JP2010004250A - Noise canceling system, noise canceling signal forming method, and noise canceling signal forming program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To automatically obtain the most suitable noise canceling performance in accordance with generated noise without troubling a user. <P>SOLUTION: In addition to an FF filter circuit 51, an adaptive filter circuit 52 of which the transfer function is adaptively changed, a filter control part 53 for controlling the adaptive filter circuit 52, a switch circuit 54, a power calculation part 55, and a synthesizing part 56 are provided. A noise canceling signal formed by the FF filter circuit 51 and a noise canceling signal formed by the adaptive filter circuit 52 controlled by the filter control part 53 are synthesized by the synthesizing part 56, and the synthesis result is used as a noise canceling signal. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a noise canceling system, a noise canceling method, and a noise canceling program applied to, for example, headphones for listening to reproduced music and a headset for reducing noise.

  Conventionally, there is an active noise canceling system (noise reduction system) mounted on headphones. The noise canceling systems that are currently in practical use can be broadly divided into two methods, a feedback method and a feedforward method.

  In general, the noise canceling system of the feedback system has a feature that a relatively large reduction is possible although a band where noise can be canceled (band where noise can be reduced) is narrow.

  On the other hand, the feed-forward noise canceling system has a wide band where noise can be canceled and is stable, but noise does not match the transfer function assumed by the positional relationship with the noise source, etc. There is a possibility that it will increase.

  For this reason, when using a stable feed-forward noise canceling system with a wide band where noise can be canceled, noise in a specific narrow band is noticeable even if the band where noise is reduced is wide. Sometimes. In such a case, the listener (user) cannot feel the reduction effect.

  As one measure for solving such problems, Patent Document 1 discloses a stronger noise reduction effect by simultaneously moving the feedforward method in addition to the feedback method in a so-called noise canceling system. An invention has been disclosed that enables this. According to the invention described in Patent Document 1, a wide band in which noise can be canceled is obtained, and a large noise reduction effect can be obtained stably.

The above-mentioned Patent Document 1 is as follows.
JP 2008-116782 A

  By the way, in the case of the invention described in Patent Document 1 described above, the filter characteristics are constant in both the feedback method and the feedforward method, and the noise canceling characteristics are constant even when the noise changes. For this reason, in the case of the invention described in Patent Document 1, it is considered that there is a case where the optimum noise canceling performance may not be realized for each noise when strictly considered. For this reason, in order to realize the optimum noise canceling performance for each noise, it is conceivable to use an adaptive filter method.

In order to put noise canceling using an adaptive filter into practical use,
(1) An expensive DSP (Digital Signal Processor) that requires high computation performance for performing adaptive filter processing is required.
(2) Since adaptive filter processing is performed on frequency components with physically large energy, even if the adaptive filter processing operates, it cannot be said that noise canceling characteristics that are optimal in terms of auditory characteristics are necessarily realized.
There is a problem.

  In view of the above, the present invention eliminates the above-mentioned problems and automatically realizes optimum noise canceling performance according to the generated noise (noise) without bothering the user. The purpose is to.

In order to solve the above problem, a noise canceling system according to claim 1 is provided.
A first signal processing means for forming a noise cancellation signal from a signal collected by an external sound collecting means provided outside a housing attached to a user's ear; and an internal collector provided inside the housing. One or both of the second signal processing means for forming a noise cancellation signal from the signal collected by the sound means;
Means for forming a noise cancellation signal from the signal collected by the external sound collecting means, and a third signal processing means capable of adaptively changing the transfer function;
Control means for controlling to change the transfer function of the third signal processing means based on the signal from the external sound collecting means and the signal from the internal sound collecting means;
Synthesis means for synthesizing a noise cancellation signal from one or both of the first signal processing means and the second signal processing means and a noise cancellation signal from the third signal processing means;

  According to the noise canceling system of the first aspect, one or both of the first and second signal processing means for forming the noise cancellation signal are provided. The first signal processing means is of a so-called feed forward type that forms a noise cancellation signal from a signal from an external sound collecting means outside the housing. The second signal processing means is of a so-called feedback system that forms a noise cancellation signal from the signal from the internal sound collecting means inside the housing.

  In addition to one or both of these, there is provided an adaptive filter type third signal processing means capable of adaptively changing the transfer function to be used. Then, the noise cancellation signal formed by one or both of the first and second signal processing means and the noise cancellation signal formed by the third signal processing means are combined by the combining means.

  Thereby, noise that cannot be canceled (removed) by the noise cancellation signal formed by one or both of the first and second signal processing means is canceled by the noise cancellation signal formed by the third signal processing means. To be able to. And since the 3rd signal processing means can change the transfer function to be used, it can form the noise cancellation signal according to the noise which has generate | occur | produced, and can reduce the residual signal of noise effectively. To be done. Therefore, the optimum noise canceling performance can be automatically realized without bothering the user according to the generated noise (noise).

A noise canceling system according to claim 2 is the noise canceling system according to claim 1,
Of the first signal processing means and the second signal processing means, the first signal processing means is provided.

  According to the noise canceling system of the second aspect of the present invention, the first signal processing means for forming the noise cancellation signal from the signal collected by the external microphone among the first and second signal processing means. Is made to work.

  As a result, noise that cannot be canceled by the noise cancellation signal formed by the first signal processing means can be effectively removed by the noise cancellation signal formed by the third signal processing. . Therefore, the optimum noise canceling performance can be automatically realized without bothering the user according to the generated noise (noise).

