JP5342521B2 - Local reproduction method, local reproduction device and program thereof - Google Patents

Description

  The present invention relates to a local reproduction method, a local reproduction apparatus, and a program thereof that can transmit sound only to people in a specific place.

  When sound is radiated using a speaker, the reproduced sound can be heard from almost all directions with respect to the speaker, although there is an influence of the directivity characteristic of the speaker. Therefore, when aiming at the construction of a local reproduction system that reproduces sound only at a specific place, it is necessary to devise a loudspeaker such as a speaker or a reproduction system.

  If the sound can be reproduced only in a specific place, it is possible to prevent the reproduced sound from becoming a noise for people other than the listener, for example, in the case of performing communication by voice amplification. Further, it is possible to prevent the communication content from leaking to the surroundings, thereby improving the confidentiality.

  As a method of realizing this local reproduction method, there is a method of using a directional speaker system that emits sound only in a certain direction. Examples of directional speakers include horn speakers, parametric speakers, analog speaker arrays, and digital filter type speaker arrays.

Furthermore, based on the boundary sound field control theory that if the sound pressure on the boundary is controlled, the outside sound field can be controlled, Patent Document 1 discloses that the sound pressure on the boundary is controlled to be zero. A technique for realizing local regeneration is described. A local reproduction apparatus 10 of Patent Document 1 will be described with reference to FIG. The local reproduction device 10 includes M (M ≧ 2) sensors 11 _m (m = 1, 2,..., M) arranged around a space A in which sound is to be reproduced. The local regeneration device 10 calculates filter coefficients of N (N ≧ 2) digital filters 12 _n (n = 1, 2,..., N) based on outputs from these M sensors. Local reproduction device 10 receives the input signal S through an input terminal 14, using N digital filters 12 _n, and the delay between mutual sound to be reproduced from the N speakers 13 _n, controls the frequency characteristic To do. With this control, the energy of the sound reproduced by the speaker 13 _n is reduced outside the space A, and the local reproduction device 10 reproduces the sound only within the space A.

JP 2006-74442 A

  However, since the filter design based on the signal processing method is performed on the computer in Patent Document 1, there is a discrepancy between the theory and the actual environment, and the local reproduction effect due to sound leakage due to control error and discontinuity of control points. There was a problem that would decrease.

  An object of the present invention is to provide a local reproduction method, an apparatus, and a program for eliminating a control error and improving a local reproduction effect.

In order to solve the above problems, the local reproduction method according to the present invention uses a speaker array including two or more speakers arranged on the same plane or in a straight line. A filter coefficient for generating an evanescent wave is calculated using the wave number k _{x in the} x direction and the wave number k _{y in the} y direction, a drive signal is generated using the input signal and the filter coefficient, and the drive signal is reproduced. .

  The present invention has an effect that local reproduction can be performed without a control error using an evanescent wave having a steep distance attenuation characteristic.

The block diagram of the conventional local reproduction | regeneration apparatus 10. FIG. The figure which shows the audible space A of the local reproduction | regeneration apparatus 100. FIG. The figure which shows the state in which the plane wave B of the wave number k exists in the half space z <0 in xyz space. The figure for demonstrating an evanescent wave. The top view of the local reproduction | regeneration apparatus 100. FIG. The figure which shows the structural example of the local reproduction | regeneration apparatus 100. FIG. The figure which shows the processing flow of the local reproduction | regeneration apparatus 100. FIG. The figure which shows the relationship at the time of reproducing | regenerating a signal using a speaker array without a filter, a frequency, and a sound pressure. The figure which shows the relationship at the time of reproducing | regenerating a drive signal using the local reproduction | regeneration apparatus 300, a frequency, and a sound pressure.

[Points of Invention]
By reproducing an evanescent wave having a characteristic of abruptly decaying in the distance direction using a speaker array, a local reproduction device 100 that can hear sound near the device as shown in FIG. (In FIG. 2, A represents an audible space). First, the evanescent wave will be described with reference to FIGS.

<Evanescent wave>
As shown in FIG. 3, in a half space z <0 in the xyz space, a plane wave B having a wave number k exists, and a sound pressure p (x, y, 0) on the z = 0 plane (xy plane) is given. The sound pressure p (x, y, z) in the half space z> 0 is obtained by the following equation (see Reference 1).
[Reference 1] EG Williams, “Fourier Acoustics Sound Radiation and Basics of Near Field Acoustic Holography”, Springer Fairlark Tokyo, 2005.

