JPH0766647A - Acoustic characteristic controller - Google Patents

Abstract

PURPOSE:To impart a desired acoustic characteristic without preliminarily determining the transmission function of acoustic space. CONSTITUTION:A filter circuit 13 imparts a prescribed filter characteristic to an acoustic signal xn and inputs the signal in an acoustic signal propagation system 11 and a delay characteristic imparting part 14. The speaker 11a of the acoustic signal propagation system outputs an acoustic signal to acoustic space 11b and a microphone 11c provided in the vicinity of a listener detects the acoustic signal propagating in acoustic space. The delay characteristic imparting part imparts the delay transmission characteristic z<-n> of the acoustic signal propagation system to the acoustic signal inputted from the filter circuit. An adaptive signal processing part 15 performs an adaptive signal processing so that the synthetic signal of the output of the delay characteristic imparting part and the output of an adaptive filter 15a may be minimum and updates the coefficient so that an adaptive filter characteristic W may be the inverse function of the transmission characteristic C (=z<-n>/C) of the acoustic signal propagation system. A coefficient setting means 17 sets the coefficient of the adaptive filter to the filter circuit by the caesura of music (intermusic) or by the instruction of a user.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an acoustic characteristic control device, and more particularly to an acoustic characteristic control device using an adaptive filter.

[0002]

2. Description of the Related Art As a method of controlling an acoustic characteristic, as shown in FIG. 3, a characteristic (transfer function) C of the controlled system is canceled before a controlled system 1 which is an acoustic space, and a desired characteristic is obtained. There is a method of inserting the control filter 2. Controlled system 1
Includes a speaker, an acoustic space, and a detection unit such as a microphone that detects an acoustic signal output from the speaker in the acoustic space at an observation point (for example, near the listener's ear). FIG. 4 is an explanatory diagram of the controlled system 1 in the case of including two speakers and two microphones, and SP
₁ and SP ₂ are speakers, MIC ₁ and MIC ₂ are microphones, ₁ a to ₁ b are power amplifiers, and 1 c to 1 d are microphone amplifiers. Acoustic signals generated from the speakers SP _{1 to} SP ₂ are acoustic spaces SSP having predetermined frequency / phase characteristics.
To reach the microphones MIC _{1 to} MIC ₂ provided at each observation point (in the vicinity of the listener's ear).

_Letting C _{ji be} the transfer characteristic of the propagation system in which the i-th acoustic signal x _i reaches the j-th microphone MIC _j , the transfer function C of the acoustic space SSP is

[Equation 1] And generally

[Equation 2] Becomes

The transfer function of the control filter 2 is determined to be H · z ⁻ⁿ / C (1). Here, H is a desired transfer function, and z ⁻ⁿ is a delay transfer function determined so that the inverse characteristic (z ⁻ⁿ / C) of the controlled system (C) satisfies the causality. Therefore, the total transfer function of the controlled system 1 and the control filter 2 is H
· Z ^-n, and the acoustic signal xn 'is ^{X' = H · z -n ·} X next to the listener hears becomes to impart transmission characteristic H and delay characteristics desired. Incidentally, a value obtained by converting xn into two is expressed as X.

The following two methods have been conventionally proposed for constructing a control filter characteristic having the characteristic of equation (1).
The first method is a method of measuring a transfer function C in an acoustic space and calculating an inverse characteristic (z ⁻ⁿ / C) of the transfer function to form a control filter, and the second method is a MEF.
This is a method of configuring a control filter using an adaptive filter that performs adaptive signal processing based on an X-LMS (Least Mean Square) algorithm. In the first method, the acoustic space S
The transfer function C of SP is measured by some method, and the measured C
The control filter 2 is calculated from For example, white noise is output from a speaker and collected by a microphone, the white noise and the collected signal are Fourier transformed, and a transfer function C is obtained using two functions obtained by the Fourier transform as an input function and an output function.

In the second method, the characteristic of the adaptive filter is determined by the adaptive signal processing using the MEFX-LMS algorithm as shown in the equation (1), the acoustic signal is input to the adaptive filter, and the output of the adaptive filter is output. Input to controlled system 1. FIG. 5 is a block diagram of a conventional example in which a control filter is realized by the second method. 1 is a controlled system, 21 is an adaptive signal processing unit, 22 is a desired characteristic imparting unit for imparting a transfer function characteristic obtained by multiplying a desired transfer characteristic H by a delay characteristic z ⁻ⁿ , and 23 is a microphone (observation point of the controlled system). The sound signal xn 'and the output signal dn of the desired characteristic imparting section 22
In the synthesizing section that outputs the error signal en of the above, the control filter 2 shown in FIG.

