JP2000165984A - Sound public-address system and method for improving its articulation - Google Patents

Abstract

PROBLEM TO BE SOLVED: To prevent deterioration in the articulation resulting from adding reverberation to a loudspeaking sound in a sound field with much reverberation such as a tunnel. SOLUTION: In the case of outputting an input sound signal 101 from a speaker 104 through a sound field control filter 102 and an amplifier 103, an impulse response 105 between N-sets of speakers 104 and M-sets of control points 106 is obtained, an acoustic transfer function matrix 108 for each frequency is calculated, a specific value decomposition calculation device 109 decomposes a specific value and outputs a specific value matrix. A specific value evaluation device 110 nullifies those whose value is less than a reference value, an inverse matrix calculation device 111 calculates an inverse matrix, a coefficient calculation device 113 calculates a coefficient 114 on the basis of an object characteristic 112 and an output of the inverse matrix calculation device 111 to set it to the sound field control filter 102.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a sound loudspeaker for loudspeaking output sound using an electroacoustic facility and a method for improving the clarity of the sound loudspeaker.

[0002]

2. Description of the Related Art Conventionally, as a method for improving intelligibility when performing loudspeaking in acoustic loudspeaking, a method disclosed in the literature (School Technical Report EA) is described.
94-38) are known. FIG. 6 shows a configuration of a conventional sound loudspeaker with improved intelligibility. In FIG. 6, reference numeral 607 denotes a sound field, which includes N speakers 604 and M control points 606. Reference numeral 605 denotes an impulse response between the speaker 604 and the control point 606, which is obtained by measurement or prediction means. The coefficient calculator 611 calculates the coefficient 612 by the least square method based on the target characteristic 609 in which a waveform with little reverberation is set, the transfer function setting unit 608 in which the impulse response 605 is set, and the divergence prevention coefficient 610. This coefficient 61
2 is set to the sound field control filter 602 and connected to the speaker 604 via the amplifier 603, so that the input audio signal 601 can be clearly transmitted to the control point 606.

[0003] In this way, even in the conventional sound loudspeaker, the clarity of the loudspeaker can be improved tentatively.

[0004]

However, in the above-described conventional method for improving intelligibility in a sound loudspeaker, if the reverberation time is long, the amount of calculation of the necessary control filter becomes enormous, and if the number of control points increases. The number of sound field control filters also increases, and the number of impulse responses that need to be measured also increases.

SUMMARY OF THE INVENTION The present invention solves such a conventional problem, and reduces the amount of operation of a sound field control filter and reduces the number of sound field control filters, thereby enabling sound to be clearly reproduced. An object of the present invention is to provide a loudspeaker and a method for improving clarity thereof.

[0006]

The sound loudspeaker and its clarity improving method of the present invention use a singular value decomposition as a method of calculating a coefficient of a sound field control filter, and use a singular value decomposition to calculate each element of a singular value matrix for each frequency. Among them, an element which is equal to or less than the reference value is set to 0, then a coefficient is calculated, a part of the impulse response between the speaker and the control point is set as a target characteristic, and a plurality of control points are three-dimensionally arranged. Means for shortening the coefficient length, using only a part of the coefficient including a tap at which the square amplitude is maximized, and performing sound field control at a plurality of places in a sound field having the same shape in one direction such as a tunnel. The speaker and the control point are arranged at the same position in each control point, the sound field control filter coefficient is calculated only in one of the plurality of control points, and the sound field control is performed in other points using this coefficient. means And means for calculating the sound field control filter coefficient for controlling the plurality of control points at the same time, those having appropriately. As a result, the amount of calculation of the sound field control filter can be reduced, and clear speech can be performed while reducing the number of sound field control filters.

[0007]

DESCRIPTION OF THE PREFERRED EMBODIMENTS According to the first aspect of the present invention, after an input audio signal is amplified through a sound field control filter, loudspeakers are output from a loudspeaker to a control point in a sound field, and a loudspeaker is output. Calculate the coefficient for the target characteristic based on the impulse response between
In the acoustic loudspeaker that sets the coefficient in the sound field control filter, as a method of calculating the coefficient of the sound field control filter, using singular value analysis, the elements of the singular value matrix for each frequency that are equal to or less than the reference value Is a sound loudspeaker provided with means for calculating a coefficient after being set to 0, which can reduce the calculation amount of the sound field control filter, and can perform clear loudspeaking while reducing the number of sound field control filters. Has an action.

According to a second aspect of the present invention, there is provided an acoustic loudspeaker according to the first aspect, further comprising means for setting a part of an impulse response between a speaker and a control point as a target characteristic. This has the effect that the amount of operation of the control filter can be reduced, and clear loudspeaking can be performed while reducing the number of sound field control filters.

According to a third aspect of the present invention, there is provided the acoustic loudspeaker according to the first or second aspect, wherein a plurality of control points are three-dimensionally arranged.
This has the effect that clear loudspeaking can be performed while reducing the number of sound field control filters.

