JP2007274640A - Acoustic processing apparatus - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an acoustic processing apparatus capable of interlocking a graphic equalizer (GEQ) with a howling canceler, and preventing a state where the howling canceler automatically suppresses the howling from becoming a state where howling occurs again due to user's operation of the GEQ. <P>SOLUTION: A level of the GEQ of a set frequency of a notch filter is held when the howling is determined by the howling canceler. When an operator of each band of the GEQ is operated, a suppression level (attenuation quantity) of the notch filter varies by the difference of level of the frequency of the GEQ and the held level. This processing is applicable to a case where a plurality of times of howling are suppressed in a plurality of frequencies. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to a sound processing apparatus, and more particularly to a sound processing apparatus in which a howling canceller that suppresses howling and an equalizer are linked.

  Conventionally, a howling canceller technology that automatically suppresses howling is known. For example, in Patent Document 1, a test signal or an acoustic signal radiated from a speaker is picked up by a microphone, howling is detected based on the level or the like, and a detected howling frequency signal is used using a filter such as a notch filter. A technique for suppressing the occurrence of howling by attenuation is disclosed. It also describes that the attenuation of the notch filter is set in conjunction with the microphone volume (paragraph 0025).

On the other hand, some types of sound processing apparatuses include a GEQ (graphic equalizer). The GEQ includes a plurality of operators (control faders) corresponding to a plurality of frequency bands of an acoustic signal, and the level of each band can be controlled by each control fader.
JP2005-102093

  By the way, there is a sound processing apparatus in which a level control unit, a GEQ, and a howling canceller unit are connected in series. Conventionally, although the level control unit and the howling canceller unit are linked, the GEQ and the howling canceller unit There was no linkage. In GEQ, the level of each band can be set arbitrarily, and it is not easy to know which band of GEQ corresponds to the frequency in which the notch filter is included in the howling canceller. There was a case where howling occurred again after lifting.

  Also, when the volume is lowered and the notch attenuation is reduced, there is a possibility that the attenuation will be 0 dB or less. There wasn't. If the peak is reached at 0 dB, the original howling suppression setting cannot be restored when the volume is increased again. In the case of combination with GEQ, it is desirable to turn the notch with attenuation of 0 to suppress howling in other bands.

  According to the present invention, a GEQ and a howling canceller can be linked, and a state in which howling is automatically suppressed by the howling canceller is changed to a state in which howling occurs again by a user's GEQ operation. It is an object of the present invention to provide a sound processing apparatus that can prevent the above.

  In order to achieve the above object, the present invention maintains the GEQ level of the frequency set by the notch filter (howling suppression means) when howling is determined by a howling canceller, and When the operator is operated, the suppression level (attenuation amount) of the notch filter is changed by the difference between the level of the frequency of the GEQ and the held level. This processing can also be applied when a plurality of howlings at a plurality of frequencies are suppressed.

  When the GEQ level is decreased and the suppression level of the notch filter falls below a predetermined value, the notch filter may be opened. When the GEQ level is increased again, information regarding suppression may be retained in order to reproduce the operation of the notch filter. It should be noted that information regarding suppression is retained, and when the level is raised again, if the attenuation of the notch filter exceeds a predetermined attenuation, it is possible to perform suppression at that notch again. It is effective not only in combination with, but also in combination with simple level control.

  According to the present invention, howling does not occur even if the gain is increased by raising the fader of the band of the GEQ containing the notch filter. Also, howling does not recur even if the frequency is adjusted without worrying about where the notch filter is located in GEQ. The attenuation amount of each notch filter can be automatically increased or decreased independently according to the band and the operation amount in which the GEQ level is operated. (There is a premise that the attenuation of the notch filter should be as shallow as possible.) Also, the notches that are no longer needed can be opened and used to suppress howling occurring at other frequencies. Can be reduced. Even when the level is once lowered and then raised again, the necessary howling suppression can be reproduced.

  Embodiments of the present invention will be described below with reference to the drawings.

  FIG. 1 shows a configuration of a sound processing apparatus according to an embodiment of the present invention. This sound processing apparatus includes a central processing unit (CPU) 101, flash memory 102, RAM (random access memory) 103, mouse 104, keyboard 105, hard disk 106, display 107, other input / output interface (I / O) 108, waveform. An input unit 110, a signal processing unit 111, a waveform output unit 112, and a bus line 114 connecting these units are provided.

