JPH1070787A - Device and method for signal correction, and device and method for adjusting coefficient of the same signal correcting device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide the signal correcting device, etc., which can easily, speedily, and surely correct characteristics of a transmission system having nonlinearity such as a speaker. SOLUTION: The signal correcting device 12 corrects and supplies an input signal (u) to the transmission system 14 so that characteristics of an integrated transmission system 10 approximate specific target characteristics. A displacement filter 16 predict virtual displacement (x) based on the input signal by using a filter coefficient determined according to the target characteristics. A correction coefficient determination part 18 determines part of the filter coefficient of a correction filter 22 according to the virtual displacement (x). A correction coefficient determination part 18 determines even the amplitude adjustment coefficient of an amplitude adjustment part 24 according to the virtual displacement (x). The correcting filter 22 is composed of a digital filter of >=2th order and corrects the input signal (u) by using the determined filter coefficient. The amplitude adjustment part 24 adjusts the amplitude of the signal by using the determined amplitude adjustment coefficient. The correcting filter 22 and amplitude adjustment part 24 correspond to a signal correcting means 20.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to a signal correction device,
The present invention relates to a signal correction method, a coefficient adjustment device of a signal correction device, and a coefficient adjustment method, and more particularly to a transmission system that corrects an input signal so that characteristics of an integrated transmission system in which a signal correction device and a transmission system are integrated approach predetermined target characteristics. About technology to give.

[0002]

2. Description of the Related Art An electrodynamic loudspeaker 2 is used for reproducing a musical sound. FIG. 23 shows a cross-sectional configuration of the electrodynamic speaker 2. When a drive current based on a tone signal is applied to the voice coil 4, the voice coil 4 receives from the magnet 6 an electromagnetic force proportional to the drive current. For this reason, the speaker diaphragm 8 connected to the voice coil 4 according to the musical tone signal
Vibrates in the x direction, and a musical tone is reproduced. In this case, the reproduced sound pressure depends on the acceleration of the voice coil 4 (speaker diaphragm 8) and the volume of air removed as the voice coil 4 (speaker diaphragm 8) moves. Therefore, for the same speaker, the reproduced sound pressure substantially depends on the acceleration and amplitude of the voice coil 4.
That is, the higher the acceleration and amplitude of the voice coil 4, the higher the reproduction sound pressure, and the lower the acceleration and amplitude of the voice coil 4, the lower the reproduction sound pressure.

On the other hand, as shown in FIG. 24, as the frequency of the reproduced musical tone becomes lower than the lowest resonance frequency f0 of the electrodynamic speaker 2, the acceleration of the voice coil 4 tends to decrease. That is, the electrodynamic loudspeaker 2 has a drawback that if the level of the musical sound signal is the same, the reproduced sound pressure in the low-frequency range decreases.

Therefore, in order to overcome such a drawback of the electrodynamic loudspeaker 2, in a conventional loudspeaker system, for example, for a frequency lower than the lowest resonance frequency f0, processing (boost) for expanding the level of the tone signal in advance is performed.
Was performed. With this configuration, it is possible to increase the amplitude of the voice coil in the low frequency range, and it is possible to alleviate a decrease in the reproduction sound pressure in the low frequency range.

[0005]

However, such a conventional speaker system has the following problems.
When the amplitude of the voice coil 4 becomes large, the movement of the voice coil 4 is restricted due to mechanical restrictions of the electrodynamic speaker 2 and the like, and as a result, distortion occurs in the reproduced musical sound. That is,
Higher harmonic components not present in the input tone signal are superimposed and appear in the reproduced tone. This is nonlinear distortion. As shown in FIG. 25, especially in the frequency range slightly lower than the lowest resonance frequency f0, the second harmonic and the third harmonic appear largely.

Further, due to the non-linearity of the excluded air volume of the edge 8e constituting the speaker diaphragm 8, even if the movement of the voice coil 4 is undistorted, non-linear distortion appears in the radiation sound pressure.

A method of correcting such non-linearity of the loudspeaker by a feedforward method to reduce distortion is disclosed in German Offenlegungsschrift DE 4111884 A1, US Pat. No. 5,386,625. This method is based on the principle of electro-dynamic conversion of a converter such as a speaker, and each seal small parameter (described later)
The method of synthesizing the nonlinear distortion is adopted. Therefore, the physical understanding of the correction principle is relatively easy. However, the actual correction algorithm is very complicated, and the adjustment method is also complicated. Furthermore, in order to use the seal small parameter itself for the correction calculation, very large and small coefficients must be mixed, which is inconvenient for high-speed processing using a DSP (Digital Signal Processor) that performs fixed-point arithmetic. It is. For this reason, the efficiency of the arithmetic processing is inferior, and the accuracy of the arithmetic operation may be reduced.

The present invention solves such a conventional problem and provides a signal correction apparatus, a signal correction method, and a signal correction method capable of easily, quickly and surely correcting characteristics of a transmission system having nonlinearity such as a speaker. It is an object of the present invention to provide a coefficient adjustment device and a coefficient adjustment method for a device.

[0009]

According to a first aspect of the present invention, there is provided a signal correcting apparatus which corrects an input signal so that a characteristic of an integrated transmission system obtained by integrating the signal correcting apparatus and a transmission system approaches a predetermined target characteristic. A signal correction device provided with a correction filter constituted by a second-order or higher-order digital filter having a feedback path, and a predetermined virtual physical quantity representing a state of an integrated transmission system based on an input signal. Virtual physical quantity predicting means for predicting, a correction coefficient determining means for determining at least a part of the filter coefficient of the correction filter based on the predicted virtual physical quantity, the correction filter,
The input signal is corrected using the determined filter coefficient.

According to a second aspect of the present invention, in the signal correction apparatus of the first aspect, the signal correction unit further includes an amplitude adjustment unit for adjusting an amplitude of the signal, and the correction coefficient determination unit determines the prediction coefficient. The amplitude adjustment unit also determines an amplitude adjustment coefficient of the amplitude adjustment unit based on the virtual physical quantity obtained, and the amplitude adjustment unit uses the determined amplitude adjustment coefficient to determine whether the signal before the correction by the correction filter or the signal after the correction. The amplitude is adjusted.

According to a third aspect of the present invention, in the signal correction apparatus according to any one of the first and second aspects, the virtual physical quantity estimating means includes a virtual digital filter of second or higher order having a feedback path. A physical quantity filter is provided, and the virtual physical quantity filter predicts a virtual physical quantity based on an input signal using a filter coefficient determined based on a target characteristic of the integrated transmission system.

According to a fourth aspect of the present invention, there is provided the signal correction apparatus according to the third aspect, wherein the correction filter includes a circulating section including a feedback path and corresponding to an upstream portion of the signal processing, and a signal processing section including the feedback path and not including the feedback path. A non-circular portion corresponding to the downstream portion,
The virtual physical quantity filter includes a cyclic part including a feedback path and a non-cyclic part without a feedback path, and the filter coefficient of the cyclic part of the correction filter is the same as the filter coefficient of the cyclic part of the virtual physical quantity filter. The filter coefficient of the non-recursive portion of the filter is a filter coefficient determined based on the virtual physical quantity.

According to a fifth aspect of the present invention, in the signal correction apparatus according to the fourth aspect, a filter coefficient of a non-cyclic portion of the correction filter includes a linear filter coefficient that does not change depending on the virtual physical quantity, and a virtual physical quantity. , And can be expressed as a sum with a non-linear filter coefficient that changes depending on.

According to a sixth aspect of the present invention, in the signal correction apparatus according to the fifth aspect, the transmission system converts a correction output of the signal correction apparatus into an analog signal and amplifies the converted output; An electrodynamic loudspeaker driven based on the amplified output, the target characteristic of the integrated transmission system is a characteristic including a characteristic of the integrated transmission system represented by a seal small parameter to be targeted, and the virtual physical quantity is , The virtual displacement of the voice coil of the electrodynamic loudspeaker calculated assuming that the integrated transmission system has target characteristics, and the virtual physical quantity filter calculates the virtual displacement in consideration of the amplification gain of the drive system. A displacement filter to be predicted, wherein the correction filter and the displacement filter are:
It is a second-order IIR (infinite impulse response) type filter.

According to a seventh aspect of the present invention, in the signal correction apparatus according to the sixth aspect, the amplifier is a voltage output type amplifier, and the three nonlinear filter coefficients of the acyclic portion of the correction filter are c (x). : Non-delay nonlinear filter coefficient d (x): first-order delay nonlinear filter coefficient e (x): second-order delay nonlinear filter coefficient, satisfying the following relationship: c (x) + e (x) = d (x) It is characterized by.

In the signal correction device according to an eighth aspect of the present invention, the amplifier is a current output type amplifier, and the three nonlinear filter coefficients of the acyclic portion of the correction filter are c (x). : Non-delay non-linear filter coefficient d (x): first-order delay non-linear filter coefficient e (x): second-order delay non-linear filter coefficient: c (x) = e (x) = d (x) / 2 Is satisfied.

According to a ninth aspect of the present invention, in the signal correction apparatus according to the sixth aspect, the non-recurring portion of the correction filter includes a third filter.
When a second-order delay nonlinear filter coefficient among the two nonlinear filter coefficients is e (x), e (x) = 0.

According to a tenth aspect of the present invention, in the signal correcting apparatus of the sixth aspect, the characteristic of the integrated transmission system represented by the target seal small parameter is such that the integrated transmission system has a characteristic that the sealing of the transmission system at the time of a low level input signal is performed. It is characterized in that it is the characteristic itself represented by the small parameter.

In the signal correction device according to an eleventh aspect of the present invention, in the signal correction device according to the sixth aspect, the characteristic of the integrated transmission system represented by the seal small parameter to be targeted is a signal of the transmission system at the time of a low-level input signal. It is a characteristic obtained by adding a desired correction to the characteristic represented by the small parameter.

According to a twelfth aspect of the present invention, in the signal correction apparatus according to the sixth aspect, a nonlinearity of a transmission system caused by a force coefficient and a stiffness that varies depending on a displacement of a voice coil of the electrodynamic speaker is reduced. It is characterized by relaxation.

According to a thirteenth aspect of the present invention, there is provided the signal correction apparatus according to the sixth aspect, wherein the edge elimination volume of the electrodynamic loudspeaker changes depending on the displacement of the voice coil of the electrodynamic loudspeaker. It is characterized in that the nonlinearity of the transmission system is reduced.

According to a fourteenth aspect of the present invention, there is provided a signal correction method for correcting an input signal and providing the same to a transmission system such that characteristics of an integrated transmission system including the transmission system approach predetermined target characteristics. Preparing a correction filter composed of a second-order or higher digital filter, predicting a predetermined virtual physical quantity representing the state of the integrated transmission system based on an input signal, and performing the correction filter based on the predicted virtual physical quantity. , And the correction filter corrects the input signal using the determined filter coefficient.

According to a fifteenth aspect of the present invention, a coefficient adjusting apparatus for a signal correcting apparatus is a coefficient adjusting apparatus for adjusting each coefficient in the signal correcting apparatus according to any one of the first to thirteenth aspects or the signal correcting method according to the fourteenth aspect. Reference signal generating means for generating a reference signal to be provided to the integrated transmission system, response signal measuring means for measuring a response signal of the integrated transmission system corresponding to the given reference signal, based on the reference signal and the response signal, In order for the characteristics of the integrated transmission system to approach the predetermined target characteristics,
Adjusting control means for adjusting the coefficients.

According to a sixteenth aspect of the present invention, in the coefficient adjusting apparatus of the fifteenth aspect, the adjustment control means is configured to control a response signal of the integrated transmission system corresponding to a given reference signal. A coefficient related to a nonlinear filter coefficient of a non-recursive portion of the correction filter is adjusted so that each harmonic distortion or cross modulation distortion is reduced.

According to a seventeenth aspect of the present invention, a method for adjusting a coefficient of a signal correcting apparatus is a coefficient adjusting method for adjusting each coefficient in the signal correcting apparatus of any one of the first to thirteenth aspects or the signal correcting method of the fourteenth aspect. Generating a reference signal to be given to the integrated transmission system, measuring a response signal of the integrated transmission system corresponding to the given reference signal, and determining a characteristic of the integrated transmission system based on the reference signal and the response signal. Each coefficient is adjusted so as to approach the target characteristic.

A storage medium according to claim 18 is a computer-readable storage medium storing a computer-executable program, and the program realizes the apparatus or method according to any one of claims 1 to 17. Characterized in that:

The concept of terms in the claims is defined as follows.

"Integrated transmission system" refers to a system in which a signal correction device for correcting an input signal and a transmission system for transmitting the corrected signal are integrated.

"Thile-Small parameter" refers to a parameter of a lumped constant model of a speaker. In the embodiment, the lowest resonance frequencies f0 and f0 ′, the sharpness of resonance Q0 and Q0 ′, and the like correspond.

The "target characteristic" refers to a target characteristic of the integrated transmission system. The characteristic includes a characteristic represented by a seal small parameter.

The "virtual physical quantity" is a predetermined physical quantity representing a state of the integrated transmission system that changes depending on an input signal, and is a virtual physical quantity when the integrated transmission system is assumed to have target characteristics. . In the embodiment, the virtual displacement x of the voice coil 4 corresponds to the virtual displacement x.

The "digital filter" refers to a digital signal processing circuit that performs digital signal processing on an input signal to obtain an output signal in order to change the frequency characteristics of the signal. In the embodiment, a correction filter and a displacement filter (both are second-order IIR filters) correspond.

The term "filter coefficient" refers to a coefficient given to change the frequency characteristics of a signal in a digital filter. In the embodiment, B1, hx0, C (x), etc. correspond. Note that, among the elements constituting C (x), c
0 is called a linear filter coefficient, and c (x) is called a nonlinear filter coefficient. Coefficient c of nonlinear filter coefficient c (x)
1, c2 and the like are called non-linear coefficients.

