JPH1097263A - Sound field control method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To eliminate a standing wave in a trunk room and to output a sound with excellent quality into a car room. SOLUTION: A signal processing part 11 obtains a frequency band that the mode and the spatial sound pressure distribution of the largest standing wave generated in the trunk room TRR become clear when the sound is outputted from an audio sound reproducing speaker SP. Then, a speaker SPC1, a microphone MC1 /the speaker SPC2, the microphone MCN-1 are respectively arranged in the trunk room in which the spatial distribution coefficient of this mode becomes positive/negative. An adaptive signal processor 13 performs adaptive signal processing so that output signals of respective microphones become the same waveform to decide the coefficients of adaptive filters 13a-1 , 13a-2 . An audio signal is inputted to the reproducing speaker SP, and is inputted to a band-pass filter 12 becoming a pass-band equal to the frequency band, and the band-pass filter 12 inputs its output signal to the adaptive filters 13a-1 , 13a-2 , and the adaptive filters 13a-1 , 13a-2 input filter outputs to respective controlling speakers SPC1, SPC2.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a sound field control method for a car audio system in which a speaker is installed on a rear tray and a trunk room is used as a speaker box.
In particular, the largest standing wave generated in the trunk room is eliminated, and the sound pressure distribution in the entire trunk room is flattened by superimposition of other modes, thereby removing the adverse effect on the speaker for audio sound reproduction, Field of the Invention The present invention relates to a sound field control method for improving the sound quality of sound output to a sound field.

[0002]

2. Description of the Related Art The sound quality of a loudspeaker is improved by increasing its rear volume. For this reason, in the case of car audio, as shown in FIG.
Is often used to use the trunk room TRR as a speaker box.

[0003]

However, since sound is reflected on the wall surface of the trunk room TRR, a standing wave is generated in the trunk room, and the transfer characteristics of the speaker are disturbed. That is, the standing wave causes resonance in the trunk room,
A peak occurs at the resonance frequency, and this peak causes a peak / dip portion in the transfer characteristic of the speaker, which adversely affects the sound reproduced in the vehicle interior. For home audio speakers, sound absorbing material such as glass wool is attached to the inside wall of the speaker box to reduce the reflection of sound to prevent the above effects, but for car audio devices, sound absorbing materials are attached because the trunk room becomes dirty. No measures have been taken.
In view of the above, an object of the present invention is to dispose a small control speaker for standing wave cancellation in a trunk room to eliminate standing waves in the trunk room and to achieve sound with good sound quality by approaching the original transfer characteristics of the speaker. That is, it is possible to output the signal to the passenger compartment.

[0004]

According to the present invention, there is provided a car audio system in which a speaker for audio sound reproduction is installed on a rear tray and a trunk room is used as a speaker box. When the output, the mode of the largest standing wave generated in the trunk room is obtained, and the frequency band in which the spatial sound pressure distribution of the obtained mode is remarkable is obtained, and in the region of the trunk room where the spatial distribution coefficient of the mode is positive. A first control speaker and a microphone are arranged, a second control speaker microphone is arranged in a negative trunk room area, and adaptive signal processing is performed so that output signals of the microphones have the same waveform in the frequency band. To determine the coefficients of the adaptive filter and pass the frequency band to the pass band This is achieved by inputting an audio signal to an adaptive filter through a band-pass filter, inputting an audio signal output from the adaptive filter to each control speaker, and inputting the audio signal to an audio sound reproduction speaker. You.

[0005]

BEST MODE FOR CARRYING OUT THE INVENTION

(A) Principle (a) Basic principle The solution of the wave equation of a one-dimensional sound field having M sound sources (speakers) inside and closed at both ends is given by the following equation. Note that the one-dimensional sound field is a sound field in which the sound pressure changes only according to a predetermined axial position x. An example of a one-dimensional sound field is an elongated pipe with a circular cross section, the diameter of which is 1/1 of the signal frequency wavelength.
8 or less pipes.

