JP2009300647A - Signal processing device, signal processing method, signal processing program and recording medium - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To suppress unnecessary deterioration of accuracy of a processing result, while securely preventing overflow in the processing result of a digital signal. <P>SOLUTION: Impulse response h(p×τ)(p=0 to N) of a processing section 170 for performing signal processing by digital processing is divided into a first part h<SB>1</SB>(q×τ)(q=0 to M(<N)), and a second part h<SB>2</SB>(r×τ)(r=M+1 to N) by an impulse response dividing section 183. After a speech data signal X(j×τ) which is read beforehand is convoluted with the first part h<SB>1</SB>(q×τ) by a convolution section 184, an absolute value Z<SB>2</SB>(j×τ) as a result of convolution is calculated by an absolute value calculation section 185. Meanwhile, an L<SB>1</SB>norm value NRM of the second part h<SB>2</SB>(r×τ) is calculated by an L<SB>1</SB>norm calculation section 186. Then, an inverse of a sum of the absolute value Z<SB>2</SB>(j×τ) and the norm value NRM are calculated by a gain control section 187, and based on the calculation result, a gain value G(T) is generated and forwarded to a multiplication section 150. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a signal processing device, a signal processing method, a signal processing program, and a recording medium on which the signal processing program is recorded.

  With the progress of digital technology in recent years, various processing processes such as frequency characteristics, phase characteristics, sound field characteristics, etc. are widely performed in a state in which acoustic signals are also digitized. In such processing, it is necessary to prevent a clip (digital clip) due to overflow (overflow) in the processing result regardless of the level of the acoustic signal to be processed.

  For this reason, techniques that do not generate digital clips have been proposed. As a first example of such a technique, a digital audio amplifier that suppresses the occurrence of overflow by controlling the attenuation rate in the variable attenuator when an overflow occurs in the processing result of the acoustic signal via the variable attenuator. A technique has been proposed (see Patent Document 1: hereinafter referred to as “Conventional Example 1”).

As a second example, the maximum value of the frequency characteristic of the impulse response of the processing means that processes the acoustic signal via the variable attenuator is calculated, and the attenuation rate in the variable attenuator is calculated based on the calculated maximum value. A graphic equalizer technique that suppresses the occurrence of overflow by controlling is proposed (see Patent Document 2: hereinafter referred to as “conventional example 2”). As a third example, the L ₁ norm (sum of absolute values) of the impulse response of the processing means that processes the acoustic signal via the variable attenuator is calculated, and the variable attenuator is calculated based on the calculated L ₁ norm. There is known a technique for suppressing the occurrence of overflow by controlling the attenuation rate in (see Non-Patent Document 1: hereinafter referred to as “Conventional Example 3”).

JP 2002-151975 A JP 2002-345075 A Hiroshi Ochi “Digital Signal Processing Learned by Simulation”, page 122, CQ Publishing Co., Ltd., 2001-7-1

  In the technique of Conventional Example 1 described above, when the occurrence of an overflow is detected, the result is fed back to control the attenuation rate. For this reason, although it may be a short time, an overflow occurs, and thus the occurrence of the overflow cannot be completely prevented.

Further, in the technique of the conventional example 2, the attenuation rate is controlled based on a value equivalent to the L _∞ norm of the impulse response. Therefore, when a narrow band signal is processed as in an equalizer, an overflow occurs. It can be completely prevented. However, when a broadband signal such as an audio signal is to be processed, an overflow may occur.

Further, in the technique of the conventional example 3, the attenuation rate is controlled based on the L ₁ norm of the impulse response, so that no overflow can occur regardless of the signal pattern. However, a signal pattern having a maximum level evaluated corresponding to the L ₁ norm of the impulse response rarely occurs, and in most cases, the maximum level generated by the processing is set to the L ₁ norm of the impulse response. A corresponding evaluation would be an overestimation. For this reason, when the attenuation rate is controlled based on the L ₁ norm of the impulse response, in most cases, a so-called bit drop occurs in the digital signal to be processed, and a large deterioration in the accuracy of the processing result occurs unnecessarily. turn into.

  For this reason, there is a demand for a technique that can reliably prevent the occurrence of overflow in the processing result of the acoustic signal and suppress unnecessary deterioration of the processing result accuracy. 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 a signal processing device and a signal processing capable of suppressing unnecessary deterioration in accuracy of a processing result while reliably preventing an overflow in the processing result of a digital signal. It aims to provide a method.

