JP5605071B2 - Coefficient setting method of digital filter, coefficient setting device, coefficient setting program, and sound field correction method using digital filter - Google Patents

Description

  The present invention relates to a coefficient setting method for a digital filter, a coefficient setting apparatus, a coefficient setting program, and a sound field correction method using a digital filter.

  Patent Document 1 discloses a technique for performing sound field correction using a digital filter. That is, Patent Document 1 discloses a technique for correcting an audio signal with a digital filter so as to cancel noise and sound distortion generated in a sound field. In this technique, a predetermined correction is performed on a converted value obtained by performing a fast Fourier transform on the impulse response of the microphone, and a value obtained by performing an inverse Fourier transform on the converted converted value is set as a coefficient of the digital filter.

JP-A-8-33092

  The technique according to Patent Document 1 has a problem that the amount of calculation is large because it is necessary to perform fast Fourier transform and inverse Fourier transform in order to calculate the filter coefficient of the digital filter.

  Further, the technique according to Patent Document 1 does not set the coefficient array of the digital filter to be even symmetric. When the coefficient array is not an even object, the phase delay of the signal output from the digital filter differs depending on the frequency. For this reason, the frequency component that should be output at the same timing is output at a timing shifted, and there is a possibility that a sound different from the original sound is output to the sound field.

  The present invention has been made in view of the above problems, and is a coefficient setting method, a coefficient setting apparatus, a coefficient setting program, and a coefficient setting program capable of setting a coefficient array of a digital filter that performs sound field correction to be evenly symmetric with a small amount of computation. It is an object to provide a sound field correction method using a digital filter.

In order to achieve the above object, a digital filter coefficient setting method according to a first aspect of the present invention includes:
A sampling step in which the speaker is a plurality of times sampling a response signal value of the signal output from the microphone for collecting test sound when input impulse signal,
Among the plurality of response signal values sampled in the sampling step, the largest response signal value is set as a reference signal value that is set as a reference coefficient of a digital filter that outputs an audio signal to the speaker that outputs the measurement sound. A reference signal value selection step to be selected; and
Of the plurality of response signal values sampled in the sampling step, a response that is sampled at a time that is a predetermined time before or after the sampling time of the reference signal value and that is smaller than the reference signal value A distortion signal value selection step of selecting a signal value as a distortion signal value representing distortion or noise of sound output from the speaker;
A first setting step of setting the reference signal value selected in the reference signal value selection step as the reference coefficient ;
Computing for said than audio signals, a first coefficient used in operation to the audio signal before by the predetermined time, the audio signal after only the predetermined time during which the digital filter is arithmetic processing using said reference factor A second setting step of setting a second coefficient used in the processing to a reciprocal of the distortion signal value selected in the distortion signal value selection step ;
It is characterized by having.

Furthermore, in the coefficient setting method of the digital filter according to the first aspect,
A signal value storage step of storing in the storage device a plurality of correspondence information in which the response signal value sampled in the sampling step and the sampling time information indicating the sampling time of the response signal value are associated;
An average value calculating step of averaging the response signal value stored in the storage device in association with the sampling time information in the signal value storing step based on the sampling time information;
In the reference signal value selection step, the largest value among the response signal values averaged in the average value calculation step is selected as the reference signal value,
In the distortion signal value selection step, a value smaller than the reference signal value is selected as the distortion signal value among the response signal values averaged in the average value calculation step.
It is also good.

In order to achieve the above object, a digital filter coefficient setting apparatus according to a second aspect of the present invention provides:
Sampling means for sampling a response signal value of a signal output from a microphone that collects a measurement sound when the speaker receives an impulse signal, a plurality of times ;
Among a plurality of response signal values sampled by the sampling means, the largest response signal value is set as a reference signal value that is set as a reference coefficient of a digital filter that outputs an audio signal to a speaker that outputs the measurement sound. A reference signal value selection means to select;
Of a plurality of response signal values sampled by the sampling means, a response sampled at a time before or after a sampling time of the reference signal value by a predetermined time and smaller than the reference signal value Distortion signal value selection means for selecting a signal value as a distortion signal value representing distortion or noise of sound output from the speaker;
First setting means for setting the reference signal value selected by the reference signal value selecting means to the reference coefficient ;
Computing for said than audio signals, a first coefficient used in operation to the audio signal before by the predetermined time, the audio signal after only the predetermined time during which the digital filter is arithmetic processing using said reference factor Second setting means for setting a second coefficient used in the processing to a reciprocal of the distortion signal value selected by the distortion signal value selection means ;
It is characterized by providing.