A noise canceling system according to claim 3 is the noise canceling system according to claim 1,
Of the first signal processing means and the second signal processing means, the second signal processing means is provided.

  According to the noise canceling system of the third aspect of the present invention, the second signal processing means for forming the noise cancellation signal from the signal collected by the internal microphone among the first and second signal processing means. Is made to work.

  Thereby, the noise that cannot be canceled by the noise cancellation signal formed by the second signal processing means can be effectively removed by the noise cancellation signal formed by the third signal processing. . Therefore, the optimum noise canceling performance can be automatically realized without bothering the user according to the generated noise (noise).

A noise canceling system according to a fourth aspect of the present invention is the noise canceling system according to the first aspect,
Both the first signal processing means and the second signal processing means are provided.

  According to the noise canceling system of the fourth aspect of the present invention, the first signal processing means for forming the noise cancellation signal from the signal collected by the external microphone, and the noise from the signal collected by the internal microphone. Both the second signal processing means for forming the cancel signal are made to function.

  Thereby, the noise that cannot be canceled by the noise cancellation signal formed by the first and second signal processing means can be effectively removed by the noise cancellation signal formed by the third signal processing. To be. Therefore, the optimum noise canceling performance can be automatically realized without bothering the user according to the generated noise (noise).

The noise canceling system according to claim 5 of the present application is the noise canceling system according to claim 1, claim 2, claim 3 or claim 4,
A plurality of each of the third signal processing means and the control means;
By providing weighting means for weighting the supplied signal for each of the plurality of control means provided,
A plurality of the third signal processing means having different movable ranges are provided.

  According to the noise canceling system of the fifth aspect of the present invention, a plurality of third signal processing means having different movable ranges are provided.

  Thereby, the noise that cannot be canceled by the noise cancellation signal formed by one or both of the first and second signal processing means is effectively removed by the noise cancellation signal formed by the third signal processing. To be able to.

  In this case, the third signal processing means is configured as a plurality of third signal processing means having different movable ranges. As a result, for example, even if noise remains in different ranges such as the low frequency side and the high frequency side, any of the frequency ranges generated by the third signal processing means in the corresponding movable range is different. The noise can be effectively canceled. Therefore, the optimum noise canceling performance can be automatically realized without bothering the user according to the generated noise (noise).

  According to the present invention, it is possible to adaptively change the transfer function to be used in addition to one or both of the first and second signal processing means for forming the noise cancellation signal according to the transfer function used in a fixed manner. The third signal processing means can also be used in combination. Thereby, it is possible to automatically realize the optimum noise canceling performance for each generated noise (noise).

  Hereinafter, an embodiment of an apparatus, a method, and a program according to the present invention will be described with reference to the drawings.

[Noise canceling system]
Currently, a system that actively reduces external noise for headphones and earphones, a so-called noise canceling system, has begun to spread. As for products that have been commercialized, most of them are constituted by analog circuits, and noise canceling methods include a feedback method and a feedforward method. There is an adaptive filter system as a principle system. Thus, the noise canceling method can be roughly divided into three methods.

  First, in order to facilitate the explanation of the present invention, with reference to FIGS. 1 to 13, each configuration example and operation principle of a conventional noise canceling system of a feedback system, a feedforward system, and an adaptive filter system will be described. Will be described.

  In addition, regarding the feedback method and the feedforward method, the system can be constructed by either digital or analog. However, with regard to the adaptive filter system, it is necessary to perform adaptive signal processing digitally. For this reason, in the following description, in order to simplify the description, it is assumed that all the filters that are the dominant factors that determine the noise cancellation characteristics are configured by digital circuits.

  In the following description and in the drawings, the feedback method may be abbreviated as FB, the feedforward method as FF, the adaptive filter method as AF, and the noise cancellation as NC, if necessary. is there.

[Feedback type noise canceling system]
First, a feedback type noise canceling system will be described. FIG. 1 is a diagram for explaining a feedback type noise canceling system. FIG. 2 is a diagram showing noise reduction characteristics of a feedback type noise canceling system, and FIG. 7 is a diagram for explaining a calculation formula showing characteristics of the feedback type noise canceling system.

  FIG. 1A shows a configuration on the right channel side when a headphone system to which a feedback type noise canceling system is applied is mounted on a user head (the head of a user (listener)) HD. . FIG. 1B shows the overall configuration of the feedback type noise canceling system.

  As shown in FIG. 1A, the feedback system generally includes a microphone (hereinafter abbreviated as a microphone) 111 inside a headphone housing (housing) HP.

  And the noise which came into headphone housing | casing HP from the outside is attenuate | damped by carrying out return control of the anti | reverse | negative phase component (noise reduction signal) of the signal (noise signal) collected with the said microphone 111.

  In this case, since the position of the microphone 111 becomes a cancellation point (control point) CP corresponding to the listener's ear position, the noise attenuation effect is taken into consideration, and the position close to the listener's ear, that is, the diaphragm of the driver 15 is usually considered. A microphone 111 is often placed on the front.

  Specifically, a feedback type noise canceling system will be described with reference to the block diagram of FIG. The feedback type noise canceling system shown in FIG. 1B includes a microphone and a microphone amplifier unit 11 including a microphone 111 and a microphone amplifier 112.

  Further, a filter circuit (hereinafter referred to as FB filter circuit) 12 designed for feedback control, a synthesis unit 13, a power amplifier 14, a driver 15 including a drive circuit 151 and a speaker 152, and an equalizer 16 are provided. I have.

