JP3175267B2 - Sound field direction information extraction method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a sound field direction information extracting method capable of extracting sound field direction information of a music hall, a theater, a gymnasium or the like.

[0002]

2. Description of the Related Art Conventionally, as a method for extracting direction information of a sound field of this kind, direction information can be obtained by using a regular tetrahedral vertex microphone as described in Japanese Patent Application No. 3-241223. I was FIG. 4 illustrates a conventional method for extracting direction information of a sound field by the regular tetrahedron vertex method. In FIG. 4, reference numeral 10 denotes an omnidirectional speaker, 11, 12, 13, 1
Reference numeral 4 denotes microphones, each of which is arranged at a vertex of a tetrahedron having a side of 17 cm, and is referred to as a tetrahedron vertex microphone. Reference numeral 20 denotes a reflected sound extraction / measuring device, which is an acoustic signal generating means 21 for emitting a reference pulse, an input signal storing means 22 for inputting and storing each recording sound of a regular tetrahedral vertex microphone, and the input signal storing means 22 By deconvolving the data of the microphones 11, 12, 13, and 14 (hereinafter, referred to as deconvolution), the characteristics of the omnidirectional speaker are cancelled, and the arrival time of the reflected sound to each microphone is clarified. A reflected sound arrival direction extracting means 23 for extracting the direction of arrival of the reflected sound from the positional relationship between the microphones, and a time series for synthesizing a specific reflected sound in a time series by extracting the direction of arrival of the reflected sound. It is constituted by the re-combining means 24. Then, the output of the time-series re-synthesis means 24 provides an impulse response 25 for each direction.

In the above configuration, the operation is as follows. First, the reference pulse generated by the audio signal generating means 21 is output from the non-directional speaker 1. The reference pulse is applied to four microphones 1 which are regular tetrahedron vertex microphones arranged at the vertices of the regular tetrahedron.
Sounds are collected by 1, 12, 13, and 14. Next, the impulse response of each microphone with respect to the passage of time is recorded by the input signal storage means 22 of the reflected sound extraction measuring device 20. In this recording, the impulse response of each microphone includes characteristics of a speaker and a microphone, and it is difficult to grasp the arrival time of a sound wave. Then, the reflected sound arrival direction extracting means 23 cancels the characteristics of the speaker and the microphone by performing the deconvolution processing, and after accurately grasping the arrival time of the sound wave,
The arrival direction of the reflected sound can be detected from a slight difference in the time structure of the impulse response of each microphone. Then, the time-series resynthesizing unit 24 can extract the time-dependent fluctuation of the specific reflected sound from the data of the reflected sound.

As described above, in the above conventional example, the microphones 11, 12, 13, and 14 are arranged at the vertices of a regular tetrahedron,
By subjecting the signals recorded by the microphones 11, 12, 13, and 14 to deconvolution processing, a specific reflected sound can be extracted.

[0005]

However, in the conventional method for extracting direction information of a sound field by the conventional tetrahedral vertex method, the obtained direction information is only a pulse train for each direction.
Therefore, there is a problem that even if a time-series reflected sound in the direction of arrival and the direction of arrival is extracted, information on a frequency change cannot be obtained.

SUMMARY OF THE INVENTION The present invention solves the above-mentioned conventional problems, and provides an excellent sound field direction information extraction method capable of measuring information of amplitude, frequency and time for a specific reflected sound. It is.

[0007]