In order to achieve the above object, a coefficient setting program for a digital filter according to a third aspect of the present invention includes:
Computer
Sampling means for sampling a response signal value of a signal output from a microphone that collects a measurement sound when the speaker receives an impulse signal, a plurality of times ,
Among a plurality of response signal values sampled by the sampling means, the largest response signal value is set as a reference signal value that is set as a reference coefficient of a digital filter that outputs an audio signal to a speaker that outputs the measurement sound. Reference signal value selection means to select,
Of a plurality of response signal values sampled by the sampling means, a response sampled at a time before or after a sampling time of the reference signal value by a predetermined time and smaller than the reference signal value Distortion signal value selection means for selecting a signal value as a distortion signal value representing distortion or noise of sound output from the speaker;
First setting means for setting the reference signal value selected by the reference signal value selection means to the reference coefficient ;
Computing for said than audio signals, a first coefficient used in operation to the audio signal before by the predetermined time, the audio signal after only the predetermined time during which the digital filter is arithmetic processing using said reference factor Second setting means for setting a second coefficient used in the processing to a reciprocal of the distortion signal value selected by the distortion signal value selection means;
It is made to function as.

In order to achieve the above object, a sound field correction method using a digital filter according to a fourth aspect of the present invention includes:
A sampling step in which the speaker is a plurality of times sampling a response signal value of the signal output from the microphone for collecting test sound when input impulse signal,
Among the plurality of response signal values sampled in the sampling step, the largest response signal value is set as a reference signal value that is set as a reference coefficient of a digital filter that outputs an audio signal to the speaker that outputs the measurement sound. A reference signal value selection step to be selected; and
Of the plurality of response signal values sampled in the sampling step, a response that is sampled at a time that is a predetermined time before or after the sampling time of the reference signal value and that is smaller than the reference signal value A distortion signal value selection step of selecting a signal value as a distortion signal value representing distortion or noise of sound output from the speaker;
A first setting step of setting the reference signal value selected in the reference signal value selection step as the reference coefficient ;
Computing for said than audio signals, a first coefficient used in operation to the audio signal before by the predetermined time, the audio signal after only the predetermined time during which the digital filter is arithmetic processing using said reference factor A second setting step of setting a second coefficient used in the processing to a reciprocal of the distortion signal value selected in the distortion signal value selection step ;
Wherein said digital filter, and the reference coefficient values set by the first setting step, and calculation processing using a first coefficient and the second coefficient set values in the second setting step An audio signal output step of outputting an audio signal to the speaker;
It is characterized by having.

  According to the present invention, the coefficient array of the digital filter that performs sound field correction can be set to be even symmetric with a small amount of calculation.

FIG. 1A is a system configuration diagram illustrating an example of an audio system including a coefficient setting device used in a sound field correction method and a coefficient setting method according to an embodiment of the present invention. FIG. 1B is a hardware configuration diagram illustrating a configuration example of the coefficient setting device. It is a figure showing the example of 1 structure of a digital filter. FIG. 3A is a flowchart illustrating an example of coefficient setting processing executed by the coefficient setting device. FIG. 3B is a flowchart illustrating an example of a response signal acquisition process executed by the coefficient setting device. FIG. 3C is a flowchart illustrating an example of correction processing executed by the coefficient setting device. FIG. 4A is a functional block diagram illustrating an example of functions that the coefficient setting device has. FIG. 4B is a diagram illustrating a configuration example of the response signal acquisition unit. FIG. 4C is a diagram illustrating a configuration example of the correction unit. FIG. 5A is a diagram illustrating an example of the average sampling signal. FIG. 5B is a diagram illustrating an example of a table in which sampling times are associated with average sampling signal magnitudes. FIG. 5C is a diagram illustrating an example of a coefficient array set in the digital filter. FIG. 6A is a diagram illustrating an example of an impulse response output from the digital filter in which the coefficient is set. FIG. 6B is a diagram illustrating an example of the impulse response of the speaker.

  Hereinafter, an audio system including a filter coefficient setting device used for implementing a filter coefficient setting method and a sound field correction method according to an embodiment of the present invention will be described with reference to the accompanying drawings.

  The audio system 100 according to the present embodiment is a car audio system that performs sound field correction on a digital audio signal output from a playback device and emits sound into a vehicle compartment that is a sound field. As shown in FIG. 1A, the audio system 100 includes a player device 10, a sound field correction device 20, an output circuit 30, a speaker 40, and a microphone 50, and is mounted on a vehicle.