Where k ′ _x , k ′ _y , and k ′ _z represent wave numbers in the x, y, and z directions at (x, y, z), respectively, and x, y, and z in (x, y, z) represent variables. . Note that P (k ′ _x , k ′ _y ) is called an angular spectrum and is obtained as follows.

For example, as shown in FIG. 3, when the plane wave B propagates from a certain direction, p (x, y, 0) is given by the following equation.

Where A (k) is an arbitrary _{constant, k x,} k _y is the (x, y, 0) in x, y directions of the wave number (constant), x of the (x, y, 0), y Represents a fixed value. At this time, when the sound pressure in the half space z> 0 is obtained from the equations (1) and (2), the following equation is obtained.

In the wave number space, the following relationship holds.

When the medium through which the sound wave propagates is uniform in the half space z ≦ 0 and the half space z> 0, k _x ≦ k and k _y ≦ k. Therefore, when equation (5) is solved for k _z , expressed.

Substituting this into equation (4) gives

Equation (7) represents a situation where the plane wave B is propagating.

On the other hand, if the medium through which the sound wave propagates is not uniform, k _x > k or k _y > k. At this time, k _z is given by equation (8).

Substituting a positive sign term into equation (4),

This represents a wave that exponentially decays in the z direction and is called an evanescent wave (see Reference 1). In the field of light and electromagnetic waves, it is known as a wave generated on a boundary surface when total reflection occurs between different media.

As shown in FIG. 4, when a wave is propagating to one of the two media (medium 1 and medium 2 in FIG. 4) having different characteristics (for example, medium 1), the propagating wave (plane wave B) is different from the medium. When total reflection occurs at the boundary between 1 and medium 2, the wave E is transmitted at the boundary. This transmitted wave is the aforementioned evanescent wave. Conditions evanescent wave is generated is to total reflection at the boundary of different media, wavenumber k _z in the z direction is to become an imaginary number. At this time, k _x > k or k _y > k. Considering the relationship k = ω / c that holds between the wave number k, the angular frequency ω, and the sound speed c, in order to realize the reproduction of the evanescent wave, the medium 1 has a slower sound speed than the medium 2 that emits the sound. It is necessary to generate a plane wave.

  However, it is difficult to prepare a medium that satisfies the above conditions. Therefore, the difference of the medium is virtually reproduced by using the delay sum type speaker array to reproduce the evanescent wave. The radiation sound of a normal speaker has the property of being attenuated in inverse proportion to the distance, but the evanescent wave has an exponential attenuation term in the z direction as shown in the equation (9), so it is a steeper distance than normal radiation. It can be seen that it has an attenuation characteristic. Therefore, if an evanescent wave reproduction device can be constructed, a local reproduction effect can be realized in which sound can be heard in the vicinity of the device, but sound cannot be heard if the device is moved a little away from the device.

<Realization of equation (1) by speaker array>
It is necessary to rewrite Equation (1) for reconstructing the sound field into a form that can be realized by a speaker array. Expression (1) has the same value relationship as the second type Rayleigh integral (see Reference 1), and an expression obtained by discretizing it is expressed as follows.

m and n are indices for speakers in the x and y directions, respectively, m = 1, 2,..., M, n = 1, 2,..., N, and the position of the speaker 13 _mn on the xy plane is (x _m , Y _n ), k is the wave number of the input signal S, the complex spectrum of the input signal S is S (k), and the filter coefficient is H _mn (k). | R−r _mn | represents a distance between r = (x, y, z) and r _mn = (x _m , y _n , 0). The delta _x and spacing of speakers arranged in the x direction, the distance between the speaker arranged delta _y in the y-direction. Hereinafter, an apparatus configuration for realizing the expression (10) will be described.

  The present invention pays attention to the theory of Fourier acoustics, and reproduces the sound pressure distribution in a three-dimensional space by reproducing the sound pressure distribution in a certain plane by a speaker array in which a plurality of speakers are arranged on a plane. Regarding the method of specifying the wave number, which is a parameter relating to the filter coefficient for determining the sound pressure distribution in the plane, the maximum value of the sound source wave number in the first embodiment, the constant multiple of the sound source wave number in the second embodiment, and the distance attenuation in the third embodiment. An example focusing on term control is shown.

  Hereinafter, embodiments of the present invention will be described in detail.