According to a fourth aspect of the present invention, the length of the coefficient is shortened, and only a part of the coefficient including a tap having the maximum square amplitude is used as the sound field control filter coefficient. The sound loudspeaker according to the present invention has an effect that the amount of calculation of the sound field control filter can be reduced, and clear loudspeaking can be performed while reducing the number of sound field control filters.

According to a fifth aspect of the present invention, in a sound field in which the same shape is continuous in one direction such as a tunnel, sound field control is performed at a plurality of places, and a speaker and a control point are located at the same position in each control place. The sound field control filter coefficient is calculated only at one of the plurality of control points, and the sound field control is performed at other points using the coefficient.
The sound loudspeaker according to any one of the above, has an effect that a clear loudspeaker can be performed while reducing the calculation amount of the sound field control filter and reducing the number of the sound field control filters.

According to a sixth aspect of the present invention, there is provided the acoustic loudspeaker according to the fifth aspect, wherein a sound field control filter coefficient for controlling a plurality of control points simultaneously is calculated. This has the effect of enabling clear voice amplification while reducing the amount and the number of sound field control filters.

According to a seventh aspect of the present invention, in the acoustic loudspeaker, singular value decomposition is performed in the frequency domain as a method of calculating the coefficient of the sound field control filter, and among the elements of the singular value matrix, a predetermined value is set. An element that is equal to or smaller than the threshold value is an intelligibility improvement method of performing an inverse matrix calculation after setting it to 0, and has an effect that a clear loudspeaker can be performed.

According to an eighth aspect of the present invention, the maximum value of each element of the singular value matrix is calculated at each frequency, and the threshold value is set based on the maximum value. This is a degree improvement method, and has an effect that divergence can be prevented for each sound field control filter.

According to a ninth aspect of the present invention, at each frequency, the maximum value of each element of the singular value matrix is 40 dB higher than the maximum value.
The clarity improving method according to claim 8, wherein a small value is set as a threshold value, and has an effect that divergence can be prevented for each sound field control filter.

According to a tenth aspect of the present invention, there is provided the method for improving clarity according to the seventh, eighth or ninth aspect, wherein a part of the impulse response between the speaker and the control point is a target characteristic. This has the effect of expanding the range in which clear loudspeaking can be performed not only at points but also around control points.

According to an eleventh aspect of the present invention, in the clarity according to the tenth aspect, the impulse response between the loudspeaker and the control point, in which a portion other than the direct sound is set to 0, is set as a target characteristic. This is an improvement method, and has an effect of expanding a range in which clear loudspeaking can be performed not only at the control point but also around the control point.

According to a twelfth aspect of the present invention, there is provided the method for improving clarity according to the eleventh aspect, wherein a plurality of control points are arranged three-dimensionally. It has the effect of expanding.

According to a thirteenth aspect of the present invention, there is provided the clarity improving method according to the twelfth aspect, wherein the control points are arranged on a three-dimensional grid point having a side length of 30 cm or less. This has the effect of stereoscopically expanding the range that can be performed.

According to a fourteenth aspect of the present invention, after calculating the coefficient, the sound field control is performed using only a part of the coefficient including a tap having the maximum squared amplitude. And has an effect of reducing the scale of a necessary sound field control filter.

According to a fifteenth aspect of the present invention, there is provided the method for improving clarity according to the fourteenth aspect, wherein a coefficient whose coefficient length is 1/10 or less of the coefficient length before shortening is used. There is an effect that the scale of the sound field control filter can be reduced.

According to a sixteenth aspect of the present invention, in a sound field having the same cross-sectional shape in one direction such as a tunnel, sound field control is performed at a plurality of locations, and a speaker and a control point are located at the same position in each control location. Claims 7 to 15
The effect of being able to perform clear loudspeaking at a plurality of locations while reducing the scale of the entire sound field control filter, and the effect of being able to reduce the number of required impulse responses Having.

According to a seventeenth aspect of the present invention, the coefficient of the sound field control filter is calculated only at one of a plurality of control points, and the sound field control is performed at other points using the coefficients. 17. The method for improving intelligibility according to claim 16, wherein a clear loudspeaker can be performed at a plurality of places while reducing the scale of the entire sound field control filter, and a function that the number of necessary impulse responses can be reduced. Having.

The invention according to claim 18 of the present invention is the method for improving clarity according to claim 17, wherein an interval between control points is set to 50 m or more, and has an effect that interference by other control points is reduced.

According to a nineteenth aspect of the present invention, there is provided the clarity improving method according to the sixteenth aspect, wherein a sound field control filter coefficient for simultaneously controlling a plurality of control points is calculated. This has the effect that clear loudspeakers can be performed at a plurality of locations while reducing the number of voices.

According to a twentieth aspect of the present invention, in the clarity improving method according to the nineteenth aspect, a filter coefficient is calculated using only an impulse response between a loudspeaker and a control point in each control point. This has the effect that clear loudspeaking can be performed at a plurality of locations while reducing the scale of the entire sound field control filter.

