JP2014204213A - Digital filter and filter characteristic modification method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a digital filter that can suppress the generation of a discontinuous change in a filtering result when a filter characteristic is modified by switching a coefficient vector.SOLUTION: When switching from a first coefficient vector to a second coefficient vector, a coefficient switching section 240A subtracts the first coefficient vector from the second coefficient vector to calculate a difference vector. Subsequently, on the basis of the calculated difference vector, an update amount vector for stepwise switching from the first coefficient vector to the second coefficient vector is calculated. On the basis of the update amount vector thus calculated, the coefficient switching section 240A switches stepwise from the first coefficient vector to the second coefficient vector.

Description

  The present invention relates to a digital filter, a filter characteristic changing method, a filter characteristic changing program, and a recording medium on which the filter characteristic changing program is recorded.

  Accompanying advances in digital technology in recent years, acoustic devices that perform digital signal processing have become widespread. An acoustic device that performs such digital signal processing generally includes a digital filter in order to realize signal processing according to a user's acoustic signal processing designation or the like (refer to Patent Document 1 below). Called "example").

  In the conventional technique, when a user gives an acoustic signal processing command, the parameter setting unit outputs a filter coefficient (coefficient vector) selection signal to a digital filter FIR (Finite Impulse Response) filter. To do. Then, the FIR filter reads a filter coefficient corresponding to the selection signal from the storage unit, and executes a filtering process with a filter characteristic determined by the read filter coefficient.

JP 2005-128174 A

  Although the details of the switching of the filter coefficients used by the FIR filter are not described in the conventional technology described above, all of the filter coefficients are changed to new values to be used at the time of switching the filter coefficients. It is common to switch.

  When such filter coefficient switching is performed to switch all the filter coefficients at once, the output signal of the FIR filter may change discontinuously greatly due to the discontinuity of the filter characteristics before and after the filter coefficient switching. is there. When such a change occurs in an acoustic signal, a so-called pop sound is generated, which makes the user feel uncomfortable in hearing.

  Therefore, it is conceivable to perform a mute process when changing the filter characteristics by switching the filter coefficient. However, if such a mute process is performed, a silent period will occur, and this will make the user feel uncomfortable in hearing.

  For this reason, when the filter characteristics of the digital filter are changed, a technique capable of suppressing a sense of incongruity caused by the discontinuity of the output signal before and after the change is desired. Meeting this requirement is one of the problems to be solved by the present invention.

  The present invention has been made in view of the above circumstances, and provides a digital filter and a filter characteristic changing method capable of suppressing the occurrence of discontinuous changes in the output signal when changing the filter characteristics. Objective.

  The invention according to claim 1 is a digital filter for performing filtering processing on an input digital signal with a characteristic determined by a coefficient vector, wherein the first coefficient is changed when switching from the first coefficient vector to the second coefficient vector. After calculating a difference vector obtained by subtracting the vector from the second coefficient vector, an update amount vector for stepwise switching from the first coefficient vector to the second coefficient vector is obtained based on the difference vector. A digital filter comprising: a calculating unit that calculates; and a switching unit that switches the coefficient vector stepwise based on the update amount vector.

  A seventh aspect of the present invention is a filter characteristic changing method used in a digital filter that performs filtering processing on an input digital signal with characteristics determined by a coefficient vector, from a first coefficient vector to a second coefficient vector. When switching, the difference vector obtained by subtracting the first coefficient vector from the second coefficient vector is calculated, and then the switching from the first coefficient vector to the second coefficient vector is stepwise based on the difference vector. A filter characteristic changing method comprising: a calculation step of calculating an update amount vector to be performed at once; and a switching step of stepwise switching of the coefficient vector based on the update amount vector.

  According to an eighth aspect of the present invention, there is provided a computer having a digital filter that performs a filtering process on an input digital signal with a characteristic determined by a coefficient vector, wherein the filter characteristic changing method according to the seventh aspect is executed. This is a filter characteristic changing program.

  According to the ninth aspect of the present invention, the filter characteristic changing program according to the eighth aspect is recorded so as to be readable by a computer provided in a digital filter that performs a filtering process with a characteristic determined by a coefficient vector on an input digital signal. It is a recording medium characterized by the above.