  In FIG. 1B, characters A, D, M, and −β described in each block are transfer functions of the power amplifier 14, the driver 15, the microphone and microphone amplifier unit 11, and the FB filter circuit 12.

  Similarly, in FIG. 1B, the letter E in the block of the equalizer 16 is a transfer function of the equalizer 16 that is multiplied by the signal S that is to be listened to. The letter H of the block placed between the driver 15 and the cancellation point CP is a transfer function of the space from the driver 15 to the microphone 111 (transfer function between the driver and the cancellation point). Each of these transfer functions is assumed to be expressed in a complex manner.

  1A and 1B, the letter N is noise that has entered the vicinity of the microphone position in the headphone housing HP from an external noise source (noise source) NS, and the letter P is a listener. It represents the sound pressure that reaches the ear (output voice).

  The cause of the noise N being transmitted into the headphone housing HP is, for example, the case where the sound pressure leaks from the gap of the ear pad portion of the headphone housing HP, or the result of the headphone housing HP vibrating due to the sound pressure. It is conceivable that sound is transmitted to the inside of the housing.

  At this time, in FIG. 1B, the sound pressure P reaching the listener's ear can be expressed as shown in the equation (1) in FIG. In the equation (1) of FIG. 7, if attention is paid to the noise N, it can be seen that the noise N is attenuated to 1 / (1 + ADHMβ). However, in order for the system of equation (1) in FIG. 7 to oscillate and operate stably as a noise canceling mechanism in the noise reduction target band, equation (2) in FIG. 7 needs to be satisfied. There is.

  In this feedback method, the designer designs the filter in consideration of human auditory characteristics after establishing the formula (2) in FIG. In many cases, the filter design considering the auditory characteristics is evaluated by the designer himself. However, many feedback-type noise-cancelling headphones have already been developed and put on the market and have been proven.

  Next, in the feedback-type noise canceling system shown in FIG. 1B, a case where necessary sound is reproduced from the headphones in addition to the above-described noise reduction function will be described.

  In addition to the music signal from the music playback device, the input voice S in FIG. 1B is a voice signal (as a headset) via communication such as a microphone sound outside the housing (when used as a hearing aid function) or telephone communication. This is a general term for audio signals that should be reproduced by a headphone driver.

  In the equation (1) in FIG. 7, when attention is paid to the input sound S, the transfer function E of the equalizer 16 can be expressed as the equation (3) in FIG. Then, considering also the transfer function E of the equalizer 16 in the equation (3) in FIG. 7, the output speech P of the noise canceling system in FIG. 1 (B) should be expressed as in the equation (4) in FIG. Can do.

  If the position of the microphone 111 is very close to the ear position, the letter H is the transfer function from the driver 115 to the microphone 111 (ear), and the letters A and D are the transfer functions of the power amplifier 114 and the driver 115, respectively. It can be seen that the same characteristics as those of headphones without a normal noise reduction function can be obtained. At this time, the transfer characteristic E of the equalizer 16 is substantially the same as the open loop characteristic seen on the frequency axis.

  The characteristics of the feedback type noise canceling system shown in FIG. 1 can be schematically shown as shown in FIG. That is, before noise cancellation, when noise as shown in FIG. 2A is generated, the function of the FB filter circuit 12 having the characteristics shown in FIG. As shown, noise can be greatly reduced.

[Feed-forward noise canceling system]
Next, a feedforward type noise canceling system will be described. FIG. 3 is a diagram for explaining a feedforward type noise canceling system. FIG. 4 is a diagram showing noise reduction characteristics of the feedforward type noise canceling system, and FIG. 8 is a diagram for explaining a calculation formula showing characteristics of the feedforward type noise canceling system. .

  FIG. 3A shows a configuration on the right channel side when a headphone system to which a feedforward type noise canceling system is applied is mounted on a user head (head of a user (listener)) HD. Yes. FIG. 3B shows the overall configuration of the feedforward type noise canceling system.

  In the feedforward method, a microphone 211 is basically installed outside the headphone housing HP as shown in FIG. It is intended that the noise collected by the microphone 211 is appropriately filtered and reproduced by the driver 25 inside the headphone housing HP, and this noise is canceled near the ear. .

  Specifically, a feedforward type noise canceling system will be described with reference to the block diagram of FIG. The feedforward type noise canceling system shown in FIG. 3B includes a microphone and a microphone amplifier unit 21 including a microphone 211 and a microphone amplifier 212.

  Furthermore, a filter circuit (hereinafter referred to as FF filter circuit) 22 designed for feedforward control, a synthesis unit 23, a power amplifier 24, a driver 25 including a drive circuit 251 and a speaker 252 are provided. .

  Also in the feedforward type noise canceling system shown in FIG. 3B, the letters A, D, and M described in each block are transmitted to the power amplifier 24, the driver 25, the microphone, and the microphone amplifier unit 21, respectively. It is a function.

  In FIG. 3, a letter N indicates an external noise source (noise source). The main reason why noise corresponding to the noise source N enters the headphone housing HP is as described in the feedback type noise canceling system.

  In FIG. 3B, the transfer function (transfer function between the noise source and the cancellation point) from the position of the external noise source N to the ear position CP is represented by the letter F.

  In FIG. 3B, the transfer function from the noise source N to the microphone 211 (transfer function between the noise source and the microphone) is represented by the letter F ′.

  In FIG. 3B, the transfer function from the driver 25 to the cancel point (ear position) CP (transfer function between the driver and the cancel point) is represented by the letter H.

