JP2714098B2 - How to correct acoustic frequency characteristics - Google Patents

Description

Description: BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a sound frequency characteristic correction method for correcting a frequency characteristic of an acoustic room when a room environment of the acoustic room changes.

[Related Art] Generally, when reproducing a sound field, not only the frequency characteristics of the radiated sound of a speaker installed in an acoustic room but also the frequency characteristics of the acoustic room are taken into consideration, and the acoustic frequency characteristics of the acoustic room are set in advance. A graphic equalizer is generally used to correct the overall acoustic frequency characteristics in consideration of such a speaker and an acoustic room.

This involves radiating a signal sound such as white noise from a speaker, collecting the signal sound with a microphone installed at a certain position in the room, measuring the frequency characteristics of the output signal of the microphone, and reversing the frequency characteristics. The level of each frequency channel of the graphic equalizer is set so as to obtain characteristics.

By the way, recently, the development of a high-fidelity sound field reproduction system has been advanced, for example, based on data such as a reflection sound measured in an actual concert hall, a characteristic of a direct sound and a reverberation characteristic, a reflection sound in a concert hall and a reverberation sound. The same reflection sound and reverberation sound are formed by digital signal processing as shown in FIG. 4, and are reproduced by a reproduction speaker (S) three-dimensionally arranged in the acoustic room (R) as shown in FIG. It is considered to reproduce the sound field. In addition,
(P) in FIG. 4 indicates an actual sound receiving point.

By the way, when a sound field is reproduced by such a high fidelity sound field reproduction system, the sound frequency characteristics of the speakers (S) are arranged in the acoustic room (R) in the same manner as in actual listening. Correction must be performed accurately.

[Problems to be Solved by the Invention] When the correction is performed by the graphic equalizer described above, when a change in the indoor environment occurs, such as a change in the material or volume of the wall or floor of the acoustic room or a change in the arrangement of each speaker. However, it is necessary to set up a microphone for each change, measure the frequency characteristics, and reset the settings of each channel of the graphic equalizer, which is very time-consuming and requires a long time for the correction work. Not only that, since the frequency band of each channel of the graphic equalizer is wide, accurate correction cannot be performed.

The present invention has been made in consideration of the above points, and has as its object to easily and accurately correct the acoustic frequency characteristics when a change in the indoor environment occurs.

Means for Solving the Problems In order to achieve the above object, according to the method for correcting acoustic frequency characteristics of the present invention, at least three reproduction speakers for reproducing a three-dimensional sound field are arranged in an acoustic room, and each of the reproduction speakers is provided. One speaker
Each of the other reproduction speakers except for the reproduction speaker being used as the sound source speaker or an attached microphone attached to each of the other reproduction speakers is used as a sound receiving microphone while the sound source speaker is used as a sound source microphone. A signal sound radiated from a reproduction speaker being used as a speaker is received by each of the sound receiving microphones, and an output signal of each of the sound receiving microphones is Fourier-transformed to obtain a sound receiving frequency response function for each of the sound receiving microphones. Calculating and calculating an average of the plurality of sound receiving frequency response functions as a measured frequency response function of the acoustic room,
Thereafter, a filter frequency response function is derived by dividing the preset frequency response function of the acoustic chamber by the measurement frequency response function, and the filter frequency response function is inverse Fourier transformed to derive filter coefficients. Is a coefficient of a correction filter provided in the preceding stage of the reproduction speaker being used as the sound source speaker.

[Operation] With the above-described configuration, when a change in the indoor environment occurs, each of the reproduction speakers arranged in the room is sequentially used as a sound source speaker one by one, and a signal sound transmitted from the sound source speaker is generated. The sound receiving frequency response function is calculated from the output signal of each of the other reproduction speakers except for the sound source speaker that received the sound or the microphone attached to each of the other reproduction speakers, and the plurality of sound reception frequency response functions are averaged. A filter coefficient for correction is derived based on the measured frequency response function thus obtained, and the coefficient is used as a coefficient for a correction filter provided in a preceding stage of a reproduction speaker being used as a sound source speaker when calculating the coefficient.