  The player device 10 is a playback device that reads a digital audio signal from a recording medium and outputs the read signal. The recording medium includes, for example, a CD (Compact Disc), a DVD (Digital Versatile Disc), a Blu-ray disc, a hard disk, and a flash memory.

  The sound field correction device 20 includes a digital filter 21 and a filter coefficient setting device (hereinafter referred to as coefficient setting device) 22. The sound field correction device 20 is a kind of noise canceller that corrects an audio signal in advance so as to cancel distortion and noise generated in the sound output from the audio system 100 to the sound field. Note that audio distortion, noise, and the like are not included in the signal reproduced from the player device 10 such as abnormal sound caused by abnormal vibration of the speaker 40, reverberation sound generated in the sound field, and the like. It means sound distortion or abnormal sound caused by the sound field itself.

The digital filter 21 performs filter processing on the digital audio signal output from the player device 10 according to the set filter coefficient, and outputs the processed signal to the speaker 40.
The coefficient setting device 22 sets filter coefficients for canceling distortion and noise generated in the sound output to the sound field in the digital filter 21.

The output circuit 30 includes a DA (Digital / Analog) converter, an output amplifier, and the like, converts the digital audio signal output from the sound field correction device 20 into an analog audio signal, and outputs the converted signal to a speaker.
The speaker 40 converts the output signal of the output circuit 30 into air vibration and outputs it to the passenger compartment (sound field).
The microphone 50 is installed at the listening position at the time of measuring distortion and noise of the sound output to the sound field. At the time of measurement, the microphone 50 acquires the sound output from the speaker 40, converts the acquired sound into an electrical signal, and outputs the converted signal. The microphone 50 is detached from the audio system 100 during normal use.

  The sound field refers to a space or place where sound exists and includes the interior of the vehicle. In addition, sound distortion and noise generated in the sound field include reverberant sound. Specifically, since a speaker 40 that outputs sound is installed in a space or place where the sound exists, sound distortion and noise generated in the sound field are vibrations generated when the speaker 40 outputs sound. Including audio distortion due to That is, sound distortion and noise generated in the sound field vary depending on the vehicle type or the type of the speaker 40.

  Next, the configuration of the digital filter 21 shown in FIG. The digital filter 21 is configured by, for example, an FIR (Finite Impulse Response) filter, and the filter coefficient is set by the coefficient setting device 22. In addition, the digital filter 21 performs arithmetic processing on a digital audio signal (hereinafter simply referred to as an audio signal) output from the player device 10 during normal use, using already set filter coefficients, and processed audio. The signal is output to the output circuit 30. This is because the noise signal included in the audio signal is filtered.

  As shown in FIG. 2, the digital filter 21 is composed of 11 taps connected in order from the input side to which the audio signal X is input to the output side to output the audio signal Y. Here, the tap located closest to the input side is designated as the 0th tap, and the remaining 10 taps are named in the same manner. Each tap includes a delay element, a multiplier, and an adder. The delay element delays the input signal by one sampling period (that is, one sampling period), and then multiplies the delayed signal by the same tap and the adjacent tap located on the output side by one. Output to. The multiplier multiplies the signal delayed by the delay element by the coefficient value set by the coefficient setting device 22. The adder multiplies the signal output from the adjacent tap located on the input side by one (that is, the tap for processing a signal whose delay time is shorter by one sampling period) by the coefficient value by the multiplier. Add the received signals. After that, the adder outputs the added signal to the adjacent tap located on the output side by one (that is, a tap for processing a signal having a delay time longer by one sampling period).

  Here, the coefficient of the multiplier included in the 0th tap is denoted as K0, and the coefficients of the multipliers included in the remaining 10 taps are sequentially named K1 to K10. Furthermore, the arrangement of the coefficient values of the multipliers respectively provided in the 0th tap to the 10th tap in order from K0 is simply referred to as the coefficient arrangement of the digital filter 21 hereinafter. That is, each element of the coefficient array will be described as being arranged in order from the coefficient used for processing the audio signal with a short delay time, but the present invention is not limited to this.

  Here, the fifth tap is a central tap located at the center of the 0th tap to the 10th tap. The sixth tap, which is one output side of the center tap, is used for arithmetic processing of the audio signal at a time one sampling period before the audio signal that is arithmetically processed with the coefficient K5 of the fifth tap. That is, the coefficients K6 to K10 included in the taps on the output side of the center tap are coefficients (hereinafter referred to as coefficients) used in the arithmetic processing for the audio signal a predetermined time before the audio signal that the digital filter 21 performs arithmetic processing using the coefficient K5. , Referred to as the first coefficient). Conversely, the coefficients K0 to K4 of the tap on the input side with respect to the center tap are coefficients (hereinafter referred to as coefficients) used in the arithmetic processing for the audio signal a predetermined time before the signal that the digital filter 21 performs arithmetic processing using the coefficient K5. , Referred to as the second coefficient).