The adaptive signal processing unit 21 is a MEFX-LMS.
An LMS signal processing unit 21a that performs adaptive signal processing according to an algorithm, an adaptive filter 21b having a digital filter configuration, and a filtered signal by convolving the acoustic signal xn with a propagation characteristic (transfer function) C of an acoustic signal propagation system from a speaker to a microphone. It has a filtered X signal producing filter 21c for producing an X signal rn. LMS signal processing unit 2
1a receives an error signal en and an acoustic signal (filtered X signal) rn input through a filter 21c, and performs adaptive signal processing using these signals so that the root mean square of the error signal en is minimized. The coefficient of the adaptive filter 21b is determined. That is, the LMS signal processing unit 21a determines the coefficient of the adaptive filter 21b so that the root mean square of the error signal en is minimized according to the multiple error filtered X LMS (MEFX-LMS) algorithm.

The adaptive filter 21b is an LMS signal processing unit 2
The acoustic signal xn is subjected to digital filter processing according to the coefficient determined by 1a, whereby the acoustic signal xn is given a transfer characteristic W and input to a speaker (not shown) of the controlled system 1. The adaptive filter 21b, as shown in FIG.
It is composed of an IR digital filter, for example, an input signal
(Acoustic signal) xn, which sequentially delays one sampling time, delay elements DL, DL ... And a coefficient w for each delay element output.
Multiplier M for multiplying ₁ (n), w ₂ (n), w ₃ (n) ... w _N (n)
Are realized by L, ML, ... And addition units AD, AD, ... Which sequentially add the outputs of the respective multiplication units.

The acoustic signal at the current time n · Ts is xn,
The coefficient of each multiplier at₁(n), w ₂(n), w₃(n) ...
w_N(n) and the output is yn, the adaptive filter 21b
formula

[Equation 3] Then, the signal yn is output.

Filter 21c for producing filtered X signal
As shown in FIG. 7, it is configured by an FIR type digital filter, and for example, delay elements DL, DL ... Which sequentially delay the input signal xn by one sampling time, and coefficients c ₁ , c ₂ , multiplication unit ML for multiplying the c ₃ ··· c _{M, M}
L, ... And an adder A for sequentially adding the outputs of the respective multipliers
It is realized by D, AD ... Coefficients c ₁ , c ₂ , c ₃ ... c _M
Is determined so as to simulate the propagation characteristics of the acoustic space propagation system (the system from the speaker to the observation point).

Output the acoustic signal at time n · Ts as xn
If (filtered X signal) is r (n), the filter 21
c is the following formula

[Equation 4] And outputs the filtered X signal r (n).

The LMS signal processor 21a uses the adaptive filter 2 at time (n + 1) · Ts after one sampling time Ts.
1b coefficients w ₁ (n + 1), w ₂ (n + 1), w ₃ (n + 1) ... w _N (n +
1) is determined by the following equation (coefficient updating equation) using the coefficient at the current time n · Ts, the combined signal en, and the filtered X signal rn.

[Equation 5] However, the j-th filter coefficient updating formula is given by w _j (n + 1) = w _j (n) + μ · r (n-j + 1) · en (5).

In equation (5), (n) is the value at the current sampling time, (n + 1) is the value one sampling time later, (n-1) is the value one sampling time earlier, and (n-2). Means the value two sampling times before, ... Further, μ is a constant (step size parameter) of 1 or less that determines the step of updating the coefficient of the adaptive filter. As a result, the characteristic W of the adaptive filter 21b finally becomes W = H · z ⁻ⁿ / C, and the output of the controlled system 1, in other words, the acoustic signal xn ′ at the listener's ear is X ′ = H · z ^{−. n} · X, and the desired acoustic characteristic H and delay characteristic are convoluted with the acoustic signal xn.

[0014]

In the first method of measuring the transfer function C of the acoustic space SSP and creating the control filter 2 from the measured C, it is necessary to emit white noise from the speaker and collect it by the microphone. There is a problem that it is uncomfortable for the listener and the measurement becomes complicated. Also, the first
The method of (1) has a problem that the amount of calculation for obtaining the inverse filter is large. Furthermore, in the first method, if the transfer characteristic C of the acoustic space changes in accordance with the number of passengers or the like, the intended control cannot be performed and a control error occurs.