[0034]

According to the first aspect of the present invention, there is provided a signal correction apparatus according to the first aspect.
In the signal correction method of No. 4, a correction filter including a second-order or higher-order digital filter having a feedback path is prepared, and at least a part of the filter coefficient of the correction filter is determined based on the predicted virtual physical quantity. The filter corrects the input signal using the determined filter coefficient.

Therefore, by using a second-order or higher-order digital filter having a feedback path, the signal processing algorithm can be standardized and simplified. In addition, adjustment of each coefficient becomes easy. Furthermore, instead of using the speaker's seal small parameters, etc., directly in the correction calculation,
Since the filter coefficients standardized to some extent are used, each coefficient value can be set within a predetermined range. Therefore, it is convenient to perform high-speed processing using a DSP that performs fixed-point arithmetic. Further, since the calculation for adjusting the range of each coefficient value can be omitted, the efficiency of the calculation processing does not decrease, and the calculation accuracy does not decrease. That is, the characteristics of the transmission system having nonlinearity can be easily, quickly, and reliably corrected.

According to a second aspect of the present invention, the signal correction apparatus further includes an amplitude adjustment unit for adjusting the amplitude of the signal, and also determines an amplitude adjustment coefficient of the amplitude adjustment unit based on the predicted virtual physical quantity. The amplitude of the signal before correction by the correction filter or the signal after the correction is adjusted using the coefficient. Therefore, it is easy to associate with the physical model of the transmission system. That is, correction of the characteristics of the transmission system
More easily.

According to a third aspect of the present invention, in the signal correction device, the virtual physical quantity predicting means includes a virtual physical quantity filter including a second-order or higher-order digital filter having a feedback path. Therefore, by using a digital filter for the virtual physical quantity prediction means together with the signal correction means,
The signal processing algorithm can be further standardized and simplified. That is, the characteristics of the transmission system having nonlinearity can be corrected more easily, quickly, and reliably.

According to a fourth aspect of the present invention, the correction filter and the virtual physical quantity filter each include a cyclic portion and a non-cyclic portion, and the filter coefficient of the cyclic portion of the correction filter is equal to that of the virtual physical quantity filter. It is characterized in that it is the same as the filter coefficient of the cyclic part. Therefore,
The cyclic portion can be shared by the correction filter and the virtual physical quantity filter, and the algorithm for signal processing can be further standardized and simplified.

According to a sixth aspect of the present invention, the transmission system includes an electrodynamic loudspeaker, and the virtual physical quantity is a virtual displacement of a voice coil of the electrodynamic loudspeaker. Therefore, it is possible to easily, quickly, and reliably correct the characteristics of the conductive speaker having large nonlinear distortion depending on the displacement.

According to a seventh aspect of the present invention, the amplifier is a voltage output type amplifier, and the three nonlinear filter coefficients of the acyclic portion of the correction filter have a relationship of c (x) + e (x) = d (x). Is satisfied. Therefore, SZ
If two non-linear filter coefficients among the three non-linear filter coefficients are determined by using the bilinear transform in the conversion, another non-linear filter coefficient can be automatically determined. That is, the algorithm of signal processing when the correction filter is constituted by a second-order IIR filter can be further simplified.

In the signal correction device according to the present invention, the amplifier is a current output type amplifier, and the three nonlinear filter coefficients of the acyclic portion of the correction filter are represented by c (x) = e (x) = d (x) / 2 is satisfied. Therefore, SZ
By using bilinear transformation in the conversion, if one nonlinear filter coefficient is determined among the three nonlinear filter coefficients, the other two nonlinear filter coefficients can be automatically determined. That is, the algorithm of signal processing when the correction filter is constituted by a second-order IIR filter can be further simplified.

According to a ninth aspect of the present invention, in the signal correction apparatus, the second-order delayed nonlinear filter coefficient e (x) among the three nonlinear filter coefficients in the acyclic portion of the correction filter is e (x) = 0. I do. Therefore, by using the backward difference approximation transform in the SZ transform, it is possible to configure so that the determination of the second-order delayed nonlinear filter coefficient e (x) among the nonlinear filter coefficients becomes unnecessary. That is, the algorithm of signal processing when the correction filter is constituted by a second-order IIR filter can be further simplified.

According to a tenth aspect of the present invention, the characteristic of the integrated transmission system represented by the target seal small parameter is the characteristic itself represented by the transmission small seal parameter at the time of a low level input signal. Features. Therefore, it is possible to reduce the non-linear distortion of the speaker while keeping the linear characteristic of the actual speaker as it is.

According to the eleventh aspect of the present invention, the characteristic of the integrated transmission system represented by the target seal small parameter is obtained by modifying the characteristic represented by the seal small parameter of the transmission system at the time of a low-level input signal. It is characterized by added characteristics. Therefore, the nonlinear characteristic of the speaker can be reduced, and the linear characteristic of the speaker can be corrected to a desired value.

According to a twelfth aspect of the present invention, there is provided a signal correction apparatus for reducing nonlinearity of a transmission system caused by a force coefficient and a stiffness which change depending on a displacement of a voice coil of an electrodynamic speaker. Therefore, non-linear distortion likely to occur in the electrodynamic speaker can be reduced.

According to a thirteenth aspect of the present invention, there is provided a signal correction apparatus for mitigating a nonlinearity of a transmission system caused by an excluded air volume at an edge of an electrodynamic loudspeaker which changes depending on a displacement of a voice coil of the electrodynamic loudspeaker. Features. Therefore, when applied to a small speaker in which the ratio of the edge area to the entire effective vibration area is relatively high, it is possible to obtain a high effect in reducing nonlinear distortion.

According to a fifteenth aspect of the present invention, there is provided a coefficient adjustment apparatus for a signal correction apparatus and a coefficient adjustment method for a signal correction apparatus according to a seventeenth aspect, wherein a response signal of an integrated transmission system corresponding to a given reference signal is measured. Each coefficient is adjusted based on the signal so that the characteristics of the integrated transmission system approach predetermined target characteristics. Therefore, each coefficient can be automatically adjusted. That is, adjustment of each coefficient can be easily performed.

According to a sixteenth aspect of the present invention, the coefficient adjustment device of the signal correction device includes a non-cyclic portion of the correction filter that reduces each harmonic distortion or cross-modulation distortion of the response signal of the integrated transmission system corresponding to the given reference signal. And adjusting a coefficient relating to the non-linear filter coefficient. Therefore, by configuring such that the adjustment is performed sequentially for each harmonic by focusing on components related to each harmonic or each cross-modulation wave, the adjustment of each coefficient can be performed more easily. it can.

A storage medium according to claim 18 is a computer-readable storage medium storing a computer-executable program, the program comprising: a signal correction device, a signal correction method, a coefficient adjustment device of the signal correction device, It is characterized by realizing any device or method of the adjustment method. Therefore, by using the present invention in a computer, it is possible to more easily, quickly, and surely correct characteristics of a transmission system having nonlinearity and adjust each coefficient.

[0050]

BEST MODE FOR CARRYING OUT THE INVENTION

[First Embodiment]-Overall Configuration of Signal Correction Device-Fig. 1 shows a signal correction device 12 according to an embodiment of the present invention.
FIG. 1 shows a block diagram of an integrated transmission system 10 including the following. The integrated transmission system 10 is a system in which the signal correction device 12 and the transmission system 14 are integrated.

The signal correction device 12 corrects an input signal u, which is an input signal, so that the characteristics of the integrated transmission system 10 approach a predetermined target characteristic, and supplies the input signal u to the transmission system 14. As the target characteristic, the characteristic of the integrated transmission system 10 represented by the target seal small parameter is given. The signal correction device 12 includes a virtual physical quantity prediction unit (virtual physical quantity filter).
, A correction coefficient determination unit 18 as a correction coefficient determination unit, and a signal correction unit 20. The signal correction means 20 includes a correction filter 22, an amplitude adjustment unit 24
It has.

The displacement filter 16 is composed of a second-order or higher-order digital filter having a feedback path, and uses a filter coefficient determined based on a target characteristic of the integrated transmission system 10 to virtually determine the input signal u. Predict displacement x.

The correction coefficient determining section 18 determines at least a part of the filter coefficients of the correction filter 22 based on the predicted virtual displacement x. The correction coefficient determining unit 18
Is based on the predicted virtual displacement x,
Is also determined.

The correction filter 22 is composed of a digital filter of second or higher order having a feedback path, and corrects the input signal u using the determined filter coefficient.

The amplitude adjustment unit 24 adjusts the amplitude of the signal before correction by the correction filter 22 or the amplitude of the signal after the correction by using the determined amplitude adjustment coefficient.

The correction filter 22 and the displacement filter 16 each include a recurring unit 22a including a feedback path and corresponding to an upstream part of signal processing, and a non-recursive unit 22b including no feedback path and corresponding to a downstream part of signal processing. 16b, and the filter coefficient of the recursive unit 22a of the correction filter 22 is the same as the filter coefficient of the recursive unit 22a of the displacement filter 16. In this embodiment, the correction filter 2
2 and the displacement filter 16 share the circulation unit 22a. The correction filter 22 and the displacement filter 16 are
Both are second-order IIR filters.

The filter coefficient of the non-repeating portion 22b of the correction filter 22 is a filter coefficient determined based on the virtual displacement x. As described later, a linear filter coefficient independent of the virtual displacement x and a virtual filter x It can be expressed as a sum with a non-linear filter coefficient that varies depending on the type.

-Hardware- FIG. 2 shows each function of the signal correction device 12 shown in FIG.
(Digital Signal Processor) 30 shows an example of a hardware configuration of the integrated transmission system 10 when it is realized using the digital signal processor 30. The DSP 30 is a microprocessor that performs high-speed operations, and has an internal memory (not shown). The internal memory stores programs, coefficients, and the like. The DSP 30 performs a signal correction process on the input signal u according to a program stored in the internal memory, and obtains a correction output signal uL as a correction output.

As in this embodiment, the amplifier 28 (described later)
Is a voltage output type, the input signal u and the corrected output signal u
L is equivalent to the voltage in the analog signal. In addition,
As will be described later, when the amplifier 28 is a current output type, the input signal and the correction output are represented by the input signal i and the correction output signal iL, and these signals are equivalent to the current in the analog signal.

The D / A converter 26 converts the corrected output signal uL (digital signal) from the DSP 30 into an analog signal. The voltage output type amplifier 28 amplifies the conversion output of the D / A converter 26 and outputs the amplified speaker 2
Drive. Note that the DSP 30 corresponds to the signal correction device 12. The D / A converter 26, the amplifier 28, and the electrodynamic speaker 2 correspond to the transmission system 14. The D / A converter 26 and the amplifier 28 correspond to a drive system.

-Derivation of Processing Equation- FIG. 3 shows an electrodynamic loudspeaker (hereinafter, referred to as a "composition") constituting the transmission system 14.
FIG. 24 shows a diagram in which a physical model (lumped parameter model) 2 (referred to simply as “speaker”) 2 (see FIG. 23) is represented. Speaker 2
Each seal small parameter of the following is represented by the following symbol. Minimum resonance frequency f0 Sharpness of resonance Q0 Sharpness of resonance Qm of pure mechanical system Force coefficient BL0, DC electric resistance Rdc Effective mass m0 Mechanical resistance Rm Stiffness K0.

Normally, a physical model of a vibration system of a speaker is approximated as a primary resonance system in a low frequency range.
Therefore, if the target ideal seal small parameter of the speaker is expressed by a symbol with “′”, the equation of motion of the ideal speaker is expressed by the second-order linear differential of the displacement x of the voice coil of the speaker as follows. BL0 '・ A0 ・ u (t) / Rdc = m0 ・ (d ² x / dt ² ) + Rm ・ (dx / dt) + K0 ′ ・ x + (BL0 ′ ² / Rdc) ・ ( dx / dt) (2) In equation (2), u (t) is an input signal to an ideal speaker, and A0 is an amplification gain of the drive system (D / A converter 26
And the total gain of the amplifier 28).

On the other hand, in the actual loudspeaker 2, the force coefficient and the stiffness of the seal small parameters show almost constant values BL0 and K0 when the displacement x of the voice coil 4 of the loudspeaker 2 is extremely small. It is known that when the displacement x of the voice coil 4 is large to some extent, the voice coil 4 tends to change depending on the displacement x. That is, the force coefficient and the stiffness have nonlinearity. This is one of the causes of the nonlinear distortion of the speaker 2. The force coefficient BL (x) and the stiffness K (x) that change depending on the displacement x are expressed by the following equation.BL (x) = BL0b (x) = BL0 ・ (1 + b1 ・^{x + b2 · x 2) K} (x) = K0 · k (x) = K0 · (1 + k1 · x + k2 · x 2 + k3 · x 3 + k4 · x 4) ··· (1) formula In (1), b1, b2, k1, k2, k3, and k4 are nonlinear coefficients.

Considering such non-linearity that changes depending on displacement, the actual equation of motion of the speaker 2 is represented by the following second-order nonlinear differential equation: BL (x) · A0 · uL ( t) / Rdc = m0 ・ (d ² x / dt ² ) + Rm ・ (dx / dt) + K (x) ・ x + BL ² (x) / Rdc ・ (dx / dt) ・ ・ ・ (3) In Equation (3), uL (t) is an actual input signal to the speaker 2.