  When the transfer function of the FF filter circuit 22 that is the core of the feedforward type noise canceling system is set to −α, the sound pressure P (output sound) reaching the listener's ear in FIG. , Can be expressed as shown in equation (1) in FIG.

  Here, considering an ideal state, the transfer function F between the noise source and the cancellation point can be expressed as shown in the equation (2) in FIG. Then, if the equation (2) in FIG. 8 is substituted into the equation (1) in FIG. 8, the first term and the second term are canceled out.

  Therefore, as a result, in the feedforward type noise canceling system shown in FIG. 3B, the output voice P can be expressed as shown in the expression (3) in FIG. Therefore, as can be seen from the expression (3) in FIG. 8, the noise is canceled, and only the music signal (or the audio signal to be listened to) remains, and the sound similar to the normal headphone operation can be heard.

  However, in reality, it is difficult to construct a complete filter having a transfer function that completely satisfies the equation (2) shown in FIG. Especially in the middle and high range, the shape of the ear varies from person to person, and the wearing state of the headphones varies. In addition, the characteristics vary depending on the position of the noise and the microphone position. For these reasons, the active noise reduction processing is usually not performed in the middle and high ranges, and passive sound insulation is often performed in the headphone housing.

  Note that the expression (2) in FIG. 8 is self-evident from the expression, but means that the transfer function from the noise source to the ear position is imitated by an electric circuit including the transfer function α.

  In this feed-forward method, the designer designs a filter in consideration of human auditory characteristics after establishing the formula (2) in FIG. In many cases, designers themselves evaluate filter designs that take this auditory characteristic into account, but many feed-forward noise-cancelling headphones have been developed and put on the market and have a proven track record. It can be said.

  Also, the cancellation point CP in the feedforward type noise canceling system shown in FIG. 3 is different from the feedback type noise canceling system in FIG. 1A as shown in FIG. It can be set at any ear position.

  However, in the normal case, the transfer function α is fixed, and at the design stage, the decision is made for some target characteristics. However, because the shape of the ear is different depending on the listener, there is a possibility that a sufficient noise cancellation effect cannot be obtained, or noise components are added in a non-reverse phase, causing abnormal noise. is there.

  From these facts, the feedforward method is generally less likely to oscillate and has high stability, but it is difficult to obtain a sufficient amount of attenuation. Attention should be paid to the stability of the system. Each of the feedback method and the feedforward method has characteristics.

  The characteristics of the feedforward type noise canceling system shown in FIG. 3 can be schematically shown as shown in FIG. That is, before noise cancellation, when noise as shown in FIG. 4A is generated, the function of the FF filter circuit 22 having the characteristics shown in FIG. As shown, noise can be greatly reduced.

[Adaptive filter type noise canceling system]
Next, an adaptive filter type noise canceling system will be described. FIG. 5 is a diagram for explaining an adaptive filter type noise canceling system. FIG. 6 is a diagram illustrating noise reduction characteristics of an adaptive filter type noise canceling system, and FIG. 9 is a diagram for explaining a calculation formula indicating characteristics of the adaptive filter type noise canceling system. .

  FIG. 5 shows the overall configuration of the adaptive filter type noise canceling system. The adaptive filter system has a configuration in which microphones are installed both inside and outside the headphone housing. Then, after applying adaptive filter processing to the noise collected by the external microphone, it is intended to be reproduced by a driver inside the housing and cancel the noise near the ear.

  Specifically, an adaptive filter type noise canceling system will be described with reference to the block diagram of FIG. The adaptive filter type noise canceling system shown in FIG. 5 includes a microphone 311 and a microphone amplifier 312, and includes an external microphone and a microphone amplifier unit 31 provided outside the headphone housing.

  Further, a filter circuit (hereinafter referred to as an adaptive filter circuit) 32 designed for adaptive control and a filter control unit 33 that controls the adaptive filter circuit 32 are provided. Further, a synthesis unit 34, a power amplifier 35, a driver 36 including a drive circuit 361 and a speaker 362 are provided.

  Furthermore, the adaptive filter type noise canceling system shown in FIG. 5 includes a microphone 371 and a microphone amplifier 372, and includes an internal microphone and a microphone amplifier unit 37 provided inside the headphone housing.

  In the adaptive filter type noise canceling system shown in FIG. 5, characters M1, A, D, and M2 described in each block are an external microphone and microphone amplifier unit 31, a power amplifier 35, a driver 36, an internal microphone, and It is each transfer function of the microphone amplifier unit 37.

  In FIG. 5, a letter N indicates an external noise source (noise source). The main reason why noise corresponding to the noise source N enters the headphone housing HP is as described in the feedback type noise canceling system.

  In FIG. 5, the transfer function (transfer function between the noise source and the cancellation point) from the position of the external noise source N to the ear position CP is represented by the letter F. In FIG. 5, the transfer function from the noise source N to the external microphone 311 (the transfer function between the noise source and the microphone) is represented by the letter F ′.

  In FIG. 5, the transfer function (transfer function between the driver and the cancellation point) from the driver 36 to the cancellation point (ear position) CP is represented by the letter H.

  Then, the adaptive filter processing in the adaptive filter circuit 32 is performed by the control of the filter control unit 33 using the signal collected by the external microphone 311 and the signal collected by the internal microphone 371 as shown in FIG. Then, filter adaptive processing is performed so that the signal from the internal microphone 371 is minimized.

  Here, if the transfer function of the adaptive filter circuit 32 serving as the core of the adaptive filter type noise canceling system is set as K, the adaptive filter type noise canceling system shown in FIG. It can be expressed as follows.