Next, the coefficient setting device 22 for setting the coefficient of the digital filter 21 will be described.
As shown in FIG. 1B, the coefficient setting device 22 includes a central processing unit (CPU) 22a, a random access memory (RAM) 22b, a read only memory (ROM) 22c, an input unit 22d, and an output unit 22e. Is done.
The CPU 22a performs overall processing of the coefficient setting device 22 by executing software processing according to a program stored in the ROM 22c. The RAM 22b is a storage device that temporarily stores data to be processed by the CPU 22a.
The input unit 22d includes, for example, an AD (Analog / Digital) converter, inputs an analog signal output from the microphone 50, and converts the input signal into a digital signal. The output unit 22e is controlled by the CPU 22a and outputs a measurement signal to be described later to the output circuit 30. The output unit 22e is controlled by the CPU 22a to set the coefficient of the digital filter 21 in the digital filter 21.

  The CPU 22a in FIG. 1B executes a coefficient setting process as shown in FIG. 3A, whereby a measurement signal output unit 221, a response signal acquisition unit 222, and a correction unit as shown in FIG. 4A. 223 and the setting unit 225, and in cooperation with the RAM 22b, functions as a coefficient array database (hereinafter simply referred to as coefficient DB) 224.

  When the execution of the coefficient setting process of FIG. 3A is started, the measurement signal output unit 221 of FIG. 4A is an impulse signal (hereinafter referred to as a measurement signal) used for measuring audio distortion and noise. Is output to the speaker 40 via the output circuit 30 (step S01).

  Next, the response signal acquisition unit 222 executes a response signal acquisition process as shown in FIG. 3B for acquiring a response signal representing a response to the measurement signal (step S02). As shown in FIG. 4B, the response signal acquisition unit 222 includes a sampling unit 222a, a signal value storage unit 222b, a signal value database (hereinafter referred to as signal value DB) 222c, and an averaging unit 222d. .

  When the execution of the response signal acquisition process illustrated in FIG. 3B is started, the sampling unit 222a in FIG. 4B performs a signal value of the response signal output from the microphone 50 (hereinafter simply referred to as a response signal value). Are sampled (that is, sampled) at the sampling period of the digital filter 21 (step S11).

  Thereafter, the signal value storage unit 222b converts the signal value sampled by the sampling unit 222a (hereinafter referred to as the sampled signal value) into sampling time information indicating the sampling time (that is, the sampling time), and the number of measurements. Correspondence information associated with the information representing the information is generated, and the generated correspondence information is stored in the signal value DB 222c (step S12). The sampling time is measured based on the measurement start time that is the time when the measurement signal output unit 221 outputs the measurement signal. According to these configurations, even when the response signal is measured for each measurement signal output over a plurality of times, the value of the response signal at each time can be compared for each sampling time.

  After step S12 is executed, the averaging unit 222d averages the sampling signal values accumulated in the signal value DB 222c in association with the sampling time information by the signal value storage unit 222b based on the sampling time information. (Step S13). As a specific example, the averaging unit 222d calculates an average value of a plurality of sampled signal values sampled at the same sampling time. Thereafter, the response signal acquisition unit 222 ends the execution of the response signal acquisition process. According to this configuration, for example, an error in signal value caused by noise can be reduced.

  After step S02 of FIG. 3A is executed, the correction unit 223 of FIG. 4A changes the coefficient value of the digital filter 21 to a coefficient value for canceling audio distortion or noise based on the measurement result. Correction processing as shown in FIG. 3C is executed (step S03). As shown in FIG. 4C, the correction unit 223 includes a reference signal value selection unit 223a, a distortion signal value selection unit 223b, a coefficient array generation unit 223c, an inverse correction unit 223d, a signal value storage unit 223e, and a normalization. Unit 223f and coefficient array storage unit 223g.

  When the execution of the correction process shown in FIG. 3C is started, the reference signal value selection unit 223a performs the sampling signal value averaged in step S13 in FIG. 3B (hereinafter, the average sampling signal value). The reference signal value is selected (step S21). Here, the reference signal value refers to a signal value used for determining an average sampled signal value (hereinafter referred to as a distorted signal value) representing distortion or noise of the measured voice among the average sampled signal value. Specifically, the reference signal value selection unit 223a selects the largest average sampling signal value as the reference signal value. Usually, the impulse response is because a plurality of distortion spectra having a signal value smaller than the signal value are generated after the largest signal value.