The invention according to claim 1 is a multiplication unit that multiplies a digital signal that changes over time by a gain value; a processing unit that applies a predetermined signal processing to a multiplication result by the multiplication unit; and an impulse response of the processing unit. Dividing means for temporally dividing the specific part into a remaining part excluding the specific part; convolution means for generating a convolution signal in which the specific part is convoluted with the digital signal; and absolute of the convolution signal An absolute value calculating means for calculating a value; a norm calculating means for calculating an L ₁ norm of the remaining portion; and a gain value based on a calculation result by the absolute value calculating means and a calculation result by the norm calculating means. Gain control means for calculating and specifying the multiplication means; and a signal processing apparatus comprising:

The invention according to claim 8 is a signal processing method used in a signal processing apparatus including processing means for performing predetermined processing on a signal obtained by multiplying a digital signal that changes over time by a gain value. A division step of temporally dividing the impulse response of the means into a specific portion and a remaining portion excluding the specific portion; a convolution step of generating a convolution signal in which the specific portion is convoluted with the digital signal; An absolute value calculating step of calculating an absolute value of the convolution signal; a norm calculating step of calculating an L ₁ norm of the residual portion; and a calculation result in the absolute value calculating step and a calculation result in the norm calculating step. And a gain calculating step for calculating the gain value.

  According to a ninth aspect of the present invention, there is provided a signal processing program that causes a calculation means to execute the signal processing method according to the eighth aspect.

  A tenth aspect of the present invention is a recording medium in which the signal processing program according to the ninth aspect of the present invention is recorded so as to be readable by an arithmetic means.

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

[Constitution]
FIG. 1 schematically shows a configuration of an acoustic device 300 including a signal processing device 100 according to an embodiment of the present invention. As shown in FIG. 1, the acoustic device 300 further includes a sound source unit 310 and a speaker unit 320 in addition to the signal processing device 100. Further, a display unit 330 and an operation input unit 340 are provided.

  The signal processing apparatus 100 processes the audio data signal ADD from the sound source unit 310 according to the signal parameter designated by the user using the operation input unit 340, and generates an output audio signal AOS. Details of the signal processing apparatus 100 will be described later.

  The sound source unit 310 generates an audio data signal ADD from the audio content. For example, when the audio device 300 is a device that reproduces audio content recorded on a recording medium such as a compact disc (CD), the sound source unit 310 includes reading means for reading the audio content from the recording medium, The audio data extraction means for extracting the audio data signal ADD from the reading result by the reading means is provided. When the audio device 300 is a device that reproduces audio content acquired by receiving broadcast waves, the sound source unit 310 includes content extraction means for extracting audio content from the reception results of broadcast waves, and the content extraction The sound data extracting means for extracting the sound data signal ADD from the extraction result by the means is provided.

  The audio data signal ADD generated by the sound source unit 310 is sent to the signal processing apparatus 100.

  The speaker unit 320 includes a speaker. The speaker unit 320 reproduces and outputs sound in accordance with the output sound signal AOS sent from the signal processing apparatus 100.

  The display unit 330 includes, for example, (i) a display device such as a liquid crystal panel, an organic EL (Electro Luminescence) panel, a PDP (Plasma Display Panel), and (ii) display control data sent from the signal processing apparatus 100. Based on this, a display controller such as a graphic renderer for controlling the entire display unit 330, and (iii) a display image memory for storing display image data are provided. The display unit 330 displays operation guidance information and the like according to display data IMD from the signal processing device 100.

  The operation input unit 340 includes a key unit provided in the main body of the audio device 300 and / 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 device of the display unit 330 can be used. In addition, it can replace with the structure which has a key part, or can also employ | adopt the structure input with a sound using a voice recognition technique in combination.

  When the user operates the operation input unit 340, the operation content of the audio device 300 is set. For example, input by a user such as designation of signal processing in the signal processing apparatus 100 is performed using the operation input unit 340. Such input contents are sent as operation input data IPD from the operation input unit 340 to the signal processing device 100.

  Next, the signal processing apparatus 100 will be described. As described above, the signal processing apparatus 100 processes the audio data signal ADD from the sound source unit 310 to generate the output audio signal AOS. The signal processing apparatus 100 having such a function includes a digital processing unit 110 and an analog processing unit 120.

  The digital processing unit 110 processes the audio data signal ADD from the sound source unit 310 according to the signal processing parameter specified by the user using the operation input unit 340, and generates a signal MFD. In addition, the digital processing unit 110 generates a volume adjustment command VLC in accordance with the volume designation designated by the user using the operation input unit 340.