<Local reproduction apparatus 100>
The local reproduction apparatus 100 according to the first embodiment will be described with reference to FIGS. The local reproduction apparatus 100 includes M × N speakers 13 _mn , M × N A / D converters 105 _mn , M × N digital filters 112 _mn , and M × N D / A conversions. Device 107 _mn , M × N speaker amplifiers 109 _mn , a filter coefficient calculation unit 130, and a wave number calculation unit 120.

The local reproduction device 100 receives the analog input signal S via the signal input terminal 14 (s1). The local reproduction device 100 reproduces an analog drive signal D (x _m , y _n , k) corresponding to the input signal S using N × M speakers 13 _mn (s5), and outputs an acoustic signal. Hereinafter, the processing content of each part is demonstrated.

<A/D_henkanki_105 _mn>
The A / D converter 105 _mn receives the analog input signal S corresponding to the reproduction signal from the signal input terminal 14, converts it into a digital input signal, and outputs it to the corresponding digital filter 112 _mn .

<Digital filter 112 _mn >
The digital filter 112 _mn sets the filter coefficient H _mn (k) calculated by the filter coefficient calculation unit 130 described later. A digital input signal S is input to each digital filter 112 _mn . Each digital filter 112 _mn converts the input digital signal S into a digital drive signal D (x _m , y _n , k) for generating an evanescent wave (s4).

In order to virtually reproduce a medium that propagates slower than the speed of sound in the air, the filter coefficient H _mn (k) is given as follows.

By specifying the _{k x} and _{k y} As this time satisfy ^{_{k x 2 + k y 2>}} k 2, the wave number _{k z} in the z-direction becomes imaginary (Equation (8), (9)), the evanescent wave It is thought that it will be regenerated. The filter coefficient calculation method will be described in the filter coefficient calculation unit 130.

The digital filter 112 _mn generates a drive signal D (x _m , y _n , k) by convolving the complex spectrum S (k) of the input digital signal with the filter coefficient H _mn (k) (s4), The drive signal D (x _m , y _n , k) is output to the corresponding D / A converter 107 _mn . Note that D (x _m , y _n , k) = S (k) H _mn (k).

<D / A converter 107 _mn and speaker amplifier 109 _mn >
The D / A converter 107 _mn converts the digital drive signal D (x _m , y _n , k) into an analog drive signal and outputs it to the corresponding speaker amplifier 109 _mn .

The speaker amplifier 109 _mn amplifies the analog drive signal D (x _m , y _n , k) and supplies the amplified signal to the speaker 13 _mn .

<Speaker 13 _mn >
The speaker 13 _mn reproduces the analog drive signal D (x _m , y _n , k) corresponding to the input signal S (s5). As shown in FIG. 5, M-number in delta _x intervals in the x direction, N pieces in delta _y intervals in the y direction, arranged a total of M × N loudspeakers 13 _mn on the same plane, constitute a speaker array. A plane formed by the speaker array is defined as an xy plane in the xyz orthogonal coordinate system. Each speaker 13 _mn uses a dipole sound source.

When k _y = 0, considering the spatial aliasing, the speaker interval Δ _{x in} the x direction needs to be equal to or less than the half wavelength (λ _x / 2) of the wave determined by the wave number k _x . Since the relationship between the wavelength λ _x and the wave number k _x is λ _x = 2π / k _x , the speaker interval Δ _x needs to _satisfy Δ _x ≦ π / k _x . If this relationship is satisfied, the smaller the speaker interval, the better. As for delta _y since no upper limit of the loudspeaker spacing, better smaller if the interval delta _y is smaller. _Similarly, when a _k x = 0, the speaker interval delta _y in the y-direction is required to be Δ _y ≦ π / _{k y.}

<Wave number calculation unit 120>
The wave number calculation unit 120 calculates the wave number k _{x in} the x direction and the wave number k _{y in the} y direction using the wave number k of the input signal S (s2), and outputs the wave number k _x to the filter coefficient calculation unit 130.

As apparent from the equation (11), the filter coefficient H _mn (k) is specified by the wave number k _{x in} the x direction and the wave number k _{y in the} y direction. _{k x} ² + k ^{y 2>} set of _{k x} and _{k y} so as to satisfy the ^{k 2} _(k x, _{k y)} By specifying a wave number _{k z} in the z-direction becomes imaginary (Equation (8), (9 )), And evanescent waves are considered to be reproduced. Therefore, the local playback apparatus 100, the wave number calculation unit 120, by using the wave number k of the input signal ^{_{S, k x 2 + k y}} 2> satisfy ^{k 2} set _(k x, _{k y)} is calculated. For example, when k _y = 0, the filter coefficient H _mn (k) that can reproduce the evanescent wave can be calculated by using k _x satisfying k _x ² > k ² , that is, k _x > k. it can. Similarly, when k _x = 0, k _y satisfying k _y > k may be used.