  Here, when an ideal state is considered, the transfer function F between the noise source and the cancellation point can be expressed as the equation (2) in FIG. Then, if the equation (2) in FIG. 9 is substituted into the equation (1) in FIG. 9, the first term and the second term are canceled out.

  Therefore, as a result, in the noise canceling system of the adaptive filter system shown in FIG. 5, the output speech P can be expressed as shown in equation (3) in FIG. Therefore, as can be seen from the equation (3) in FIG. 9, the noise is canceled and only the music signal (or the audio signal to be listened to) remains, and the sound similar to the normal headphone operation can be heard.

  However, in practice, it is difficult to construct a complete filter having a transfer function that completely satisfies the equation (2) shown in FIG. Especially in the middle and high range, the shape of the ear varies from person to person, and the wearing state of the headphones varies. In addition, the characteristics vary depending on the position of the noise and the microphone position. For these reasons, the active noise reduction processing is usually not performed in the middle and high ranges, and passive sound insulation is often performed in the headphone housing.

  It should be noted that equation (2) in FIG. 9 is self-evident when looking at the equation, but imitating the transfer function from the noise source to the ear position with an electric circuit including the transfer function K of the adaptive filter circuit 32. I mean.

  Here, the cancellation point CP is different from that in the feed-forward method shown in FIG. 3, and the sound at the microphone position becomes the control point as in the feedback method shown in FIG. For this reason, in consideration of the noise attenuation effect, a microphone is often placed near the normal ear, that is, in front of the diaphragm of the driver 36.

  Since the transfer function K of the adaptive filter circuit 32 is sequentially updated by the filter control unit 33, it is not a fixed value but variable, unlike the transfer function β of the FB filter circuit 12 and the transfer function α of the FF filter circuit 22. Value.

  As this adaptive algorithm, the LMS algorithm is generally used. The LMS algorithm monitors the error signal at the cancellation point (= reduced but remaining noise signal) and reduces the mean square of this error signal based on the steepest descent method to the input signal from the external microphone. The transfer function K of the adaptive filter circuit is updated.

  This LMS algorithm is a typical adaptive algorithm due to the advantages of being stable and having a small amount of calculation. The basics of the LMS algorithm are described in detail in the following papers, for example.

Paper on LMS algorithm
Adaptive noise cancelling: Principles and applications
Widrow, B .; Glover, JR, Jr .; McCool, JM; Kaunitz, J .; Williams, CS; Hearn, RH; Zeidler, JR; Eugene Dong, Jr .; Goodlin, RC
Proceedings of the IEEE
Volume 63, Issue 12, Dec. 1975 Page (s): 1692-1716

  In the adaptive filter type noise canceling system, due to the variability of the transfer function K of the adaptive filter circuit 32, the shape of the ear differs depending on the person, so that a sufficient noise canceling effect cannot be obtained, or the noise characteristics There is an advantage that the system can be automatically adjusted to be close to the optimum state when the change occurs.

  However, the optimal state in this case refers to a state where the physical energy of noise is minimum. However, human auditory characteristics are not flat in frequency characteristics (so-called loudness characteristics), and energy is divided and evaluated for each band having a bandpass-like independence of a frequency component as a critical band. For this reason, the state with the minimum physical energy is not necessarily the optimum state for human hearing.

Regarding the above loudness characteristics, the isosensitive curve of human hearing is shown in the international standard ISO226. In addition, the address of the Web page regarding the isometric curve is as shown below.
http://www.aist.go.jp/aist_j/press_release/pr2003/pr20031022/pr20031022.html

Moreover, the following are typical documents relating to the above critical band.
Critical Band Width in Loudness Summation
E. Zwicker, G. Flottorp, and SS Stevens
Psycho-Acoustic Laboratory, Harvard University, Cambridge, Massachusetts
The Journal of the Acoustical Society of America
May 1957 Volume 29, Issue 5, pp. 548-557

  The characteristics of the adaptive filter type noise canceling system shown in FIG. 5 can be schematically shown in FIG. As shown in FIG. 6, in the case of the adaptive filter method, the transfer function K of the adaptive filter circuit 32 is updated so that the largest portion (peak) of the frequency component of the noise signal is reduced. Therefore, in FIG. 6, various operation modes are possible as shown in the left, middle, and right diagrams.

  That is, before noise cancellation, when large noise is generated on the low frequency side as shown in FIG. 6A, the function of the adaptive filter circuit 32 having the characteristics shown in FIG. As shown in FIG. 6C, noise can be reduced.

  Before noise cancellation, when large noise is generated on the high frequency side as shown in FIG. 6D, the function of the adaptive filter circuit 32 having the characteristics shown in FIG. As shown in FIG. 6F, noise can be reduced.

  Furthermore, before noise cancellation, when a large noise is generated only in the high range as shown in FIG. 6G, the function of the adaptive filter circuit 32 having characteristics as shown in FIG. As shown in FIG. 6I, high frequency noise can be reduced.

  However, it cannot be said that all the states shown in FIG. For this reason, it is necessary to weight the LMS algorithm on the frequency axis in consideration of human auditory characteristics (frequency characteristics / critical band).

  FIG. 10 shows a configuration example of an adaptive filter type noise canceling system in consideration of this weighting. This method requires filter processing for weighting. In this weighting filtering process, it is necessary to apply the same to the input signal to the filter control unit 33 and the error signal necessary for updating the coefficient (transfer function) K in the adaptive filter 32.