  Here, as shown in FIGS. 5A and 5B, the magnitude (level) of the average sampling signal value at the time of the first sample is equal to the average sampling signal at the time of the second to sixth samples. Since the value is larger than the value, the reference signal value selection unit 223a selects the average sampling signal value of the first sample as the reference signal value. In FIG. 5A, the horizontal axis represents time, and the vertical axis represents the average response signal value (level). FIG. 5B is a table that displays the sampling time and the average sampling signal value in association with each other.

  After step S21 in FIG. 3C is executed, the distortion signal value selection unit 223b in FIG. 4C performs the magnitude of the reference signal value selected by the reference signal value selection unit 223a and the average sampling signal value. One or a plurality of distortion signal values are selected from the average sampled signal values based on the magnitude of (step S22). Usually, when the voice is distorted, the volume of the voice is distorted. Therefore, according to this configuration, the distortion signal value can be accurately selected.

  More specifically, the distortion signal value selection unit 223b determines an average sampling signal value smaller than the reference signal value as the distortion signal value. That is, as shown in FIGS. 5A and 5B, the average sampled signal values after the second sample from the level “1.0” of the average sampled signal value at the time of the first sample as the reference signal value. Therefore, the distortion signal value selection unit 223b determines the average sampling signal value after the second sample as the distortion signal value. This is because the volume of distorted audio is usually smaller than the volume of undistorted audio.

  After step S22 of FIG. 3C is executed, the coefficient array generation unit 223c of FIG. 4C generates a coefficient array of the digital filter 21 as shown in FIG. 5C (step S23). . The coefficient array in FIG. 5C has element numbers A0 to A10. In the element identified by the element number A0, the coefficient value set to the coefficient K0 of the digital filter 21 shown in FIG. 2 is stored. Similarly, coefficient values respectively set for the coefficients K1 to K10 are stored in the elements identified by the element numbers A1 to A10.

  After step S23 in FIG. 3C is executed, the inverse correction unit 223d in FIG. 4C reverses the sign of the distortion signal value and corrects it to the inverse of the value before correction. Execute (step S24). This is because sound distortion and noise are canceled out using the inverse of the distortion signal value. Note that a certain reciprocal number is a number in which a value obtained by adding the certain number to the value “0”.

  Next, the signal value storage unit 223e stores the reference signal in the coefficient array, and converts the distortion signal value (hereinafter simply referred to as a corrected distortion signal) that has been reversely corrected by the inverse correction unit 223d into the coefficient array of the reference signal value. Is stored in a symmetrical position with respect to a position (referred to as a reference position) (step S25).

  Specifically, the signal value storage unit 223e stores the average sampled signal values in the coefficient array in order from the earliest sampling time in order from the element A5 of the coefficient array. Here, since the sampling time of the reference signal value shown in FIG. 5 (a) is the earliest among the average sampling signal values, the signal value storage unit 223e has a filter array as shown in FIG. 5 (c). Is stored in the A5th position, which is the center position. That is, the A5th position of the filter array is the reference position, and the coefficient K5 of the digital filter 21 in which the value of the element stored at the A5th position is set is referred to as the reference coefficient.

  Further, since the sampling time of the distortion signal value shown in FIG. 5A is later than the sampling time of the reference signal value, the signal value storage unit 223e has a filter array as shown in FIG. The corrected distortion signal value is stored in the A6th to A10th elements. Further, the signal value storage unit 223e stores the corrected distortion signal value at a position that is symmetric with respect to the A5th position, which is the reference position, in order to store not only the A6th but also the A4th array element. The inverse of the average sampling signal value of the second sample (that is, the corrected distortion signal value) is stored. Similarly, the signal value storage unit 223e supplies the corrected distortion signal values of the third to sixth samples, not only to the A7th to A10th elements, but also to the A3th to A0th elements of the coefficient array, respectively. Store.

  After step S25 of FIG. 3C is executed, the normalization unit 223f of FIG. 4C converts the element of the coefficient array stored by the signal value storage unit 223e into the signal value ( That is, normalization is performed so that the magnitude of the reference signal value is 1 (step S26). As a specific example, the normalization unit 223f divides each element of the coefficient array by the absolute value of the reference signal value.

  Next, the coefficient array storage unit 223g stores the coefficient array whose values are stored by the signal value storage unit 223e in the coefficient DB 224 (step S27). Thereafter, the correction unit 223 in FIG. 4B ends the execution of the correction process.