  As shown in FIG. 2, the digital processing unit 110 having such a function includes a signal delay unit 130 as a delay unit, a prefetch buffer unit 140 as a buffer storage unit, and a multiplication unit 150 as a multiplication unit. . The digital processing unit 110 includes a signal selection unit 160, a processing unit 170 as a processing unit, and a gain value generation unit 180. Further, the digital processing unit 110 includes a control unit 190.

  The signal delay unit 130 receives the audio data signal ADD (= X (t)) from the sound source unit 310. Then, the signal delay unit 130 generates a signal DAD (= X (T)) by giving a predetermined delay to the audio data signal ADD, and sends the signal DAD to the multiplication unit 150.

Here, the delay time T _D (= t−T) given to the audio data signal ADD in the signal delay unit 130 is equal to or longer than the generation time of the signal CEF (= G (T)) in the gain value generation unit 180 described later. The time is determined in advance. In the present embodiment, when the sampling period is τ, the times t and T are expressed by the following equations (1) and (2).
t = j · τ (j = 0, 1, 2,...) (1)
T = k · τ (k = 0, 1, 2,...) (2)

Here, the relationship between the value j and the value k is expressed by the following equation (3).
(J−k) = T _D / τ = constant value (3)

Additional prefetch buffer unit 140 temporarily stores the voice data signals ADD of a predetermined amount corresponding to the delay time T _D. Then, the temporarily stored content is sent from the prefetch buffer unit 140 to the gain value generation unit 180 as a signal BFD (= X (t)) in response to a read request from the gain value generation unit 180.

  The multiplication unit 150 receives the signal DAD (= X (T)) from the signal delay unit 130 and the signal CEF (= G (T)) from the gain value generation unit 180. Then, the multiplication unit 150 calculates a product of the signal DAD and the signal CEF at each time point. The multiplication result by the multiplication unit 150 is sent from the multiplication unit 150 to the signal selection unit 160 as a signal GCD (= Y (T) = X (T) · G (T)).

  The signal selection unit 160 receives the signal GCD from the multiplication unit 150 and the impulse signal IMP from the gain value generation unit 180. Then, the signal selection unit 160 selects either the signal GCD or the impulse signal IMP in accordance with the signal selection command SLC from the control unit 190, and sends it to the processing unit 170 as a signal SLD.

  Processing unit 170 receives signal SLD from signal selection unit 160. Then, the processing unit 170 performs processing of the signal SLD according to the values of various parameters included in the processing control command MDC from the control unit 190. Such processing includes frequency characteristics, phase characteristics, sound field characteristics, and the like.

  Such processing by the processing unit 170 is equivalent to processing for convolving the impulse response of the processing unit 170 measured as described later with the signal SLD. A processing result by the processing unit 170 is sent to the gain value generation unit 180 and the analog processing unit 120 as a signal MFD.

The gain value generation unit 180 generates a signal CEF (= G (T)) to be supplied to the multiplication unit 150. As shown in FIG. 3, the gain value generation unit 180 having such a function includes an impulse signal generation unit 181 as an impulse signal generation unit and an impulse response collection unit 182 as a collection unit. The gain value generation unit 180 includes an impulse response division unit 183 as a division unit, a convolution unit 184 as a convolution unit, and an absolute value calculation unit 185 as an absolute value calculation unit. Furthermore, the gain value generation unit 180 includes an L ₁ norm calculation unit 186 as a norm calculation unit and a gain control unit 187 as a gain control unit.

  When the impulse signal generation unit 181 receives the impulse signal generation command CCR from the control unit 190, the impulse signal generation unit 181 generates the impulse signal IMP. The impulse signal IMP generated in this way is sent from the impulse signal generator 181 to the signal selector 160. In addition, the impulse signal generation unit 181 sends, to the impulse response collection unit 182 as a signal IGS, simultaneously with the generation start of the impulse signal IMP, that the impulse signal generation has started.

  When the impulse response collection unit 182 receives the start of impulse signal generation as the signal IGS from the impulse signal generation unit 181, the subsequent signal MFD from the processing unit 170 is an impulse signal generated by the processing unit 170. Collected as an impulse response of the processing unit 170, which is an IMP processing result. An example of such an impulse response waveform is shown in FIG.

  This impulse response is obtained as a data string [h (0), h (τ),..., H (p · τ),. The impulse response collection time is determined in advance from the viewpoint of convergence of the impulse response h (p · τ) based on experiments, simulations, experiences, and the like.

  When the collection of the impulse response h (p · τ) is completed, the impulse response collection unit 182 reports the completion of collection to the control unit 190 as a signal AQE. Then, the impulse response collection unit 182 sends the collected impulse response to the impulse response division unit 183 as a signal IMR.