1 is a block diagram schematically showing a configuration of an audio device including a digital filter according to a first embodiment of the present invention. It is a block diagram which shows the structure of the digital filter which concerns on 1st Embodiment. It is a flowchart for demonstrating the coefficient switching process by the coefficient switching part of FIG. It is a block diagram which shows the structure of the digital filter which concerns on 2nd Embodiment. It is a flowchart for demonstrating the coefficient switching process by the coefficient switching part of FIG. It is a block diagram which shows the structure of the digital filter which concerns on 3rd Embodiment. It is a flowchart for demonstrating the coefficient switching process by the coefficient switching part of FIG.

  Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings. In the following description and drawings, the same or equivalent elements are denoted by the same reference numerals, and redundant description is omitted.

[First Embodiment]
First, a first embodiment of the present invention will be described with reference to FIGS.

<Configuration>
FIG. 1 shows an example of positioning of the digital filter 120A according to the first embodiment. In addition, as shown in FIG. 1, the following description will be given by exemplifying a case where the digital filter 120 </ b> A is one of the elements constituting the acoustic device 100.

  As shown in FIG. 1, a sound source unit 310 and a sound output unit 320 are connected to the acoustic device 100.

  The sound source unit 310 is configured to output a source sound signal SRD. The source sound signal SRD output from the sound source unit 310 is supplied to the acoustic device 100.

  As the sound source unit 310, for example, a unit for reading audio content data of a two-channel stereo system recorded on a recording medium such as a DVD (Digital Versatile Disk), or audio content data from a reception result of a stereo broadcast wave is used. Examples include a unit to be extracted.

  The sound output unit 320 includes a speaker SP. The sound output unit 320 receives the output sound signal SOD sent from the acoustic device 100. The sound output unit 320 reproduces and outputs the sound corresponding to the output sound signal SOD from the speaker SP.

<< Configuration of Acoustic Device 100 >>
The acoustic device 100 includes a pre-processing unit 110 and a post-processing unit 130 in addition to the digital filter 120A. The acoustic device 100 includes an input unit 140 and a coefficient calculation unit 150.

  The pre-processing unit 110 receives the source sound signal SRD sent from the sound source unit 310. Then, the pre-processing unit 110 performs pre-processing on the source sound signal SRD to generate a signal PRD. The signal PRD generated in this way is sent to the digital filter 120A.

  The digital filter 120A receives the signal PRD sent from the pre-processing unit 110. Then, the digital filter 120A performs a filtering process on the signal PRD according to the coefficient switching designation CSC sent from the coefficient calculation unit 150 to generate the signal FLD. The signal FLD generated in this way is sent to the post-processing unit 130.

  The configuration of the digital filter 120A will be described later.

  The post-processing unit 130 receives the signal FLD sent from the digital filter 120A. Then, the post-processing unit 130 performs post-processing on the signal FLD to generate an output sound signal SOD. The output sound signal SOD generated in this way is sent to the sound output unit 320.

  The input unit 140 is configured by a key unit (including a slide key or the like) provided in the main body unit of the audio device 100, or a remote input device including the key unit. Here, as a key part provided in the main body, a touch panel provided in a display unit (not shown) can be used. Moreover, it can replace with the structure which has a key part, and the structure which inputs voice can also be employ | adopted.

  By operating the input unit 140, the user can input a filter characteristic change command specifying the content of the filtering process for the signal PRD. The filter characteristic change command input to the input unit 140 is sent to the coefficient calculation unit 150 as the input command IPD.

  The coefficient calculation unit 150 receives the input command IPD sent from the input unit 140. Then, the coefficient calculation unit 150 calculates a new coefficient vector based on the content of the new filtering process specified in the filter characteristic change command sent as the input command IPD.

  When calculating the coefficient vector, the coefficient calculation unit 150 refers to a plurality of coefficient vector templates stored in a storage unit (not shown) based on the contents of the designated filtering process, and appropriately performs interpolation or the like. A coefficient vector corresponding to the content of the designated filtering process is calculated. A coefficient switching designation CSC designating the coefficient vector thus calculated is sent to the digital filter 120A.

(Configuration of digital filter 120A)
Next, the configuration of the digital filter 120A will be described. The digital filter 120A is an IIR (Infinite Impulse Response) filter.

As illustrated in FIG. 2, the digital filter 120 </ b> A includes _Q delay output units 211 _{1 to} 211 _Q and R delay output units 212 _{1 to} 212 _R. The digital filter 120A includes (Q + 1) individual multipliers 221 _{0 to} 221 _Q and R individual multipliers 222 _{1 to} 222 _R. Furthermore, the digital filter 120A includes an adding unit 230 and a coefficient switching unit 240A.