According to a twenty-first aspect of the present invention, there is provided the method for improving clarity according to the twentieth aspect, wherein the interval between the control points is set to 50 m or more, and has an effect that interference by other control points is reduced.

Hereinafter, an embodiment of the present invention will be described with reference to FIG.
This will be described with reference to FIG. (Embodiment 1) FIG. 1 shows a first embodiment of the present invention. In FIG. 1, an input audio signal 101 is
It is connected to N sound field control filters 102. The output of the sound field control filter 102 is N amplifiers 103
To the N speakers 104 in the sound field 107. In the sound field 107, an impulse response 105 between the N speakers 104 and the M control points 106 is measured, and an acoustic transfer function matrix 108 is calculated for each frequency based on the measured impulse responses 105. Acoustic transfer function matrix 108
Is subjected to singular value decomposition by the singular value decomposition calculator 109.
One output of the singular value decomposition calculator 109 is input to the inverse matrix calculator 111 via the singular value evaluator 110. Another output of the singular value decomposition calculator 109 is an inverse matrix calculator 1
11, an input audio signal is output. The coefficient calculator 113 calculates a coefficient 114 based on the output of the target characteristic 112 and the output of the inverse matrix calculator 111. Coefficient 114 is sound field control filter 1
02 is set.

Next, the operation of the first embodiment will be described. In the sound field 107, N speakers 10
N × M impulse responses 105 between 4 and M control points are obtained by measurement or simulation.
The impulse response 105 is converted into a frequency domain by a method such as FFT, and an acoustic transfer function matrix 108 is calculated at each frequency. Here, the acoustic transfer function matrix 108 is represented by C (M, N). Each element of the matrix C is a complex number. Note that the suffixes M and N mean that the matrix is M rows and N columns. In the singular value decomposition calculator 109, equation (1)
The singular value decomposition of the matrix C is performed as follows. C = UWV ^T Equation (1) U (M, N): M × N column orthogonal matrix ((i = 1 to M) ΣUikUjn = δkn, 1 <k <N, 1 <n <N)) W
(N, N): N × N diagonal matrix (diagonal component wj [j = 1 to N]
V ^T (N, N): transposed matrix of N × N orthogonal matrix V

When the W obtained by the equation (1) is used as it is, if there is an element having a value of 0 or a value close to 0 among the elements of W, the sound field control filter corresponding to the element has a value of 0. There is a problem that the coefficient 114 diverges. Therefore, the singular value evaluator 110 prevents divergence for each sound field control filter by the following method.

The singular value evaluator 110 obtains the maximum value Wmax of each element of the singular value matrix W. Next, the maximum value Wm of each element
Find the ratio to ax. If the ratio is equal to or less than the reference value, the value of the element wj is set to 0. Standard value is -60
The range may be from -20 dB to -20 dB, and -40 dB is appropriate.

The inverse matrix calculator 111 calculates the inverse matrix of C according to the equation (2). ^{C -1 = V [diag (1} / wj)] U T ··· formula (2) diag (1 / wj ): elements 1 / wj [j = 1~N] of N × N diagonal matrix

An impulse response with a small reverberation component is set as a target at the M control points, and each impulse response is converted into a frequency domain. At each frequency, a matrix B (M) is set as a target characteristic 112. The coefficient calculator 113 calculates the solution X (N) according to equation (3). X = C ^-1 B Equation (3)

X is obtained for each frequency, and the frequency characteristics of the N sound field control filters are obtained.
Conversion into the time domain is performed by a method such as that described above, and the coefficient 114 is obtained. By setting this coefficient in the sound field control filter 102, it is possible to clearly loudspeak at the control point 106. In this way, clear loudspeaking can be performed while preventing divergence for each sound field control filter.

(Embodiment 2) FIG. 2 shows a second embodiment of the present invention. In FIG. 2, an input audio signal 201 is connected to N sound field control filters 202. The output of the sound field control filter 202 is connected to N speakers 204 in a sound field 207 via N amplifiers 203 respectively. In the sound field 207, an impulse response 205 between N speakers 204 and M control points 206 is measured. Impulse response 20
5, an acoustic transfer function matrix 208 is calculated for each frequency. Further, the impulse response 205 is also input to the target characteristic setter 215. The acoustic transfer function matrix 208 is input to the singular value decomposition calculator 209. The singular value decomposition calculator 209 performs singular value decomposition of the acoustic transfer function matrix 208. One of the outputs of the singular value decomposition calculator 209 is input to the inverse matrix calculator 211 via the singular value evaluator 210. Another output of the singular value decomposition calculator 209 is input to the inverse matrix calculator 211. Coefficient calculator 21
3, the target characteristic 2 which is the output of the target characteristic setter 215
The coefficient 214 is calculated on the basis of the output of the inverse matrix calculator 211 and 12. The coefficient 214 is set in the sound field control filter 202.