The impulse response dividing unit 183 receives the impulse response h (p · τ) (p = 0 to N) reported as the signal IMR from the impulse response collecting unit 182. Then, the impulse response dividing unit 183 divides the impulse response h (p · τ) into a first part and a second part. In the present embodiment, the impulse response dividing unit 183 converts the impulse response h (p · τ) into a first portion h ₁ (q · τ) (q = 0 to M) and a second portion h ₂ (r · τ). ) (R = M + 1 to N). Examples of waveforms of the first part h ₁ (q · τ) and the second part h ₂ (r · τ) generated by this division are shown in FIGS.

  In the present embodiment, the value M is a calculation amount related to the later-described expression (4), no occurrence of overflow due to calculations according to the later-described expressions (4) to (8), and the signal DAD (= X ( T)) and the signal GCD (= Y (T) = X (T) · G (T)) from the viewpoint of maintaining the similarity between the signals and the signals GCD (= Y (T) = X (T) · G (T)). Yes.

When such division is completed, the impulse response dividing unit 183 sends the first part h ₁ (q · τ) to the convolution unit 184 as the signal IMR1. Further, the impulse response division unit 183 sends the second part h ₂ (r · τ) to the L ₁ norm calculation unit 186 as the signal IMR2.

The convolution unit 184 receives the first part h ₁ (q · τ) as the signal IMR 1 from the impulse response division unit 183. The convolution unit 184 receives the signal BFD (= X (j · τ)) from the prefetch buffer unit 140. When no data is stored in the prefetch buffer unit 140, a data signal having a value “0” is transmitted from the prefetch buffer unit 140 as the signal BFD.

Then, the convolution unit 184 performs a calculation according to the following equation (4), thereby convolving the first part h ₁ (q · τ) of the impulse response with the data string X (j · τ).
Z ₁ (j · τ) = h (0) · X (j · τ) + h (τ) · X ((j−1) · τ)) +.
+ H (q · τ) · X ((j−q) · τ) +... + H (M · τ) · X ((j−M) · τ)
(4)

FIG. 8 shows an example of a waveform obtained by convolving the first portion h ₁ (q · τ) with the waveform data string X (j · τ) shown in FIG. The convolution result Z ₁ (j · τ) obtained in this way is sent to the absolute value calculation unit 185 as a signal CVD.

The absolute value calculation unit 185 calculates the absolute value Z ₂ (j · τ) of the convolution result Z ₁ (j · τ) by performing calculation according to the following equation (5).
Z ₂ (j · τ) = | Z ₁ (j · τ) | (5)

An example of the calculation result of the absolute value Z ₂ (j · τ) calculated in this way is shown in FIG. The absolute value Z ₂ (j · τ) calculated in this way is sent from the absolute value calculation unit 185 to the gain control unit 187 as a signal ABD.

The L ₁ norm calculation unit 186 receives the second part h ₂ (r · τ) of the impulse response as the signal IMR2 from the impulse response division unit 183. Then, the L ₁ norm calculation unit 186 calculates the L ₁ norm value NRM of the second portion h ₂ (r · τ) by the following equation (6).
NRM = | h ((M + 1) · τ) | + | h ((M + 2) · τ) | + ... + | h (N · τ) |
(6)

The L ₁ norm value NRM obtained in this way is sent from the L ₁ norm calculation unit 186 to the gain control unit 187.

The gain control unit 187 receives the absolute value Z ₂ (j · τ) as the signal ABD from the absolute value calculation unit 185 and the L ₁ norm value NRM of the L ₁ norm calculation unit 186. Then, gain control section 187 calculates evaluation value E (j · t) by the following equation (7).
E (j · t) = Z ₂ (j · τ) + NRM (7)
An example of the evaluation value E (j · t) calculated in this way is shown in FIG.

Subsequently, the gain control unit 187 calculates the gain value G (j · t) by the following equation (8).
G (j · t) = 1 / E (j · t) (8)

  The gain control unit 187 adjusts the output timing of the gain value G (j · t), and multiplies the gain value synchronized with the signal DAD (= X (T)) output from the signal delay unit 130 by the multiplication unit 150. Is generated as a signal CEF (= G (T)) to be specified. The signal CEF generated in this way is sent to the multiplier 150.

  Returning to FIG. 2, the control unit 190 will be described next. The control unit 190 performs the following control on the digital processing unit 110 and the analog processing unit 120.