Incidentally, the ideal lowest resonance angular frequency ω0 ′
(= 2 · π · f0 ′) and the resonance sharpness Q0 ′, and the actual lowest resonance angular frequency ω0 (= 2 · π · f0), the resonance sharpness Q0 and the resonance sharpness Qm of the pure mechanical system are respectively can be expressed by the formula, ω0 '= (K0' / m0) 1/2 Q0 '= 1 / (Rm + (BL0' 2 / Rdc)) · (m0 · K0 ') 1/2 ω0 = (K0 / ^{m0) 1/2 Q0 = 1 / (} Rm + (BL0 2 / Rdc)) · (m0 · K0) 1/2 Qm = 1 / Rm · (m0 · K0) 1/2 ··· (8).

Here, assuming that G0 = BL0 · A0 / (Rdc · m0) ≒ ω0 · A0 / (Q0 · BL0), the equation (2) becomes G0 · u (t) = ( d ² x / dt ² ) + ω0 ′ / Q0 ′ · (dx / dt) + ω0 ′ ² · x (4).

The equation (3) is obtained from the equations (1) and (8) as follows: b (x) · G0 · uL (t) = (d ² x / dt ² ) + (1+ (1-Q0) / Qm) · (b ² (x) −1)) · (ω0 / Q0) · (dx / dt) + k (x) · ω0 ² · x (5).

As can be seen from Equation (4), the displacement x (t) of the voice coil of an ideal speaker shows linearity with respect to the input signal u (t), and does not cause distortion. The solution of the linear differential equation represented by the equation (4) is expressed as follows: x (t) = L ⁻¹ {G0 / (s ² + s · ω0 ′ / Q0 ′ + ω0 ′ ² )} * u (t (18) However, in this specification, unless otherwise specified, s is a Laplace operator L ^-1 {P} is a Laplace inverse transformation of P P * Q is a relation between P and Q Represents the convolution integral.

From equation (18), the velocity v (t) and acceleration a (t) of the ideal voice coil of the speaker are expressed by the following equations: v (t) = dx / dt = L ⁻¹ { (s ・ G0) / (s ² + s ・ ω0 '/ Q0' + ω0 ' ² )} * u (t) ・ ・ ・ (6) a (t) = d ² x / dt ² = L ^-1 { (s ² · G 0) / (s ² + s · ω 0 '/ Q 0' + ω 0 ' ² )} * u (t) (7) On the other hand, as can be seen from the nonlinear differential equation (5), The displacement x (t) of the voice coil of the speaker 2 is represented by the input signal uL
(t) shows non-linearity and causes distortion. Therefore, the actual displacement x (t) of the speaker 2 in Expression (5) is expressed by Expression (1).
8) Displacement of voice coil of ideal speaker shown in 8)
the actual input signal to speaker 2 so that it equals x (t)
If uL (t) is corrected, the actual speaker 2 should have no linear distortion.

Equation (5) is replaced by equations (18), (6) and (7)
, Then uL (t) = 1 / b (x) · [L ^-1 {s ² / (s ² + s · ω0 '/ Q0' + ω0 ' ² )} * u (t) + (1 + (1-Q0 / Qm) ・ (b ² (x) -1)) ・ ω0 / Q0 ・ L ^-1 {s / (s ² + s ・ ω0 '/ Q0' + ω0 ' ² )} * u ( t) + k (x) · ω0 ² · L ⁻¹ {1 / (s ² + s · ω0 ′ / Q0 ′ + ω0 ′ ² )} * u (t)] (9).

As shown in equation (9), the actual speaker 2
The signal uL (t) (corrected output) obtained by subjecting the signal u (t) (input signal) to predetermined correction as a signal to the speaker 2, so that the loudspeaker 2 exhibiting non-linearity is in a linearly distortion-free state. Can be. That is, uL (t) shown in Expression (9) means a non-linear correction signal for making the speaker 2 exhibiting non-linearity have no linear distortion.

In order to execute the above nonlinear correction by the DSP 30, it is necessary to convert a continuous system into a discrete system. Bilinear transformation, Euler approximation (backward difference approximation), and the like are used as SZ transformation for transforming a continuous system into a discrete system.
In this embodiment, a bilinear transformation is used. Assuming that the sampling period is T and the sampling frequency is Fs (= 1 / T), the conversion formula of the bilinear conversion is expressed as follows: s = 2 / T · (1-z ⁽⁻¹⁾ ) / (1 + z ^(-1) ) = 2 · Fs · (1-z ^(-1) ) / 1 + z ^(-1) ) (10).

Using equation (10), the displacement, velocity, and acceleration are converted as follows: Displacement: 1 / (s ² + s · ω0 ′ / Q0 ′ + ω0 ′ ² ) = (hx0 + hx1 ・ z ^-1 + hx2 ・ z ^-2 ) / (1 + B1 ・ z ^-1 + B2 ・ z ^-2 ) ・ ・ ・ (100) where hx0 = (1 / (4 ・ Fs ² )) / ( 1 + π / Q0 '・ (f0' / Fs) + π ²・ (f0 '/ Fs) ² ) hx1 = (1 / (2 ・ Fs ² )) / (1 + π / Q0' ・ (f0 '/ Fs) + π ²・ (f0 '/ Fs) ² ) hx2 = (1 / (4 ・ Fs ² )) / (1 + π / Q0' ・ (f0 '/ Fs) + π ²・ (f0' / Fs ) ² ) ・ ・ ・ (20) Also, B1 = (-2 + 2 ・ π ²・ (f0 '/ Fs) ² ) / (1 + π / Q0' ・ (f0 '/ Fs) + π ²・ ( f0 '/ Fs) ² ) B2 = (1-π / Q0' ・ (f0 '/ Fs) + π ²・ (f0' / Fs) ² ) / (1 + π / Q0 '・ (f0' / Fs) + π ² · (f0 '/ Fs) ² ) (12).

Speed: s / (s ² + s · ω0 ′ / Q0 ′ + ω0 ′ ² ) = (hv0 + hv1 · z ⁻¹ + hv2 · z ⁻² ) / (1 + B1 · z ⁻¹ + B2・ Z ^-1 ) where hv0 = (1 / 2Fs) / {1+ (f0 '/ Fs ・ π / Q0') + (f0 '/ Fs ・ π) ² } hv1 = 0 hv2 =-(1 / 2Fs ) / {1+ (f0 ′ / Fs · π / Q0 ′) + (f0 ′ / Fs · π) ² } (101).

Acceleration: s ² / (s ² + s · ω0 ′ / Q0 ′ + ω0 ′ ² ) = (ha0 + ha1 · z ⁻¹ + ha2 · z ⁻² ) / (1 + B1 · z ⁻¹ + B2 ・ z ^-2 ) where ha0 = 1 / {1+ (f0 '/ Fs ・ π / Q0') + (f0 '/ Fs ・ π) ² } ha1 = -2 / {1+ (f0' / Fs・ Π / Q0 ') + (f0' / Fs ・ π) ² } ha2 = 1 / {1+ (f0 '/ Fs ・ π / Q0') + (f0 '/ Fs ・ π) ² } ・ ・ ・ ( 102).

[0076] When the bilinear transform, coefficients in each equation above has a following relationship, hx0 = hx2 = hx1 / 2 = (1 / 4Fs 2) / {1+ (f0 '/ Fs · π / Q0 ') + (f0' / Fs ・ π) ² } hv0 = -hv2 = (1 / 2Fs) / {1+ (f0 '/ Fs ・ π / Q0') + (f0 '/ Fs ・ π) ² } = 2Fs ・ hx0 hv1 = 0 ha0 = ha2 =-(ha1 / 2) = 1 / {1+ (f0 '/ Fs ・ π / Q0') + (f0 '/ Fs ・ π) ² } = 4Fs ²・ hx0 ... (103).

Expressions (100), (20), (12), (1)
01) and (102), equations (18), (6), and (7) representing the displacement x (t), velocity v (t), and acceleration a (t) of the voice coil of the ideal speaker described above. ) Is converted as follows: x (n) = G0 · Z ⁻¹ {(hx0 + hx1 · z ⁻¹ · hx2 · z ⁻² ) / (1 + B1 · z ⁻¹ + B2 · z ^-2 )} * u (n) ・ ・ ・ (19) v (n) = G0 ・ Z ^-1 ((hv0 + hv1 ・ z ^-1・ hv2 ・ z ^-2 ) / (1 + B1 ・ z ^-1) + B2 ・ z ^-2 )} * u (n) ・ ・ ・ (104) a (n) = G0 ・ Z ^-1 ((ha0 + ha1 ・ z ^-1・ ha2 ・ z ^-2 ) / (1 + B1 · Z ^-1 + B2 · z ^-2 )} * u (n) (105) In this specification, unless otherwise specified, Z ^-1 {P} is the inverse Z transformation of P. Express.

The displacement x (n) represented by the equation (19) is the virtual displacement x of the voice coil 4.

Equations (19), (104) and (105) are
If applied to equation (9) indicating the nonlinear correction signal, uL (n) = 1 / b (x) · {Z ⁻¹ {(ha0 + ha1 · z− ¹ · ha2 · z− ² ) / (1+ B1 ・ z ^-1 + B2 ・ z ^-2 )} * u (n) + (1+ (1-Q0 / Qm) ・ (b (x) ² -1)) ・ ω0 / Q0 ・ Z ^-1 (( hv0 + hv1 ・ z ^-1 + hv2 ・ z ^-2 ) / (1 + B1 ・ z ^-1 + B2 ・ z ^-2 )} * u (n) + k (x) ・ ω0 ²・ Z ^-1 {( hx0 + hx1 · z ⁻¹ + hx2 · z ⁻² ) / (1 + B1 · z ⁻¹ + B2 · z ⁻² )} * u (n)} (106).

Here, hv0 · ω0 / Q0 = (f0 / Fs · π / Q0) / (1 + f0 ′ / Fs · π / Q0 ′ + (f0 ′ / Fs · π) ² ) hx0 · ω0 ² = Since (f0 / Fsππ) ² / (1 + f0 '/ Fs ・ π / Q0' + (f0 '/ Fs ・ π) ² ), using this relationship to rearrange the equation (106), UL (n) = 1 / b (x) ・ Z ^-1 ((C (x) + D (x) ・ z ^-1 + E (x) ・ z ^-2 ) / (1+ B1 · z ⁻¹ + B2 · z ⁻² )} * u (n) (11).

[0081] formula (11), C (x) = c0 + c (x) = c0 + c1 · x + c2 · x 2 + c3 · x 3 + c4 · x 4 , however, c0 = ha0 + ω0 / Q0・ Hv0 + ω0 ²・ hx0 = (1+ (f0 / Fs ・ π / Q0) + (f0 / Fs ・ π) ² ) / (1+ (f0 '/ Fs ・ π / Q0') + (f0 '/ Fs ・ π) ² ) c (x) = (1-Q0 / Qm) ・ (b (x) ² -1) ・ ω0 / Q0 ・ hv0 + (k (x) -1) ・ ω0 ²・ hx0 = {( 1-Q0 / Qm) ・ (b (x) ² -1) ・ f0 / Fs ・ π / Q + (k (x) -1) ・ (f0 / Fs ・ π) ² } / (1 + f0 '/ Fs • π / Q0 ′ + (f0 ′ / Fs · π) ² ) (13). D (x) = d0 + d (x) = d0 + d1 ・ x + d2 ・ x ² + d3 ・ x ³ + d4 ・ x ⁴ , where d0 = ha1 + ω0 / Q0 ・ hv1 + ω0 ²・ hx1 = ( -2 + 2 (f0 / Fs ・ π) ² ) / (1 + f0 '/ Fs' ・ π / Q0 '+ (f0' / Fs ・ π) ² ) d (x) = (1-Q0 / Q1)・ (B (x) ² -1) ・ ω0 / Q0 ・ hv1 + ω0 ²・ (k (x) -1) ・ hx1 = (2 ・ (f0 / Fs ・ π) ²・ (k (x) -1 )) / (1+ (f0 ′ / Fs · π / Q0 ′) + (f0 ′ / Fs · π) ² ) (14). E (x) = e0 + e (x) = e0 + e1 ・ x + e2 ・ x ² + e3 ・ x ³ + e4 ・ x ⁴ where e0 = ha2 + ω0 / Q0 ・ hv2 + ω0 ²・ hx2 = ( 1- (f0 / Fs ・ π / Q0) + (f0 / Fs ・ π) ² ) / (1 + f0 '/ Fs ・ π / Q0' + (f0 '/ Fs ・ π) ² ) e (x) = (1-Q0 / Qm) ・ (b (x) ² -1) ・ ω0 / Q0 ・ hv2 + ω0 ²・ (k (x) -1) ・ hx2 = {-(1-Q0 / Qm) ・ (b (x) ² -1) ・ f0 / Fs ・ π / Q0 + (k (x) -1) ・ (f0 / Fs ・ π) ² } / (1+ (f0 '/ Fs ・ π / Q0') + ( f0 '/ Fs · π) ² ) (15). From the equations (13), (14), and (15), the following relationship is derived for the nonlinear filter coefficients c (x), d (x), and e (x): d (x) = c (x) + e (x) (16) Accordingly, d (x) is represented by c (x) and e (x).

On the other hand, nonlinear coefficient filter coefficients b (x), c
The relationship between (x), d (x) and e (x) is c (x) = d (x) / 2 + (b ² (x) -1) ・ ((c0-e0) / 2) ・(1-Q0 / Qm) e (x) = d (x) / 2- (b ² (x) -1) ・ ((c0-e0) / 2) ・ (1-Q0 / Qm) Therefore, b ( x) = {(c (x) -e (x)) / (c0-e0) Qm / (Q0-Qm) +1} ^1/2・ ・ ・ (17) Therefore, b (x) is also c (x) and e (x).