Next, the operation of the second embodiment will be described. In the sound field 207, the N speakers 20
N × M impulse responses 205 between 4 and M control points are obtained by measurement or simulation.
The M control points are arranged on a straight line. Control point interval d is 3
It is preferably within 0 cm, and in practice, it is appropriate to be 10 cm. The impulse response 205 is converted into a frequency domain by a method such as FFT, and an acoustic transfer function matrix 208 is calculated at each frequency. The acoustic transfer function matrix 208 is input to the singular value decomposition calculator 209. Here, the acoustic transfer function matrix 208 is represented by C (M, N). Each element of the matrix C is a complex number. The subscripts M and N are M
This means that the matrix is a row-N column matrix. Singular value decomposition calculator 2
At 09, the singular value decomposition of the matrix C is performed as in the equation (1).

When the W obtained by the equation (1) is used as it is, if there is an element having a value of 0 or a value close to 0 among the elements of W, the sound field control filter corresponding to the element has a value of 0. There is a problem that the coefficient 214 diverges. Therefore, the singular value evaluator 210 prevents divergence for each sound field control filter by the following method.

The singular value evaluator 210 obtains the maximum value Wmax of each element of the singular value matrix W. Next, the maximum value Wm of each element
Find the ratio to ax. If the ratio is equal to or less than the reference value, the value of the element wj is set to 0. Standard value is -60
The range may be from -20 dB to -20 dB, and -40 dB is appropriate.

The target characteristic setting unit 215 receives the impulse response 205 as an input,
05 is calculated. The direct sound portion of the impulse response is calculated as follows. One of the N speakers 204 is selected, and taps at which the squared amplitude of the impulse response between the speaker and the M listening points is maximized are obtained. The window function is multiplied around the tap, and a range of taps ± T at which the square amplitude becomes maximum is defined as a direct sound portion of the impulse response. The shape of the window function may be a rectangular window or a Hanning window. The optimum value of T is set to a range in which the response of the direct sound of the speaker to be used falls. Generally, it is in the range of 10 to 50 taps. The direct sound portions of the M impulse responses obtained in this way are converted into a frequency domain by a method such as FFT, and a matrix B (M) is set as a target characteristic 212 at each frequency. The inverse matrix calculator 211 calculates the inverse matrix of C according to the equation (2).

The coefficient calculator 213 calculates the solution X (N) according to equation (3). By obtaining X for each frequency, the frequency characteristics of the N sound field control filters are calculated. The frequency characteristic of each sound field control filter is converted into a time domain by a method such as inverse FFT, and a coefficient 2
Get 14. By setting this coefficient in the sound field control filter 202, the control point 206 and its surroundings can be clearly louder. In this way, clear loudspeaking can be performed while preventing divergence for each sound field control filter.

(Embodiment 3) FIG. 3 shows a third embodiment of the present invention. In FIG. 3, an input audio signal 301 is connected to N sound field control filters 302. The output of the sound field control filter 302 is connected to N speakers 304 in the sound field 307 via N amplifiers 303 respectively. In the sound field 307, an impulse response 305 between the N speakers 304 and the M control points 306 is measured. Impulse response 30
5, an acoustic transfer function matrix 308 is calculated for each frequency. Further, the impulse response 305 is also input to the target characteristic setter 315. The acoustic transfer function matrix 308 is input to the singular value decomposition calculator 309. The singular value decomposition calculator 309 performs singular value decomposition of the acoustic transfer function matrix 308. One of the outputs of the singular value decomposition calculator 309 is input to the inverse matrix calculator 311 via the singular value evaluator 310. Another output of the singular value decomposition calculator 309 is input to the inverse matrix calculator 311. Coefficient calculator 31
3, the target characteristic setter 315 outputs the target characteristic 3
A coefficient 314 is calculated on the basis of the output of the inverse matrix calculator 311 and 12. The calculated coefficient 314 is the maximum value detector 3
At 16, a tap at which the square amplitude is maximized is obtained, and at a coefficient length changer 317, a window function is multiplied around the tap to obtain a range of tap ± T at which the square amplitude is maximum as a new coefficient, and the sound field control filter is obtained. Set to 302.

Next, the operation of the third embodiment will be described. In the sound field 307, N speakers 30
N × M impulse responses 305 between 4 and M control points 306 are obtained by measurement or simulation. The M control points 306 are arranged three-dimensionally. The distance d between the control points 306 is preferably within 30 cm, and 10 c in practical use.
It is appropriate to set the length from m to 20 cm. The practical arrangement of the control points 306 is as follows: 27 control points are arranged in a cube having a side length of 20 cm with a control point interval d of 10 cm.
This is 1.5 m above the floor.

Each of the impulse responses 305 is
The data is converted into a frequency domain by a method such as FT, and an acoustic transfer function matrix 308 is calculated at each frequency. The acoustic transfer function matrix 308 is input to the singular value decomposition calculator 309. Here, the acoustic transfer function matrix 308 is represented by C (M, N). Each element of the matrix C is a complex number. Note that the suffixes M and N mean that the matrix is M rows and N columns. The singular value decomposition calculator 309 performs singular value decomposition of the matrix C as shown in Expression (1).