  When the signal processing mode designation by digital processing input to the operation input unit 340 by the user is received as the signal IPD, the control unit 190 calculates parameter values for signal processing in the designated mode and includes them. A machining control command MDC is generated. The control unit 190 sends the generated machining control command MDC to the machining processing unit 170 and sends a signal selection command SLC to the signal selection unit 160 indicating that the impulse signal IMP from the gain value generation unit 180 should be selected. .

  Subsequently, the control unit 190 sends an impulse signal generation command CCR to the gain value generation unit 180. When the control unit 190 receives from the gain value generation unit 180 a signal AQE that the impulse response collection is complete, the signal selection command SLC indicating that the signal GCD from the multiplication unit 150 should be selected is sent to the signal selection unit 160. send.

  In the present embodiment, the control unit 190 generates a volume adjustment command VLC whose output volume should be “0” for the period from immediately before the issuance of the impulse signal generation command CCR to the reception of the signal AQE, The data is sent to the analog processing unit 120. Further, when the control unit 190 receives the signal AQE, the control unit 190 sends a buffer clear command BCL to the effect that the stored contents should be cleared to the prefetch buffer unit 140.

  In addition, when the specification of the volume adjustment mode by analog processing input to the operation input unit 340 by the user is received as the signal IPD, the control unit 190 generates a volume adjustment command VLC for the specified volume adjustment. Then, the control unit 190 sends the generated volume adjustment command VLC to the analog processing unit 120.

  Returning to FIG. 1, the analog processing unit 120 receives the signal MFD from the digital processing unit 110. Then, the analog processing unit 120 generates an output audio signal AOS under the control of the digital processing unit 110 and sends it to the speaker unit 320. As illustrated in FIG. 11, the analog processing unit 120 having such a function includes a DA (Digital to Analogue) conversion unit 121, a volume adjustment unit 122, and a power amplification unit 123.

  The DA converter 121 includes a DA converter. The DA converter 121 receives the digital signal MFD from the digital processor 110. The DA conversion unit 121 converts the signal MFD into an analog signal. The analog signal ACS, which is the conversion result of the DA converter 121, is sent to the volume controller 122.

  In the present embodiment, the volume adjusting unit 122 includes an electronic volume element and the like. The volume adjustment unit 122 receives the signal ACS from the DA conversion unit 121. Then, the volume adjustment unit 122 performs volume adjustment processing on the signal ACS in accordance with the volume adjustment command VLC from the digital processing unit 110. A signal VCS, which is an adjustment result by the volume adjustment unit 122, is sent to the power amplification unit 123.

  The power amplification unit 123 includes a power amplifier. The power amplification unit 123 receives the signal VCS from the volume adjustment unit 122. The power amplifying unit 123 power amplifies the signal VCS. An output audio signal AOS that is an amplification result by the power amplifier 123 is sent to the speaker unit 320.

[Operation]
Next, the operation of the acoustic apparatus 300 configured as described above will be described mainly focusing on the generation process of the gain value G (T) by the gain value generation unit 180 in the signal processing apparatus 100.

<Processing of Control Unit 190 in Gain Value Generation Processing>
First, the process of the control unit 190 in the gain value generation process will be described. In the gain value generation processing, as shown in FIG. 12, first, in step S11, the control unit 190 has designated a signal processing mode by new digital processing by an input to the operation input unit 340 by the user. Determine whether or not. If the result of this determination is negative (step S11: N), the process of step S11 is repeated.

  On the other hand, if the result of the determination in step S11 is affirmative (step S11: Y), the process proceeds to step S12. In step S12, the control unit 190 calculates parameter values for signal processing of the designated mode, and generates a processing control command MDC including them. The control unit 190 sends the generated machining control command MDC to the machining processing unit 170 and sends a signal selection command SLC to the signal selection unit 160 indicating that the impulse signal IMP from the gain value generation unit 180 should be selected. (See FIG. 2).

  As described above, in the present embodiment, in step S12, the control unit 190 generates the volume adjustment command VLC for setting the output volume to “0” before issuing the impulse signal generation command CCR. The data is sent to the processing unit 120.

  When the process of step S12 is completed in this way, the control unit 190 determines whether or not the signal AQE indicating the end of impulse response collection has been received from the gain value generation unit 180 in step S13. It is determined whether collection of impulse responses by 180 has ended. If the result of this determination is negative (step S13: N), the process of step S13 is repeated.

  On the other hand, if the result of the determination in step S13 is affirmative (step S13: Y), the process proceeds to step S14. In step S <b> 14, the control unit 190 sends a signal selection command SLC indicating that the signal GCD from the multiplication unit 150 should be selected to the signal selection unit 160.