  The CPU 101 controls the overall operation of this sound processing apparatus. The flash memory 102 is a nonvolatile memory that stores various programs and data used by the CPU 101 and the signal processing unit 111 (DSP group). The RAM 103 is a volatile memory used for a load area and a work area for programs executed by the CPU 101. A mouse 104 and a keyboard 105 are operators for giving various instructions to the apparatus and inputting data. The hard disk 106 is an external storage device that stores programs and data. The display 107 is a display device for displaying various information. The other I / O 108 is an interface for connecting various external devices to this apparatus.

  The waveform input unit 110 is an interface for inputting an acoustic signal from an acoustic signal generation source 109 such as a microphone. For example, an analog acoustic signal is input and converted into a digital signal, which is passed to the signal processing unit 111 (A / D (analog) A digital) conversion function and a digital signal input function for inputting a digital sound signal and passing it to the signal processing unit 111 are realized. The signal processing unit 111 is composed of several DSPs (digital signal processors) and the like. These DSPs execute various micro programs based on instructions from the CPU 101 to perform howling canceller processing (howling suppression processing), equalizer processing, and the like, and output the processed waveform signal to the waveform output unit 112. . The waveform output unit 112 converts the digital sound signal output from the signal processing unit 111 into an analog sound signal and emits the sound via the sound system 113.

  FIG. 2 shows a process flow and function sharing between the howling canceller and the GEQ realized by the sound processing apparatus of FIG. An input 201 and an output 203 indicate an input by the waveform input unit 110 and an output by the waveform output unit 112 in FIG. Reference numeral 202 denotes processing performed by the signal processing unit 111 in FIG. 1, and reference numeral 204 denotes processing performed by the CPU 101 in FIG. In the signal processing unit 202, other processing 221, waveform analysis and notch (N) filter group processing 222, bandpass filter group processing 223, and other processing 224 are performed on the input signal, and output 203 is output. . In the CPU 204, determination and notch filter control processing 242, level designation processing 243 for each band, and other processing 241 are executed. 222 and 242 are processes for realizing a howling canceller, and 223 and 243 are processes for realizing GEQ.

  Of the processing 222, the waveform analysis is a processing for determining whether or not a howling component is included in the frequency band by sweeping from the lower one (or higher one) for each frequency band. The frequency band can be divided arbitrarily. For example, the range from 20 Hz to 20 kHz is divided into 31 frequency bands at equal intervals on a logarithmic scale. Whether or not a howling component is included in each frequency band is determined by dividing an input signal for each frequency band by a bandpass filter and determining whether or not a howling condition is satisfied in each band. The howling condition is a condition that “a state in which the level value is higher than a predetermined threshold and the pitch fluctuation is lower than the predetermined threshold continues for a predetermined time”. When this condition is satisfied, it is determined that howling has occurred.

  Of the processing 222, in the processing by the notch filter group, first, the notch filter corresponding to the frequency band determined to cause howling is operated to gradually lower the gain. At the same time, detection of howling by waveform analysis is continued, and when howling can be suppressed in a frequency band in which gain is reduced by the notch filter, the decrease in gain by the notch filter is stopped. At this time, it is determined which howl can be suppressed by decreasing to which gain in which frequency band. Thereafter, howling is suppressed by continuing the signal attenuation process using the notch filter. Note that the notch filter is connected in series with the input signal, and normally allows the signal to pass therethrough. When howling is detected, a notch filter is assigned to the frequency band, and howling is suppressed by the above processing.

  The process 222 of the signal processing unit 202 is executed based on an instruction and control from the process 242 of the CPU 204.

  A process 243 of the GEQ CPU 204 is a process of detecting operations of a plurality of faders for setting levels of each GEQ band (frequency band) and sending the set level values to the bandpass filter group 223. . The band-pass filter group includes a band-pass filter that controls the gain of the band of the input signal based on the level value set for the band.

  FIG. 3 shows a processing procedure of a howling suppression instruction event. This process is executed by the CPU 101 and the signal processing unit 111 in cooperation when the user instructs the operation start of the howling canceller, and is a process included in the processes 222 and 242 in FIG.

