S26). [0043] [0044] From the above, [0045] [0046] したがって、 [0047] [0048] The above process is repeated until the calculation of all the filter coefficients Hn (ω) is completed (YES in step S27). [0049] When there is a constraint condition (YES in step S21), the control amount at the control point 11-04-2019 10 provided with the constraint condition in the control point is expressed as C1 = [1, 0, 0, ..., 0] <T>, and the control point The following linear equation holds in [0050] [0051] A linear equation that holds at control points other than the control point for which the constraint condition is provided is [0052] [0053] Let H be the complex conjugate transpose of the complex matrix H and the space correlation matrix K be K = G2G2 <T> under the condition that the constraint at the constraint control point holds. The squared solution [0054] [0055] A filter coefficient Hn (ω) solved for the filter coefficient Hn (ω) is determined. Such a conditional minimization problem can be solved by using Lagrange's undetermined method. Therefore, the evaluation function Q (H) to be minimized is defined as follows based on this method. [0056] 11-04-2019 11 [0057] Here, Λ is a column vector consisting of M undetermined coefficients. The filter coefficient can be determined by placing the result of differentiating the evaluation function by H equal to zero. [0058] [0059] Substituting G1H = C1 to determine Λ, [0060] [0061] By substituting equation (16) into equation (15), filter coefficients are given as follows (step S28). [0062] [0063] Although there has been described an example of changing the formula for obtaining the complex matrix H having the filter coefficient Hn (ω) as an element by the magnitude relationship between the total number N of speakers and the number M of control points when there is no constraint condition. If there is no constraint condition, the filter coefficient may not be generated (NO in step S21 to the dashed line at the end). In addition, a filter coefficient generation device may be configured not to generate a filter coefficient when there is a constraint condition (YES in step S21-dashed line in end). 11-04-2019 12 That is, the filter coefficient generation device of the present invention may be configured as a dedicated device corresponding to the presence or absence of the constraint condition. [0064] [Local Reproduction Device] FIG. 3 shows an example of a functional configuration of the local reproduction device 200. As shown in FIG. The local reproduction device 200 has two or more speakers a 1,..., A N (where N is a positive number of 2 or more) whose sound emission directions are also arranged at arbitrary positions. Let each speaker be an. And a signal supply unit 40 for generating a supply signal for convoluting in the time domain the digital filter coefficient corresponding to each speaker an in the input digital sound signal to each speaker an. Then, three or more suppression control points arranged at different positions from the respective speakers an whose value of the complex amplitude is 0, which is a control point specifying the value of the complex amplitude at any position, are 0 (see FIG. The suppression control point in 中 is attached, and the reproduction control point ◎, which is a control point for setting the value of complex amplitude to 1, at an arbitrary position between each speaker an and the suppression control point The coefficients are filter coefficients generated by the filter coefficient generation apparatus 100 described above. The suppression control point x may be provided to surround the speaker an and the reproduction control point 点. The speaker an may be disposed at any position on the two-dimensional plane, or may be disposed at any position on the three-dimensional space. 11-04-2019 13 When the number of speakers an is two, the number of suppression control points may be three or more, and the number of suppression control points may be three or more even when the number of speakers an is N. [0065] The signal supply unit 40 is configured such that the serial configuration of the A / D converter 41 n, the digital filter 42 n, the D / A converter 43 n, and the AMP 44 n connected to each speaker an is prepared in parallel by the number of speakers an. Be done. Then, one acoustic signal is input to the plurality of A / D converters 41 n. The digital filter coefficients Hn (ω) calculated by the above-described filter coefficient generation device 100 are set in each of the digital filters 42 n. Each digital filter 42 n generates a digital drive signal Dn (ω) for driving each speaker an by convoluting the filter coefficient Hn (ω) with the input digital sound signal, and D corresponding to this drive signal is generated. / A converter 43 n. The drive signal Dn (ω) can be expressed by the following equation. [0066] [0067] Digital filter coefficients Hn (ω) are generated for each of Q angular frequency regions ω (q). Assuming that the number of frequency band divisions is Q and q = 0, 1,..., Q-1, the angular frequency band can be expressed by the following equation. 