  This is because, in principle, the coefficient update is performed using the correlation between the input signal and the error signal when the LMS algorithm minimizes the error signal. Therefore, as shown in FIG. 10, a weighting filter 38 a is provided between the external microphone and microphone amplifier unit 31 and the filter control unit 33. A weighting filter 38 b is provided between the internal microphone / microphone amplifier unit 37 and the filter control unit 33.

  Each of the weighting filters 38a and 38b has a transfer function of W. Further, in the noise canceling system shown in FIG. 10, each part other than the weighting filters 38a and 38b is configured in the same manner as the adaptive filter type noise canceling system shown in FIG.

  For this reason, in FIG. 10, the same reference numerals are given to the parts configured in the same manner as the noise canceling system shown in FIG. 5, and the description of the parts is omitted.

  FIG. 11 is a schematic diagram of characteristics of the adaptive filter type noise canceling system using the weighting shown in FIG. In FIG. 11, as shown in FIG. 11A, it is assumed that the signal before weighting has a higher peak in the high band than the low band, but the peak in the low band is more important auditoryly.

  Here, due to the effect of the weighting filters 38a and 38b having the characteristics as shown in FIG. 11B, the signal to be processed is apparently converted into the frequency characteristics as shown in FIG. The adaptive filter circuit 32 performs coefficient updating so as to minimize energy for the converted signal. In this case, the characteristic of the adaptive filter circuit 32 is as shown in FIG.

  For this reason, the characteristic of the signal to be processed obtains an apparent noise canceling result as shown in FIG. However, the result shown in FIG. 11E is a result in a state where the weighting filters 38a and 38b are applied.

Therefore, in reality, an apparent result obtained by multiplying the coefficient W ^-1 opposite to the coefficient W of the weighting filters 38a and 38b is obtained, as shown in FIG. 11 (F). In other words, the physical energy is not minimized by the weighting filters 38a and 38b, but auditory energy is minimized.

  However, as described with reference to FIG. 8, there is a problem that even if weighting is successful, the amount of calculation becomes very large. Further, if the amount of delay generated by the filtering process is large, there is a risk that the LMS algorithm will diverge. For this reason, there is a problem that system design is very difficult.

  In addition, the adaptive filter processing using the LMS algorithm does not always reach the minimum solution (Global minimum), but may reach the local minimum solution (Local minimum).

  Of course, there is a difference in the amount of noise reduction when reaching the minimal solution and when reaching the local minimum solution. There may be a problem that the amount cannot be obtained.

  As a result, although the adaptive filter method generally has a system automatic adjustment function, there is no guarantee that a sufficient reduction amount can be obtained. Is almost never used.

[Noise canceling system that combines feedback method and feedforward method]
It is also conceivable to superimpose the feedback type noise canceling system and the feedforward type noise canceling system described above. FIG. 12 shows a noise canceling system system in which the feedback method and the feedforward method are overlapped.

  12, parts that are configured in the same manner as the feedback-type noise canceling system shown in FIG. 1 are denoted by the same reference numerals as the corresponding parts of the noise canceling system in FIG. Further, in FIG. 12, the same reference numerals as those in the corresponding part of the noise canceling system in FIG. 3 are attached to the same parts as those in the feedforward type noise canceling system shown in FIG. Therefore, the details of each part will be omitted because they are redundantly described.

  In this case, it is assumed that one power amplifier and one driver are installed in each case for feedback and feedforward. For this reason, in FIG. 12, the transfer functions of the feedforward power amplifier 24 and the driver 25 are indicated by letters A1 and D1, and the transfer functions of the feedback power amplifier 14 and the driver 15 are indicated by letters A2 and D2. .

  Similarly, the transfer function between the feedforward driver and the cancellation point is indicated by the letter H1, and the transfer function between the feedback driver and the cancellation point is indicated by the letter H2.

  In the noise canceling system having the configuration shown in FIG. 12, first, external sound is captured by a feedforward type component. However, depending on the sound source and the nature of the sound wave (behavior like a spherical wave or a plane wave), a band where noise is reduced can be obtained inside the headphones as described above, while a band where noise occurs. Can also happen.

  The same problem occurs in the wearing state of the headphones and the shape of the individual ear. The feedback system components (feedback mechanism) are effective for the noise in the housing that remains or is generated by this feedforward system, and by moving the two simultaneously, the above effects can be obtained. be able to.

  That is, the noise canceling system shown in FIG. 12 includes the feedback type noise canceling system part shown in FIG. 1 and the feedforward type noise canceling system part shown in FIG. is there. In other words, the noise canceling system shown in FIG. 12 is a twin-type noise canceling system.

  In the case of the twin system, the characteristics of the feedback type noise canceling system and the characteristics of the feed forward type noise canceling system are combined to have an attenuation characteristic.

  The characteristics of the twin-type noise canceling system shown in FIG. 12 can be schematically shown as shown in FIG. That is, before noise cancellation, when noise as shown in FIG. 13A is generated, the FB filter circuit 12 and the FF filter circuit 22 having characteristics as shown in FIG. As shown in FIG. 13C, noise can be greatly reduced.

  As described above, since the noise canceling system of the feedback method and the feedforward method has a constant canceling characteristic, the noise can be sufficiently reduced depending on the characteristics of the generated noise. It may be impossible.

  In the case of an adaptive filter type noise canceling system, it is considered that an expensive DSP may be required or an optimal canceling characteristic may not always be realized in terms of auditory characteristics.

  Therefore, the invention according to this application solves these problems and makes it possible to always cancel noise with high accuracy.