  In step 301, the analysis result by the waveform analysis processing 222 of the signal processing unit 202 is received. In step 302, it is determined whether howling has occurred based on the analysis result. If not, the process is terminated. If so, a notch filter for suppressing the howling is assigned at step 303. Here, the i-th notch filter is assigned. In step 304, the frequency causing howling is set in the assigned i-th notch filter. Let NF (i) be the frequency set for the i-th notch filter. In step 305, the attenuation amount NL (i) of the i-th notch filter is gradually increased until no howling is detected. If howling is suppressed, the level GL of the GEQ at the howling frequency is calculated in step 306, and the attenuation amount NL (i) of the i-th notch filter when no howling is detected in step 307 is defined as NLo (i). The GEQ level GL at that time is stored in GLo (i).

  FIG. 4 shows a flow of processing executed when there is an operation event of the fader (b) corresponding to the frequency band b of GEQ. This process is executed by the CPU 101 and the signal processing unit 111 in cooperation with each other when the fader is operated by the user, and is a process included in the processes 223 and 243 in FIG.

  In step 401, the level control of the band b is performed according to the set value of the fader (b). In step 402, an active notch filter whose | logNF-logBF | is equal to or smaller than a predetermined value ε is detected. NF is a frequency set in the notch filter, and BF indicates the center frequency of the band b. Therefore, a notch filter having | log NF−log BF | equal to or less than a predetermined value ε means that the frequency NF attenuated to suppress howling is within a predetermined vicinity range (on a logarithmic scale) from the center frequency of band b. The notch filter as shown in FIG. This means that the frequency NF is in a range affected by the level control of the band b.

  If there is no such notch filter in step 403, the process is terminated. If there is, in step 404, the first notch filter detected is set to i. In step 405, the GEQ level GL of the frequency NF (i) of the i-th notch filter is calculated. In step 406, GL-GLo (i) is substituted into Δ. Also, NLo (i) + Δ is substituted for NL (i). In short, Δ indicates the amount of change in the EQ level value (a value based on the level value GLo (i) when howling is suppressed) due to the operation of the fader. The attenuation amount NL (i) of the notch filter is stored as NLo (i) when the howling is suppressed. If the fader is not operated, howling is reduced if this attenuation amount NLo (i) is attenuated. It can be suppressed. However, since the level is changed by the change amount Δ by operating the fader, the notch filter attenuation amount NL (i) is assumed to be obtained by adding the change amount Δ to the reference attenuation amount NLo (i). This suppresses howling.

  After step 406, step 407 determines whether NL (i) is greater than zero. If larger, in step 408, the attenuation of the notch filter (i) is set to NL (i). When it is not large, it means that the GEQ level GL is narrowed down so that howling does not occur even if the notch filter is not used. The notch filter attenuates the input signal and does not amplify the level. Therefore, in step 409, the notch filter (i) is set to through. After Steps 408 and 409, the next notch filter is set to i in Step 410, and when there is no next notch filter in Step 411, the process ends. When there is a next notch filter i, the process returns to step 405 and continues.

  With the above processing, when the band b has a frequency affecting the notch filter, the fader is operated in the band GEQ b, and when the gain is increased, the attenuation amount of the notch filter is increased by the increase. Conversely, when the gain is reduced, howling is suppressed by reducing the attenuation amount of the notch filter by the reduced amount. However, when the gain is reduced and the amount of attenuation of the notch filter reaches zero, the notch filter is turned through. In this case, since the values of GLo (i) and NLo (i) are held as they are, attenuation processing for suppressing howling by the notch filter is appropriately performed when the gain is increased again by the fader.

  Note that step 420 in FIG. 4 is another example of setting of the notch filter through (step 409). Instead of setting to through, the notch filter (i) may be opened so that it can be assigned for other use.

  In the above-described embodiment, signal processing is performed by the signal processing unit (DSP), but signal processing may be performed by the CPU instead of the DSP. Further, howling suppression processing and equalizer processing may be performed by dedicated hardware that is not programmable. The signal processing performed by the signal processing unit may include one or more of level control processing, compressor processing, mixing processing, reverberation applying processing, and the like.