When the W obtained by the equation (1) is used as it is, if there is an element having a value of 0 or a value close to 0 among the elements of W, the sound field control filter corresponding to the element has a value of 0. There is a problem that the coefficient 314 diverges. Therefore, the singular value evaluator 310 prevents divergence for each sound field control filter by the following method.

The singular value evaluator 310 determines the maximum value Wmax of each element of the singular value matrix W. Next, the maximum value Wm of each element
Find the ratio to ax. If the ratio is equal to or less than the reference value, the value of the element wj is set to 0. Standard value is -60
The range may be from -20 dB to -20 dB, and -40 dB is appropriate.

The target characteristic setting unit 315 receives the impulse response 305 as an input,
05 is calculated. Impulse response 3
The method of calculating the direct sound part of 05 is performed as follows. One of the N speakers 304 is selected, and taps that maximize the squared amplitude of the impulse response between the speaker and the M listening points are obtained. The window function is multiplied around the tap, and a range of taps ± T at which the square amplitude becomes maximum is defined as a direct sound portion of the impulse response. The shape of the window function may be a rectangular window or a Hanning window. The optimum value of T is set to a range in which the response of the direct sound of the speaker to be used falls. Generally, it is in the range of 10 to 50 taps.
The direct sound portions of the M impulse responses obtained in this way are converted into a frequency domain by a method such as FFT, and a matrix B (M) is set as a target characteristic 312 at each frequency. In the inverse matrix calculator 311, C
Is calculated.

The coefficient calculator 313 calculates the solution X (N) according to equation (3). Find X for each frequency,
The frequency characteristics of the N sound field control filters are obtained, and each is converted to a time domain by a method such as inverse FFT, thereby obtaining N pre-shortening coefficients. Maximum value detector 316
Then, taps that maximize the square amplitude of the coefficient before shortening are obtained. The coefficient length changer 317 multiplies a window function centering on the tap at which the square amplitude of the coefficient before shortening becomes the maximum, and sets the range of the tap ± T at which the square amplitude becomes the maximum as a new coefficient. The shape of the window function may be a rectangular window or a Hanning window. It is appropriate that the size of T is 1/10 of the length of the coefficient before shortening, but it may be further shortened.

By setting the coefficient 314 calculated in this way in the sound field control filter 302, the control point 306 and its surroundings can be clearly loudspeaked. It can be performed.

(Embodiment 4) FIG. 4 shows a fourth embodiment of the present invention. In FIG. 4, an input audio signal 401 is connected to N sound field control filters 402. The output of the sound field control filter 402 is passed through N amplifiers 402 to control points A 41 in the sound field 407.
6 N speakers 404 and N speakers 418 at a control location B 417. The sound field 407 is a sound field such as a tunnel in which the same cross-sectional shape is continuous in one direction. In the sound field 407, the speaker 40 is controlled at the control point A416.
4 and an impulse response 405 between the control points 406,
At a control point B417, an impulse response 419 between the speaker 418 and the control point 420 is defined. Based on the impulse response 405, an acoustic transfer function matrix 408 is calculated for each frequency. Further, the impulse response 405 is also input to the target characteristic setter 415. Acoustic transfer function matrix 408
Is input to the singular value decomposition calculator 409. The singular value decomposition calculator 409 performs singular value decomposition of the acoustic transfer function matrix 408. One output of the singular value decomposition calculator 409 is input to an inverse matrix calculator 411 via a singular value evaluator 410. Another output of the singular value decomposition calculator 409 is input to the inverse matrix calculator 411. In the coefficient calculator 413,
A coefficient 414 is calculated based on a target characteristic 412 output from the target characteristic setting device 415 and an output from the inverse matrix calculator 411. The coefficient 414 is set in the sound field control filter 402.

Next, the operation of the fourth embodiment will be described. In the sound field 407, N × N × N × N × N × N × N × N × N
M impulse responses 405 are obtained by measurement or simulation. In this case, the control point B
The N × M impulse responses 419 between the N speakers 418 at 417 and the M control points 420 are not used.

The sound field 407 has a shape such as a tunnel in which the same cross-sectional shape is continuous in one direction. Therefore, the cross-sectional shapes at the control point A 416 and the control point B 417 are the same. Interval L between control point A 416 and control point B 417
Should be as wide as possible and need to be 50m or more.
m or more.

At control point A 416, M control points 406 are arranged three-dimensionally. The distance d between the control points 406 is 3
It is good to be within 0 cm, and in practice, it is appropriate to make it 10 cm to 20 cm. The practical arrangement of the control points 406 is such that the distance d between the control points is 10 c in a cube having a side length of 20 cm.
m, 27 control points are arranged, and the control points are set at 1.m from the floor surface.
The height is 5 m. Further, at the control point A416,
The speaker 404 is arranged. The speakers 404 are arranged so as to cover the control points 406, respectively, and are arranged close to each other so that the distance between the speakers is as small as possible.