In order to achieve the above object, the present invention provides a four-channel microphone in which microphones are arranged at four vertices of a regular tetrahedron at designated measurement points in a sound field, and a non-directional microphone. A loudspeaker, a microphone arranged on a spherical surface which can pick up an impulse response on a spherical surface at an arbitrary radius with the center of gravity of the tetrahedron as a center point, and a direction information extracting device for extracting direction information of a sound field. In the direction information extraction device, the omnidirectional speaker emits an acoustic signal by an impulse into a sound field from the omnidirectional speaker, and the acoustic signal radiated from the omnidirectional speaker is transmitted by the four-channel microphone. First input signal storage means for picking up and storing; and a second input signal storing means for picking up and storing an acoustic signal radiated from the omnidirectional speaker by the microphone arranged on the spherical surface. An input signal storage means, said first
A reflected sound arrival direction extracting means for extracting a direction of arrival of the reflected sound from the information of the input signal storing means, and a second reflected sound signal extracted by the reflected sound arrival direction extracting means. After correcting the time difference of each microphone of the microphone arranged on the spherical surface with respect to the reflected sound from the information,
Spatial separation means for synchronous addition, reflected sound characteristic time series separation means for temporally separating based on the arrival time of the reflected sound from the result obtained by the spatial separation means, and reflected sound characteristic time series separating means Time-series re-synthesis means for re-synthesizing only a specific reflected sound from the results obtained by
It is characterized in that direction information is output to the output of the time-series recombining means.

[0008]

According to the present invention, the amplitude and time information of a specific reflected sound can be obtained from the data of the four-channel microphone by using the four-channel microphone, the microphone arranged on the spherical surface, the omnidirectional speaker, and the direction information extracting device. Further, by obtaining frequency information and time information from the data of the microphone arranged on the spherical surface for the specific reflected sound, the signal output from the direction information extracting device has an amplitude and a frequency corresponding to the specific reflected sound. Directional information with time information can be extracted.

[0009]

FIG. 1 is a schematic block diagram showing the configuration of an embodiment of the present invention. In FIG. 1, 1 is a direction information extracting device, 10 is a non-directional microphone, 11, 12, 13,
14 is a microphone, each microphone is 17 cm on a side
And is called a four-channel microphone. 15, 16, 17, 18
Is a microphone having a center point at the center of gravity of a tetrahedron in which the microphones of the four-channel microphone are arranged, and each microphone being arranged at a different radius and being freely movable, on a spherical surface having the center point at the center. That can pick up the impulse response at
It is called a microphone arranged on a spherical surface. Further, the positional relationship between the microphones 15, 16, 17, and 18 is such that the distance from the center point is a geometric progression. For example, the distances from the center of the four microphones are 0.25 m and 0.50.
m, 1.00 m, 2.00 m, etc. The directional information extraction device 1 includes a non-directional speaker 10
Signal generating means 2 for radiating an acoustic signal due to an impulse into a sound field, input signal storing means 3 for collecting and storing an acoustic signal radiated from an omnidirectional speaker 10 with a 4-channel microphone, and omnidirectional speaker An input signal storage means 5 for collecting and storing the acoustic signal radiated from 10 by a microphone arranged on a spherical surface, a reflected sound arrival direction extraction means 4 for extracting an arrival direction of a reflected sound from information of the input signal storage means 3, A spatial separation circuit 6 that corrects the time difference of each microphone of the microphones arranged on the spherical surface with respect to the reflected sound from the information of the input signal storage means 5 for the specific reflected sound extracted by the reflected sound arrival direction extracting means 4 and then adds the synchronized sound. A reflected sound characteristic time series separating means 7 for temporally separating based on the arrival time of the reflected sound from the result obtained by the spatial separating circuit 6, A time-series re-synthesizing means 8 for re-synthesizing only a specific reflected sound from the results obtained by the sex-time-separating means 7; and a spatial separation circuit 6 for each microphone of the microphones arranged on the spherical surface for the reflected sound. Correction means 6a for correcting the time difference between
a synchronous addition means 6 for synchronously adding the result obtained by a
b. Thus, a direction-specific impulse response 9 including frequency information is obtained from the output of the time-series re-synthesis unit 8 of the direction information extraction device 1.