The delay output units 211 _{1 to} 211 _Q are connected in series, and each of the delay output units 211 _q (q = _{1 to Q} ) delays the input signal X _q-1 (T−τ) by a unit delay time τ. And output as a signal X _q (T). Here, the signal X ₀ (T) is the signal PRD sent from the pre-processing unit 110. As a result, the relationship between the signal X _q (T) and the signal X ₀ (T) is expressed by the following equation (1).
X _q (T) = X ₀ (T−q · τ) (1)

In the first embodiment, each of the delay output units 211 _q samples the signal X _q-1 (T) in synchronization with a reference clock (not shown) having a unit delay time (one sample time) τ as one cycle. And output as a signal X _q (T + τ). For this reason, during the unit delay time τ, the sampling result is held and output from the delay output unit 211 _q . Here, the unit delay time τ is ¼ of the signal input period of the signal X ₀ (T).

The signal X _q (T) generated by the delay output unit 211 _q is sent to the individual multiplication unit 221 _q . Note that the signal X ₀ (T) is sent to the individual multiplier 221 ₀ .

The delay output units 212 _{1 to} 212 _R are connected in series, and each of the delay output units 211 _r (r = _{1 to R} ) delays the input signal Y _r-1 (T−τ) by a unit delay time τ. And output as a signal Y _r (T). Here, the signal Y ₀ (T) is the addition result sent from the addition unit 230. As a result, the relationship between the signal Y _r (T) and the signal Y ₀ (T) is expressed by the following equation (2).
Y _r (T) = Y ₀ (T−r · τ) (2)

In the first embodiment, each of the delay output units 212 _r samples the signal Y _r-1 (T) in synchronization with a reference clock (not shown) having a unit delay time (one sample time) τ as one cycle. And output as a signal Y _r (T + τ). Therefore, during the unit delay time τ, the sampling result is held and output from the delay output unit 212 _q .

The signal Y _r (T) generated by the delay output unit 212 _q is sent to the individual multiplication unit 222 _r .

Each of the individual multiplication units 221 _k (k = 0 to Q) receives the signal X _k (T) and the coefficient KX _k sent from the coefficient switching unit 240A. Then, the individual multiplier 221 _k multiplies the signal X _k (T) by the coefficient KX _k . The result of this multiplication (= KX _k · X _k (T)) is sent to the adding unit 230.

Each of the individual multipliers 222 _r receives the signal Y _r (T) and the coefficient KY _r sent from the coefficient switching unit 240A. Then, the individual multiplier 222 _r multiplies the signal Y _r (T) by the coefficient KY _r . The result of this multiplication (= KY _r · Y _r (T)) is sent to the adding unit 230.

The adder 230 includes the multiplication results (KX ₀ · X ₀ (T)) to (KX _Q · X _Q (T)) by the individual multipliers 221 _{0 to} 221 _Q and the individual multipliers 222 _{1 to} 222 _R. The multiplication results (KY ₁ · Y ₁ (T)) to (KY _R · Y _R (T)) are received. Then, the adding unit 230 calculates the signal Y ₀ (T) by the following equation (3).

The signal Y ₀ (T) thus calculated is sent to the delay output unit 212 ₁ and also sent to the post-processing unit 130 as the signal FLD.

The coefficient switching unit 240A receives the coefficient switching designation CSC sent from the coefficient calculation unit 150. Then, the coefficient switching unit 240A executes coefficient switching processing based on the coefficient vectors (KXA ₀ ,..., KXA _Q ; KYA ₁ ,..., KYA _R ) after switching designated by the coefficient switching designation CSC. The coefficient switching unit 240A functions as a calculation unit and a switching unit.

  Details of the coefficient switching process by the coefficient switching unit 240A will be described later.

<Operation>
Next, the operation of the acoustic device 100 configured as described above will be described mainly focusing on the coefficient switching process by the coefficient switching unit 240A in the digital filter 120A.

  It is assumed that the sound source unit 310 has already started operation, and that the source sound signal SRD is sent from the sound source unit 310 to the acoustic device 100.

  In the acoustic device 100, the pre-processing unit 110 receives the source sound signal SRD sent from the sound source unit 310. Then, the pre-processing unit 110 performs pre-processing on the source sound signal SRD to generate a signal PRD, and sends the generated signal PRD to the digital filter 120A (see FIG. 1).