[Embodiments of the present invention]
FIG. 14 is a block diagram illustrating an example of a twin-type noise canceling system. That is, the twin block diagram shown in FIG. 12 can be modified as shown in FIG.

  In the configuration shown in FIG. 14, a DSP (Digital Signal Processor) / CPU (Central Processing Unit) unit 40 is considered as a system that performs only noise canceling except for the equalizer 30 in view of realizing an adaptive filter circuit. This is equivalent to the adaptive filter type noise canceling system shown in FIG.

  For this reason, in the noise canceling system shown in FIG. 14, the same reference numerals are given to the parts configured in the same manner as the noise canceling system shown in FIG. 5, and the description of the parts is omitted.

  In the noise canceling system system of FIG. 14, the FF filter circuit 41 and the FB filter circuit 42 realized by the DSP / CPU unit 40 are time-invariant systems. In contrast, the adaptive filter circuit 32 of the noise canceling system of FIG. 5 is a time-varying system.

  In the case of the noise canceling system of FIG. 14, it is not necessary to perform an operation for sequential updating of the FF filter circuit 41 and the FB filter circuit 42 which are noise canceling filters, and the stability of the filter is constant, but a high noise canceling effect The point is that can be obtained.

  However, since the system is digitized, it may be considered that the block diagram is equivalent to the adaptive filter method. In the DSP / CPU 40, it is possible to add an adaptive filter that performs sequential update processing, and many new advantages arising from this point can be considered. These advantages are not possible when considering the entire noise cancellation system on an analog basis just a year or two ago, and can be finally taken into account by the remarkable performance improvement of digital technology in recent years. It has become.

  The invention of this application cancels noise with high accuracy in consideration of the characteristics of the generated noise by variously changing the configuration of the part shown by the DSP / CPU 40 shown in FIG. Is to be able to.

  Therefore, each noise canceling system to which one embodiment of the present invention described below is applied basically has the same configuration as the noise canceling system shown in FIG.

  For this reason, in the figure used in the embodiment described below, the same reference numerals are given to the parts configured in the same manner as the noise canceling system shown in FIG. To avoid this, it will be omitted. Hereinafter, the present invention will be described in detail.

[First Embodiment]
FIG. 15 is a block diagram for explaining a noise canceling system according to the first embodiment to which one embodiment of the present invention is applied. In the case of the noise canceling system shown in FIG. 15, the DSP / CPU unit 50 realizes the function of the FF filter circuit 51 and the function of the adaptive filter circuit 52.

  That is, as shown in FIG. 15, the DSP / CPU 50 includes an FF filter circuit 51, an adaptive filter circuit 52, a filter control circuit 53 that controls the adaptive filter circuit 52, a switch circuit 54, and a power calculation unit 55. And a combining unit 56 is provided.

  As shown in FIG. 15, the output signal from the external microphone and microphone amplifier unit 31 is supplied to the FF filter circuit 51, the adaptive filter circuit 52, and the filter control unit 53. The output signal from the internal microphone and microphone amplifier unit 37 is supplied to the input terminals of the adaptive filter circuit 52, the power calculation unit 55, and the switch circuit 54.

  As will be described later, immediately after the operation, the power calculation unit 55 confirms whether or not the power of the output signal from the internal microphone and the microphone amplifier unit 37 has been reduced by the function of the feedforward type noise canceling system. .

  When the output power from the internal microphone and the microphone amplifier unit 37 is reduced, the noise can be expected to be further reduced by the adaptive filter type noise canceling system. Circuit 54 is turned on. As a result, the output signal from the internal microphone and microphone amplifier unit 37 is supplied to the filter control unit 53.

  The filter control unit 53 operates when both the output signal from the external microphone and microphone amplifier unit 31 and the output signal from the internal microphone and microphone amplifier unit 37 are supplied. The filter control unit 53 controls the transfer function (coefficient) K of the adaptive filter circuit 52 based on the output signal from the external microphone and microphone amplifier unit 31 and the output signal from the internal microphone and microphone amplifier unit 37. .

  When the power calculation unit 55 confirms that the output power from the internal microphone and the microphone amplifier unit 37 has decreased to a predetermined value or less, the power calculation unit 55 turns off the switch circuit 54. Thus, the adaptive filter circuit 52 holds the adaptively changed transfer function K so that the adaptive filter circuit 52 appropriately performs noise reduction processing.

Therefore, in the case of the noise canceling system shown in FIG.
(1) First, an input signal from the external microphone and microphone amplifier unit 31 is processed by a normal feedforward type noise canceling system portion, and noise is reduced.
(2) The transfer function (coefficient) K of the adaptive filter 52 is updated by the filter control unit 53 using signals from the external microphone and microphone amplifier unit 31 and the internal microphone and microphone amplifier unit 37.
(3) The input to the internal microphone and microphone amplifier unit 37 is minimized (not necessarily minimal).
The noise cancellation process is performed by the procedure described above.

  Thereby, in the case of the noise canceling system shown in FIG. 15, a large reduction amount can be realized as compared with the case where noise canceling is performed only by the feedforward method. (Advantage 1) can be realized.

  In the twin type noise canceling system shown in FIGS. 12 and 14, the transfer function β of the FB filter circuit is fixed, whereas in the case of the noise canceling system shown in FIG. The transfer function (coefficient) K of the circuit 52 is variable. As a result, the noise can be minimized by automatically following the ever-changing noise that cannot be completely erased by the function of the FF filter circuit 51. (Advantage 2) can be realized.