On the other hand, the second method based on adaptive signal processing has a problem that the transfer function C of the acoustic space SSP must be measured in advance as in the first method. In the second method, since adaptive signal processing is performed according to the MEFX-LMS algorithm, it is necessary to convolve the transfer function C of the acoustic space with the acoustic signal xn to create the filtered X signal rn. For this reason, the calculation shown in the equation (3) is required, and there is a problem that the calculation amount increases. Further, in an automobile, the acoustic system (controlled system) in the vehicle compartment is likely to change, so that the characteristic of the adaptive filter constantly changes following it, and there is a problem that it sounds unnatural depending on the music being listened to. Furthermore, since the transfer function C of the acoustic space changes according to the number of passengers and the riding position,
There is a case where the transfer function measured in advance differs from the actual transfer function, and in such a case, the characteristic W of the adaptive filter does not become exactly W = H · z ⁻ⁿ / C, and the desired acoustic characteristic cannot be given. There is. From the above, it is an object of the present invention to provide an acoustic characteristic control device capable of imparting desired acoustic characteristics without having to obtain a transfer function of an acoustic space and without frequently changing the characteristics of an adaptive filter. .

[0016]

According to the present invention, the above object is to provide an acoustic signal propagation system and an acoustic signal having a speaker, an acoustic space, and a detection means for detecting an acoustic signal output to the acoustic space. A first filter circuit that imparts a predetermined filter characteristic to output to the acoustic signal propagation system, a delay characteristic imparting unit that imparts delay transfer characteristic of the acoustic signal propagation system, and a detection signal detected by the acoustic signal detecting means are input. An adaptive filter, a delay characteristic applying section output, a combined signal of the adaptive filter output and the detection signal are input, and an adaptive signal processing section for performing adaptive signal processing to determine a coefficient of the adaptive filter; And a means for setting the first filter.

[0017]

The first filter circuit imparts a predetermined filter characteristic to the acoustic signal xn and inputs it to the acoustic signal propagation system (controlled system) and the delay characteristic imparting section. The speaker of the acoustic signal propagation system outputs the acoustic signal to the acoustic space, and the microphone provided near the listener detects the acoustic signal propagated in the acoustic space. The delay characteristic imparting unit imparts the delay transfer characteristic z ⁻ⁿ for satisfying the causality to the acoustic signal input from the first filter circuit. The adaptive signal processing unit performs adaptive signal processing using the combined signal and the microphone detection signal so that the root mean square of the combined signal of the delay characteristic adding unit output and the adaptive filter output is minimized, and the adaptive filter characteristic W is acoustic. The coefficient is updated so that it is an inverse function (= z ⁻ⁿ / C) of the transfer characteristic of the signal propagation system. The coefficient setting means may be, for example, a music break (between songs).
Alternatively, according to the user's instruction
Set in the filter circuit of.

The acoustic signal (acoustic signal heard by the listener) xn 'detected by the microphone is X' = ^W'.C.X = (z- ^{n, where W'is} the characteristic of the first filter circuit. / C) · C · X = z ⁻ⁿ · X. Therefore, if a desired characteristic imparting section for imparting the desired characteristic H is arranged in the preceding stage of the first filter circuit, the acoustic signal indicated by X '= H · z ⁻ⁿ · X can be heard by the listener. That is, according to the present invention, without obtaining the transfer function of the acoustic space,
A desired acoustic characteristic can be provided without frequently changing the characteristic of the adaptive filter.

[0019]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS (a) Overall Configuration FIG. 1 is a configuration diagram of an acoustic control device according to the present invention, and FIG. 2 is a general configuration of the acoustic control device of the present invention. 11 in the figure
Is an acoustic signal propagation system, 11a is a speaker, 11b is an acoustic space such as a passenger compartment, and 11c is a microphone for detecting an acoustic signal output from the speaker to the acoustic space. 11d is a power amplifier that amplifies an audio signal (acoustic signal).

Reference numeral 12 denotes a desired characteristic imparting section for imparting a desired acoustic characteristic (transfer characteristic) H to an audio signal (acoustic signal) xn output from an audio device, and 13 an acoustic signal obtained by appropriately copying the characteristic of an adaptive filter described later. A filter circuit for imparting a predetermined filter characteristic W'to a signal, 14 a delay characteristic imparting section for imparting a delay transfer characteristic z ^-n for satisfying a causal law to the output signal xn "of the filter circuit, and 15 an adaptive signal processing section. Note that the delay of Z ⁻ⁿ is (delay transfer characteristic of acoustic signal propagation system) + (delay of half the tap length of the inverse filter), and 1/2 of the length of the inverse filter is preferable. Although not theoretically proven, it is empirically known.