When receiving the signal PRD (= X ₀ (T)) sent from the pre-processing unit 110, the digital filter 120A is the one before the switching designated at that time by the coefficient switching designation CSC sent from the coefficient calculation unit 150. coefficient vector _{(KXB 0, ..., KXB Q} ; KYB 1, ..., KYB R) coefficient vector _{(KX 0, ..., KX Q} ; KY 1, ..., KY R) as was described above (1) to (3) A filtering process using the equation is applied to the signal PRD. Then, the digital filter 120A sends the result of the filtering process as a signal FLD (= Y ₀ (T)) to the subsequent processing unit 130 (see FIGS. 1 and 2).

  Upon receiving the signal FLD sent from the digital filter 120A, the post-processing unit 130 performs post-processing on the signal FLD to generate an output sound signal SOD. Then, the post-processing unit 130 sends the generated output sound signal SOD to the sound output unit 320 (see FIG. 1).

  When receiving the output sound signal SOD sent from the acoustic device 100, the sound output unit 320 reproduces and outputs the sound corresponding to the output sound signal SOD from the speaker SP. As a result, a sound corresponding to the filtering process by the digital filter 120A is reproduced and output from the speaker SP.

When the user operates the input unit 140 and inputs a filter characteristic change command specifying the content of the filtering process by the digital filter 120A during the sound reproduction output performed as described above, the filter characteristic change command Is sent to the coefficient calculation unit 150 as an input command IPD. Upon receipt of this input command IPD, the coefficient calculation unit 150 generates a new coefficient vector (KXA ₀ , KXA ₀ , K) based on the content of the new filtering process specified in the filter characteristic change command sent as the input command IPD. _{..., KXA Q; KYA 1,} ..., and calculates the KYA _R). Then, the coefficient calculation unit 150 sends a coefficient switching specification CSC specifying the calculated coefficient vector (KXA ₀ ,..., KXA _Q ; KYA ₁ ,..., KYA _R ) to the digital filter 120A (see FIG. 1).

In the first embodiment, as described above, the coefficient calculation unit 150 refers to a plurality of coefficient vector templates stored in a storage unit (not shown) based on the contents of the designated filtering process, and appropriately A coefficient vector (KXA ₀ ,..., KXA _Q ; KYA ₁ ,..., KYA _R ) corresponding to the contents of the designated filtering process is calculated by performing interpolation or the like.

  In the digital filter 120 </ b> A, the coefficient switching unit 240 </ b> A receives the coefficient switching designation CSC sent from the coefficient calculation unit 150. In response to the coefficient switching designation CSC, the coefficient switching unit 240A executes a coefficient switching process.

<< Coefficient switching process by coefficient switching unit 240A >>
In the coefficient switching process, as shown in FIG. 3, first, in step S11, the coefficient switching unit 240A determines whether or not a new coefficient switching designation CSC has been received. If the result of the determination in step S11 is negative (step S11: N), the process of step S11 is repeated.

  When the new coefficient switching designation CSC sent from the coefficient calculation unit 150 is received and the result of the determination in step S11 is affirmative (step S11: Y), the process proceeds to step S12. In step S12, the coefficient switching unit 240A sets the count value p held therein to “1”.

Next, in step S13, the coefficient switching unit 240A calculates the number of updates N. When calculating such update count N, coefficient switching section 240A, first, the switching coefficient vector of before and which has been used up to the present time _{(KXB 0, ..., KXB Q} ; KYB 1, ..., KYB R) and, newly designated and switching coefficient vector of換後 _{(KXA 0, ..., KXA Q} ; KYA 1, ..., KYA R) difference vector between the _{(ΔKX 0, ..., ΔKX Q} ; ΔKY 1, ..., ΔKY R) is calculated.

Next, the coefficient switching unit 240A extracts the maximum value | Δ | _MAX among the absolute values of the components of the difference vector. The coefficient switching unit 240A, the maximum value | delta | after dividing the _MAX at a predetermined value delta _REF, by rounding up the following division result-point, to calculate the number of updates N.

Note that the “predetermined value Δ _REF ” is a predetermined value even if the absolute values of the values indicated by the signals X _k (T) (k = 0 to Q) and Y _r (T) (r = 1 to R) increase. From the viewpoint that the discontinuity generated in the signal Y ₀ (T) can be made sufficiently small if the coefficient change is about Δ _REF , it is predetermined based on experiments, simulations, experiences, and the like.