11-04-2019 14 [0068] [0069] Here, fs is a sampling frequency of the A / D converter 41 n. [0070] The filter coefficient generation device 100 calculates the filter coefficient Hn (ω (q)) in the angular frequency band ω (q), and performs inverse Fourier transform of it to set as the filter coefficient in each digital filter 42 n. If the value of the frequency band division number Q is increased, the accuracy as a filter is increased but the calculation cost is increased. The division number Q can be set arbitrarily, and is, for example, a value of several hundreds. In addition, the operation can be speeded up by setting to a power of two. It should be noted that the A / D converter 41 n may not be provided if the input acoustic signal is a digital signal. Also, the frequency band may not be an angular frequency. [0071] FIG. 4 shows an example of the functional configuration of the local reproduction device 300. As shown in FIG. Three or more speakers a1, ..., aN (where N is a positive number of 3 or more). 11-04-2019 15 Let each speaker be an. ) Are arranged in such a way that a straight line forms a plane. Four or more suppression control points x are provided on the circumference of a predetermined radius R from an origin α of a coordinate system in which three or more speakers an are arranged. Then, a reproduction control point ◎ is provided between three or more speakers an and the suppression control point. Four or more suppression control points x are arranged on a threedimensional space. The suppression control point x is provided so as to surround the speaker an and the reproduction control point 抑 圧. [0072] For example, as shown in FIG. 5, the three speakers an may be arranged at corners of the xy plane of the three-dimensional coordinate system. In that case, the suppression control point x may be provided at an arbitrary position on the predetermined radius r from the origin α on the same xy plane. Further, the reproduction control point ◎ is provided on the same xy plane. In this example, it is possible to improve the control accuracy of the sound pressure of the reproduction control point on the same xy plane. [0073] Simulation Results In order to confirm the performance of the local reproduction apparatus of the present invention, computer simulation was performed. The experimental conditions are shown in Table 1 and FIG. [0074] The centers of the speaker units, each of which consists of 12 speakers, are (0.4, 0), (-0.4, 0), (0, 11-04-2019 16 0.4), (0) in the xy coordinate system. , -0.4), with the sound emission direction directed to the origin. And, the reproduction control point 設 け was provided at the position of (0.1, 0.25). FIG. 7 shows the result of measuring the sound pressure on y = 0.25. The horizontal axis is the position on the x axis, and the vertical axis is the sound pressure [dB]. It can be seen that the sound pressure on the x coordinate of the reproduction control point ◎: 0.1 is controlled as high as 0 [dB]. [0075] Thus, according to the local reproduction apparatus of the present invention, it is possible to limit the reproduction area of the acoustic signal to a specific area. In the filter coefficient generation device 100 and the specific area sound reproduction devices 200, 300, and 400, a predetermined program is read into a computer including, for example, a ROM, a RAM, and a CPU, and the CPU executes the program. It may be configured to be realized. [0076] In that case, the program in which the processing content is described can be recorded in any computer-readable recording medium. Examples of the computer readable recording medium include a magnetic recording device, an optical disc, a magneto-optical recording medium, and a semiconductor memory. More specifically, for example, as a magnetic recording device, a hard disk drive, a flexible disk, a magnetic tape or the like as an optical disk, a DVD (Digital Versatile Disc), a DVD-RAM (Random Access Memory), a CD-ROM (Compact Disc Read) (Only Memory), CD-R (Recordable) / RW (Rewritable), etc. as a magneto-optical recording medium, MO (Magneto Optical disc), etc. as a semiconductor memory, EEP-ROM (Electronically Erasable and Programmable Only Readable Memory), etc. Can be used. [0077] Further, the distribution of this program is carried out, for example, by selling, transferring, lending, etc. a portable recording medium such as a DVD, a CD-ROM, etc. in which the program is recorded. Furthermore, the program may be stored in a storage device of a server computer, and the program may be distributed by transferring the program from the server computer to another computer via a network. 11-04-2019 17 [0078] Further, each means may be configured by executing a predetermined program on a computer, or at least a part of the processing content may be realized as hardware. 11-04-2019 18

1/--страниц