At control point B 417, control point 420
Are arranged the same as the control point 406 at the control point A 416. Also, the speaker 418 is arranged in the same manner as the speaker 404 at the control point A416. Arranging the speaker and the control point identically in each of the control point A 416 and the control point B 417 has an effect of increasing the correlation between the impulse response 405 at the control point A 416 and the impulse response 419 at the control point B 417.

N speakers 404 at control location A 416
And M control points 406 are converted into N × M impulse responses 405 in the frequency domain by a method such as FFT, respectively.
8 is calculated. The acoustic transfer function 408 is input to a singular value decomposition calculator 409. Here, the acoustic transfer function matrix 40
8 is represented by C (M, N). Each element of the matrix C is a complex number. Note that the suffixes M and N mean that the matrix is M rows and N columns. The singular value decomposition calculator 409 performs singular value decomposition of the matrix C as shown in Expression (1).

When W obtained by the equation (1) is used as it is, if there is an element having a value of 0 or a value close to 0 among the elements of W, the sound field control filter corresponding to the element has a value of 0. There is a problem that the coefficient 414 diverges. Therefore, the singular value evaluator 410 prevents divergence for each sound field control filter by the following method.

The singular value evaluator 410 obtains the maximum value Wmax of each element of the singular value matrix W. Next, the maximum value Wm of each element
Find the ratio to ax. If the ratio is equal to or less than the reference value, the value of the element wj is set to 0. Standard value is -60
The range may be from -20 dB to -20 dB, and -40 dB is appropriate.

The target characteristic setting unit 415 receives the impulse response 405 as an input,
05 is calculated. Impulse response 4
The method of calculating the direct sound part of 05 is performed as follows. One of the N speakers 404 is selected, and taps that maximize the squared amplitude of the impulse response between the speakers and the M listening points are obtained. The window function is multiplied around the tap, and a range of taps ± T at which the square amplitude becomes maximum is defined as a direct sound portion of the impulse response. The shape of the window function may be a rectangular window or a Hanning window. The optimum value of T is set to a range in which the response of the direct sound of the speaker to be used falls. Generally, it is in the range of 10 to 50 taps.
The direct sound portions of the M impulse responses obtained in this manner are converted into a frequency domain by a method such as FFT, and a matrix B (M) is set as a target characteristic 412 at each frequency. In the inverse matrix calculator 411, C
Is calculated.

The coefficient calculator 413 calculates the solution X (N) according to equation (3). Find X for each frequency,
The frequency characteristics of the N sound field control filters are obtained, and each is converted into a time domain by a method such as inverse FFT, so that N coefficients 414 are obtained. By setting the coefficient 414 calculated in this way in the sound field control filter 402, the sound is clearly loudspeaked at two points: the control point 406 at the control point A416 and its surroundings, and the control point 420 at the control point B417 and its surroundings. It is possible to perform clear loudspeaking at a plurality of locations while reducing the scale of the entire sound field control filter. Although the number of control points is set to two here, it is apparent that the same result can be obtained even when the number of control points is two or more.

(Embodiment 5) FIG. 5 shows a fifth embodiment of the present invention. In FIG. 5, an input audio signal 501 is connected to N sound field control filters 502. The output of the sound field control filter 502 passes through N amplifiers 503 and the control point A 51 in the sound field 507.
6 N speakers 504 and N speakers 518 at a control location B517. The sound field 507 is a sound field such as a tunnel in which the same cross-sectional shape is continuous in one direction. In the sound field 507, the speaker 50 is controlled at the control point A516.
Impulse response 505 between 4 and control point 506,
At a control point B517, an impulse response 519 between the speaker 518 and the control point 520 is defined. Also, control point A 516 and control point B
Between 517, an impulse response 521 and an impulse response 522 are defined. Based on the impulse response 505 and the impulse response 519,
An acoustic transfer function matrix 508 is calculated for each frequency.
Further, the impulse response 505 and the impulse response 522 are also input to the target characteristic setter 515.
The acoustic transfer function matrix 508 is input to the singular value decomposition calculator 509. The singular value decomposition calculator 509 performs singular value decomposition of the acoustic transfer function matrix 508. Singular value decomposition calculator 5
One of the outputs 09 is input to an inverse matrix calculator 511 via a singular value evaluator 510. Another output of the singular value decomposition calculator 509 is input to an inverse matrix calculator 511.
The coefficient calculator 513 calculates a coefficient 514 based on the output of the target characteristic 512 which is an output of the target characteristic setting device 515 and the output of the inverse matrix calculator 511. The coefficient 514 is set in the sound field control filter 502.

Next, the operation of the fifth embodiment will be described. In the sound field 507, N × N × N × N × N × N
M impulse responses 505 are obtained by measurement or simulation. Further, N × N × N × N × N × N
M impulse responses 519 are obtained by measurement or simulation. Here, the impulse response 521 and the impulse response 522 between the control point A 516 and the control point B 517 are not used.