  In this embodiment, in step S14, the control unit 190 sends a buffer clear command BCL indicating that the stored contents should be cleared to the prefetch buffer unit 140, and sets the output volume immediately before the issuance of the impulse signal generation command CCR. Similarly, a volume adjustment command VLC to be adjusted is generated and sent to the analog processing unit 120.

  When the process of step S14 is thus completed, the process returns to step S11. In this way, the processing of the control unit 190 in the gain value generation processing is executed by repeating the processing of steps S11 to S14.

<Processing of Gain Value Generation Unit 180 in Gain Value Generation Processing>
Next, the process of the gain value generation unit 180 in the gain value generation process will be described. In the gain value generation process, as shown in FIG. 13, the gain value generation unit 180 first determines whether or not an impulse generation command CCR is received from the control unit 190 in step S21. If the result of this determination is negative (step S21: N), the process proceeds to step S26. The processing after step S26 will be described later.

  On the other hand, when the result of the determination in step S21 is affirmative (step S21: Y), the process proceeds to step S22. In step S <b> 22, the impulse signal generator 181 generates an impulse signal IMP and sends it to the signal selector 160. Further, the impulse signal generation unit 181 sends, to the impulse response collection unit 182 as a signal IGS, simultaneously with the start of generation of the impulse signal IMP, as a signal IGS (see FIG. 3).

  When the signal IGS indicating that the impulse signal generation has started is received, in step S23, the impulse response collection unit 182 indicates the subsequent signal MFD from the processing unit 170 as a result of processing the impulse signal IMP by the processing unit 170. Collected as impulse response h (p · τ) (p = 1 to N) of the processing unit 170.

  When the collection of the impulse response h (p · τ) is completed, the impulse response collection unit 182 reports the completion of collection to the control unit 190 as a signal AQE. Then, the impulse response collection unit 182 sends the collected impulse response to the impulse response division unit 183 as a signal IMR (see FIG. 3).

Receiving the impulse response h (p · τ) collection result from the impulse response collection unit 182, the impulse response dividing unit 183 converts the impulse response h (p · τ) to the first part h ₁ (q · τ) (q = 0 to M) and the second part h ₂ (r · τ) (r = M + 1 to N). When the division ends in this way, the impulse response division unit 183 sends the divided first part h ₁ (q · τ) as the new first part to the convolution unit 184 as the signal IMR1. Further, the impulse response division unit 183 sends the second part h ₂ (r · τ) as the signal IMR2 to the L ₁ norm calculation unit 186 (see FIG. 3).

In step S25, the L ₁ norm calculation unit 186 that has received the second part h ₂ (r · τ) calculates the L ₁ norm value NRM of the second part h ₂ (r · τ) by the above-described equation (6). calculate. The L ₁ norm value NRM calculated in this way is sent to the gain control unit 187 as a new L ₁ norm value (see FIG. 3).

In parallel with the processing of step S25, in step S26, the convolution unit 184 reads the data X (t) from the prefetch buffer unit 140, and the data X (j · τ) is expressed by the equation (4) as follows: The first part h ₁ (q · τ) (q = 0 to M) reported from the impulse response division unit 183 at that time is convolved. The convolution result Z ₁ (j · τ) calculated by the convolution unit 184 is sent to the absolute value calculation unit 185 as a signal CVD (see FIG. 3).

  When no data is stored in the prefetch buffer unit 140, that is, when the audio data X (t) is not supplied from the sound source unit 310, all the data from the prefetch buffer unit 140 is “ Therefore, the convolution result calculated by the equation (4) is “0”.

The absolute value calculation unit 185 that has received the convolution result Z ₁ (j · τ) calculates the absolute value Z ₂ (j · τ) of the convolution result Z ₁ (j · τ) in step S27. The absolute value Z ₂ (j · τ) calculated in this way is sent as a signal ABD from the absolute value calculation unit 185 to the gain control unit 187 (see FIG. 3).

The gain control unit 187 that has received the L ₁ norm value NRM from the L ₁ norm calculation unit 186 and the absolute value Z ₂ (j · τ) from the absolute value calculation unit 185 performs gain value control processing in step S27. . In the gain value control process, the gain control unit 187 first calculates the evaluation value E (j · t) by the above-described equation (7).

Subsequently, the gain control unit 187 calculates the gain value G (j · t) by the above-described equation (8). If all the data acquired by the impulse response division unit 183 in step S25 described above becomes “0” because no data is stored in the prefetch buffer unit 140, the evaluation value E (j T) is the L ₁ norm value NRM itself of the second part h ₂ (r · τ), so that the gain value calculation is performed based only on the second part h ₂ (r · τ). .