Embodiment An embodiment will be described with reference to FIG. 1 to FIG.

In FIG. 1 showing a measurement system of frequency characteristics in an acoustic room, (1) is an acoustic room, (2a), (2b), (2c), (2d).
Are four reproduction speakers for reproducing a three-dimensional sound field. The reproduction speakers (2a) to (2d) are arranged at respective predetermined positions in the acoustic room (1). When the frequency characteristics are corrected, each of the reproduction speakers (2a) to (2d) is sequentially placed one by one. While being used as a sound source speaker, each reproduction speaker other than the reproduction speaker being used as the sound source speaker is used as a sound receiving microphone.

Further, in FIG. 1, (3) is an impulse signal generator, (4) is an amplifier for amplifying an impulse signal from the signal generator (3), and (5a) to (5c) are reproduction speakers as sound receiving microphones. Voltage limiting circuit connected to (6
a) to (6c) are A / D converters that A / D convert the output signals of each playback speaker as a sound receiving microphone, respectively, (7a)
To (7c) are FFs that calculate the sound receiving frequency response function by Fourier transforming the outputs of the A / D converters (6a) to (6c), respectively.
T, (8) is an arithmetic unit for averaging the sound receiving frequency response functions by the FFTs (7a) to (7c) and deriving a corrected filter frequency response function, and (9) is an inverse Fourier transform unit.
A filter coefficient is derived by performing an inverse Fourier transform on the filter frequency response function, and this filter coefficient is used as a coefficient of a correction filter provided in a preceding stage of the sound source speaker.

Next, a signal processing procedure will be described with reference to FIG.

Now, for example, the case where the reproduction speaker (2a) is used as a sound source speaker will be described. Except for the reproduction speaker (2a) used as a sound source speaker, three other reproduction speakers (2b) to (2d) receive sound. Used as

As shown in FIG. 2, the impulse signal from the signal generator (3) is amplified by an amplifier (4), supplied to a speaker (2a) being used as a sound source speaker, and radiated from the speaker (2a). You. The impulse signal radiated from the speaker (2a) is received by the speakers (2b) to (2d) being used as sound receiving microphones, and the output signals of the speakers (2b) to (2d) are subjected to voltage limiting. A / D converter (6a) through circuit (5a) through (5c) respectively
Input to (6c) and converted to digital signal.

At this time, when the output level of each of the speakers (2b) to (2d) is lower than a predetermined level, the noise is transmitted to the A / D converters (6a) to (6c) by the voltage limiting circuits (5a) to (5c) as background noise. Input is cut off.

Next, the outputs of the A / D converters (6a) to (6c) are Fourier-transformed by the FFTs (7a) to (7c) to calculate a sound receiving frequency response function. The average of each sound receiving frequency response function is calculated, and the measured frequency response random number H1 (z) of the acoustic room (1) is calculated. Thereafter, the filter frequency response function H3 (z) (= H2 (z) / H1 (z) is obtained by dividing the preset frequency response function H2 (z) of the acoustic room (1) by the measured frequency response function H1 (z). )) Is derived,
The filter frequency response function H3 (z) is inverse Fourier transformed by the inverse Fourier transformer (9) to derive filter coefficients.

Then, for example, in the high-fidelity sound field reproduction system shown in FIG. 3, a reproduction speaker (2a) being used as a sound source speaker
The filter coefficients derived as described above are set as the coefficients of the correction filter (10a) provided at the preceding stage of the above, and the other reproduction speakers (2b) to (2d) are sequentially output to the sound source one by one. The above-described measurement is repeated in the same manner in the case of a speaker, and as shown in FIG. 3, the coefficients of the correction filters (10b) to (10d) provided before the reproduction speakers (2b) to (2d), respectively. Is set.