FIGS. 4 and 5 are signal processing circuit diagrams of the signal correction devices 12 and 52, respectively, when expressed by the equation (11). In the correction coefficient determination unit 18 of the signal correction device 12 shown in FIG. 4, the nonlinear filter coefficients c (x) and e (x)
, The nonlinear filter coefficients d (x) and b (x) are automatically determined. On the contrary, the correction coefficient determination unit 18 of the signal correction device 52 shown in FIG. Non-linear filter coefficients based on filter coefficients d (x), b (x)
c (x) and e (x) are automatically determined.

-Configuration of Coefficient Adjustment Device- FIG. 6 illustrates an overall configuration of a coefficient adjustment device 32 for adjusting each coefficient of the signal correction device 12 shown in FIG. The coefficient adjustment device 32 includes an oscillator 34 as a reference signal generation unit,
An A / D converter 36, a microphone 38 and a measuring device 40 as a response signal generation unit, and a controller 42 as an adjustment control unit are provided.

The oscillator 34 generates a reference signal to be given to the integrated transmission system 10. The A / D converter 36 converts the reference signal provided as an analog signal into a digital signal and provides the digital signal to the integrated transmission system 10. The measuring device 40 measures, via the microphone 38, a response signal of the integrated transmission system 10 corresponding to the given reference signal.

The controller 42 adjusts each coefficient of the DSP 30 (signal adjusting device 12) based on the reference signal and the response signal so that the characteristic of the integrated transmission system 10 approaches a predetermined target characteristic. In this embodiment, the controller 42 controls the integrated transmission system 1 corresponding to the given reference signal.
It is configured to adjust a non-linear coefficient such as a non-linear filter coefficient of the acyclic portion 22b (see FIG. 1) of the correction filter 22 so that each harmonic distortion of the response signal of 0 is reduced. As the controller 42, a computer or a microcomputer can be used.

The measuring resistor R1 is connected in series to the DC resistor Rdc of the speaker 2. Measuring resistor R
The resistance value of 1 is sufficiently smaller than the DC resistance Rdc of the speaker 2.

-Coefficient Adjustment Process- FIG. 7 illustrates a flow of a process when the coefficient adjustment process of the signal correction device 12 shown in FIG. 4 is performed using the coefficient adjustment device 32 of FIG. First, the minimum resonance frequency f0 and resonance sharpness Q0 of the speaker 2 are measured (Step S).
2). A sweep frequency is given to the speaker 2, and a measuring resistor R1
Is measured (the admittance of FIG. 24 is minimized). This frequency is the lowest resonance frequency f0. The resonance sharpness Q0 is obtained as the resonance sharpness at the lowest resonance frequency f0 (the kurtosis at f0 of the admittance curve in FIG. 24).

Further, in this step, the ratio A0 / BL0 between the total gain A0 of the D / A converter 26 and the amplifier 28 and the force coefficient BL0 of the speaker 2 is determined. When the volume (not shown) of the amplifier 28 is variable,
In a signal correction process to be described later, a configuration may be adopted in which the total gain A0 is sequentially detected. In this case, the total gain A0 is obtained by sequentially measuring the position of the volume, the input / output ratio of the amplifier 28, and the like. A0 / BL0 obtained in this manner is used when calculating the virtual displacement x in the signal correction processing.

Next, referring to the measured lowest resonance frequency f0 and resonance sharpness Q0, the lowest resonance frequency f
0 ′ and resonance sharpness Q0 ′ (target characteristic) are determined (step S4).

Next, the measured minimum resonance frequency f0 and resonance sharpness Q0, and the minimum resonance frequency f0 to be realized
0 'and the resonance sharpness Q0', the displacement filter 16
Then, the filter coefficients B1, B2, hx0, hx1, hx2 and the linear filter coefficients c0, d0, e0 of the correction filter 22 are calculated (step S6). Filter coefficients B1, B2, hx of displacement filter 16
0, hx1 and hx2 are calculated according to the above equations (12) and (20). Linear filter coefficient c0, d of the correction filter 22
0 and e0 are calculated according to the above equations (13), (14) and (15). Note that the filter coefficients B1 and B2 of the correction filter 22 are the filter coefficients B1 and B2 of the displacement filter 16.
Same as 1, B2.

Next, the nonlinear filter coefficient shown in FIG.
Non-linear coefficients c1 to c4 and e1 to e4 constituting c (x) and e (x) are determined (step S8). As described above, in the correction coefficient determination unit 18 of the signal correction device 12 shown in FIG. 4, based on the nonlinear filter coefficients c (x) and e (x), the nonlinear filter coefficients d (x) and b (x ) Is automatically determined.
Therefore, if the nonlinear coefficients c1 to c4 and e1 to e4 are determined, the filter coefficients C (x), D (x), E
(x) and the amplitude adjustment coefficient 1 / b (x) are calculated based on the virtual displacement x, as described later.

A method of determining the nonlinear coefficients c1 to c4 and e1 to e4 using the controller 42 shown in FIG. 6 will be described below. The controller 42 firstly outputs the signal correction device 12 (D
The values of the coefficients c2 to c4 and e2 to e4 (see FIG. 4) of SP30) are set to “0”. Next, a reference signal of a predetermined frequency is
Integrated transmission system 1 via oscillator 34 and A / D converter 36
0, the response signal (sound pressure, acceleration) of the integrated transmission system 10 corresponding to the reference signal is taken in through the microphone 38 and the measuring device 40. The controller 42 calculates a coefficient c1,
The signal level of the second harmonic of the response signal is monitored while appropriately changing the value of e1. The controller 42 determines the values of the coefficients c1 and e1 when the signal level of the second harmonic of the response signal has the smallest value as the values of the nonlinear coefficients c1 and e1.

Next, the controller 42 calculates the coefficients c1, e1
The values of the previously determined nonlinear coefficients c1 and e1 are used, and the values of the undetermined coefficients c3 to c4 and e3 to e4 (see FIG. 4) are set to “0”. Coefficient c when the signal level of the third harmonic has the smallest value
2. The values of e2 are determined as the values of the nonlinear coefficients c2 and e2.

By repeating the same procedure, the values of the coefficients c3 and e3 and the coefficients c4 and e4 when the signal levels of the fourth harmonic and the fifth harmonic of the response signal show the smallest values are calculated as follows: The nonlinear coefficients c3 and e3, respectively, and the nonlinear coefficient
Determine the values of c4 and e4. Thus, the nonlinear coefficient c1
~ C4, e1 ~ e4 are determined.

On the other hand, the signal correction device 52 shown in FIG.
In the correction coefficient determination unit 18 of FIG. 4, the nonlinear filter coefficient d
Based on (x), b (x), nonlinear filter coefficients c (x), e (x)
Is automatically determined. Therefore, in this case, in step S8, the nonlinear coefficients d1 to d4 and b1 to b4 may be determined instead of the nonlinear coefficients c1 to c4 and e1 to e4.

In step S8, the nonlinear coefficient
The method for determining c1 to c4, e1 to e4, and the like is not limited to the method described above. For example, the coefficients other than the non-linear coefficients to be determined may be set to “0”. In this case, for example, when trying to determine the values of the nonlinear coefficients c2 and e2, the coefficients c1 and e1
And the values of the coefficients c3 to c4 and e3 to e4 are set to “0” regardless of whether they have been determined or not, and the signal level of the third harmonic of the response signal indicates the smallest value. Coefficient c
2. The values of e2 are determined as the values of the nonlinear coefficients c2 and e2.

Further, nonlinear coefficients (for example, c1 to c4, e1 to e
When determining 4), first determine the nonlinear coefficients c1 and e1,
Thereafter, the order is determined in the order of c2 and e2, c3 and e3,..., But the order of the determination is not limited to this. Further, the configuration may be such that the non-linear coefficients c1 to c4 and e1 to e4 are determined at once.

When the values of the nonlinear coefficients c1 and e1 are determined, for example, the controller 42 monitors the signal level of the second harmonic of the response signal while appropriately changing the values of the coefficients c1 and e1. The configuration is such that the values of the coefficients c1 and e1 when the signal level of the second harmonic of the signal has the smallest value are automatically determined as the values of the nonlinear coefficients c1 and e1. Some or all of the operations may be manually performed by an operator.

-Signal Correction Processing- An example of a processing flow when signal correction processing is performed using the signal correction apparatus 12 that has been subjected to coefficient adjustment processing is shown in a flowchart of FIG. As shown in FIG. 4, after the input signal u (digital signal) is adjusted to a predetermined level by the attenuator ATT (step S10), the high-pass filter H
The DC offset component is cut by the PF (step S12).

Based on the input signal u which has been subjected to the pre-processing in this way, the displacement filter 16
Is calculated (step S14). The virtual displacement x is calculated according to the aforementioned equation (19).

On the basis of the calculated virtual displacement x, the correction coefficient determining unit 18 determines the filter coefficients C (x), D (x), E (x) of the correction filter 22 and the amplitude adjustment coefficient 1 / b ( x) is determined (step S16).

In this step, the filter coefficient C (x)
Is the formula (13) of the corresponding formula C (x) = c0 + c (x) = c0 + c1 · x + c2 · x 2 + c3 · x 3 + c4 · x 4 ··· (13-1) Therefore, it is calculated. Also, the filter coefficient E (x) is
Calculated in accordance with the applicable formula E of formula (15) (x) = e0 + e (x) = e0 + e1 · x + e2 · x 2 + e3 · x 3 + e4 · x 4 ··· (15-1) Is done. The filter coefficient D (x) is calculated by the equation (1)
This is calculated according to the expression D (x) = d0 + d (x) (14-1) and expression (16) in 4).

Further, the amplitude adjustment coefficient 1 / b (x) is obtained from b (x) = 1 + b1 × x + b2 × ² (1-1 ′) derived from equation (1) and It is calculated according to (17).

Next, the input signal is corrected in the correction filter 22, and a corrected output signal uL is calculated (step S18). The corrected output signal uL is obtained by calculating filter coefficients B1 and B2 calculated in advance (step S6) and filter coefficients C (x), D (x), E (x),
It is obtained by correcting the input signal u according to the equation (11) based on the amplitude adjustment coefficient 1 / b (x).

The obtained corrected output signal uL (digital signal) is converted into an analog signal by the D / A converter 26 (step S20) and then amplified by the amplifier 28 to a predetermined level, as shown in FIG. (Step S2
2) The speaker 2 is driven (step S24).

As described above, the desired output signal uL is obtained by performing desired linear correction and non-linear correction on the input signal u, and the loudspeaker 2 is driven based on the obtained corrected output signal uL. With low frequency characteristics of
An integrated transmission system 10 with less nonlinear distortion can be realized.

FIG. 21 shows the frequency response of the sound pressure when the linear correction is not performed (the lowest resonance frequency f0 and the resonance sharpness Q0), and the case where the linear correction according to this embodiment is performed (f0 '=
f0 / 2, Q0 '= Q0). It can be seen that the sound pressure in the low frequency range is increased by performing the linear correction for reducing the lowest resonance frequency.

FIG. 22 shows data obtained by measuring the level of a harmonic appearing in the output sound pressure using a sine wave (50 Hz) as an input signal to the integrated transmission system 10. FIG. 22A
FIG. 22B is data indicating a harmonic level when non-linear correction is not performed, and FIG. 22B is data indicating a harmonic level when non-linear correction according to this embodiment is performed. It can be seen that by performing the non-linear correction, harmonics of 100 Hz or more are reduced.

[Second Embodiment] FIG. 9 shows a signal processing circuit of a signal correction device 62 according to another embodiment of the present invention. The overall configuration of the signal correction device, the hardware, the configuration of the coefficient adjustment device for adjusting each coefficient of the signal correction device, the coefficient adjustment process, and the signal correction process using the signal correction device are shown in FIG. There are many points in common with the device 12.

However, the above-described signal correction device 12 shown in FIG. 4 can be applied to the case where both the linear correction and the non-linear correction are performed, but the signal correction device 62 shown in FIG. It is applied when implementing. Therefore, the signal correction device 62 has the following features in the content of signal processing and the like.

The signal correction device 62 does not perform linear correction. Therefore, in the example of the signal correction device 12 described above, the speaker small parameters f0 ′, Q0 ′, Qm ′, K0 ′, BLO ′, m0 ′, Rdc ′, Rm ′ representing the target characteristics of the integrated transmission system 10 are used. The seal small parameters f0, Q0, Qm, K0, BLO, m0, Rdc, and Rm representing the actual measurement characteristics of No. 2 may be used.

As a result, the filter coefficients used in equation (11) representing the corrected output signal uL are expressed by equations (12) to (1).
B1 = (-2 + 2 ・ π ²・ (f0 / Fs) ² ) / (1 + π / Q0 ・ (f0 / Fs) + π ²・ (f0 / Fs) ² ) B2 = (1-π / Q0 ・ (f0 / Fs) + π ²・ (f0 / Fs) ² ) / (1 + π / Q0 ・ (f0 / Fs) + π ²・ (f0 / Fs ) ² ) ... (22).