Next, in step S14, the coefficient switching unit 240A calculates an update amount vector (δKX ₀ ,..., ΔKX _Q ; δKY ₁ ,..., ΔKY _R ). In calculating the update amount vector, the coefficient switching unit 240A calculates the update amount vector by dividing each component of the difference vector by the number of updates N and then rounding off the least significant digit of the division result. To do.

Next, in step S15, the coefficient switching unit 240A calculates an updated vector obtained by adding the update amount vector to the current coefficient vector, and the value of each component of the calculated updated vector is calculated as a corresponding individual multiplication unit 221. _The coefficients are updated by sending them to _{0 to} 221 _Q and the individual multipliers 222 _{1 to} 222 _R. Subsequently, in step S16, the coefficient switching unit 240A increments the count value p to obtain a new count value p (p ← p + 1).

Next, in step S17, the coefficient switching unit 240A determines whether or not the count value p has reached “N + 1”, so that the current coefficient vector and the above-described newly designated coefficient vector (KXA ₀ , .., KXA _Q ; KYA ₁ ,..., KYA _R ) is determined whether or not it can be said that the difference is small. If the result of the determination in step S17 is negative (step S17: N), the process returns to step S15. Thereafter, the processes in steps S15 to S17 are repeated until the result of the determination in step S17 becomes affirmative.

When the count value p is “N + 1” and the determination result in step S17 is affirmative, the process proceeds to step S18. In this step S18, the coefficient switching unit 240A converts the coefficient vector into a newly designated coefficient vector (KXA ₀ ,..., KXA _Q ; KYA ₁ ,..., KYA _R ) in order to compensate for the above-described slight difference. Update and complete the current coefficient vector switching. Then, the process returns to step S11.

  Thereafter, the processes of steps S11 to S18 are repeated. As a result, every time it is necessary to switch to a new coefficient vector, coefficient switching is executed by the coefficient switching unit 240A.

  As described above, in the digital filter 120A of the first embodiment, when the coefficient switching unit 240A switches from the pre-switching coefficient vector (first coefficient vector) to the post-switching coefficient vector (second coefficient vector), before switching. After calculating a difference vector obtained by subtracting the coefficient vector from the post-switching coefficient vector, an update amount vector for performing stepwise switching from the pre-switching coefficient vector to the post-switching coefficient vector is calculated based on the difference vector. . Then, the coefficient switching unit 240A performs switching of the coefficient vectors in a stepwise manner based on the calculated update amount vector.

  Therefore, according to the digital filter 120A of the first embodiment, it is possible to suppress the occurrence of a discontinuous change in the output signal when the filter characteristic is changed by switching the coefficient vector.

  As a result, in the acoustic device 100, since the built-in digital filter 120A operates as described above, the generation of so-called pop sounds accompanying the change in filter characteristics is suppressed without performing mute processing. Therefore, in the acoustic device 100, when the filter coefficient of the built-in digital filter 120A is switched, it is possible to suppress an uncomfortable sense of hearing for the user.

  In the first embodiment, the coefficient switching unit 240A determines the number of updates based on the difference vector, and then divides the difference vector by the determined number of updates to calculate an update amount vector. For this reason, the number of updates can be optimized, and the coefficient vector can be updated quickly.

  In the first embodiment, after the coefficient update based on the update amount vector is completed, the coefficient switching unit 240A further performs the update to the second vector coefficient. For this reason, the coefficient vector after switching can be made to coincide with the second vector coefficient accurately.

[Second Embodiment]
Next, a second embodiment of the present invention will be described mainly with reference to FIGS.

<Configuration>
FIG. 4 shows a configuration of a digital filter 120B according to the second embodiment. As shown in FIG. 4, the digital filter 120B is different from the digital filter 120A of the first embodiment described above in that a coefficient switching unit 240B is provided instead of the coefficient switching unit 240A. Hereinafter, this difference will be mainly described.

The coefficient switching unit 240B is different from the coefficient switching unit 240A described above in that it does not make a determination based on the calculation of the update count N and maintains a predetermined constant update count N therein. Yes. The “update count N” held by the coefficient switching unit 240B is obtained by dividing each component of the difference vector (ΔKX ₀ ,..., ΔKX _Q ; ΔKY ₁ ,..., ΔKY _R ) by the update count N and then rounding off. From the viewpoint that the absolute value of each component of the obtained update amount vector (δKX ₀ ,..., ΔKX _Q ; δKY ₁ ,..., ΔKY _R ) does not become larger than the predetermined value Δ _REF described above, experiment, simulation, experience Etc., based on the above.