[0006]

(Equation 1) Where ω _n = nπc ₀ / L ω: angular frequency of the emitted signal p (x, ω): sound pressure, ξ _n : damping ratio in the n-th mode q _m : input signal to speaker m x _m : position of speaker m N: number of modes L: length of one-dimensional sound field.

The above equation shows that the sound pressure p (x,
ω) is given by the sum of the modes unique to the sound field represented by cos (nπx / L). cos (nπ
x / L) is a coefficient indicating the spatial distribution of the mode (referred to as a spatial distribution coefficient). In equation (1),

(Equation 2) Is a coefficient indicating the frequency localization of the n-th mode, for example, as shown in FIG. As can be seen from FIG. 1, the ratio of the superimposed modes changes with the frequency. Therefore, at the frequency where the first mode is higher than the other modes, the coefficients (n = 1, 2,.
Is as shown in FIG.

[0008]

(Equation 3) In FIG. 2, a portion above the dotted line (= 0) is plus, a portion below the dotted line is minus, and a minus portion is inverted in phase with a plus portion.

[0009]

(Equation 4) Is the component of the n-th mode of the sound emitted from the speaker into the space, multiplied by the coefficient of the same n-th mode expressed by equation (3), and added to the other modes to express the sound pressure in the space It can be seen from equation (1).

When one of the modes becomes significantly larger than the other mode in a certain frequency band, it appears as a standing wave, and the standing wave causes resonance in the trunk room, and the resonance frequency causes A peak occurs, and this peak causes a peak / dip portion in the transfer characteristic of the speaker, which adversely affects the acoustic characteristic in the vehicle interior. therefore,
The present invention counteracts this prominent mode so as to obtain a more uniform sound field by superposition of other modes, thereby eliminating the adverse effects on the transfer characteristics of the loudspeaker.

Here, consider the case where the same signal q (ω) is input to both sound sources when the number of sound sources (speakers) is M = 2. At this time, equation (1) is

(Equation 5) And can be transformed. From this equation, cos (nπx ₁ / L) = − c
It can be seen that the n-th mode becomes 0 by arranging the sound source at a position satisfying os (nπx ₂ / L). At this time, as in the n-th mode, the mode in which the sign of cos is inverted is canceled, and conversely, the mode in which the sign of the value of cos is the same becomes larger.

Signals q are applied to both sound sources via filters W ₁ and W _2.
When you enter (ω),

(Equation 6) Becomes By adjusting the filters W ₁ and W ₂ , cos (nπx ₁ / L) W ₁ (ω) = − cos (nπx ₂ / L)
By setting W ₂ (ω), the mode can be canceled without being affected by the sound source position.

In the mode control of the speaker inside the enclosure, it is most preferable that after the sound is output from the speaker, the sound pressure in the entire inside of the enclosure is uniformly changed so that the sound pressure difference depending on the position is eliminated. That is,
It is most desirable that only the zero-order mode remains and the other modes disappear. This requires solving the following simultaneous equations:

[0014]

(Equation 7) However, since there are an infinite number of expressions and only _two variables W ₁ and W _2, there is no complete solution. To get a complete solution,
You need an infinite number of sound sources with filters. Here, attention is paid to the frequency localization of each mode. As described above, since each mode has frequency localization according to equation (2), sufficient control is possible even with two sound sources by focusing on the frequency band F where one mode is extremely large. Becomes That is, even in the case of the two sound sources, the components in the frequency band F of the mode can be removed to make the overall sound pressure in the trunk room substantially uniform.