By the way, in FIG. 3, (11) is an acoustic source such as a tuner or a CD player, (12) is a preamplifier, and (13)
Is a reverberation circuit.

Therefore, according to the above-described embodiment, when a change in the indoor environment such as a change in the material of the wall and floor of the acoustic room (1) or a change in the speaker arrangement occurs, the microphone is placed in the room every time the environment changes as in the related art. It is not necessary to measure the frequency characteristics by installing the existing speakers, and measure the frequency response function of the acoustic room using the existing reproduction speakers (2a) to (2d) to correct the filter (10a) to
Since the coefficient of (10d) can be derived, and the Fourier transform is used to derive the coefficients of the correction filters (10a) to (10d), processing that subdivides the frequency band can be performed. , It is possible to easily and accurately correct sound frequency characteristics in a shorter time than before,
It is possible to easily cope with changes in the indoor environment.

In the above-described embodiment, the case where the reproduction speakers themselves except the sound source speakers are used as the sound receiving microphones is described. However, the microphones attached to these reproduction speakers may be used as the sound receiving microphones.

[Effects of the Invention] According to the present invention, sound frequency characteristics can be corrected easily and in a short time by using only existing equipment such as a reproduction speaker, and furthermore, a sound source speaker is sequentially output one by one in an acoustic room. After obtaining a plurality of sound receiving frequency response functions for each of the reproducing speakers by using the Fourier transform for each of the reproducing speakers, averaging the plurality of sound receiving frequency response functions and performing measurement in the acoustic room. Since the frequency response function is calculated, the measured frequency response is close to the frequency response function when actually listening at the sound receiving point in the acoustic room where there are various conditions that affect the acoustic frequency characteristics such as reflected sound and reverberant sound A function can be obtained, and the finally obtained filter coefficient can be used for a correction filter provided in front of a reproduction speaker used as a sound source speaker when calculating the coefficient. Since the calculation of such coefficients is performed by sequentially using each reproduction speaker one by one as a sound source speaker, the acoustic frequency characteristics are corrected in consideration of the arrangement environment of each reproduction speaker in the acoustic room. Can be accurately reflected individually for each reproduction speaker.

[Brief description of the drawings]

1 to 3 show a first embodiment of a method for correcting an acoustic frequency characteristic according to the present invention, FIG. 1 is a block diagram of a frequency characteristic measuring system, FIG. 2 is an operation explanatory diagram of FIG. FIG. 3 is a block diagram of a high-fidelity sound field reproduction system, and FIG. 4 is a perspective view showing an example of an indoor arrangement of speakers for reproducing a general three-dimensional sound field. (1) Acoustic room, (2a)-(2d) ... Reproduction speaker,
(7a) to (7c): FFT, (8): Operation unit, (9):
... Inverse Fourier transform unit.

Claims (1)

(57) [Claims] 1. A reproduction speaker in which at least three reproduction speakers for reproducing a three-dimensional sound field are arranged in an acoustic room, and each of the reproduction speakers is used one by one as a sound source speaker, while being used as the sound source speaker. Each of the other reproduction speakers except for the above or an attached microphone attached to each of the other reproduction speakers is used as a sound reception microphone, and the signal sound radiated from the reproduction speaker being used as the sound source speaker is received by the respective sound reception microphones. The output signal of each of the sound-receiving microphones is Fourier transformed to calculate a sound-receiving frequency response function for each of the sound-receiving microphones, and the average of the plurality of sound-receiving frequency response functions is measured in the acoustic chamber. Calculated as a frequency response function, and then the preset frequency response function of the acoustic room is set to the measured frequency response. The filter frequency response function is derived by dividing by a number, a filter coefficient is derived by performing an inverse Fourier transform of the filter frequency response function, and the filter coefficient is used as a correction filter provided in a stage preceding a reproduction speaker being used as the sound source speaker. A method for correcting acoustic frequency characteristics, characterized by using a coefficient of