In the case of k _y = 0, the wave number calculation unit 120 uses the maximum value k _{max of the} wave number included in the input signal S, multiplies the predetermined value α by the maximum value k _max , and sets the wave number k _x to k _x = Calculated as αk _max . Alternatively, when k _x = 0, the wave number calculation unit 120 calculates the wave number k _y as k _y = αk _max . α is a predetermined value larger than 1. For example, the value of α is about 1 to several tens, and if the value is changed, the position of the frequency that does not attenuate tends to change, so that the attenuation characteristic can be controlled according to the frequency characteristic of the input signal. Note that k _max can be obtained from the wave number included in the input signal S.

<Filter coefficient calculation unit 130>
The filter coefficient calculation unit 130 calculates a filter coefficient used in the digital filter 112 _mn using the wave number k _{x in the} x direction and the wave number k _{y in the} y direction (s3), and outputs the filter coefficient to the digital filter 112 _mn .

The filter coefficient calculation unit 130 uses the wave number k _{x in} the x direction and the wave number k _{y in the} y direction,

Thus, the filter coefficient is calculated. Note that the positions (x _m , y _n ) of N × M speakers 13 _mn are stored in advance in the filter coefficient calculation unit 130 or a storage unit (not shown). Equation (11) represents the filter coefficient in the frequency domain. The relationship between the wave number k and the frequency f of the input signal S is expressed as k = 2πf / c, where c is the speed of sound.

When the sampling frequency of the input signal in the A / D converter 105 _mn is f _s , the frequency band division number is Q, and q = 0, 1,.

It expresses. The filter coefficient calculation unit 130 calculates a filter coefficient in each frequency band f (q), and sets a result obtained by _performing inverse Fourier transform as a filter coefficient in each digital filter 13 _mn . That is, Q frequency domain filter coefficients are calculated, converted to time domain filter coefficients by inverse Fourier transform, and set in each digital filter 13 _mn . If the value of the frequency band division number Q is increased, the accuracy as a filter increases, but the calculation cost increases. Q can be arbitrarily set, for example, a value of about several hundred. Further, if the power is set to a power of 2, the calculation can be speeded up.

<Effect>
With such a configuration, there is an effect that local reproduction can be performed without a control error using an evanescent wave having a steep distance attenuation characteristic. That is, by putting the physical property into the constraint condition when designing the filter, the control error is eliminated and the local regeneration effect is improved.

<Modification>
Each speaker 13 _mn may use a monopole sound source. When a monopole sound source is used, it is necessary to rewrite equation (1) with the first type Rayleigh integration (see Reference 1), and is expressed as follows instead of equation (10).

In this case, the filter coefficient calculation unit 130 calculates the filter coefficient according to the following expression instead of the expression (11).

In addition, k _z is

Calculated by

The local reproduction apparatus 100 according to the first embodiment is configured to include M × N A / D converters 105 _mn , but includes only one A / D converter 105 and converts the converted digital input signal s (k) into each digital signal. It is good also as a structure output to filter _112mn .

Speaker array, N pieces, a total of M × N loudspeakers 13 _mn are disposed on the same plane necessarily the M in delta _x intervals in an arrangement direction (x direction as shown in FIG. 5, in delta _y intervals in the y-direction The speaker array may be any speaker array including two or more speakers arranged on the same plane or in a straight line.

Local reproduction device _{^{100, k x 2 + k y 2}} > k 2 satisfy the set _(k x, _{k y)} are set in advance, is stored in the unillustrated storage unit such as the filter coefficient calculation unit 130, A configuration may be _{adopted in} which the set (k _x , k _y ) is read from the storage unit or the like, the digital filter 112 _mn is calculated, and the digital filter 112 _mn is set. In that case, the wave number calculation unit 120 may not be provided.

The α used in the wave number calculation unit 120 may be set to different values α _x and α _y when used when calculating the wave numbers k _x and k _y . At this time, the wave number k _x is calculated as k _x = α _x k _max , and the wave number k _y is calculated as k _y = α _y k _max . α _x and α _y are predetermined values larger than 1, respectively.