In this case, as shown in FIG. 3, the spatial distribution coefficient cos (nπx / L) of the mode to be the cancellation target
Sound pressure control speakers SP at positions x ₁ and x ₂ having different signs
When C1 and SPC2 are arranged and the same signal q (ω) is applied to these speakers, the sound cancels out in that mode. That is, the sound pressure control speaker SP is set so that the same desired sound pressure is simultaneously obtained on the plus side and the minus side of the mode.
By driving C1 and SPC2, the spatial distribution coefficient cos (nπ
In the mode in which the sign of (x / L) is inverted, the modes cancel each other, and eventually a mode in which the sign is not inverted (for example, the mode of n = 0 in FIG. 2) remains, and a desired sound pressure can be obtained in a state where the mode is excited. For example, the microphones MC1 and MC1 are located at arbitrary positions x ₃ and x ₄ having different signs of the spatial distribution coefficient cos (nπx / L).
MC2 is provided, and the microphones MC1 and MC2 provide a sound source S.
When the sounds from PC1 and SPC2 are detected, ideally, the sounds in the mode in which the sign of the spatial distribution coefficient cos (nπx / L) is inverted cancel each other out, and only the sound in the mode in which the sign is not inverted is obtained. Become.

As described above, each control speaker SPC1, SPC
When the same signal q (ω) is input to PC2, each microphone MC
1, MC2 W ₁ as the sound pressure detected is q (omega) in
(Ω) and W ₂ (ω) may be controlled. That is,

(Equation 8) What is necessary is just to find the least square solution of the filter that satisfies.

[0017] If the input to this filter W1 _1, W ₂ each control speakers via SPC1, SPC2 to the signal q (omega), inverted mode code as described above is canceled,
In particular, a large mode is canceled preferentially, and a mode having the same code becomes larger. In a frequency band in which a certain mode is very large, the other modes are almost the same size. Therefore, the modes are made uniform to each other, and the sound pressure distribution of the entire sound field becomes almost flat.

[0018] It should be noted that, since the actual audio sound reproducing speaker SP that does not contain a filter in the trunk room is present, if the position and x _5, the following equation

(Equation 9) , A least-squares solution of filters W ₁ and W ₂ that satisfies To find this filter, an adaptive filter or the like may be used.

(B) Mode Analysis By the way, in order to apply the above control, it is necessary to check in advance the modes that have an adverse effect on the audio sound reproduction speaker and the frequency band in which the modes are localized. Multiplying both sides of equation (1) by cos (n'πx / L) (n 'is an integer), integrating with respect to x, and replacing n' with n,

(Equation 10) Becomes Thus, only the n-th mode component in the spatial sound pressure distribution can be extracted.

Further, when the left side of the equation (10) is discretized,

[Equation 11] Becomes That is, mode analysis can be performed by installing a plurality of microphones at positions in (K + 1) trunk rooms of x _{0 to} x _K and performing discrete Fourier transform on their outputs with respect to positions. With this, it is possible to check which mode has an adverse effect on the audio sound reproduction speaker SP, and the sound pressure control speakers SPC1 and SP
The position of C2 and the position of the control point (the positions of the microphones MC1 and MC2) can be determined.

The following equation is obtained from the equation (11), and the frequency band in which the mode of interest is located is determined.

(Equation 12) Therefore, a band-pass filter that passes only this band is designed, an audio signal is input to the band-pass filter, and an output signal of the band-pass filter is provided to filters W ₁ and W ₂ provided in front of the control speakers SPC 1 and SPC _2. To control only the target frequency band.

The above description has been made with reference to a one-dimensional sound field for the sake of simplicity.

(Equation 13) There is no essential difference.

In summary, the mode to be controlled is determined by analyzing the mode in the trunk room, and then a frequency band in which the mode is remarkable is determined, and a band-pass filter that passes the frequency band is designed. Then, the spatial distribution coefficient cos (nπx /
The speaker SPC1 and the microphone MC1 for sound pressure control are arranged in the trunk room area on the plus side of L), and the speaker SPC2 and microphone MC2 for sound pressure control are arranged in the trunk room on the minus side. In this state, each microphone M
The coefficients of the adaptive filters (filters W ₁ and W ₂ ) are adjusted so that the difference between the signal obtained from C1 and MC2 and the signal obtained by adding a time delay to the input signal to the adaptive filter becomes small. Then, the output signal of the band-pass filter is input to the adaptive filter, and the output signal of the adaptive filter is input to each of the sound pressure control speakers SPC1 and SPC2. Thereafter, if this process is repeated, the signal obtained from each of the microphones MC1 and MC2 approaches the same waveform as a signal obtained by adding a time delay to the input signal to the adaptive filter.