[0114] C (x) = c0 + c (x) = c0 + c1 · x + c2 · x 2 + c3 · x 3 + c4 · x 4 , however, c0 = (1 + π / Q0 · (f0 / Fs ) + π ²・ (f0 / Fs) ² ) / (1 + π / Q0 ・ (f0 / Fs) + π ²・ (f0 / Fs) ² ) = 1 c (x) = {(1-Q0 / Qm ) ・ (B (x) ² -1) ・ f0 / Fs ・ π / Q + (k (x) -1) ・ (f0 / Fs ・ π) ² } / (1 + f0 / Fs ・ π / Q0 + (f0 / Fs · π) ² ) (23). D (x) = d0 + d (x) = d0 + d1 · x + d2 · x 2 + d3 · x 3 + d4 · x 4 , however, d0 = (- 2 + 2 · π 2 · (f0 / Fs) 2) / (1 + π / Q0 ・ (f0 / Fs) + π ²・ (f0 / Fs) ² ) = B1 d (x) = (2 ・ (f0 / Fs ・ π) ²・ (k (x) -1)) / (1+ (f0 / Fs · π / Q0) + (f0 / Fs · π) ² ) (24). E (x) = e0 + e (x) = e0 + e1 ・ x + e2 ・ x ² + e3 ・ x ³ + e4 ・ x ⁴ , where e0 = (1-π / Q0 ・ (f0 / Fs) + π ²・ (f0 / Fs) ² ) / (1 + π / Q0 ・ (f0 / Fs) + π ²・ (f0 / Fs) ² ) = B2 e (x) = {-(1-Q0 / Qm) ・(b (x) ² -1) ・ f0 / Fs ・ π / Q0 + (k (x) -1) ・ (f0 / Fs ・ π) ² } / (1+ (f0 / Fs ・ π / Q0) + ( f0 / Fs · π) ² ) (25). Expression (19) representing the virtual displacement x of the voice coil 4
The filter coefficient used for is expressed by the following equation instead of equation (20): hx0 = (1 / (4 · Fs ² )) / (1 + π / Q0 · (f0 / Fs) + π ² · ( f0 / Fs) ² ) = hx2 = hx1 / 2 hx1 = (1 / (2 ・ Fs ² )) / (1 + π / Q0 ・ (f0 / Fs) + π ²・ (f0 / Fs) ² ) hx2 = (1 / (4 · Fs ² )) / (1 + π / Q0 · (f0 / Fs) + π ² · (f0 / Fs) ² ) (26).

FIG. 10 shows a signal processing circuit of the signal correction device 72 when expressed by the equation (11). In the signal correction device 72, similarly to the signal correction device 52 shown in FIG. 5, the nonlinear filter coefficients c (x) and e (x) are automatically set based on the nonlinear filter coefficients d (x) and b (x). The signal correction device 62 differs from the signal correction device 62 in that the signal correction device 62 is determined.

The signal correction device 62 and the signal correction device 72
The characteristics of the integrated transmission system 10 to be targeted are the characteristics (measured characteristics) themselves represented by the lowest resonance frequency f0, the resonance sharpness Q0, and the like of the transmission system 14 at the time of a low-level input signal. Therefore, it is possible to reduce the nonlinear distortion of the speaker 2 while keeping the actual linear characteristic of the speaker 2 as it is.

[Third Embodiment] FIG. 11 shows a signal processing circuit of a signal correction device 82 according to still another embodiment of the present invention. The overall configuration of the signal correction device, the hardware, the configuration of the coefficient adjustment device for adjusting each coefficient of the signal correction device, the coefficient adjustment process, and the signal correction process using the signal correction device are shown in FIG. There are many points in common with the device 12.

However, the signal correction device 12 shown in FIG. 4 is applied when the amplifier 28 of the transmission system 14 is a voltage output type amplifier.
Is applied when the amplifier 28 of the transmission system 14 is a current output type amplifier. Therefore, the signal correction device 82 has the following features in the content of signal processing and the like.

In the signal correction device 82, the input signal is represented by the input signal i, and the correction output is represented by the correction output signal iL. Both the input signal i and the corrected output signal iL are equivalent to the current in the analog signal.

In this embodiment, the equation of motion of the ideal speaker and the equation of motion of the actual speaker are expressed by the following equations instead of the equations (2) and (3). BL0 ′ · A0 · i (t) = m0 ・ (d ² x / dt ² ) + Rm ・ (dx / dt) + k0 '・ x ・ ・ ・ (27) BL (x) ・ A0 ・ iL (t) = m0 ・ (d ² x / dt ² ) + Rm · (dx / dt) + k (x) · x (28).

Here, if G0 ′ = BL0 · A0 / m0 (29), equations (27) and (28) are G0 ′ · i (t) = (d ² x / dt ² ) + ω0 '/ Qm ・ (dx / dt) + ω0' ²・ x ・ ・ ・ (107) b (x) ・ G0 ′ ・ iL (t) = (d ² x / dt ² ) + ω0 / Qm ・ (dx / dt) + k (x) · ω0 2 · x ··· (108).

The solution (virtual displacement x (t) with respect to the input signal i (t)) of the linear differential equation represented by the equation (107) is given by the equation (1)
X (t) = L ⁻¹ {G0 ′ / (s ² + s · ω0 ′ / Qm + ω0 ′ ² )} * i (t) (36) .

The velocity v (t) and acceleration a (t) are given by the following equation (6).
In place of (7), v (t) = dx (t) / dt = L ⁻¹ {s · G0 ′ / (s ² + s · ω0 ′ / Qm + ω0 ′ ² )} * i (t) ・ ・ ・ (109) a (t) = d ² x (t) / dt ² = L ^-1 (s ²・ G0 '/ (s ² + s ・ ω0' / Qm + ω0 ' ² ) } * i (t) ... (110).

The signal iL (t) representing the nonlinear correction signal
Is obtained by adding Expressions (36), (109) and (11) to Expression (108).
0), iL (t) = 1 / b (x) · {L ^-1 {s ² / (s ² + ω0 '/ Qm · s + ω0 ' ² )} * i (t) + L ^-1 {ω0 / Qm ・ s / (s ² + ω0' / Qm ・ s + ω0 ' ² )} * i (t) + k (x) ・ ^{ω0 2 · L -1 {1 /} (s 2 + ω0 '/ Qm · s + ω0' 2)} * i (t)} ··· (111).

On the other hand, when SZ transformation (bilinear transformation) is performed using equation (10), equation (3) representing virtual displacement x (t) is obtained.
6) is converted into the following equation: x (n) = G0 ′ · Z ⁻¹ {(hx0 + hx1 · z ⁻¹ · hx2 · z ⁻² ) / (1 + B1 · z ⁻¹ + B2 · z ^-2 )} * i (n) (37).

However, in equation (37), hx0 = (1 / (4 · Fs ² )) / (1 + π / Qm · (f0 ′ / Fs) + π ² · (f0 ′ / Fs) ² ) = hx2 = hx1 / 2 hx1 = (1 / (2 ・ Fs ² )) / (1 + π / Qm ・ (f0 '/ Fs) + π ²・ (f0' / Fs) ² ) hx2 = (1 / (4 Fs ² )) / (1 + π / Qm · (f0 ′ / Fs) + π ² · (f0 ′ / Fs) ² ) (38) Note that the filter coefficients B1 and B2 are expressed by the following equations (30) and (30). (Described later).

Similarly, SZ conversion is performed for the velocity v (t) and acceleration a (t) expressed by the equations (109) and (110). By rewriting equation (111) using these results, the corrected output signal iL (n) and filter coefficient on the Z plane are expressed by the following equations instead of equations (11) to (15). iL (n) = 1 / b (x (n)) Z- ¹ ((C (x) + D (x) z- ¹ + E (x) z- ² ) / (1 + B1z ^-1 + B2 · z ^-2 )} * i (n) (30).

In the equation (30), B1 = (− 2 + 2 · π ² · (f0 ′ / Fs) ² ) / (1 + π / Qm · (f0 ′ / Fs) + π ² · (f0 ′ / Fs) ² ) B2 = (1-π / Qm ・ (f0 '/ Fs) + π ²・ (f0' / Fs) ² ) / (1 + π / Qm ・ (f0 '/ Fs) + π ²・ ( f0 '/ Fs) ² ) (31).

[0129] C (x) = c0 + c (x) = c0 + c1 · x + c2 · x 2 + c3 · x 3 + c4 · x 4 , however, c0 = (1 + π / Qm · (f0 / Fs ) + π ²・ (f0 / Fs) ² ) / (1 + π / Qm ・ (f0 '/ Fs) + π ²・ (f0' / Fs) ² ) c (x) = ((f0 / Fs ・ π ) ²・ (k (x) -1)) / (1+ (f0 ′ / Fs ・ π / Qm) + (f0 ′ / Fs ・ π) ² ) = e (x) = d (x) / 2 ・・ ・ (32). D (x) = d0 + d (x) = d0 + d1 ・ x + d2 ・ x ² + d3 ・ x ³ + d4 ・ x ⁴ , where d0 = (-2 + 2 ・ π ²・ (f0 / Fs) ² ) / (1 + π / Qm ・ (f0 '/ Fs) + π ²・ (f0' / Fs) ² ) d (x) = (2 ・ (f0 / Fs ・ π) ²・ (k (x) −1)) / (1+ (f0 ′ / Fs · π / Qm) + (f0 ′ / Fs · π) ² ) = 2 · c (x) (33). E (x) = e0 + e (x) = e0 + e1 ・ x + e2 ・ x ² + e3 ・ x ³ + e4 ・ x ⁴ , where e0 = (1-π / Qm ・ (f0 / Fs) + π ²・ (f0 / Fs) ² ) / (1 + π / Qm ・ (f0 '/ Fs) + π ²・ (f0 ′ / Fs) ² ) e (x) = ((f0 / Fs ・ π) ²・(k (x) −1)) / (1+ (f0 ′ / Fs · π / Qm) + (f0 ′ / Fs · π) ² ) (34). Therefore, the relationship between the nonlinear filter coefficients c (x), d (x), and e (x) is expressed by the following equation instead of equation (16): c (x) = d (x) / 2 = e (x) (35) In performing coefficient adjustment processing of the signal correction device 82, each pair of nonlinear coefficients, that is, c1 and b1, c2 and b2, c3 and b3, c4 and b4
Is determined in the same manner as in the case of the signal correction device 12 described above, the output signal (sound pressure, acceleration) from the speaker 2
It is sufficient to sequentially search for a value that reduces the second, third, fourth, and fifth harmonic distortion. Note that the nonlinear filter coefficient d (x)
And e (x) are automatically determined from equation (35).

As described above, the signal correction device 82 is applied when the amplifier 28 of the transmission system 14 is a current output type amplifier. Therefore, it is not necessary to consider the nonlinear distortion of the electromagnetic braking, so that the content of the correction processing can be simplified.

[Fourth Embodiment] FIG. 12 shows a signal processing circuit of a signal correction device 92 according to still another embodiment of the present invention. The overall configuration of the signal correction device, the hardware, the configuration of the coefficient adjustment device for adjusting each coefficient of the signal correction device, the coefficient adjustment process, and the signal correction process using the signal correction device are described in FIG. Device 82;
There are many common points.

However, the above-described signal correction device 82 shown in FIG. 11 can be applied to the case where both the linear correction and the non-linear correction are performed, but the signal correction device 92 shown in FIG. It is applied when implementing. Therefore, the signal correction device 92 has the following features in the content of signal processing and the like.

The signal correction device 92 does not perform linear correction. Therefore, in the example of the signal correction device 82 described above, the speaker small parameters f0 ′, Q0 ′, Qm ′, K0 ′, BLO ′, m0 ′, Rdc ′, Rm ′ representing the target characteristics of the integrated transmission system 10 are used. The seal small parameters f0, Q0, Qm, K0, BLO, m0, Rdc, and Rm representing the actual measurement characteristics of No. 2 may be used.

As a result, the filter coefficients used in equation (30) representing the corrected output signal iL on the Z plane are given by equations (31) to (31).
In place of (34), B1 = (− 2 + 2 · π ² · (f0 / Fs) ² ) / (1 + π / Qm · (f0 / Fs) + π ² · (f0 / Fs) ² ) B2 = (1-π / Qm ・ (f0 / Fs) + π ²・ (f0 / Fs) ² ) / (1 + π / Qm ・ (f0 / Fs) + π ²・ (f0 / Fs) ² ) (39).

[0135] C (x) = c0 + c (x) = c0 + c1 · x + c2 · x 2 + c3 · x 3 + c4 · x 4 , however, c0 = (1 + π / Qm · (f0 / Fs ) + π ²・ (f0 / Fs) ² ) / (1 + π / Qm ・ (f0 / Fs) + π ²・ (f0 / Fs) ² ) = 1 c (x) = ((f0 / Fs ・ π ) ²・ (k (x) -1)) / (1+ (f0 / Fs ・ π / Qm) + (f0 / Fs ・ π) ² ) = e (x) = d (x) / 2 ・ ・ ・(40). D (x) = d0 + d (x) = d0 + d1 ・ x + d2 ・ x ² + d3 ・ x ³ + d4 ・ x ⁴ , where d0 = (-2 + 2 ・ π ²・ (f0 / Fs) ² ) / (1 + π / Qm ・ (f0 / Fs) + π ²・ (f0 / Fs) ² ) = B1 d (x) = (2 ・ (f0 / Fs ・ π) ²・ (k (x) −1)) / (1+ (f0 / Fs · π / Qm) + (f0 / Fs · π) ² ) = 2 · c (x) (41). E (x) = e0 + e (x) = e0 + e1 ・ x + e2 ・ x ² + e3 ・ x ³ + e4 ・ x ⁴ , where e0 = (1-π / Qm ・ (f0 / Fs) + π ²・ (f0 / Fs) ² ) / (1 + π / Qm ・ (f0 / Fs) + π ²・ (f0 / Fs) ² ) = B2 e (x) = ((f0 / Fs ・ π) ²・(k (x) -1)) / (1+ (f0 / Fs · π / Qm) + (f0 / Fs · π) ² ) (42). Expression (37) representing the virtual displacement x of the voice coil 4
The filter coefficient used for is expressed by the following equation instead of equation (38): hx0 = (1 / (4 · Fs ² )) / (1 + π / Qm · (f0 / Fs) + π ² · ( f0 / Fs) ² ) = hx2 = hx1 / 2 hx1 = (1 / (2 ・ Fs ² )) / (1 + π / Qm ・ (f0 / Fs) + π ²・ (f0 / Fs) ² ) hx2 = (1 / (4 · Fs ² )) / (1 + π / Qm · (f0 / Fs) + π ² · (f0 / Fs) ² ) (43).