Further, the filter coefficient H _mn may be obtained prior to the reproduction process of the local reproduction apparatus 100 and set in the digital filter 112 _mn . For example, the filter coefficient H _mn may always be calculated using the input signal S when performing the reproduction process. Alternatively, the filter coefficient H _mn may be calculated in the initial stage of the regeneration process, and then the same filter coefficient H _mn may be used. Prior to the reproduction processing of the desired input signal, the filter coefficient calculation unit 130 may obtain the filter coefficient H _mn in advance using the sample input signal and set it in the digital filter 112 _mn . With such a configuration, it is possible to reduce the calculation cost when reproducing a desired input signal.

<Local reproduction apparatus 200>
The local reproduction device 200 according to the second embodiment will be described with reference to FIG. Only parts different from the first embodiment will be described. The local reproduction device 200 is different from the local reproduction device 100 in that a wave number calculation unit 220 is provided.

<Wave number calculation unit 220>
The wave number calculation unit 220 calculates the wave number k _{x in} the x direction and the wave number k _{y in the} y direction using the wave number k of the input signal S, and outputs the wave number k _{x in} the y direction to the filter coefficient calculation unit 130.

When the wave number included in the input signal S in the frequency band f (q) of Equation (12) is k (q), the wave number calculation unit 220 sets the wave number in the x direction in each frequency band f (q) to k _x (q ) = Βk (q) or the wave number in the y direction is calculated as k _y (q) = βk (q). Note that β is a predetermined value larger than 1. The wave number k (q) included in the input signal S in the frequency band f (q) is obtained from the input signal S.

In the first embodiment, k _x and k _y are obtained regardless of the frequency band f (q). However, in this embodiment, the wave numbers k _x (q) and k _y (q are obtained for each frequency band f (q). ) Note that β is a value in which the number of changes in attenuation due to frequency decreases when the value is decreased, and the number of changes in attenuation increases when the value is increased. Therefore, k _x (q) = βk (q)> k ( q), k _y (q) = βk (q)> k (q) is preferably as small as possible.

In addition, the filter coefficient calculation unit 130 calculates the filter coefficient of each frequency band f (q) using the wave numbers k _x (q) and k _y (q) for each frequency band by the following formula.

The filter coefficient calculation unit 130 calculates a filter coefficient in each frequency band f (q), and sets a result obtained by _performing inverse Fourier transform as a filter coefficient in each digital filter 13 _mn .

<Effect>
By adopting such a configuration, the same effects as those of the first embodiment are obtained. Furthermore, since the wave number and filter coefficient can be set for each frequency, the local reproduction device can be configured more flexibly.

<Modification>
Β used in the wave number calculation unit 220 may be set to different values β _x and β _y when used when calculating the wave numbers k _x and k _y .

<Local reproduction device 300>
The local reproduction device 300 according to the third embodiment is described with reference to FIG. Only parts different from the first embodiment will be described. The local reproduction device 300 is different from the local reproduction device 100 in that a wave number calculation unit 320 is provided.

If the attenuation term in the z direction of p (x, y, z) of Evanescent wave equation (9) is B (z), for example, when k _y = 0,

It can be expressed as. Expression (13) indicates that the wave number k _{x in the} x direction can be obtained from an arbitrary attenuation amount and the wave number k of the input signal. For ^{example, B (z) = e -kaz} ( where, _{k a} is an arbitrary constant, to identify any attenuation ^{e -kaz)} If you want to play the evanescent wave having an attenuation characteristic of,

Can be specified. Similarly, if k _x = 0,

Can be specified. The wave number calculation unit 320 calculates the wave number k _x or k _y using the equation (14) or the equation (15).

<Wave number calculation unit 320>
Wave number calculation unit 320, by using the wave number k and an arbitrary constant _{k a} of the input signal S, by the equation (14), or to calculate the wave number _{k x} the x-direction, or, wavenumber in the y direction by the formula (15) k _y is calculated. Further, wave number calculation section 320 outputs to filter coefficient calculation section 130. Note that arbitrary constant k _a is calculated in advance from any attenuation, it may be input to the filter coefficient calculator 130. Also, it may be stored in the storage unit (not shown) or the like an arbitrary constant k _a, may be configured to read.

<Effect>
By adopting such a configuration, the same effects as those of the first embodiment are obtained. Furthermore, since the wave number and filter coefficient can be set for each frequency, the local reproduction device can be configured more flexibly. Further, since the wave numbers k _x and k _y can be calculated from an arbitrary attenuation amount, there is an effect that the control at the time of performing the local reproduction is easy.