(B) Embodiment (a) Configuration FIG. 4 is a configuration diagram of a sound field control system according to the present invention. In the figure, TRR is a trunk room, CRM is a vehicle compartment, and SP is a speaker for reproducing audio sound, which is also used for performing mode analysis of a sound field. SPC1 and SPC2 are first and second sound pressure control speakers arranged at positions on the plus side and the minus side, respectively, of the spatial distribution coefficient of the mode (the mode that adversely affects the sound field) obtained by the mode analysis. , MC _{1 to} MC _N are N microphones provided at appropriate positions in the trunk room in the x-axis direction. Note that the microphones MC ₁ and MC _N-1 are arranged at positions corresponding to the plus side and the minus side of the spatial distribution coefficient, respectively.

Reference numeral 11 denotes a signal processing unit, which performs a mode analysis process in the trunk room, a process of obtaining a frequency band in which a mode adversely affecting the sound field in the trunk room becomes remarkable, and the like. Reference numeral 12 denotes a band-pass filter having a pass band in a frequency band in which a mode that adversely affects the sound field becomes noticeable. Reference numeral 13 denotes adaptation when the number of signals = 1, the number of speakers = 2, and the number of observation points (the number of microphones) = 2. It is a signal processing device. Reference numeral 14 denotes a first delay unit for accurately approximating the inverse characteristic of the acoustic system, which has a delay time Δ2 which is about half the tap length of the adaptive filter, and 15 denotes a _second delay unit. And a delay unit for adjusting the delay time of the audio signal through system to the delay time of the adaptive signal processing system. 16 the arithmetic unit for outputting a difference between the first output signal and the first output signal of the microphone MC ₁ delay portion 14 as an error signal e _1, the output signal and the second microphone of the first delay section 14 17 the difference between the MC _N-1 of the output signal is an arithmetic unit for outputting as an error signal e _2.

In the adaptive signal processor 13, 13a-1, 13a
Reference numeral a-2 denotes first and second adaptive filters (W ₁ , W ₂ ) constituted by FIR digital filters, and 13b denotes a signal processing filter (filter for producing a filtered X signal) for controlling each sound pressure. First and second speakers SPC1 and SPC2
Characteristics C ₁₁ , C up to microphones MC ₁ , MC _N-1
21 , C ₁₂ , and C ₂₂ are convolved with the band-pass filter output signal, and 13c-1 to 13c-2 perform adaptive signal processing based on the Filtered-X LMS algorithm to perform adaptive filter 13a-1,13
The first and second adaptive signal processing units determine the coefficient of a-2.

(B) The signal q (ω) is input to a speaker (audio sound reproducing speaker) SP for analyzing the operation mode, and the sound output from the speaker is output to each of the microphones MC ₁ to MC ₁ .
The signal is detected by the _MCN , and the detection signal is input to the signal processing unit 11. The signal processing unit obtains the components of each mode based on equation (11), and determines the mode that has the most adverse effect on the sound field.
Next, a frequency band F in which the mode is remarkable is obtained by the equation (12), and the bandpass filter 12 is designed so that the frequency band becomes a pass band. Thereafter, the first sound pressure control speaker SPC1 and the microphone MC ₁ is disposed on the positive side of the area of the spatial distribution coefficients of the target mode, the second sound pressure control speaker SPC2 and the microphone MC _N to the negative side of region Place _-1 . Next, the propagation elements C ₁₁ , from the first and second sound pressure control speakers SPC1 and SPC2 to the first and second microphones MC ₁ and MC _N-1
C ₂₁ , C ₁₂ and C ₂₂ are measured and set in the signal processing filter 13 b of the adaptive signal processing device 13.