  After the execution of step S03 shown in FIG. 3A is completed, the first setting unit 225a included in the setting unit 225 in FIG. 4A is a reference at the reference position A5 of the coefficient array stored in the coefficient DB 224. The signal value is set to the reference coefficient K5 of the digital filter 21 (step S04).

  Next, the second setting unit 225b uses the corrected distortion signal values at the A0 to A4 positions and the A6 to A10 positions of the coefficient array stored in the coefficient DB 224 as the second coefficients of the digital filter 21. The coefficients K0 to K4 and the first coefficient K6 to K10 are set (step S05). Thereafter, the execution of the coefficient setting process is terminated.

  The first setting unit 225a and the second setting unit 225b calculate an average value for each element of a plurality of coefficient arrays stored in the coefficient DB 224 (that is, a plurality of coefficient arrays calculated over a plurality of times). In addition to the calculation, a configuration in which a coefficient array having the calculated average value as a coefficient element is set in the coefficient array of the digital filter 21 can be employed. According to this configuration, it is possible to reduce the influence of noise generated during measurement of the sound field on the coefficient value.

  As another specific example, the first setting unit 225a and the second setting unit 225b are configured to set the coefficient array selected by the user in the digital filter 21 among the plurality of coefficient arrays stored in the coefficient DB 224. Can be adopted.

Next, the impulse response of the digital filter 21 in which the coefficient array is set to be even symmetric will be described.
When an impulse signal is input, the digital filter 21 in which the coefficient array is set outputs an even symmetric impulse response as shown in FIG. Among the responses, the signal of the first sample shown in the fourth quadrant of FIG. 6A output by the digital filter 21 in the first sample is the signal in the second sample shown in the first quadrant of FIG. It is a signal that cancels the distortion or noise of the voice. That is, when the speaker 40 outputs the sound of the first sample shown in FIG. 5A to the sound field based on the signal output from the digital filter 21 at the time of the 0th sample, it starts to vibrate by outputting the sound. To do. Next, when the sound is output based on the signal output from the digital filter 21 at the time of the first sample, the speaker 40 performs an operation of canceling the vibration generated at the time of the 0th sample. For this reason, as shown in FIG. 6B, the distortion or noise in the sound field of the voice shown in the second sample in FIG. 5A is eliminated.

Similarly, the signal in FIG. 6A output from the second sample to the fifth sample by the digital filter 21 is a signal that cancels the distortion or noise of the sound in the third to sixth samples in FIG. 5A, respectively. is there.
For this reason, as shown in FIG. 6B, the impulse response result of the speaker 40 is the distortion shown in the first quadrant and the fourth quadrant in the impulse response result of the digital filter 21 shown in FIG. 6A. Alternatively, a response result in which noise is canceled is obtained. Therefore, the sound field correction apparatus 20 can output a sound that cancels the distortion or noise of the sound output from the speaker 40 based on the measurement result of the measurement sound in the sound field.

  According to the above configuration, since the sign of the measured distortion signal value is reversed and set to the coefficient array of the digital filter 21, the digital filter 21 that reduces distortion or noise of the sound collected by the microphone 50 requires less calculation. Can be generated in quantity. In addition, according to these configurations, the coefficient array of the even symmetric digital filter can be generated with a smaller amount of calculation than in the prior art.

  In particular, the coefficient setting device 22 sets the coefficient array so that the digital filter 21 becomes a linear phase FIR filter by setting the coefficient array of the digital filter 21 so that the filter length is odd and evenly symmetric. Here, the phase delay of the audio signal output from the linear phase FIR filter becomes a substantially constant value even if the frequency of the input audio signal differs. For this reason, even if audio signals having different frequencies such as low frequency and high frequency are simultaneously input to the digital filter 21, the phase delay of each audio signal becomes substantially the same, so that audio signals having different frequencies are output from the digital filter 21. Output almost simultaneously. For this reason, even if audio signals having different frequencies are input to the digital filter 21 at the same time, sounds having different frequencies reach the position where the microphone 50 is installed (that is, the listening position of the user during normal use) almost simultaneously. That is, since it is possible to prevent the frequency components that should be output at the same timing from being output at a shifted timing, the original sound represented by the audio signal read by the player device 10 can be output to the sound field.

  Furthermore, according to these configurations, the coefficient array of the digital filter 21 that reduces the difference between the measurement signal output to the speaker 40 and the response signal input to the microphone 50 without using, for example, Fourier analysis and inverse Fourier analysis. It can be calculated.

  Furthermore, according to these configurations, for example, a coefficient array that reduces the difference can be calculated without using an LMS method (Least Mean Square). Therefore, not only can the amount of calculation for calculating the coefficient array be reduced, but also convergence can be achieved. There is no problem.