  When the calculation of the gain value G (j · t) based on the predetermined amount of data acquired from the prefetch buffer unit 140 is thus completed, the process returns to step S21. Thereafter, the processes of steps S21 to S27 are repeated, and the calculation of the gain value G (j · t) is continuously executed.

  The gain control unit 187 adjusts the output timing of the gain value G (j · t) calculated in this manner, and synchronizes with the signal DAD (= X (T)) output from the signal delay unit 130. A signal CEF for designating T) to the multiplication unit 150 is generated and sent to the multiplication unit 150.

<Volume adjustment processing>
Next, the volume adjustment process will be described.

  When the specification of the volume adjustment mode by analog processing input to the operation input unit 340 by the user is received as the signal IPD, the control unit 190 generates a volume adjustment command VLC for the specified volume adjustment. Then, the control unit 190 sends the generated volume adjustment command VLC to the analog processing unit 120. Receiving the volume adjustment command VLC, the analog processing unit 120 performs volume adjustment in a mode specified by the volume adjustment command VLC.

The above gain value generation processing and signal processing processing are executed in the digital processing unit 110 in response to the user's designation, and a signal MFD is generated. An example of the signal MFD generated in this way is shown by a solid line in FIG. In FIG. 14, the signal processing result when the inverse of the L ₁ norm value of the impulse response h (p · t) is adopted as the gain value is indicated by a broken line. As can be seen by comparing the two, the control of the gain value according to the present embodiment can increase the change in the signal level.

  Further, the sound volume adjustment process described above is executed in the signal processing apparatus 100 in response to the user's designation, and the output audio signal AOS based on the audio data from the sound source unit 310 is supplied to the speaker unit 320. Then, the playback sound based on the output sound signal AOS is output from the speaker unit 320.

As described above, in this embodiment, the impulse response h (p · τ) (p = 0 to N) of the processing unit 170 that performs signal processing by digital processing is measured. The impulse response dividing unit 183 calculates the measured impulse response h (p · τ), the first portion h ₁ (q · τ) (q = 0 to M (<N)), and the second portion h ₂ (r Divide into τ) (r = M + 1 to N). Subsequently, after the convolution unit 184 convolves the first portion h ₁ (q · τ) with the pre-read voice data signal X (j · τ), the absolute value calculation unit 185 performs the absolute value Z of the convolution result. ₂ Calculate (j · τ). On the other hand, the L ₁ norm calculation unit 186 calculates the L ₁ norm value NRM of the second portion h ₂ (r · τ). Then, gain control section 187 calculates the reciprocal of the sum of absolute value Z ₂ (j · τ) and L ₁ norm value NRM, generates gain value G (T) based on the calculation result, and multiplies it. Send to section 150.

For this reason, in this embodiment, the audio data signal GCD adjusted to a level that can reliably prevent the occurrence of a clip due to the occurrence of an overflow in the processing by the processing unit 170 can be supplied to the processing unit 170. Further, as in the case of the present embodiment, the level value is compared with the level adjustment of the supply signal to the processing unit 170 by the L ₁ norm value of the impulse response h (p · τ) that can reliably prevent the occurrence of the clip. The size can be secured.

  Therefore, according to the present embodiment, unnecessary deterioration in the accuracy of the processing result can be suppressed while reliably preventing overflow in the processing result of the digital signal.

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

  For example, in the above embodiment, when dividing the impulse response into the first part and the second part, the continuous predetermined period starting from the initial stage of the impulse response is set as the first part, but is continuous. Any part of the impulse response, whether it is discrete or discrete, may be the first part.

In the above embodiment, the gain value G (T) supplied to the multiplier 150 after adjusting the output timing of the reciprocal of the sum of the absolute value Z ₂ (j · τ) and the L ₁ norm value NRM. Was generated. On the other hand, in a manner that does not exceed the value of the reciprocal of the sum, an envelope regarding the temporal change of the value of the reciprocal of the sum is obtained, and the gain value G (T) supplied to the multiplier 150 based on the envelope May be generated.

  In the above-described embodiment, the present invention is applied to the acoustic device. However, it is needless to say that the present invention can also be applied to an apparatus that performs digital signal processing other than the acoustic device.

  The digital processing unit 110 in the above embodiment is configured as a computer system including a central processing unit (CPU: Central Processor Unit) and a DSP (Digital Signal Processor), and the function of the digital processing unit 110 can be achieved by executing a program. Can be realized. These programs may be acquired in the form recorded on a portable recording medium such as a CD-ROM or DVD, or may be acquired in the form of delivery via a network such as the Internet. Good.