The signal compensator 92 determines that the target characteristic of the integrated transmission system 10 is the characteristic (actual measurement characteristic) itself represented by the lowest resonance frequency f0 of the transmission system 14 and the sharpness of resonance Q0 at the time of a low-level input signal. It is. Therefore, similarly to the signal correction device 62 and the signal correction device 72 described above, it is possible to reduce the nonlinear distortion of the speaker 2 while keeping the actual linear characteristics of the speaker 2 as they are.

[Fifth Embodiment] FIG. 13 shows a signal processing circuit of a signal correction apparatus 102 according to still another embodiment of the present invention. The overall configuration of the signal correction device, hardware,
Regarding the configuration of the coefficient adjustment device for adjusting each coefficient of the signal correction device, the coefficient adjustment process, and the signal correction process using the signal correction device, the signal correction device 12 shown in FIG.
There are many common points.

However, in the signal correction device 12 shown in FIG. 4 described above, the expression (9) of the continuous system on the S plane is expressed by Z
When converting to a discrete system in a plane, S
Although the −Z conversion is performed, the signal correction device 102 is configured to perform the SZ conversion by backward difference approximation (Euler approximation). Therefore, the signal correction device 102
Has the following features in signal processing contents and the like.

In the signal correction device 102, the conversion equation of the SZ conversion is replaced by the conversion equation (10) of the bilinear conversion.
The following equation (conversion equation of backward difference approximation) is used: s = (1-z ^(-1) ) / T = Fs (1-z ^(-1) ) (44).

As a result, the corrected output signal uL and the filter coefficient on the Z plane are expressed by the following equation instead of the equations (11) to (15). UL (n) = (1 / b (x))・ Z ^-1 ((C (x) + D (x) ・ z ^(-1) + E0 ・ z ^(-2) ) / (1 + B1 ・ z ^(-1) + B2 ・ z ^(-2) ) } * u (n) (45).

In the equation (45), B1 = (− 2 + 2 · π / Q0 ′ · (f0 ′ / Fs)) / (1 + 2 · π / Q0 ′ · (f0 ′ / Fs) + 4 · π ²・ (f0 '/ Fs) ² ) B2 = 1 / (1 + 2 ・ π / Q0' ・ (f0 '/ Fs) +4 ・ π ²・ (f0' / Fs) ² ) ・ ・ ・ (46) .

[0142] Moreover, C (x) = c0 + c (x) = c0 + c1 · x + c2 · x 2 + c3 · x 3 + c4 · x 4 , however, c0 = (1 + 2 · π / Q0 · (f0 / Fs) +4 ・ π ²・ (f0 / Fs) ² ) / (1 + 2 ・ π / Q0 '・ (f0' / Fs) +4 ・ π ²・ (f0 '/ Fs) ² ) c (x) = (-2 ・ π / Q0 ・ (f0 / Fs) ・ (1-Q0 / Qm) ・ (b (x) ² -1) +4 ・ π ²・ (f0 / Fs) ²・ (k (x) −1)) / (1 + 2 · π / Q0 ′ · (f0 ′ / Fs) + 4 · π ² · (f0 ′ / Fs) ² ) (47). D (x) = d0 + d (x) = d0 + d1 ・ x + d2 ・ x ² + d3 ・ x ³ + d4 ・ x ⁴ , where d0 = (-2 + 2 ・ π ・ (f0 / Fs)) / (1 + 2 ・ π / Q0 '・ (f0' / Fs) +4 ・ π ²・ (f0 '/ Fs) ² ) d (x) =-2 ・ π / Q0 ・ (f0 / Fs) ・ ( 1-Q0 / Qm) ・ (b (x) ² -1) / (1 + 2 ・ π / Q0 '・ (f0' / Fs) +4 ・ π ²・ (f0 '/ Fs) ² ) (48). E0 = e0 = B2 = 1 / (1 + 2 · π / Q0 ′ · (f0 ′ / Fs) + 4 · π ² · (f0 ′ / Fs) ² ) (49). Therefore, the mutual relationship between the nonlinear filter coefficients c (x) and d (x) is expressed by the following equation instead of the equation (16): d (x) = (2 · B2 + d0) · (1-Q0 / Qm) ・ (b (x) ² -1) c (x) = d (x) +4 ・ π ²・ (f0 '/ Fs) ²・ (k (x) -1) / (1 + 2 ・π / Q0 ′ · (f0 ′ / Fs) + 4 · π ² · (f0 ′ / Fs) ² ) = d (x) + c ′ (x) (50). The virtual displacement x and the like of the voice coil 4 are given by Expression (19),
Instead of (20), it is expressed by the following equation.

X (n) = (A0 · ω0 / Q0 / BL0) · Z ⁻¹ {hx0 / (1 + B1 · z ^(-1) + B2 · z ^(-2) )} * u (n)・ ・ (51) where hx0 = (1 / (Fs ² )) / (1 + 2 ・ π / Q0 ′ ・ (f0 ′ / Fs) +4 ・ π ²・ (f0 ′ / Fs) ² ) ・ ・・ (52).

In performing coefficient adjustment processing of the signal correction device 102, each pair of nonlinear coefficients, ie, b1 and c'1, and b2 and c '
2, b3 and c'3, and b4 and c'4, respectively, in order to determine the secondary, tertiary, and output signals (sound pressure and acceleration) from the speaker 2 as in the case of the signal correction device 12 described above. It is sufficient to sequentially search for a value at which the fourth and fifth harmonic distortions decrease. Note that the nonlinear filter coefficients d (x) and c (x) are automatically determined from Expression (50).

As described above, the signal correction device 102 uses the S-
At the time of Z conversion, SZ conversion is performed by simple backward difference approximation. Therefore, the content of the correction processing can be simplified.

[Sixth Embodiment] FIG. 14 shows a signal processing circuit of a signal correction device 112 according to still another embodiment of the present invention. The overall configuration of the signal correction device, hardware,
The configuration of the coefficient adjustment device for adjusting each coefficient of the signal correction device, the coefficient adjustment process, and the signal correction process using the signal correction device are described with reference to the signal correction device 82 shown in FIG.
And there are many common points.

However, the signal correction device 82 shown in FIG. 11 is configured to use the bilinear transform when performing the SZ conversion.
When performing the Z conversion, the system is configured to use backward difference approximation. Therefore, the signal correction device 112 has the following features in the content of signal processing and the like.

In the signal correction device 112, the conversion formula of the SZ conversion is changed to the conversion formula (10) of the bilinear conversion.
Equation (44) described above is used.

As a result, the corrected output signal iL and the filter coefficient on the Z plane are represented by the following equation instead of the equations (30) to (34): iL (n) = (1 / b (x))・ Z ^-1 {(C (x) + D0 ・ z ^(-1) + E0 ・ z ^(-2) ) / (1 + B1 ・ z ^(-1) + B2 ・ z ^(-2) )} * i (n) (53).

In the equation (53), B1 = (− 2 + 2 · π / Q0 ′ · (f0 ′ / Fs)) / (1 + 2 · π / Q0 ′ · (f0 ′ / Fs) + 4 · π ²・ (f0 '/ Fs) ² ) B2 = 1 / (1 + 2 ・ π / Q0' ・ (f0 '/ Fs) +4 ・ π ²・ (f0' / Fs) ² ) ・ ・ ・ (54) .

[0151] Moreover, C (x) = c0 + c (x) = c0 + c1 · x + c2 · x 2 + c3 · x 3 + c4 · x 4 , however, c0 = (1 + 2 · π / Q0 · (f0 / Fs) +4 ・ π ²・ (f0 / Fs) ² ) / (1 + 2 ・ π / Q0 '・ (f0' / Fs) +4 ・ π ²・ (f0 '/ Fs) ² ) c (x) = 4 ・ π ²・ (f0 / Fs) ²・ (k (x) -1) / (1 + 2 ・ π / Q0 '・ (f0' / Fs) +4 ・ π ²・ (f0 ' / Fs) ² ) (55). D0 = d0 = (-2 + 2 ・ π ・ (f0 / Fs)) / (1 + 2 ・ π / Q0 '・ (f0' / Fs) +4 ・ π ²・ (f0 '/ Fs) ² ) ・・ ・ (56). E0 = e0 = B2 = 1 / (1 + 2 · π / Q0 ′ · (f0 ′ / Fs) + 4 · π ² · (f0 ′ / Fs) ² ) (57). The nonlinear filter coefficients d (x) and e (x) do not exist (d (x) = 0,
e (x) = 0). Therefore, the equation (35) representing the mutual relationship between the nonlinear filter coefficients and the like is not applied.

The virtual displacement x and the like of the voice coil 4 are as follows.
Expressions (37) and (38) are replaced by the following expression.

X (n) = (A0 · ω0 / Q0 / BL0) · Z ⁻¹ {hx0 / (1 + B1 · z ^(-1) + B2 · z ^(-2) )} * i (n)・ ・ (58) where hx0 = (1 / (Fs ² )) / (1 + 2 ・ π / Q0 '・ (f0' / Fs) +4 ・ π ²・ (f0 '/ Fs) ² ) ・ ・・ (59).

In performing the coefficient adjustment process of the signal correction device 112, each pair of nonlinear coefficients, ie, b1 and c1, b2 and c2,
To determine b3 and c3 and b4 and c4, respectively, as in the case of the above-described signal correction device 12, the second, third, fourth and fifth harmonics of the output signal (sound pressure and acceleration) from the speaker 2 are determined. It is only necessary to sequentially search for a value that reduces the wave distortion.

As described above, the signal correction device 112 is configured to perform SZ conversion by simple backward difference approximation when converting the description system of the continuous system on the S plane into the discrete system on the Z plane. . Therefore, the content of the correction processing can be simplified.

[Seventh Embodiment] FIG. 16 shows a signal processing circuit of a signal correction device 122 according to still another embodiment of the present invention. The overall configuration of the signal correction device, hardware,
Regarding the configuration of the coefficient adjustment device for adjusting each coefficient of the signal correction device, the coefficient adjustment process, and the signal correction process using the signal correction device, the signal correction device 12 shown in FIG.
There are many common points.

However, in the signal correction device 12 shown in FIG. 4, the non-linearity of the force coefficient and the stiffness is focused on as the cause of the nonlinear distortion. The non-linearity of the excluded air volume at the edge 8e (see FIG. 15) of the speaker diaphragm 8 is also taken into consideration. Therefore, the signal correction device 122 has the following features in signal processing contents and the like.

As described above, the reproduction sound pressure of the speaker 2 is
It also depends on the volume of air eliminated with the movement of the speaker diaphragm 8. Here, in an ideal speaker, the effective vibration area Sa of the speaker is always constant. Therefore, the density of air is ρ0, the acceleration of the voice coil 4 is a = d ² x / dt ² , and the air volume exclusion amount of the speaker is Wa (= Sa ·
x), the sound pressure Pa of the speaker 2 at the distance r
(T) is expressed by the following equation: Pa (t) = ρ0 / (2 · π · r) · Sa · a = ρ0 / (2 · π · r) · Sa · (d ² x / dt ² ) ... (65).

= Ρ0 / (2 · π · r) · (d ² (Wa) / dt ² ) As can be seen from the above equation, the sound pressure Pa shows linearity with respect to the displacement x of the cone 8d. It is distortion.

However, in an actual speaker, distortion occurs due to the non-linearity of the excluded air volume at the edge. FIG.
As shown in FIG. 7, a cone 8d constituting the speaker diaphragm 8
Is moved from P0 to P1, and from P0 to P2
, The excluded air volume We of the edge 8e is different even if the moving distance is the same (We1 ≠ We).
2).

As described above, the excluded air volume W of the edge 8e is obtained.
e indicates that there is little difference depending on the moving direction of the voice coil 4 when the displacement x of the voice coil 4 of the speaker 2 is extremely small, but when the displacement x of the voice coil 4 is large to some extent, It is known that differences due to directions appear. That is, the excluded air volume at the edge has a non-linearity. This is one of the causes of the nonlinear distortion of the speaker 2.

The displacement air volume We (x), which varies depending on the displacement x, and the force coefficient that varies depending on the displacement x
BL (x) and stiffness K (x) are expressed by the following equation instead of the above equation (1). BL (x) = BL0 · b (x) = BL0 · (1 + b1 · x + b2 ・ x ² ) K (x) = K0 ・ k (x) = K0 ・ (1 + k1 ・ x + k2 ・ x ² ) We (x) = We0 + W1 ・ x + W2 ・ x ²・ ・ ・(60) In equation (60), b1, b2, k1, k2, W1, and W2 are nonlinear coefficients.

Assuming that the area of the cone 8d is Sd and the air volume exclusion of the cone 8d is Wd (= Sd · x), the sound pressure Pd from the cone 8d at the point of the distance r from the speaker 2 is expressed by the following equation. Pd (t) = ρ0 / (2 ・ π ・ r) ・ Sd ・ a = ρ0 / (2 ・ π ・ r) ・ Sd ・ (d ² x / dt ² ) = ρ0 / (2 ・ π ・ r ) · (D ² (Wd) / dt ² ) (61).

As can be seen from the above equation, the sound pressure Pd from the cone 8d is proportional to the second-order time derivative of the excluded air volume Wd.

On the other hand, the sound pressure Pe radiated from the edge 8e at a point at a distance r from the speaker 2 is expressed by the following equation: Pe (t) = ρ0 / (2 · π · r) · (d ² (We ( x)) / dt ² ) (62).

The sound pressure Pa at a distance r from the speaker 2 is represented by the sum of the sound pressure Pd from the cone 8d and the sound pressure Pe radiated from the edge 8e. Assuming (x), equation (61),
According to (62), Pa (t) = Pd (t) + Pe (t) = ρ0 / (2 · π · r) · (d ² (Wd + We (x)) / dt ² ) = ρ0 / (2 · Π · r) · (d ² (Wa (x)) / dt ² ) (63).