Configuration diagram of sound processing apparatus of embodiment Diagram showing how the canceling canceller and GEQ process flow and sharing of functions Flow chart showing processing procedure of howling suppression instruction event Flow chart of processing executed when there is a fader operation event

Explanation of symbols

  DESCRIPTION OF SYMBOLS 101 ... Central processing unit (CPU), 102 ... Flash memory, 103 ... RAM (random access memory), 104 ... Mouse, 105 ... Keyboard, 106 ... Hard disk, 107 ... Display, 108 ... Other input / output interface (I / O) , 109 ... microphone, 110 ... waveform input section, 111 ... signal processing section, 112 ... waveform output section, 113 ... sound system, 114 ... bus line.

Claims (4)

An input means for inputting an acoustic signal;
Howling determination means for analyzing the acoustic signal and determining whether howling has occurred;
A notch filter that attenuates the frequency component centered on the frequency indicated by the frequency data of the input acoustic signal by the attenuation indicated by the attenuation data;
When the howling determination means determines the occurrence of the howling, the frequency data indicating the detected howling frequency and the attenuation amount data indicating a predetermined attenuation amount are started to be supplied to the notch filter. The amount of attenuation indicated by the supplied attenuation amount data is gradually increased, and when the howling determination means stops determining the occurrence of the howling, the increase is stopped, thereby suppressing the detected howling. Control means;
A plurality of controls respectively corresponding to a plurality of frequency bands of the acoustic signal;
A plurality of level control means corresponding to the plurality of operation elements, and a plurality of level control means for controlling the level of the corresponding frequency band of the acoustic signal in accordance with each operation of the plurality of operation elements;
The level of the frequency indicated by the frequency data generated by the level control of the corresponding level control means in response to the operation of the frequency band including the frequency indicated by the frequency data among the plurality of operators being operated. When the level increases, the control means increases the attenuation indicated by the attenuation data supplied to the notch filter by the amount of change, and when the level decreases, the attenuation data Interlocking means for reducing the amount of attenuation indicated by the change amount,
An acoustic processing apparatus comprising: output means for outputting an acoustic signal processed by the notch filter and the plurality of level control means. The sound processing apparatus according to claim 1 has a plurality of the notch filters connected in series,
The howling determination means can sequentially determine the occurrence of a plurality of howlings at a plurality of different frequencies, and the control means sequentially applies one notch filter to each of the sequentially determined howlings. Allocation, and starts supplying frequency data indicating the frequency of the howling and the attenuation amount data indicating a predetermined attenuation amount to the assigned notch filter, and the attenuation amount indicated by the supplied attenuation amount data. By gradually increasing, by stopping the increase when the howling determination means no longer determines the occurrence of the howling, sequentially suppress the plurality of sequentially determined howling,
When one of the plurality of operating elements is operated, the interlocking unit includes a frequency band in which the frequency indicated by the frequency data supplied to the notch filter corresponds to the operating element. If there are one or more notch filters included in the signal, one or more frequencies indicated by frequency data supplied to the one or more notch filters generated by level control of the level control means corresponding to the operator The amount of change in level at each of the first and second notches is calculated, and the amount of attenuation indicated by the amount of attenuation data supplied to the one or more notch filters is increased when the level increases by the corresponding amount of change. And reducing the level when the level is reduced. The sound processing apparatus according to claim 1 or 2,
The level control means reduces the level of the predetermined frequency band of the acoustic signal, and the interlocking means reduces the attenuation amount indicated by the attenuation amount data accordingly, and supplies it to one of the plurality of notch filters. The sound processing apparatus further comprising means for releasing the notch filter from the howling and releasing the notch filter when the attenuation value indicated by the attenuation data is less than a predetermined value. . The sound processing apparatus according to claim 1 or 2,
The level control means reduces the level of any frequency band of the acoustic signal, and the interlocking means reduces the attenuation amount indicated by the attenuation amount data accordingly, so that any one of the plurality of notch filters is provided. Even if the attenuation amount indicated by the supplied attenuation data falls below a predetermined value, the frequency data and the information regarding the attenuation data are retained, and the level of the frequency band is increased again. Means for restarting howling suppression by the frequency data and the attenuation amount data again in any one of the plurality of notch filters when the attenuation amount indicated by the attenuation amount data increases to the predetermined value or more. An acoustic processing apparatus comprising:

Images