<Modification>
_{K a} is used in the wave number calculation section 320, when used to calculate the wave number _k x, a _{k y,} different values _{k ax,} may be set _{k ay.}

<Experimental result>
FIG. 8 shows the sound pressure attenuation amount for each frequency of the reproduction apparatus without a filter, and FIG. 9 shows the sound pressure attenuation amount for each frequency of the local reproduction apparatus 300. In the local reproduction device 300 in FIG. 9, k _a = 10, k _x (k, 10) = √ (k ² +10 ² ), and k _y = 0. In either case, a flat speaker array is used in which 100 speakers are arranged so that M = N = 10 at a speaker interval Δ _x = Δ _y = 4.8 cm. 2 shows how the sound pressure is attenuated by the wave number (frequency) of the input signal when moving away from each playback device in the z direction in FIG. The local reproduction device 300 in FIG. 9 has a larger attenuation amount of sound pressure at the same distance than the reproduction device without a filter in FIG. It can be seen that the local reproduction device 300 has a local reproduction effect. It can also be seen that there is little difference in the amount of attenuation depending on the frequency of the input signal, and the effect of attenuation term control appears. Also, the value of k _{a to be} specified is designed to be designed so that k _a = 1 to several tens, because if the value is too large, the attenuation immediately occurs in the vicinity of the apparatus.

<Program>
The local reproduction device described above can also be operated by a computer. In this case, each process of a program for causing a computer to function as a target device (device having the functional configuration shown in the drawings in various embodiments) or a processing procedure thereof (shown in each embodiment) is processed by the computer. A program to be executed by the computer may be downloaded from a recording medium such as a CD-ROM, a magnetic disk, or a semiconductor storage device or via a communication line into the computer, and the program may be executed.

100, 200, 300 Local reproduction device 13 _mn speaker 112 _mn digital filter 120, 220, 320 Wave number calculation unit 130 Filter coefficient calculation unit

Claims (7)

A local reproduction method using a speaker array composed of two or more speakers arranged on the same plane or in a straight line,
The plane formed by the said speaker array and xy plane in xyz orthogonal coordinate system, with the wave number k _x and y directions of the wave number k _y in the x-direction, and a filter coefficient calculating step of calculating the filter coefficients for generating the evanescent wave ,
A drive signal generating step for generating a drive signal using the input signal and the filter coefficient;
A reproduction step of reproducing the drive signal in the speaker array,
A local reproduction method characterized by the above. The local regeneration method according to claim 1, wherein
k _max is the maximum wave number included in the input signal, α is a predetermined value greater than 1, and
A wave number calculating step of calculating the wave number k _x as k _x = αk _max or calculating the wave number k _y as k _y = αk _max ;
A local reproduction method characterized by the above. The local regeneration method according to claim 1, wherein
k (q) is the wave number of the input signal in the frequency band f (q), β is a predetermined value greater than 1,
A wave number calculating step of calculating the wave number k _x as k _x (q) = βk (q) or calculating the wave number k _y as k _y (q) = βk (q);
A local reproduction method characterized by the above. The local regeneration method according to claim 1, wherein
Attenuation specified by arbitrary values _{k a} and ^{e -kaz,} k a (q) and the wave number of the input signals in the frequency band f (q),
The wave number _{k x k x (k (q} ), k a) = √ ((k (q) 2 + k a 2) or is calculated as, or, the wave number _{k y k y (k (q} ), k the wave number calculation step of calculating as ^{_{a) = √ ((k (}} q) 2 + k a 2), further comprising,
A local reproduction method characterized by the above. The local regeneration method according to claim 1, wherein
The wave number _k x, _{k y} is ^{_{k x 2 + k y 2>}} k satisfy the ^second relationship set _(k x, _{k y)} is,
A local reproduction method characterized by the above. A speaker array composed of two or more speakers arranged on the same plane or in a straight line;
And said plane formed by the speaker array and xy plane in xyz orthogonal coordinate system, x-direction and y-direction of the wave number k _x, with k _y, the filter coefficient calculation unit for calculating a filter coefficient for generating the evanescent wave,
A drive signal generation unit that generates a drive signal using the input signal and the filter coefficient;
A local reproduction apparatus.   A program for causing a computer to execute the local reproduction method according to any one of claims 1 to 5.