The sound field 507 has a shape such as a tunnel in which the same cross-sectional shape is continuous in one direction. Therefore, the cross-sectional shapes at the control point A516 and the control point B517 are the same. Interval L between control point A516 and control point B517
Should be as wide as possible and need to be 50m or more.
m or more.

At control point A 516, M control points 506 are arranged three-dimensionally. The distance d between the control points 506 is 3
It is good to be within 0 cm, and in practice, it is appropriate to make it 10 cm to 20 cm. The practical arrangement of the control points 506 is such that the distance d between the control points is 10 c in a cube having a side length of 20 cm.
m, 27 control points are arranged, and the control points are set at 1.m from the floor surface.
The height is 5 m. Also, at control point A516,
The speaker 504 is arranged. The speakers 504 are arranged so as to cover the control points 506, respectively, and are arranged close to each other so that the distance between the speakers is as small as possible.

At control point B 517, control point 520
Are arranged the same as the control point 506 at the control point A516. Also, the speaker 518 is arranged the same as the speaker 504 at the control point A516. Arranging the speaker and the control point identically at each of the control point A516 and the control point B517 has an effect of increasing the correlation between the impulse response 505 at the control point A516 and the impulse response 519 at the control point B517.

N speakers 504 at control point A516
Then, the N × M impulse responses 505 between and the M control points 506 are converted into the frequency domain by a method such as FFT, and the acoustic transfer function matrix A is calculated. Also,
N × M impulse responses 519 between N speakers 518 and M control points 520 at control point B 517
Are respectively converted into a frequency domain by a method such as FFT, and an acoustic transfer function matrix B is calculated. The acoustic transfer function matrices A (N, M) and B (N, M) are complex matrices each having N rows and M columns. Here, the acoustic transfer function matrix C (N, 2M)
Is calculated by combining A and B as in equation (4).

The matrix C obtained in this manner is defined as an acoustic transfer function matrix 508. The acoustic transfer function matrix 508 is input to the singular value decomposition calculator 509. The singular value decomposition calculator 509 performs singular value decomposition of the matrix C as shown in Expression (1).

When the W obtained by the equation (1) is used as it is, if there is an element having a value close to 0 or almost 0 among the elements of W, the sound field control filter corresponding to the element has a value of 0. There is a problem that the coefficient 514 diverges. Therefore, the singular value evaluator 510 prevents divergence for each sound field control filter by the following method.

The singular value evaluator 510 obtains the maximum value Wmax of each element of the singular value matrix W. Next, the maximum value Wm of each element
Find the ratio to ax. If the ratio is equal to or less than the reference value, the value of the element wj is set to 0. Standard value is -60
The range may be from -20 dB to -20 dB, and -40 dB is appropriate.

The target characteristic setter 515 receives the impulse response 505 and the impulse response 522 as inputs, and calculates the direct sound portion of the impulse response 505 and the impulse response 522. The direct sound portion of the impulse response is calculated as follows. One of the N speakers 504 is selected, and taps that maximize the squared amplitude of the impulse response between the speaker and the M listening points are obtained. The window function is multiplied around the tap, and a range of taps ± T at which the square amplitude becomes maximum is defined as a direct sound portion of the impulse response. The shape of the window function may be a rectangular window or a Hanning window. The optimum value of T is set to a range in which the response of the direct sound of the speaker to be used falls. Generally, it is in the range of 10 to 50 pp.
The direct sound portions of the M impulse responses obtained in this manner are converted into a frequency domain by a method such as FFT, and a matrix B (2M) is set as a target characteristic 512 at each frequency. The inverse matrix calculator 511 calculates the inverse matrix of C according to the equation (2).

The coefficient calculator 513 calculates the solution X (N) according to equation (3). Find X for each frequency,
The frequency characteristics of the N sound field control filters are obtained, and each is converted to the time domain by a method such as inverse FFT, so that N coefficients 514 are obtained. By setting the coefficient 514 calculated in this way in the sound field control filter 502, the two points, namely, the control point 506 at the control point A516 and its surroundings and the control point 520 at the control point B517 and its surroundings, are clearly amplified. It is possible to perform clear loudspeaking at a plurality of locations while reducing the scale of the entire sound field control filter. Although the number of control points is set to two here, it is apparent that the same result can be obtained even when the number of control points is two or more.

[0070]

As described above, according to the present invention, the singular value decomposition is performed in the frequency domain as a coefficient calculation method for the sound field control filter, and among the elements of the singular value matrix, the elements that are equal to or less than the predetermined threshold value are set. Is a sound loudspeaker that performs an inverse matrix calculation after setting it to 0, and a method for improving its clarity, and can perform clear loudspeaking while reducing the number of sound field control filters and the amount of calculation.

[Brief description of the drawings]

FIG. 1 is a schematic block diagram of a sound loudspeaker according to a first embodiment of the present invention.

FIG. 2 is a schematic block diagram of a sound loudspeaker according to a second embodiment of the present invention.

FIG. 3 is a schematic block diagram of a sound loudspeaker according to a third embodiment of the present invention.