It is a block diagram which shows roughly the structure of an audio equipment provided with the signal processing apparatus which concerns on one Embodiment of this invention. It is a block diagram which shows the structure of the digital processing part of FIG. It is a block diagram which shows the structure of the gain value production | generation part of FIG. It is a figure which shows the example of the waveform of an impulse response. It is a figure which shows the example of the waveform of the 1st part of an impulse response. It is a figure which shows the example of the waveform of the 2nd part of an impulse response. It is a figure which shows the example of the waveform of an audio | voice data signal. It is a figure which shows the example of a convolution result. It is a figure which shows the example of an absolute value calculation result. An absolute value calculation result is a diagram showing an example of a calculation result of the sum of the L ₁ norm of the second part. It is a block diagram which shows the structure of the analog process part of FIG. It is a flowchart for demonstrating the process by the control part of FIG. 3 in the case of the gain value production | generation process in a digital signal part. It is a flowchart for demonstrating the process by the gain value generation part of FIG. 3 in the case of the gain value generation process in a digital signal part. It is a figure which shows the example of the signal processing result by a digital signal part.

Explanation of symbols

DESCRIPTION OF SYMBOLS 100 ... Signal processing apparatus 130 ... Signal delay part (delay means)
140 ... Prefetch buffer section (buffer storage means)
150 ... Multiplication unit (multiplication means)
170 ... Processing section (processing means)
181 ... Impulse signal generator (impulse signal generator)
182 ... Impulse response collection unit (collection means)
183 ... Impulse response dividing unit (dividing means)
184 ... Convolution part (convolution means)
185 ... Absolute value calculation unit (absolute value calculation means)
186 ... L ₁ norm calculation unit (norm calculation means)
187 ... Gain control section (gain control means)

Claims (10)

Multiplication means for multiplying a digital signal that changes over time by a gain value;
Processing means for performing predetermined signal processing on the multiplication result of the multiplication means;
Dividing means for dividing the impulse response of the processing means into a specific part and a remaining part excluding the specific part in time;
Convolution means for generating a convolution signal obtained by convolving the specific portion with the digital signal;
Absolute value calculating means for calculating an absolute value of the convolution signal;
Norm calculating means for calculating the L ₁ norm of the remaining portion;
Gain control means for calculating the gain value based on the calculation result by the absolute value calculation means and the calculation result by the norm calculation means and designating the gain value;
A signal processing apparatus comprising: The convolution means generates the convolution signal by convolving the specific part with the digital signal up to the time of input to the multiplication means,
The gain control means calculates the gain value based on a sum of a calculation result by the absolute value calculation means and a calculation result by the norm calculation means;
The signal processing apparatus according to claim 1.   The signal processing apparatus according to claim 2, wherein the gain control unit calculates an envelope of the time change of the sum, and obtains the gain value using the envelope.   A delay equal to or greater than the sum of the time from the initial point of the impulse response to the end point of the specific portion, the calculation time of the convolution means, the calculation time of the absolute value calculation means, and the calculation time of the gain control means The signal processing apparatus according to claim 2, further comprising a delay unit that applies the digital signal to the multiplication unit.   5. The signal processing apparatus according to claim 2, further comprising a buffer storage unit that temporarily stores the digital signal for a predetermined time and supplies the digital signal to the convolution unit.   The signal processing apparatus according to claim 1, wherein the specific portion is a portion corresponding to a predetermined period from an initial stage in the impulse response. An impulse signal generating means for generating an impulse signal in a digital format and supplying the impulse signal to the processing means;
Collecting means for collecting the processing result of the impulse signal by the processing means and sending it to the dividing means;
The signal processing apparatus according to claim 1, further comprising: A signal processing method used in a signal processing apparatus comprising processing means for performing predetermined processing on a signal obtained by multiplying a digital signal that changes over time by a gain value,
A division step of temporally dividing the impulse response of the processing means into a specific portion and a remaining portion excluding the specific portion;
A convolution step of generating a convolution signal obtained by convolving the specific portion with the digital signal;
An absolute value calculating step of calculating an absolute value of the convolution signal;
A norm calculating step of calculating an L ₁ norm of the remaining portion;
A gain calculation step of calculating the gain value based on a calculation result in the absolute value calculation step and a calculation result in the norm calculation step;
A signal processing method comprising:   A signal processing program for causing a calculation means to execute the signal processing method according to claim 8.   10. A recording medium in which the signal processing program according to claim 9 is recorded so as to be readable by an arithmetic means.

Images