In this state, the audio signal is supplied to the terminal Ta.
Further, the audio signal is input to the audio sound reproduction speaker SP via the second delay unit 15, and the audio signal component of the predetermined frequency band F is input to the adaptive signal processing device 13 via the band pass filter 12. The adaptive signal processing device 13
Adaptive signal processing is performed so that the power of the error signals e ₁ and e ₂ output from the arithmetic units 16 and 17 is minimized (so that each microphone output signal and the band pass filter output are the same). Second adaptive filters 13a-1, 13a-2
To determine the coefficients W _1, W ₂ of.

The first and second adaptive filters 13a-1 and 13a-2 apply the coefficients W ₁ and W ₂ to the output signal of the band-pass filter 12.
, And outputs the adaptive filter output from the first and second sound pressure control speakers SPC1 and SPC
Enter 2 Thereafter, if the above processing is repeated, the signal obtained from each of the microphones MC ₁ and MC _N-1 approaches the same waveform as the output waveform of the band-pass filter 12, and a uniform frequency characteristic is obtained regardless of the position in the x direction. Become like That is, the frequency band F,
, The adverse effect can be removed over the entire space. After the adaptive filter coefficients W ₁ and W ₂ have converged, the coefficients W ₁ and W ₂ can be set in the FIR fill, and this FIR filter can be used in place of the adaptive signal processing device 13.

(C) Function of First Delay Unit A simple system shown in FIG. 5A (one sound source SPK,
(One microphone MIC and one filter FIL)
Now, consider a case where the first delay unit 14 has no delay.
If the transfer characteristic of the adaptive filter W is W (z) = 1 / C (z) (14), W (z) · C (z) = (1 / C (z)) · C (z) = 1 (15) The goal is achieved.

However, here, it is a big problem whether the characteristic of W (z) = 1 / C (z) is possible. For example, consider the model shown in FIG. 5B as the indoor characteristic C (z). In this model, the following equation holds: y (k) = u (k) -cu (k-1) (16) When both sides of this equation are z-transformed, Y (z) = U (z) -cz ^-1 U (z) = (1-cz ^-1 ) U (z) (17) Accordingly, the indoor transfer characteristic is as follows: C (z) = Y (z) / U (z) = 1−cz ⁻¹ (18)

Therefore, the inverse characteristic of C (z) is as follows: W (z) = 1 / C (z) = 1 / (1-cz ⁻¹ ) (19) The input / output relationship of a system having this transfer characteristic is as follows: U (z) = W (z) .X (z) = X (z) / (1-cz ^-1 ) (20) By transforming this equation, U (z) = X (z) + cz ^-1 U (z) (21), and applying the inverse z transform, u (k) = x (k) + cu (k-1) ( 22) FIG. 5C is a block diagram showing the equation (22). Here, the right side of Expression (19) is as follows: W (z) = 1 + cz ⁻¹ + c ² z ⁻² + c ³ z ⁻³ (23) This impulse response is as shown in FIG. 6A, and naturally coincides with the impulse response in FIG. 5C. Here, the condition is different between | c | <1 and | c | ≧ 1.

If | c | <1, as shown in FIG.
Since the impulse response converges over time,
This filter is stable. However, if | c | ≧ 1, the impulse response diverges over time as shown in FIG. 6B, resulting in an unstable filter. In the case of | c | ≧ 1, the expression (19) is rewritten as W (z) = 1 / (1-cz ⁻¹ ) = 1 / (− cz ⁻¹ +1) (24) Performing division by denominator yields W
(z) by the following equation ^{W (z) = - c -1} z 1 -c -2 z 2 -c -3 z 3 - a.. (25). This impulse response is as shown in FIG.