FIG. 4 is a schematic block diagram of a sound loudspeaker according to a fourth embodiment of the present invention.

FIG. 5 is a schematic block diagram of a sound loudspeaker according to a fifth embodiment of the present invention.

FIG. 6 is a schematic block diagram of a conventional sound speaker.

[Description of sign]

101, 201, 301, 401, 501 Input audio signals 102, 202, 302, 402, 502 Sound field control filters 103, 203, 303, 403, 503 Amplifiers 104, 204, 304, 404, 504, 418, 5
18 speakers 105, 205, 305, 405, 505, 419, 5
19, 521, 522 impulse response 106, 206, 306, 406, 506, 420, 5
20 Control points 107, 207, 307, 407, 507 Sound field 108, 208, 308, 408, 508 Acoustic transfer function matrix 109, 209, 309, 409, 509 Singular value decomposition calculators 110, 210, 310, 410, 510 Singular value evaluator 111, 211, 311, 411, 511 Inverse matrix calculator 112, 212, 312, 412, 512 Target characteristic 113, 213, 313, 413, 513 Coefficient calculator 114, 214, 314, 414, 514 Coefficient 215, 315, 415, 515 Target characteristic setter 316 Maximum value detector 317 Coefficient length changer 416, 516 Control location A 417, 517 Control location B

Claims (21)

[Claims] After amplifying an input audio signal through a sound field control filter, a speaker outputs loudspeaker sound to a control point in a sound field, and a target characteristic based on an impulse response between the speaker and the control point. In the acoustic loudspeaker that calculates a coefficient for the sound field control filter, the coefficient calculation method of the sound field control filter uses a singular value analysis, and among the elements of the singular value matrix for each frequency, An audio loudspeaker comprising: means for calculating a coefficient after setting an element which is equal to or less than a reference value to 0. 2. The sound loudspeaker according to claim 1, further comprising means for setting a part of an impulse response between the speaker and the control point as a target characteristic. 3. The sound speaker according to claim 1, wherein a plurality of control points are three-dimensionally arranged. 4. The sound loudspeaker according to claim 3, wherein the length of the coefficient is shortened, and only a part of the coefficient including a tap having the maximum square amplitude is used as the sound field control filter coefficient. 5. In a sound field where the same shape is continuous in one direction such as a tunnel, sound field control is performed at a plurality of locations, a speaker and a control point are arranged at the same position at each control location, and one of the plurality of control locations is controlled. The sound loudspeaker according to any one of claims 1 to 4, wherein a sound field control filter coefficient is calculated only at a point, and the sound field control is performed at another point using the coefficient. 6. The sound loudspeaker according to claim 5, wherein a sound field control filter coefficient for simultaneously controlling a plurality of control points is calculated. 7. An acoustic loudspeaker performs singular value decomposition in the frequency domain as a method of calculating a coefficient of a sound field control filter, and among elements of a singular value matrix, elements below a preset threshold are set to 0. After that, an intelligibility improvement method of performing an inverse matrix calculation. 8. The clarity improvement method according to claim 7, wherein a maximum value of each element of the singular value matrix is calculated at each frequency, and a threshold value is set based on the maximum value. 9. The clarity improvement method according to claim 8, wherein at each frequency, a threshold value is set to a value smaller by 40 dB than the maximum value of each element of the singular value matrix. 10. A part of an impulse response between a speaker and a control point is set as a target characteristic.
Or the clarity improving method according to 9. 11. The intelligibility improving method according to claim 10, wherein, of the impulse response between the speaker and the control point, a part other than the direct sound is set to 0 as a target characteristic. 12. The method according to claim 11, wherein a plurality of control points are arranged three-dimensionally. 13. The clarity improving method according to claim 12, wherein the control points are arranged on a three-dimensional lattice point having a side length of 30 cm or less. 14. The clarity improvement method according to claim 7, wherein after calculating the coefficient, the sound field control is performed using only a part of the coefficient including a tap having the maximum square amplitude. 15. A method for comparing a coefficient length with a coefficient length before shortening by 1
15. The method for improving clarity according to claim 14, wherein a coefficient set to / 10 or less is used. 16. The sound field control in a plurality of places in a sound field having the same cross-sectional shape in one direction such as a tunnel, and a speaker and a control point are arranged at the same position in each control place. Clarity improvement method described in Crab. 17. The clarity improvement according to claim 10, wherein a coefficient of the sound field control filter is calculated only at one of the plurality of control points, and the sound field control is performed at other points using the coefficients. Method. 18. The method for improving clarity according to claim 17, wherein an interval between control points is set to 50 m or more. 19. The method according to claim 16, wherein a sound field control filter coefficient for simultaneously controlling a plurality of control points is calculated. 20. The clarity improving method according to claim 19, wherein the filter coefficient is calculated using only an impulse response between a speaker inside the control point and the control point at each control point. 21. The clarity improving method according to claim 20, wherein an interval between control points is set to 50 m or more.