<Operation>
Next, the operation of the digital filter 120B configured as described above will be described mainly focusing on the coefficient switching process by the coefficient switching unit 240B. It is assumed that the pre-processing unit 110 has already started operation, and a signal PRD is sent from the pre-processing unit 110.

Upon receiving the signal PRD (= X ₀ (T)) sent from the pre-processing unit 110, the digital filter 120B, like the above-described digital filter 120A, is the coefficient vector before switching specified at that time. (KXB ₀ ,..., KXB _Q ; KYB ₁ ,..., KYB _R ) is a coefficient vector (KX ₀ ,..., KX _Q ; KY ₁ ,..., KY _R ) and the above-described equations (1) to (3) are used. The used filtering process is performed on the signal PRD. Then, the digital filter 120B sends the result of the filtering process as a signal FLD (= Y ₀ (T)) to the post-processing unit 130 (see FIG. 4).

  In the digital filter 120B, the coefficient switching unit 240B receives the coefficient switching designation CSC sent from the coefficient calculating unit 150. Corresponding to the coefficient switching designation CSC, the coefficient switching unit 240B executes coefficient switching processing.

In the coefficient switching process, as shown in FIG. 5, the same processes as those in steps S11 and S12 (see FIG. 3) described above are executed in steps S21 and S22. In step S23, the coefficient switching unit 240B calculates an update amount vector (δKX ₀ ,..., ΔKX _Q ; δKY ₁ ,..., ΔKY _R ). In calculating the update amount vector, the coefficient switching unit 240B divides each component of the difference vector by the update count N stored therein, and then performs rounding processing by rounding off the least significant digit of the division result. Thus, an update amount vector is calculated.

  Thereafter, in steps S24 to S27, processing similar to the processing in steps S15 to S18 described above is executed. As a result, every time it is necessary to switch to a new coefficient vector, coefficient switching is executed by the coefficient switching unit 240B.

  As described above, in the digital filter 120B of the second embodiment, when the coefficient switching unit 240B switches from the pre-switching coefficient vector to the post-switching coefficient vector, the difference obtained by subtracting the pre-switching coefficient vector from the post-switching coefficient vector. After calculating the vector, based on the difference vector, an update amount vector for stepwise switching from the pre-switching coefficient vector to the post-switching coefficient vector is calculated. Then, the coefficient switching unit 240B performs the coefficient vector switching step by step based on the calculated update amount vector.

  Therefore, according to the digital filter 120B of the second embodiment, it is possible to suppress the occurrence of discontinuous changes in the output signal when changing the filter characteristics by switching the coefficient vector.

  Since the digital filter 120B operates as described above, an acoustic device incorporating the digital filter 120B suppresses the generation of a so-called pop sound that accompanies a change in the filter characteristics of the digital filter 120B without performing mute processing. Therefore, in the acoustic device, when the filter coefficient of the built-in digital filter 120B is switched, it is possible to suppress a sense of discomfort on the user's hearing.

  In the second embodiment, the coefficient switching unit 240B calculates an update amount vector by dividing the difference vector by an update count that is a predetermined constant value, and the calculated update amount vector is calculated by the update count. Update coefficient based on. Therefore, the coefficient vector can be reliably updated with a simple configuration.

  In the second embodiment, after the coefficient update based on the update amount vector is completed, the coefficient switching unit 240B further updates the second vector coefficient. For this reason, the coefficient vector after switching can be made to coincide with the second vector coefficient accurately.

[Third Embodiment]
Next, a third embodiment of the present invention will be described mainly with reference to FIGS.

<Configuration>
FIG. 6 shows the configuration of a digital filter 120C according to the third embodiment. As shown in FIG. 6, the digital filter 120C is different from the digital filter 120B of the second embodiment described above in that a coefficient switching unit 240C is provided instead of the coefficient switching unit 240B. Hereinafter, this difference will be mainly described.

Compared with the coefficient switching unit 240B described above, the coefficient switching unit 240C performs the current coefficient vector (KX ₀ ,..., KX _Q ; KY ₁ ,..., KY _R ) each time the coefficient is updated based on the update amount vector. The difference is that fine difference determination is performed to determine whether or not the difference from the new coefficient vector designated by the coefficient calculation unit 150 is a small difference. The coefficient switching process by the coefficient switching unit 240C will be described later.

<Operation>
Next, the operation of the digital filter 120C configured as described above will be described mainly focusing on the coefficient switching process by the coefficient switching unit 240C. It is assumed that the pre-processing unit 110 has already started operation, and a signal PRD is sent from the pre-processing unit 110.