In the adaptive signal processing unit 15, 15a is an adaptive filter to which the acoustic signal xn 'detected by the microphone 11c is input, 15b is adaptive signal processing based on the LMS algorithm and the coefficient of the adaptive filter 15a,
That is, the LMS signal processing unit determines the characteristic W. The adaptive filter 15a has the configuration shown in FIG. LM
The S signal processing unit 15b receives the error signal en and the acoustic signal xn 'which are the combined signal of the output signal dn of the delay characteristic adding unit 14 and the output signal yn of the adaptive filter 15a, and uses these signals to output the error signal Adaptive signal processing based on the LMS algorithm is performed so that the root mean square of en is minimized, and the coefficient of the adaptive filter 15a (characteristic W of the adaptive filter) is determined. That is, the adaptive signal processing unit 15b
Let w be the coefficient of the adaptive filter 15a at the current time n · Ts.
_{If 1} (n), w ₂ (n), w ₃ (n) ... w _N (n), the coefficient w ₁ (of the adaptive filter at time (n + 1) · Ts after one sampling time Ts (n + 1) n + 1), w ₂ (n + 1), w ₃ (n + 1) ... w _N (n + 1)
Is determined by the following equation (coefficient updating equation).

[0022]

[Equation 6] However, en is an error signal, xn 'is a microphone detection signal, and the j-th filter coefficient update formula is _wj (n + 1) = _wj (n) + μx (n-j + 1)' ・ en It is given in (7).

When the LMS adaptive signal processing is continued using the equation (6), the root mean square error (E [en ² ]) finally becomes the minimum, and the characteristic W of the adaptive filter 15a is W = z ⁻ⁿ / C ( 9) Reference numeral 16 is a combining unit (adding unit) that outputs a combined signal en of the output signal dn of the delay characteristic adding unit 14 and the output signal yn of the adaptive filter 15a, and 17 is a filter coefficient copying unit that adapts based on the copy enable signal CES. What sets the coefficient of the filter 15a in the filter circuit 13, 1
A copy enable signal generator 8 generates a copy enable signal CES in response to, for example, a music break (between songs) or a user instruction (key operation). The desired characteristic imparting unit 12, the filter circuit 13, the delay characteristic imparting unit 14, the adaptive signal processing unit 15, the synthesizing unit 16, and the filter coefficient copying unit 17 constitute an acoustic characteristic control unit, for example, one DSP.
(Desital Signal Processor).

The overall desired operation characteristic imparting section 12 imparts a desired desired acoustic characteristic (transfer characteristic) H to an audio signal (acoustic signal) xn output from an audio device (not shown). Then, the filter circuit 13 applies a predetermined filter characteristic W ′ to the output of the desired characteristic imparting section and inputs it to the acoustic signal propagation system 11 and the delay characteristic imparting section 14. The initial characteristic of the characteristic W'of the filter circuit 13 is 1. The speaker 11a of the acoustic signal propagation system 11 outputs the acoustic signal xn ″ output from the filter circuit 13 to the acoustic space 11b, and the microphone 11c provided near the listener detects the acoustic signal propagating in the acoustic space. The delay characteristic imparting unit 14 imparts the delay transfer characteristic z 1 of the acoustic signal propagation system to the acoustic signal xn ″ input from the filter circuit 13.

The acoustic signal (microphone detection signal) xn 'detected by the microphone 11c is supplied to the adaptive filter 15a and LM.
It is input to the S signal processing unit 15b. Adaptive filter 15a
Outputs the acoustic signal (microphone detection signal) xn 'with the characteristic W (the initial characteristic is 1) determined by the adaptive signal processing, and the combining unit 16 outputs the output signal dn of the delay characteristic adding unit 14.
And a composite signal en of the output signal yn of the adaptive filter 15a is output. LMS signal processor 15b of adaptive signal processor 15
Uses the combined signal en and the microphone detection signal xn 'so as to minimize the root mean square of the combined signal en, and performs the calculation of the equation (6) at every predetermined sampling time Ts to adaptive filter 1
Determine the coefficient of 5a.

When the adaptive signal processing by the equation (6) is continued,
Finally, the root mean square error (E [en ² ]) becomes the minimum, and the characteristic W of the adaptive filter 15a becomes W = z ⁻ⁿ / C. When the characteristic W of the adaptive filter 15a becomes z ⁻ⁿ / C, the copy key (not shown) is operated to output the copy enable signal CES from the copy enable signal generator 18, and the filter coefficient copy unit 17
Are the coefficients w ₁ (n), w ₂ (n), w ₃ (n) of the adaptive filter 15a.
.. Set w _N (n) in the filter circuit 13. Consequently, thereafter, the acoustic signal xn 'at the listener's ear, which is the observation point (microphone installation position), becomes xn' = H * W '* C' * z- ⁿ * xn (10). In the case of W '= 1 / C', xn '= ^H.z -n.xn (11), and the desired acoustic characteristic H and delay characteristic are added to the acoustic signal xn. In addition, W '= 1 /
If it does not become C ', the coefficient of the adaptive filter 15a is copied to the filter circuit 13 at the appropriate time to finally obtain W'.
= 1 / C ', and the acoustic signal xn' at the listener's ear is as shown in equation (11).