The operation of the above configuration will be described with reference to FIGS. First, in FIG. 1, the input signal storage means 3 receives an impulse response picked up by a four-channel microphone.
Is input with an impulse response collected by a microphone arranged on a spherical surface. Next, in the reflected sound arrival direction extraction means 4,
The direction of arrival of the reflected sound is extracted from the data obtained by the input signal storage means 3 to calculate a plurality of virtual sound sources such as the virtual sound sources 30 to 39 in FIG. The spatial separation circuit 6 spatially separates the area A in FIG. 2 from other areas by performing directional sound reception on a specific reflected sound extracted by the reflected sound arrival direction extracting means 4. The signal waveform is processed from FIG. 3B to FIG. 3C. Next, the reflected sound characteristic time series separating means 7 separates temporally based on the arrival time of the reflected sound. In the area A of FIG. 2, the virtual sound source in the area B is separated by the time window Δt. The extracted signal waveform is shown in FIG.
(D). The time series resynthesizing means 8 resynthesizes only a specific reflected sound from the results obtained by the reflected sound characteristic time series separating means 7, and outputs the impulse response for each direction including frequency information. 9 is output.

A four-channel microphone, a microphone arranged on a spherical surface, an omnidirectional speaker 10, and a direction information extracting device 1 can obtain specific amplitude and time information of reflected sound from data of the four-channel microphone. By obtaining frequency information and time information from the data of the microphone arranged on the spherical surface for a specific reflected sound, the signal output from the direction information extraction device includes amplitude, frequency, and time information for the specific reflected sound. Direction information can be extracted.

[0012]

As is apparent from the above embodiment, the present invention uses a four-channel microphone, a microphone arranged on a spherical surface, an omnidirectional speaker, and a direction information extracting device to specify a specific reflection from data of the four-channel microphone. The amplitude and time information of the sound can be obtained, and the frequency information and time information can be obtained from the data of the microphone arranged on the spherical surface for the specific reflected sound. Amplitude, frequency,
Direction information provided with time information can be extracted. Since the deconvolution processing is performed on the impulse response of the four channels recorded at the vertices of the regular tetrahedron, the reflected sound in each direction can be extracted in a short time.

[Brief description of the drawings]

FIG. 1 is a schematic block diagram showing a sound field direction information extraction method according to the present invention.

FIG. 2 is a conceptual diagram showing a sound field direction information extracting method according to the present invention.

FIG. 3 is a signal waveform diagram illustrating a sound field direction information extraction method according to the present invention.

FIG. 4 is a conventional method for extracting direction information using a regular tetrahedral vertex microphone.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Direction information extraction device 2 Acoustic signal generation means 3, 5 Input signal storage means 4 Reflected sound arrival direction extraction means 6 Spatial separation circuit 6a Phase correction means 6b Synchronous addition means 7 Reflected sound characteristic time series separation means 8 Time series resynthesis Means 9 Impulse response in each direction 10 Non-directional speaker 11-18 Microphone

──────────────────────────────────────────────────続 き Continued on the front page (58) Field surveyed (Int. Cl. ⁷ , DB name) H04R 1 / 40,3 / 00,5 / 027 G01H 3 / 00,17 / 00 G01S 3/00

Claims (1)

(57) [Claims] 1. A four-channel microphone in which microphones are arranged at four vertices, which are vertices of a tetrahedron, at specified measurement points in a sound field, an omnidirectional speaker, and a center point at the center of gravity of the tetrahedron, A microphone arranged on a spherical surface capable of picking up an impulse response on the spherical surface with a radius of
A direction information extracting device for extracting direction information of a sound field, wherein the direction information extracting device includes an acoustic signal generating means for emitting an acoustic signal by an impulse from the omnidirectional speaker into a sound field; First input signal storage means for collecting and storing the acoustic signal radiated from the speaker with the four-channel microphone, and collecting and storing the acoustic signal radiated from the omnidirectional speaker with the microphone arranged on the spherical surface A second input signal storage unit;
A reflected sound arrival direction extracting means for extracting a direction of arrival of the reflected sound from the information of the input signal storing means, and a second reflected sound signal extracted by the reflected sound arrival direction extracting means. After correcting the time difference of each microphone of the microphone arranged on the spherical surface with respect to the reflected sound from the information,
Spatial separation means for synchronous addition, reflected sound characteristic time series separation means for temporally separating based on the arrival time of the reflected sound from the result obtained by the spatial separation means, and reflected sound characteristic time series separating means Time-series re-synthesis means for re-synthesizing only a specific reflected sound from the results obtained by
A direction information extraction method for a sound field, wherein direction information is output to an output of the time-series resynthesis unit.