  Furthermore, since the audio system 100 can generate a coefficient array with a small amount of computation, instead of the coefficient setting device 22, representative function units (measurement signal output unit 221, sampling unit 222a, A configuration including a circuit that realizes a reference signal value selection unit 223a, a distortion signal value selection unit 223b, an inverse correction unit 223d, a normalization unit 223f, and a setting unit 225) can be employed. According to this configuration, for example, the coefficient array of the digital filter 21 can be set to be even symmetric with a simple and inexpensive configuration that does not use a DSP (Digital Signal Processor) or a microcomputer.

  In the present embodiment, the sound field correcting device 20 outputs the measurement signal to the speaker 40 once, and also based on the sampled response signal value of the microphone 50 that collected the measurement sound (that is, sampling). A configuration in which the coefficient array of the digital filter 21 is set (without averaging the signal values) can be employed. In addition, the sound field correcting device 20 inputs the measurement signal to the digital filter 21 in which the coefficient array has already been set, and again, based on the response signal value of the microphone 50 that has collected the measurement sound, the coefficient of the digital filter 21. A configuration for setting an array can be employed.

  In the present embodiment, the audio system including the sound field correction device 20 may be built in the car navigation device, or may be used indoors or outdoors, for example, without being mounted on the vehicle. Furthermore, the sound field correction apparatus 20 may be incorporated in a CD player, a DVD player, a Blu-ray disc player, or a portable audio player.

In the present embodiment, the reference signal value selection unit 223a has been described as selecting the reference signal value based on the magnitude of the response signal value. However, the reference signal value selection unit 223a is based on the sampling time of the response signal value. The reference signal value may be selected. Specifically, the reference signal value selection unit 223a may first select a response signal value that exceeds a predetermined threshold as the reference signal value. This is because the distortion of the sound caused by the vibration of the speaker 40 often occurs after the sound output causing the vibration.
Similarly, the distortion signal value selection unit 223b may select a distortion signal value based on the sampling time of the response signal value and the sampling time of the reference signal value. Specifically, the distortion signal value selection unit 223b can adopt a configuration in which a response signal value sampled at a time later than the sampling time of the reference signal value is selected as a distortion signal value. Usually, the impulse response is because a plurality of distortion spectra having a signal value smaller than the signal value are generated after the largest signal value. According to these configurations, it is possible to accurately specify audio distortion caused by vibration of the speaker 40 caused by audio output.

  It should be noted that the present invention can be realized by a sound field correction device having the same functions and configurations as those of the sound field correction device 20 of the above-described embodiment, and also by applying a program to an existing sound field correction device. It is also possible to function as a sound field correcting device used for implementing the sound field correcting method using the digital filter and the coefficient setting method of the digital filter. In this case, a program for causing a computer (a control unit such as a CPU) having the same configuration as that of the sound field correction apparatus 20 exemplified in the above embodiment to realize the same function as the function of the sound field correction apparatus 20 described above. Can be made to function as a device used in the coefficient setting method according to the present invention.

  The application method of such a program is arbitrary, for example, it can be stored and distributed in a recording medium such as a memory card, CD-ROM, or DVD-ROM, and can also be distributed via a communication medium such as the Internet. .

DESCRIPTION OF SYMBOLS 10 Player apparatus 20 Sound field correction apparatus 21 Digital filter 22 Coefficient setting apparatus 30 Output circuit 40 Speaker 50 Microphone 22a CPU
22b RAM
22c ROM
22d Input unit 22e Output unit 221 Measurement signal output unit 222 Response signal acquisition unit 222a Sampling unit 222b Signal value storage unit 222c Signal value DB
222d Averaging unit 223a Reference signal value selection unit 223b Distortion signal value selection unit 223c Coefficient array generation unit 223d Reverse correction unit 223e Signal value storage unit 223f Normalization unit 223g Coefficient array storage unit 224 Coefficient DB
225 setting unit 225a first setting unit 225b second setting unit

Claims (5)