The excluded air volume Wa (x) of the entire speaker 2
Indicates a property that changes with the displacement x due to the presence of the edge 8e. Therefore, the total excluded air volume Wa
(X) is the effective vibration area Sa with the displacement x as a parameter.
Considering the product of (x) and displacement x, Wa (x) = Sa (x) · x where Sa (x) = Sa · s (x) = Sa (1 + s1 · x + s2 · x ² ) ... (112).

Therefore, from equations (63) and (112), Pa (t) = ρ0 / (2 · π · r) · (d ² (Sa (x) · x) / dt ² ) = ρ0 / (2 • π · r) · Sa · (d ² (s (x) · x) / dt ² ) (64).

As can be seen from the equation (64), even if the displacement x of the voice coil 4 is strain-free, the effective vibration area Sa (x) is reduced due to the nonlinearity of the excluded air volume We (x) of the edge 8e. Changes, and the sound pressure Pa causes distortion.

By the way, assuming that a non-linear correction for distorting the displacement x is introduced, the above equations (19) and (1)
1) Similarly, virtual displacement x (n) and correction output signal uL
(N) is represented by the following equation: x (n) = G0 · Z ⁻¹ {(hx0 + hx1 · z ⁻¹ + hx2 · z ⁻² ) / (1 + B1 · z ⁻¹ + B2 · z ^{− 2} )} * u (n) ・ ・ ・ (66) uL (n) = 1 / b (x) ・ Z ^-1 {(C (x) + D (x) ・ z ^-1 + E (x) ・z− ² ) / (1 + B1 · z− ¹ + B2 · z− ² )} * u (n) (67).

In order to perform the nonlinear correction of the edge 8e, the correction output signal uL (n) represented by the equation (67) is not sufficient, and it is necessary to perform the correction further. The correction output signal uL '(n) in the case where the correction is further performed can be expressed by the following equation in consideration of equations (65) and (64). UL' (n) = 1 / s (x)・ UL (n) = 1 / (b (x) ・ s (x)) ・ Z ^-1 ((C (x) + D (x) ・ z ^-1 + E (x) ・ z ^-2 ) / ( 1 + B1 · z ⁻¹ + B2 · z ⁻² )} * u (n) (69).

That is, in order to perform nonlinear correction in consideration of the nonlinearity of the edge 8e, the correction output signal uL
(N) and the corrected output signal uL ′ (n)
4 (see FIG. 2).

Here, n (x) = {1+ (1-Q0 / Qm) · (b (x) ² −1} = 1 = 1 + n1 · x + n2 · x ² g (x) = (1 / b (x)) · (1 / s (x)) = 1 + g1 · x + g2 · x 2 ··· (6
8) Equation (69) is expressed as follows: uL ′ (n) = g (x) · Z ⁻¹ ｛(C (x) + D (x) · z ⁻¹ + E
(X) .z- ² ) / (1 + B1.z- ¹ + B2.z- ² ) @ * u (n) (70).

In the equation (70), B1 = (− 2 + 2 · π ² · (f0 ′ / Fs) ² ) / (1 + π / Q0 ′ · (f0 ′ / Fs) + π ² · (f0 ′ / Fs) ² ) B2 = (1-π / Q0 '・ (f0' / Fs) + π ²・ (f0 '/ Fs) ² ) / (1 + π / Q0' ・ (f0 '/ Fs) + π ² · (f0 ′ / Fs) ² ) (71).

[0175] Moreover, C (x) = c0 + c (x) = c0 + c1 · x + c2 · x 2 + c3 · x 3 + c4 · x 4 except c0 = (1 + π / Q0 · (f0 / Fs) + π ²・ (f0 / Fs) ² ) / (1 + π / Q0 ′ ・ (f0 ′ / Fs) + π ²・ (f0 ′ / Fs) ² ) c (x) = ((n (x ) -1) / Q0 ・ (f0 / Fs) + π ²・ (f0 / Fs) ²・ (k (x) -1)) / (1 + π / Q0 '・ (f0' / Fs) + π ² (F0 '/ Fs) ² ) (72). D (x) = d0 + d (x) = d0 + d1 ・ x + d2 ・ x ² + d3 ・ x ³ + d4 ・ x ⁴ , where d0 = (-2 + 2 ・ π ²・ (f0 / Fs) ² ) / (1 + π / Q0 '・ (f0' / Fs) + π ²・ (f0 '/ Fs) ² ) d (x) = (2 ・ π ²・ (f0 / Fs) ²・ (k ( x) -1)) / (1 + π / Q0 ′ · (f0 ′ / Fs) + π ² · (f0 ′ / Fs) ² ) = c (x) + e (x) (73). E (x) = e0 + e (x) = e0 + e1 ・ x + e2 ・ x ² + e3 ・ x ³ + e4 ・ x ⁴ , where e0 = (1-π / Q0 ・ (f0 / Fs) + π ²・ (f0 / Fs) ² ) / (1 + π / Q0 '・ (f0' / Fs) + π ²・ (f0 '/ Fs) ² ) e (x) = (-(n (x) -1 ) / Q0 ・ (f0 / Fs) + π ²・ (f0 / Fs) ²・ (k (x) -1)) / (1 + π / Q0 '・ (f0' / Fs) + π ²・ (f0 '/ Fs) ² ) ... (74). On the other hand, the relationship among the nonlinear parameters c (x), d (x), and e (x) is expressed by the following equation from equations (72) to (74): d (x) = c (x) + e (x) (75) d (x) is proportional to (k (x) -1). The virtual displacement x and the like of the voice coil 4 are calculated by the equations (19), (20), and ( 12)
Can be used as it is.

In performing the coefficient adjustment processing of the signal correction device 122 (see FIG. 16), each combination of nonlinear coefficients, that is,
c1 and e1 and g1, c2 and e2 and g2, c3 and e3 and g3, c4 and e4 and g4,
In order to determine each of them, as in the case of the signal correction device 12 described above, the output signal (sound pressure, acceleration) 2
Next, third, fourth, and fifth harmonic distortion values may be sequentially searched for. Note that as shown in equation (75), c
d (x) can be automatically determined from two parameters (x) and e (x).

On the other hand, based on equations (72) to (74),
The signal correction device 132 shown in FIG. 17 can also be configured. In performing the coefficient adjustment process of the signal correction device 132, each combination of nonlinear coefficients, that is, k1, n1, and g1, k2 and n2,
To determine g2, k3, n3, and g3, and k4, n4, and g4, respectively, the second, third, fourth, and fifth order of the output signal (sound pressure and acceleration) from the speaker 2 are determined in the same manner as described above. What is necessary is just to search in order for the value at which the harmonic distortion decreases. Note that equation (72)
As shown in (74), c (x), d (x), and e (x) can be automatically determined from two parameters k (x) and n (x).

As described above, in the signal correction device 122 and the signal correction device 132, the correction is further performed in consideration of the non-linearity of the excluded air volume at the edge 8e of the speaker diaphragm 8 of the speaker 2. Therefore, when applied to a small speaker in which the ratio of the edge area to the entire effective vibration area is relatively high, it is possible to obtain a high effect in reducing nonlinear distortion.

In this embodiment, the force coefficient,
Although attention is paid to the three nonlinearities of the stiffness and the excluded air volume of the edge, and all the nonlinearities are reduced, any one or more of these nonlinearities may be reduced. For example, it may be configured to reduce only the non-linearity of the excluded air volume at the edge.

[Eighth Embodiment] FIG. 18 shows a signal processing circuit of a signal correction device 142 according to still another embodiment of the present invention. The overall configuration of the signal correction device, hardware,
The configuration of the coefficient adjustment device for adjusting each coefficient of the signal correction device, the coefficient adjustment process, and the signal correction process using the signal correction device are described with reference to the signal correction device 132 shown in FIG.
And there are many common points.

However, in the signal correction device 132 shown in FIG. 17 described above, when converting the description formula of the continuous system on the S plane into the discrete system on the Z plane, the S-
Although the Z conversion is performed, the signal correction device 142
SZ conversion is performed by backward difference approximation (Euler approximation). Therefore, the signal correction device 142 has the following features in the content of signal processing and the like.

In the signal correction device 142, the conversion equation of the SZ conversion is changed to the conversion equation (10) of the bilinear conversion.
Equation (44) described above is used. As a result, the corrected output signal uL ′ and the filter coefficient on the Z plane are calculated by the equation (70).
UL ′ (n) = g (x) · Z ⁻¹ {(C (x) + D (x) · z ⁽⁻¹⁾ + E0 · z ^{( -2)} ) / (1 + B1z ^(-1) + B2z ^(-2) )} * u (n) (76).

In the equation (76), B1 = (− 2 + 2 · π / Q0 ′ · (f0 ′ / Fs)) / (1 + 2 · π / Q0 ′ · (f0 ′ / Fs) + 4 · π ²・ (f0 '/ Fs) ² ) B2 = 1 / (1 + 2 ・ π / Q0' ・ (f0 '/ Fs) +4 ・ π ²・ (f0' / Fs) ² ) ・ ・ ・ (77) .

[0184] Moreover, C (x) = c0 + c (x) = c0 + c1 · x + c2 · x 2 + c3 · x 3 + c4 · x 4 , however, c0 = (1 + 2 · π / Q0 · (f0 / Fs) +4 ・ π ²・ (f0 / Fs) ² ) / (1 + 2 ・ π / Q0 '・ (f0' / Fs) +4 ・ π ²・ (f0 '/ Fs) ² ) c (x) = (2 ・ π / Q0 ・ (f0 / Fs)) ・ (n (x) -1) +4 ・ π ²・ (f0 / Fs) ²・ (k (x) -1)) / ( 1 + 2 · π / Q0 ′ · (f0 ′ / Fs) + 4 · π ² · (f0 ′ / Fs) ² ) (78). D (x) = d0 + d (x) = d0 + d1 ・ x + d2 ・ x ² + d3 ・ x ³ + d4 ・ x ⁴ , where d0 = (-2 + 2 ・ π ・ (f0 / Fs)) / (1 + 2 ・ π / Q0 '・ (f0' / Fs) +4 ・ π ²・ (f0 '/ Fs) ² ) d (x) = (-2 ・ π / Q0 ・ (f0 / Fs)) (N (x) -1) / (1 + 2 · π / Q0 ′ · (f0 ′ / Fs) + 4 · π ² · (f0 ′ / Fs) ² ) (79). E0 = e0 = B2 = 1 / (1 + 2 · π / Q0 ′ · (f0 ′ / Fs) + 4 · π ² · (f0 ′ / Fs) ² ) (80). As can be seen from equations (78) and (79), d (x) is (n (x) −
In proportion to 1), c (x) is represented by the sum of d (x) and that proportional to (k (x) -1).

Note that the virtual displacement x and the like of the voice coil 4 are as follows:
Equations (51) and (52) can be used as they are.

In performing the coefficient adjustment processing of the signal correction device 142, each combination of nonlinear coefficients, that is, k1, n1, g1, and k2
And n2 and g2, k3 and n3 and g3, and k4 and n4 and g4, respectively, are determined in the same manner as described above by the second, third and fourth order of the output signal (sound pressure and acceleration) from the speaker 2. A value at which the fifth harmonic distortion is reduced may be searched in order. Note that, as described above, k (x), n according to equations (78) and (79).
c (x) and d (x) can be automatically determined from the two parameters (x).

As described above, in the signal correction device 142, when the description system of the continuous system on the S plane is converted into the discrete system on the Z plane, SZ conversion is performed by simple backward difference approximation. I have. Therefore, when the correction is performed in consideration of the non-linearity of the excluded air volume at the edge 8e of the speaker diaphragm 8 of the speaker 2, the content of the correction processing can be simplified.

[Other Embodiments] In each of the above-described embodiments, as shown in FIG.
The displacement filter 16 includes a return section 22a that includes a return path and corresponds to an upstream portion of signal processing, and a non-return section 22b that does not include a return path and corresponds to a downstream portion of signal processing.
6b, and the correction filter 22 and the displacement filter 16 share the repetition unit 22a, but the filter coefficient of the repetition unit of the correction filter 22 and the filter coefficient of the repetition unit of the displacement filter 16 are the same. Then, there is no need to share the recursive parts of these filters. Further, for the displacement filter 16, the recursive section does not necessarily need to be arranged in the upstream part of the signal processing. FIG. 19 illustrates an example in which the correction filter 22 and the displacement filter 16 do not share a cyclic portion. In the example of FIG. 19, the non-repeating portion 16b is disposed in the upstream portion of the signal processing of the displacement filter 16, and the recurring portion 16a is disposed in the downstream portion.

The case where the filter coefficient of the cyclic portion of the correction filter and the filter coefficient of the cyclic portion of the displacement filter are the same has been described as an example, but the filter coefficient of the cyclic portion of the correction filter and the filter coefficient of the cyclic portion of the displacement filter are described. What are filter coefficients?
It need not be the same.

Further, in each of the above embodiments, FIG.
As shown in FIG.
In addition to the above, an amplitude adjusting unit 24 is further provided,
8 determines not only the filter coefficient of the correction filter 22 but also the amplitude adjustment coefficient of the amplitude adjustment unit 24 based on the predicted virtual displacement x, and the amplitude adjustment unit 24 uses the determined amplitude adjustment coefficient. , The amplitude of the signal is adjusted, but the present invention is not limited to this. FIG.
As shown in (1), only the correction filter 22 may be provided as a signal correction unit, and the correction coefficient determination unit 18 may be configured to determine only the filter coefficient of the correction filter 22 based on the predicted virtual displacement x. In the case of such a configuration, the filter coefficient C (x) of the correction filter 22,
D (x) and E (x) are coefficients including adjustment of the signal amplitude.