By the above, the filter becomes stable, but becomes an acausal filter.
The non-causality is a filter that knows what input is input at the time before the input to the filter and outputs it in advance, and does not actually exist. However, if a delay is allowed in the output time, FIG.
By adding a time delay to the impulse response and shifting it to a positive time, as shown in (a), some approximation of this filter is possible.

In an actual acoustic system, the above-mentioned components are synthesized in a more complicated manner, and the inverse characteristic is as shown in FIG. 7 (b). In FIG.
If the filter is shifted (delayed) to the center of a filter having a finite tap length that can be used as shown in (c), the inverse characteristic can be accurately approximated. In other words, the first delay unit 14 in FIG. 4 has a time delay of about half the tap length of the adaptive filter used to accurately approximate the inverse characteristic of the acoustic system. As described above, the present invention has been described with reference to the embodiments. However, the present invention can be variously modified in accordance with the gist of the present invention described in the claims, and the present invention does not exclude these.

[0036]

As described above, according to the present invention, the mode of the largest standing wave generated in the trunk room when sound is output from the audio sound reproducing speaker is determined, and the spatial sound pressure in the determined standing wave is determined. A frequency band where the distribution is remarkable is determined, a first control speaker and a microphone are arranged in a trunk room area where the spatial distribution coefficient of the mode is positive, and a second control speaker and microphone are arranged in a trunk room area where the mode is negative.
The control speaker microphones are arranged, adaptive signal processing is performed so that the output signals of the microphones have the same waveform, the coefficients of the adaptive filters are determined, and the adaptive filter processing is performed via a band-pass filter having the frequency band as a pass band. The audio signal is input to the filter, the audio signal output from the adaptive filter is input to each control speaker, and the audio signal is input to the audio sound reproduction speaker. A standing wave that has an adverse effect on the transfer characteristics is eliminated, and a sound with good sound quality can be output into the vehicle cabin by approaching the original transfer characteristics of the speaker.

[Brief description of the drawings]

FIG. 1 is a frequency characteristic diagram of each mode.

FIG. 2 is a diagram showing a mode of spatial distribution of a mode at a certain frequency.

FIG. 3 is a diagram illustrating the principle of the present invention.

FIG. 4 is a configuration diagram of a sound field control system according to the present invention.

FIG. 5 is an explanatory diagram of a function of a first delay unit in FIG. 4;

FIG. 6 is an explanatory diagram of an impulse response for explaining a function of a first delay unit.

FIG. 7 is an explanatory diagram of an impulse response delay for explaining a function of a first delay unit.

FIG. 8 is an explanatory diagram of a speaker arrangement in the car audio system.

[Explanation of symbols]

TRR .. trunk room CRM .. casing SP · audio sound reproducing speaker SPC1, SPC2 ... sound first pressure control, the second speaker MC ₁ to MC _N-1 ... microphone 11 ... signal processing unit 12. Bandpass filter 13. Adaptive signal processing device 13a-1, 13a-2. Adaptive filter 14, 15, First and second delay unit 16, 17, Synthesis unit.

Claims (1)

[Claims] 1. A car audio system in which a speaker for audio sound reproduction is installed on a rear tray and a trunk room is used as a speaker box, wherein a sound is output from the speaker for audio sound reproduction.
The mode of the largest standing wave generated in the trunk room is determined, and the frequency band in which the spatial sound pressure distribution of the determined mode is remarkable is determined. The first control is performed in the trunk room region where the spatial distribution coefficient of the mode is positive. Speaker and microphone, a second control speaker microphone is arranged in a negative trunk room area, and adaptive signal processing is performed so that output signals of the microphones have the same waveform in the frequency band. An audio signal is input to an adaptive filter via a band-pass filter having a pass band in the frequency band, an audio signal output from the adaptive filter is input to each control speaker, and an audio signal is output. Field control method comprising inputting sound to a speaker for audio sound reproduction