A sampling step in which the speaker is a plurality of times sampling a response signal value of the signal output from the microphone for collecting test sound when input impulse signal,
Among the plurality of response signal values sampled in the sampling step, the largest response signal value is set as a reference signal value that is set as a reference coefficient of a digital filter that outputs an audio signal to the speaker that outputs the measurement sound. A reference signal value selection step to be selected; and
Of the plurality of response signal values sampled in the sampling step, a response that is sampled at a time that is a predetermined time before or after the sampling time of the reference signal value and that is smaller than the reference signal value A distortion signal value selection step of selecting a signal value as a distortion signal value representing distortion or noise of sound output from the speaker;
A first setting step of setting the reference signal value selected in the reference signal value selection step as the reference coefficient ;
Computing for said than audio signals, a first coefficient used in operation to the audio signal before by the predetermined time, the audio signal after only the predetermined time during which the digital filter is arithmetic processing using said reference factor A second setting step of setting a second coefficient used in the processing to a reciprocal of the distortion signal value selected in the distortion signal value selection step ;
A coefficient setting method for a digital filter, comprising: A signal value storage step of storing in the storage device a plurality of correspondence information in which the response signal value sampled in the sampling step and the sampling time information indicating the sampling time of the response signal value are associated;
An average value calculating step of averaging the response signal value stored in the storage device in association with the sampling time information in the signal value storing step based on the sampling time information;
In the reference signal value selection step, the largest value among the response signal values averaged in the average value calculation step is selected as the reference signal value,
In the distortion signal value selection step, a value smaller than the reference signal value is selected as the distortion signal value among the response signal values averaged in the average value calculation step.
The coefficient setting method for a digital filter according to claim 1 . Sampling means for sampling a response signal value of a signal output from a microphone that collects a measurement sound when the speaker receives an impulse signal, a plurality of times ;
Among a plurality of response signal values sampled by the sampling means, the largest response signal value is set as a reference signal value that is set as a reference coefficient of a digital filter that outputs an audio signal to a speaker that outputs the measurement sound. A reference signal value selection means to select;
Of a plurality of response signal values sampled by the sampling means, a response sampled at a time before or after a sampling time of the reference signal value by a predetermined time and smaller than the reference signal value Distortion signal value selection means for selecting a signal value as a distortion signal value representing distortion or noise of sound output from the speaker;
First setting means for setting the reference signal value selected by the reference signal value selecting means to the reference coefficient ;
Computing for said than audio signals, a first coefficient used in operation to the audio signal before by the predetermined time, the audio signal after only the predetermined time during which the digital filter is arithmetic processing using said reference factor Second setting means for setting a second coefficient used in the processing to a reciprocal of the distortion signal value selected by the distortion signal value selection means ;
A coefficient setting device for a digital filter, comprising: Computer
Sampling means for sampling a response signal value of a signal output from a microphone that collects a measurement sound when the speaker receives an impulse signal, a plurality of times ,
Among a plurality of response signal values sampled by the sampling means, the largest response signal value is set as a reference signal value that is set as a reference coefficient of a digital filter that outputs an audio signal to a speaker that outputs the measurement sound. Reference signal value selection means to select,
Of a plurality of response signal values sampled by the sampling means, a response sampled at a time before or after a sampling time of the reference signal value by a predetermined time and smaller than the reference signal value Distortion signal value selection means for selecting a signal value as a distortion signal value representing distortion or noise of sound output from the speaker;
First setting means for setting the reference signal value selected by the reference signal value selection means to the reference coefficient ;
Computing for said than audio signals, a first coefficient used in operation to the audio signal before by the predetermined time, the audio signal after only the predetermined time during which the digital filter is arithmetic processing using said reference factor Second setting means for setting a second coefficient used in the processing to a reciprocal of the distortion signal value selected by the distortion signal value selection means;
A coefficient setting program for a digital filter, which is made to function as: A sampling step in which the speaker is a plurality of times sampling a response signal value of the signal output from the microphone for collecting test sound when input impulse signal,
Among the plurality of response signal values sampled in the sampling step, the largest response signal value is set as a reference signal value that is set as a reference coefficient of a digital filter that outputs an audio signal to the speaker that outputs the measurement sound. A reference signal value selection step to be selected; and
Of the plurality of response signal values sampled in the sampling step, a response that is sampled at a time that is a predetermined time before or after the sampling time of the reference signal value and that is smaller than the reference signal value A distortion signal value selection step of selecting a signal value as a distortion signal value representing distortion or noise of sound output from the speaker;
A first setting step of setting the reference signal value selected in the reference signal value selection step as the reference coefficient ;
Computing for said than audio signals, a first coefficient used in operation to the audio signal before by the predetermined time, the audio signal after only the predetermined time during which the digital filter is arithmetic processing using said reference factor A second setting step of setting a second coefficient used in the processing to a reciprocal of the distortion signal value selected in the distortion signal value selection step ;
Wherein said digital filter, and the reference coefficient values set by the first setting step, and calculation processing using a first coefficient and the second coefficient set values in the second setting step An audio signal output step of outputting an audio signal to the speaker;
A sound field correction method using a digital filter characterized by comprising:

Images