Upon receiving the signal PRD (= X ₀ (T)) sent from the pre-processing unit 110, the digital filter 120C, like the case of the digital filter 120B described above, receives the coefficient switching designation CSC sent from the coefficient calculation unit 150. , The coefficient vector (KXB ₀ ,..., KXB _Q ; KYB ₁ ,..., KYB _R ) designated at that time as the coefficient vector (KX ₀ ,..., KX _Q ; KY ₁ ,..., KY _R ) The filtering process using the equations (1) to (3) is performed on the signal PRD. Then, the digital filter 120C sends the result of the filtering process as a signal FLD (= Y ₀ (T)) to the post-processing unit 130 (see FIG. 6).

  In the digital filter 120C, the coefficient switching unit 240C receives the coefficient switching designation CSC sent from the coefficient calculation unit 150. Corresponding to the coefficient switching designation CSC, the coefficient switching unit 240C executes a coefficient switching process.

In the coefficient switching process, as shown in FIG. 7, the same processes as those in steps S21 to 23 (see FIG. 5) described above are executed in steps S31 to S33. Then, in step S34, the coefficient switching unit 240C causes the current coefficient vector (KX ₀ ,..., KX _Q ; KY ₁ ,..., KY _R ) and the new coefficient vector (KXA ₀ ) designated by the coefficient calculation unit 150. _{, ..., KXA Q; KYA 1} , ..., a difference between the KYA _R) makes a differential refinement determination is the determination of whether a differential refinement.

In determining the slight difference, the coefficient switching unit 240C first calculates the difference between the current coefficient vector and the new coefficient vector designated by the coefficient calculation unit 150, and calculates the difference between the absolute values of the components in the calculated vector. Extract the maximum value. The coefficient switching unit 240C, the maximum value which is the extracted, by determining or less than a predetermined value delta _REF as described above, performing the fine difference determination.

  If the result of the determination in step S34 is affirmative (step S34: Y), the process proceeds to step S38 described later. On the other hand, when the result of the determination in step S34 is negative (step S34: N), the process proceeds to step S35.

In step S35, the coefficient switching unit 240C calculates an updated vector obtained by adding the update amount vector to the current coefficient vector, and the value of each component of the calculated updated vector is assigned to the corresponding individual multiplying unit 221 _{0 to} 221. _The coefficients are updated by sending them to the _Q and individual multipliers 222 _{1 to} 222 _R. Subsequently, in step S36, the coefficient switching unit 240C increments the count value p to obtain a new count value p (p ← p + 1).

Next, in step S37, the coefficient switching unit 240C determines whether or not the count value p has reached “N + 1”, so that the current coefficient vector and the newly specified coefficient vector (KXA ₀ , .., KXA _Q ; KYA ₁ ,..., KYA _R ) is determined whether or not it can be said that the difference is small. If the result of the determination in step S37 is negative (step S37: N), the process returns to step S34. Thereafter, the processes in steps S34 to S37 are repeated until the result of determination in step S34 or the result of determination in step S37 becomes affirmative.

If the result of determination in step S34 or the result of determination in step S37 is affirmative (step S34: Y or step S37: Y), the process proceeds to step S38. In this step S38, the coefficient switching unit 240C changes the coefficient vector to a newly designated coefficient vector (KXA ₀ ,..., KXA _Q ; KYA ₁ ,..., KYA _R ) in order to compensate for the above-described slight difference. Update and complete the current coefficient vector switching. Then, the process returns to step S31.

  Thereafter, the processes of steps S31 to S38 are repeated. As a result, every time it is necessary to switch to a new coefficient vector, coefficient switching is executed by the coefficient switching unit 240C.

  As described above, in the digital filter 120C of the third embodiment, when the coefficient switching unit 240C switches from the first coefficient vector to the second coefficient vector, the difference obtained by subtracting the first coefficient vector from the second coefficient vector. After calculating the vector, an update amount vector for performing stepwise switching from the first coefficient vector to the second coefficient vector is calculated based on the difference vector. Then, the coefficient switching unit 240C performs the switching of the coefficient vectors step by step based on the calculated update amount vector.

  Therefore, according to the digital filter 120C of the third embodiment, it is possible to suppress the occurrence of a discontinuous change in the output signal when the filter characteristic is changed by switching the coefficient vector.