In each of the above embodiments, the coefficient relating to the nonlinear filter coefficient of the acyclic portion of the correction filter is reduced so that each harmonic distortion of the response signal of the integrated transmission system corresponding to the given reference signal is reduced. However, when two signals are input, the cross-modulation distortion of the response signal is reduced.
The coefficient may be adjusted. Further, the coefficient may be adjusted so that both harmonic distortion and cross modulation distortion are reduced.

In each of the above embodiments, the case where the correction filter is a second-order IIR type filter has been described as an example. However, the correction filter may be a third-order or higher-order IIR type filter. The correction filter is not necessarily an IIR
The filter need not be a type filter, but may be a digital filter of second or higher order having a feedback path.

Also, a case has been described as an example where a displacement filter, which is a second-order IIR filter, is used as the virtual physical quantity prediction means. However, the virtual physical quantity prediction means is not limited to this. Further, the virtual physical quantity prediction means does not necessarily need to be a second-order or higher-order digital filter having a feedback path.

Although the virtual displacement of the voice coil of the electrodynamic loudspeaker is used as the virtual physical quantity, for example, the virtual velocity, the virtual acceleration,
The current value and voltage value of the voice coil can also be used.

In each of the above embodiments, DS
P30 executes a correction process according to a correction program (including data) stored in an internal memory (not shown). The correction program is written into the internal memory by using the controller 42. The correction program may be stored in a flexible disk (FD) and installed in the internal memory of the DSP via the controller 42 or the like as necessary. In addition,
In addition to the flexible disk, a computer-readable storage medium storing a correction program such as a CD-ROM or an IC card may be installed in the internal memory of the DSP. Furthermore, you may make it download using a communication line.

In each of the above embodiments, the controller 42 performs coefficient adjustment processing according to an adjustment program stored in a hard disk (not shown). This adjustment program is read from a flexible disk storing the adjustment program via a flexible disk drive (not shown) and installed on the hard disk. Note that, other than the flexible disk, a hard disk may be installed from a computer-readable storage medium storing an adjustment program such as a CD-ROM or an IC card. Furthermore, you may make it download using a communication line.

In each of the above embodiments, the adjustment program is installed from the flexible disk to the hard disk, so that the adjustment program stored in the flexible disk is indirectly executed by the computer (controller 42). . However, without being limited to this, the adjustment program stored in the flexible disk may be directly executed from the FDD 14. In addition, as an adjustment program that can be executed by a computer, not only a program that can be directly executed by simply installing it as it is, but also
This includes those that need to be converted to another form once (for example, decompressing data that has been compressed) and those that can be executed in combination with other module parts.

Further, in the above-described embodiment, the case where each function of the signal correction device 12 and the like is realized by using the DSP 30 has been described as an example, but a part or all of each function is replaced by a computer other than the DSP. It can also be configured to be realized by using. Further, a part or all of the functions may be configured to be realized using hardware logic.

In each of the above embodiments, the case where the transmission system includes an electrodynamic speaker has been described as an example, but the present invention is not limited to this. For example, if the transmission system has an actuator,
The present invention can be applied to correction of an input signal to an actuator so as to reduce nonlinearity of the actuator and obtain a desired operation. That is, the present invention can be applied to signal correction devices in general.

[Brief description of the drawings]

FIG. 1 shows a signal correction device 12 according to an embodiment of the present invention.
FIG. 2 is a block diagram of an integrated transmission system 10 including:

FIG. 2 is a diagram showing an example of a hardware configuration of the integrated transmission system 10 when each function of the signal correction device 12 is realized using a DSP 30.

FIG. 3 is a drawing in which an electrodynamic loudspeaker 2 constituting a transmission system 14 is represented by a physical model (lumped parameter model).

FIG. 4 is a signal processing circuit diagram showing the content of signal processing of the signal correction device 12;

FIG. 5 is a signal processing circuit diagram showing the contents of signal processing of the signal correction device 52.

FIG. 6 is a diagram illustrating an overall configuration of a coefficient adjustment device 32 for adjusting each coefficient of the signal correction device 12.

FIG. 7 shows a signal correction device 12 using a coefficient adjustment device 32;
9 is a flowchart illustrating a flow of a process when the coefficient adjustment process is performed.

FIG. 8 is a flowchart illustrating an example of a processing flow when performing signal correction processing using the signal correction device 12 that has been subjected to coefficient adjustment processing.

FIG. 9 is a signal correction device 6 according to another embodiment of the present invention.
2 is a diagram illustrating a second signal processing circuit.

FIG. 10 is a diagram showing a signal processing circuit of the signal correction device 72.

FIG. 11 is a diagram showing a signal processing circuit of a signal correction device 82 according to still another embodiment of the present invention.

FIG. 12 is a diagram showing a signal processing circuit of a signal correction device 92 according to still another embodiment of the present invention.

FIG. 13 is a diagram showing a signal processing circuit of a signal correction device 102 according to still another embodiment of the present invention.

FIG. 14 is a diagram showing a signal processing circuit of a signal correction device 112 according to still another embodiment of the present invention.

FIG. 15 is a drawing for explaining the non-linearity of the excluded air volume at the edge 8e of the speaker diaphragm 8;

FIG. 16 is a diagram showing a signal processing circuit of a signal correction device 122 according to still another embodiment of the present invention.

FIG. 17 is a diagram showing a signal processing circuit of the signal correction device 132.

FIG. 18 is a diagram showing a signal processing circuit of a signal correction device 142 according to still another embodiment of the present invention.

FIG. 19 is a view schematically showing a signal processing circuit of a signal correction device according to still another embodiment of the present invention.

FIG. 20 is a drawing schematically showing a signal processing circuit of a signal correction device according to still another embodiment of the present invention.

FIG. 21 is a diagram showing data obtained by comparing the frequency response of sound pressure when linear correction is not performed with the frequency response of sound pressure when linear correction according to the present invention is performed.

FIG. 22 shows a case where Si is used as an input signal to the integrated transmission system 10.
5 is a diagram showing data obtained by measuring the level of a harmonic appearing in output sound pressure using n waves. FIG. 22A is data showing a harmonic level when non-linear correction is not performed, and FIG. 22B is data showing a harmonic level when non-linear correction is performed according to the present invention.

FIG. 23 is a drawing showing a cross-sectional configuration of the electrodynamic speaker 2.

FIG. 24 shows a voice coil 4 in the electrodynamic speaker 2.
3 is a diagram showing data of frequency response such as acceleration of the vehicle.

FIG. 25 shows a voice coil 4 in the electrodynamic speaker 2.
3 is a diagram showing data on the frequency response of the harmonic component of the sound pressure of FIG.

[Explanation of symbols]

 10 Integrated transmission system 12 Signal correction device 14 Transmission system 16 Displacement filter 18 Correction coefficient determination unit 20 Signal correction means 22... Correction filter 24... Amplitude adjustment unit

[Procedure amendment]

[Submission date] October 2, 1996

[Procedure amendment 1]

[Document name to be amended] Drawing

[Correction target item name] All figures

[Correction method] Change

[Correction contents]

FIG. 3

FIG. 7

FIG.

FIG.

FIG. 23

FIG.

FIG. 2

FIG. 6

FIG. 4

FIG. 5

FIG. 8

FIG. 21

FIG. 24

FIG. 9

FIG. 10

FIG. 11

FIG.

FIG. 13

FIG. 14

FIG. 16

FIG.

FIG.

FIG.

FIG.

FIG. 25

Claims (18)

[Claims] 1. A signal correction device for correcting an input signal and providing the same to a transmission system such that characteristics of an integrated transmission system in which the signal correction device and the transmission system are integrated approaches a predetermined target characteristic, comprising a feedback path. Signal correction means having a correction filter constituted by a digital filter of second or higher order; virtual physical quantity prediction means for predicting a predetermined virtual physical quantity representing the state of the integrated transmission system based on an input signal; Correction coefficient determining means for determining at least a part of the filter coefficient of the correction filter based on the correction coefficient, wherein the correction filter corrects the input signal using the determined filter coefficient. apparatus. 2. The signal correction device according to claim 1, wherein the signal correction unit further includes an amplitude adjustment unit that adjusts an amplitude of the signal, and wherein the correction coefficient determination unit determines, based on the predicted virtual physical quantity, Also determine the amplitude adjustment coefficient of the amplitude adjustment unit, the amplitude adjustment unit, using the determined amplitude adjustment coefficient,
The amplitude of a signal before correction by the correction filter or a signal after the correction is adjusted. 3. The signal correction device according to claim 1, wherein the virtual physical quantity predicting unit includes a virtual physical quantity filter configured by a second-order or higher-order digital filter having a feedback path. The physical quantity filter predicts a virtual physical quantity based on an input signal using a filter coefficient determined based on a target characteristic of the integrated transmission system. 4. The signal correction apparatus according to claim 3, wherein the correction filter includes a return path and corresponds to an upstream portion of the signal processing, and a non-cyclic filter including no return path and corresponds to a downstream portion of the signal processing. The virtual physical quantity filter includes a cyclic section including a feedback path, and a non-cyclic section including no feedback path. The filter coefficient is the same, and the filter coefficient of the non-recursive portion of the correction filter is a filter coefficient determined based on the virtual physical quantity. 5. The signal correction device according to claim 4, wherein a filter coefficient of a non-cyclic portion of the correction filter is a linear filter coefficient that does not change depending on the virtual physical quantity, and a non-linear filter coefficient that changes depending on the virtual physical quantity. The characteristic that can be expressed by the sum with the filter coefficient. 6. The signal correction device according to claim 5, wherein the transmission system converts a correction output of the signal correction device into an analog signal and amplifies the converted output;
An electrodynamic loudspeaker driven based on the amplified output, wherein the target characteristic of the integrated transmission system is a characteristic including a characteristic of the integrated transmission system represented by a seal small parameter to be targeted; Is a virtual displacement of the voice coil of the electrodynamic loudspeaker calculated assuming that the integrated transmission system has target characteristics, and the virtual physical quantity filter is a virtual displacement in consideration of an amplification gain of the drive system. Wherein both the correction filter and the displacement filter are second-order IIR (infinite impulse response) filters. 7. The signal correction device according to claim 6, wherein the amplifier is a voltage output type amplifier, and c (x) is a non-delay nonlinear filter coefficient d. (X): a first-order delay nonlinear filter coefficient e (x): a second-order delay nonlinear filter coefficient, where c (x) + e (x) = d (x). 8. The signal correction device according to claim 6, wherein the amplifier is a current output type amplifier, and c (x): a non-delay nonlinear filter coefficient d (X): a first-order delay nonlinear filter coefficient e (x): a second-order delay nonlinear filter coefficient, satisfying the following relationship: c (x) = e (x) = d (x) / 2 thing. 9. The signal correction apparatus according to claim 6, wherein e (x) = 0 when e (x) is a second-order delay nonlinear filter coefficient among three nonlinear filter coefficients of a non-cyclic portion of the correction filter. Characterized by the following. 10. The signal correction device according to claim 6, wherein the characteristic of the integrated transmission system represented by the target seal small parameter is the characteristic itself represented by the seal small parameter of the transmission system at the time of a low level input signal. It is characterized by being. 11. The signal correction device according to claim 6, wherein the characteristic of the integrated transmission system represented by the target seal small parameter is a characteristic represented by the seal small parameter of the transmission system at the time of a low level input signal. Features characterized by desired modifications. 12. The signal correction device according to claim 6, wherein a nonlinearity of a transmission system caused by a force coefficient and a stiffness that changes depending on a displacement of a voice coil of the electrodynamic speaker is reduced. thing. 13. The signal correction device according to claim 6, wherein the nonlinearity of the transmission system due to the excluded air volume at the edge of the electrodynamic loudspeaker, which varies depending on the displacement of the voice coil of the electrodynamic loudspeaker, is reduced. What is characterized by doing. 14. A signal correction method for correcting an input signal so as to bring a characteristic of an integrated transmission system including a transmission system closer to a predetermined target characteristic and giving the input signal to the transmission system, wherein a digital filter of second or higher order having a feedback path. And a predetermined virtual physical quantity representing the state of the integrated transmission system is predicted based on the input signal. Based on the predicted virtual physical quantity, at least a part of the filter coefficient of the correction filter is provided. Wherein the correction filter corrects an input signal using the determined filter coefficient. 15. A signal adjustment device for adjusting each coefficient in the signal correction device according to any one of claims 1 to 13 or the signal correction method according to claim 14, wherein a reference signal provided to the integrated transmission system is provided. Reference signal generating means for generating a response signal of the integrated transmission system corresponding to the given reference signal, and measuring the response signal of the integrated transmission system, based on the reference signal and the response signal, the characteristic of the integrated transmission system to a predetermined target characteristic. An adjustment control means for adjusting each of the coefficients so as to approach each other. 16. A coefficient adjusting device for a signal correction device according to claim 15, wherein said adjustment control means reduces each harmonic distortion or cross modulation distortion of a response signal of an integrated transmission system corresponding to a given reference signal. In this manner, a coefficient relating to a nonlinear filter coefficient of a non-recursive portion of the correction filter is adjusted. 17. A coefficient adjustment method for adjusting each coefficient in the signal correction device according to any one of claims 1 to 13 or the signal correction method according to claim 14, wherein a reference signal provided to the integrated transmission system is provided. The response signal of the integrated transmission system corresponding to the given reference signal is measured, and the coefficients are adjusted based on the reference signal and the response signal such that the characteristics of the integrated transmission system approach predetermined target characteristics. A coefficient adjustment method for a signal correction device. 18. A computer-readable storage medium storing a computer-executable program, wherein the program implements the apparatus or method according to any one of claims 1 to 17. Storage medium.