Further, when the acoustic signal propagation system 11 is a vehicle interior space of an automobile, the transfer characteristic C'of the acoustic signal propagation system changes depending on the number of occupants, the seating position and the like. In such a case, the listener presumes that the transfer characteristic C ′ has changed, and operates the copy key to set the coefficient of the adaptive filter 15a to the filter circuit 1.
Copy to 3. In this way, the acoustic signal propagation system 1
Even if the characteristic of No. 1 changes, the listener can hear the acoustic signal to which the desired characteristic is added.

Modifications In the above description, the case where the copy enable signal is generated by operating the copy key has been described. However, the music enable of the audio signal (acoustic signal) is detected, and the copy enable signal CES is generated every time the music interval is detected. Then, the coefficient of the adaptive filter 15a may be copied to the filter circuit 13. By doing this, the filter circuit 13 is automatically
The characteristic W'of 1 can be set to 1 / C ', and since the characteristic of the filter circuit 13 is fixed during one song, no unnaturalness is given to the listener. Further, in the above, the desired acoustic characteristic imparting unit is provided in the front stage of the first filter circuit, but it may be provided in the rear stage. Further, the desired acoustic characteristic H can be set to 1 and the desired acoustic characteristic imparting section can be omitted. In this case, it is possible to enjoy a sound that is not affected by the propagation characteristics of the acoustic space. Further, although the case where there is one speaker and one microphone has been described above, the present invention can be applied to the case where there are two or more speakers. Although the present invention has been described above with reference to the embodiments, the present invention can be variously modified according to the gist of the present invention described in the claims, and the present invention does not exclude these.

[0029]

As described above, according to the present invention, since the characteristic W of the adaptive filter can be set to Z ^-n / C while outputting music, it is not necessary to measure the transfer characteristic C of the acoustic signal propagation system in advance.
Further, according to the present invention, since the coefficient of the adaptive filter is determined based on the adaptive signal processing by the LMS algorithm,
The calculation for convolving the transfer function C of the acoustic signal propagation system with the acoustic signal is not necessary, and the number of calculations can be reduced. Further, according to the present invention, the characteristics of the filter circuit change only when the copy key is operated or only between songs, and the filter characteristics are fixed other than that, so that the sound heard by the listener changes. Without giving the listener an unnatural feeling.

[Brief description of drawings]

FIG. 1 is a configuration diagram of an acoustic characteristic control device of the present invention.

FIG. 2 is a configuration diagram of a general acoustic characteristic control device of the present invention.

FIG. 3 is an explanatory diagram of a conventional acoustic characteristic control method.

FIG. 4 is an explanatory diagram of a controlled system.

FIG. 5 is a configuration diagram of a conventional control filter using adaptive signal processing.

FIG. 6 is a configuration diagram of an adaptive filter.

FIG. 7 is a configuration diagram of a filter for generating a filtered X signal.

[Explanation of symbols]

 11 ... Acoustic signal propagation system 12 ... Desired characteristic imparting unit 13 ... Filter circuit 14 ... Delay characteristic imparting unit 15 ... Adaptive signal processing unit 15a ... Adaptive filter 17 ... Filter coefficient copying unit

Claims (2)

[Claims] 1. An acoustic signal propagation system comprising a speaker, an acoustic space, and a detection means for detecting an acoustic signal output from the speaker to the acoustic space, wherein a predetermined filter characteristic is added to the acoustic signal. A first filter circuit that outputs to the acoustic signal propagation system, a delay characteristic imparting unit that imparts a delay transfer characteristic of the acoustic signal propagation system to the output signal of the first filter circuit, and the acoustic signal detected by the detection unit is input. An adaptive filter, an input signal of the output of the delay characteristic adding unit and the output of the adaptive filter, and the detected acoustic signal are input, and an adaptive signal processing unit that performs adaptive signal processing to determine the coefficient of the adaptive filter, An acoustic characteristic control device having means for setting the first filter. 2. The acoustic characteristic control device according to claim 1, wherein a desired acoustic characteristic imparting section is provided in a stage before or after the first filter.