  Since the digital filter 120C operates as described above, an acoustic device incorporating the digital filter 120C suppresses the generation of a so-called pop sound that accompanies a change in the filter characteristics of the digital filter 120C without performing a mute process. Therefore, in the acoustic device, when the filter coefficient of the built-in digital filter 120C is switched, it is possible to suppress a sense of discomfort in the sense of hearing for the user.

  In the third embodiment, the coefficient switching unit 240C performs the above-described fine difference determination every time the coefficient is updated based on the update amount vector. Therefore, the coefficient vector can be updated quickly.

  In the third embodiment, after the coefficient update based on the update amount vector is completed, the coefficient switching unit 240C further updates the second vector coefficient. For this reason, the coefficient vector after switching can be made to coincide with the second vector coefficient accurately.

[Modification of Embodiment]
The present invention is not limited to the above-described embodiment, and various modifications are possible.

  For example, in the first embodiment, the fine difference determination in the third embodiment is not performed. On the other hand, prior to updating the coefficient based on the update amount vector (immediately after step S14 in FIG. 3), the slight difference determination is performed. If the result of the small difference determination is affirmative, You may make it perform the process of step S18.

  In the first to third embodiments, the update amount vector is constant when the coefficient is updated based on the update amount vector in one coefficient switching. On the other hand, the update amount vector may be changed in one coefficient switching.

  In the first to third embodiments, the present invention is applied to the IIR filter. However, the present invention may be applied to the FIR filter.

  Moreover, in said 1st-3rd embodiment, applying the digital filter of this invention to the filtering process of an acoustic digital signal was assumed. On the other hand, you may apply the digital filter of this invention also to the filtering process of digital signals other than an acoustic digital signal.

  In addition, the coefficient calculation unit and the coefficient switching unit in the above embodiment are configured as a computer as a calculation unit including a central processing unit (CPU), a DSP (Digital Signal Processor), and the like, which are prepared in advance. You may make it perform one part or all part of the process in said embodiment by running a program with the said computer. This program is recorded on a computer-readable recording medium such as a hard disk, CD-ROM, or DVD, and is read from the recording medium and executed by the computer. The program may be acquired in a form recorded on a portable recording medium such as a CD-ROM or DVD, or may be acquired in a form distributed via a network such as the Internet. Also good.

120A to 120C Digital filter 240A to 240C Coefficient switching unit (calculation unit, switching unit)

Claims (9)

A digital filter that performs filtering on an input digital signal with characteristics determined by a coefficient vector,
Upon switching from the first coefficient vector to the second coefficient vector, a difference vector obtained by subtracting the first coefficient vector from the second coefficient vector is calculated, and then based on the difference vector, the first coefficient vector is changed from the first coefficient vector to the second coefficient vector. A calculation unit for calculating an update amount vector for performing switching to the second coefficient vector stepwise;
A switching unit that performs stepwise switching of the coefficient vector based on the update amount vector;
A digital filter comprising:   The digital filter according to claim 1, wherein the calculation unit divides the difference vector by an update count to calculate the update amount vector. The calculation unit determines the number of updates based on the difference vector,
The switching unit updates the coefficient vector based on the update amount vector, and performs the update count.
The digital filter according to claim 2.   The digital filter according to claim 2, wherein the update count is a predetermined constant.   5. The digital filter according to claim 3, wherein the switching unit further updates to the second vector coefficient after performing the update based on the update amount vector the number of times of update. 6.   The switching unit updates the second vector coefficient when the absolute value of each component in the difference between the coefficient vector being updated and the second coefficient vector becomes a predetermined value or less. The digital filter according to claim 4. A filter characteristic changing method used in a digital filter that performs filtering processing on an input digital signal with characteristics determined by a coefficient vector,
Upon switching from the first coefficient vector to the second coefficient vector, a difference vector obtained by subtracting the first coefficient vector from the second coefficient vector is calculated, and then based on the difference vector, the first coefficient vector is changed from the first coefficient vector to the second coefficient vector. A calculation step of calculating an update amount vector for performing switching to the second coefficient vector stepwise;
A switching step of switching the coefficient vector stepwise based on the update amount vector;
A filter characteristic changing method comprising:   A filter characteristic changing program that causes a computer provided in a digital filter that performs filtering processing on an input digital signal with characteristics determined by a coefficient vector to execute the filter characteristic changing method according to claim 7.   9. A recording medium in which the filter characteristic changing program according to claim 8 is recorded so as to be readable by a computer provided in a digital filter that performs a filtering process with a characteristic determined by a coefficient vector on an input digital signal. .

Images