  In addition, compared with a simple adaptive filter type noise canceling system, the function of the FF filter circuit 51 can be used in advance to design an acoustic noise canceling characteristic to some extent. Since it is only necessary to focus on the residual signal, an auditory design is easier than a noise canceling system using only the adaptive filter method. (Advantage 3) can be realized.

  FIG. 16 is a flowchart for explaining the operation of the noise canceling system in which the feedforward method and the adaptive filter method shown in FIG. 15 are combined. The processing in FIG. 16 is executed in the DSP / CPU unit 50 when the noise canceling system is powered on or when an execution instruction is received from the user through an operation unit (not shown).

  First, in the DSP / CPU unit 50, the first noise canceling system (in this example, a feedforward type noise canceling system) is operated (step S101).

  Then, the power calculation unit 55 of the DSP / CPU unit 50 determines whether or not the output power from the internal microphone and microphone amplifier unit 37 has decreased (step S102). The power deriving method here is obtained by, for example, a method of obtaining the sum of squares while holding data for a certain section. It is possible to determine whether or not the output power has decreased according to the change in power obtained by the method.

  If it is determined in step S102 that the output power from the internal microphone and microphone amplifier unit 37 has not decreased, it indicates that the first noise canceling system is not performing noise cancellation. In this case, since the user cannot enjoy the benefit of noise (noise), which is the original purpose, the process shown in FIG. 16 ends.

  If it is determined in step S102 that the output power from the internal microphone and microphone amplifier unit 37 has decreased as originally designed, the second noise canceling system is operated (step S103). In the first embodiment, the second noise canceling system is an adaptive filter type noise canceling system.

  In this case, the power calculation unit 55 turns on the switch circuit 54, causes the filter control unit 53 to function, and starts processing for adaptively updating the transfer function (coefficient) K of the adaptive filter circuit 52 (step S104).

  Then, the power calculation unit 55 determines whether or not the output power from the internal microphone and the microphone amplifier unit 37 has decreased and the power has fallen below a predetermined reference value Plow (step S105).

  When it is determined in step S105 that the output power from the internal microphone and microphone amplifier unit 37 is not lower than the reference value Plow, the processing from step S105 is repeated. As a result, the transfer function (coefficient) K of the adaptive filter circuit 52 is continuously updated and waits for the output power to fall below the reference value Plow.

  When it is determined in step S105 that the output power from the internal microphone and microphone amplifier unit 37 has fallen below the reference value Plow, this means that the noise has already been reduced to the target level. In this case, the power calculator 55 turns off the switch circuit 54 (step S106), and ends the process shown in FIG.

  As a result, the update of the transfer function K of the adaptive filter circuit 52 by the filter control unit 53 is stopped. Even if updating of the transfer function K of the adaptive filter circuit 52 is stopped, the transfer function K of the adaptive filter circuit 52 is not emptied, and the value of the transfer function K is maintained as it is, so that the noise level is reduced. Is also kept constant.

  17 and 18 are diagrams for explaining the noise reduction characteristics of the noise canceling system according to the first embodiment shown in FIG.

  First, as shown in FIG. 17A, it is assumed that noise power having a relatively large peak on the low frequency side and a small peak on the high frequency side is generated. In this case, in the noise canceling system having the configuration shown in FIG. 15, noise canceling is performed by the feedforward first noise canceling system having the characteristics shown in FIG.

  As a result, the noise power after the noise canceling can be reduced on the low frequency side as shown in FIG. 17C, but the peak on the low frequency side is still large. For this reason, the noise reduction shown in FIG. 17D (FIG. 17C) can be further reduced by the adaptive filter type second noise canceling system.

  Since the adaptive filter type noise canceling system automatically adapts to a frequency component having a large noise power, the noise canceling characteristic is as shown in FIG. 17E, and the noise canceling of the characteristic is also performed. Is called.

  As a result, as described with reference to FIG. 15, the noise canceling system in which the feedforward method and the adaptive filter method are sequentially connected obtains a noise canceling effect as shown in FIG. It becomes possible. That is, the low-frequency noise (noise) that has a large power can be reduced to the reference value Plow or less.

  As is clear from the noise reduction characteristics shown in FIG. 17, a noise canceling effect larger than that of the noise canceling system of the feedforward method alone is obtained.

  In addition, an adaptive filter type single noise canceling system simply adapts to a frequency component having a large power, and there is a high degree of uncertainty that this may have an audibly sufficient noise canceling effect.

  However, the characteristics of the feed-forward type noise canceling system are designed in advance according to the frequency band in which the auditory effect is high. In this way, a certain noise canceling effect can be ensured by the feedforward first noise canceling system. In addition, the residual component can be further reduced by an adaptive filter type noise canceling system.

  Therefore, the noise canceling system having the configuration shown in FIG. 15 can achieve a higher noise canceling effect than the noise canceling system of the feed forward method alone or the adaptive filter method alone, both physically and auditorily. it can.

  Further, as shown in FIG. 18A, it is assumed that noise power having a relatively large peak on the high frequency side and a small peak on the low frequency side is generated. In this case, in the noise canceling system having the configuration illustrated in FIG. 15, noise canceling is performed by the feedforward first noise canceling system having the characteristics illustrated in FIG. 18B.

  As a result, the noise power after the noise canceling has a large peak on the high frequency side, although the noise power on the low frequency side is reduced as shown in FIG. 18C. For this reason, the noise reduction shown in FIG. 18D (FIG. 18C) can be further reduced by the adaptive filter type second noise canceling system.

  Since the adaptive filter type noise canceling system automatically adapts to a frequency component having a large noise power, the noise reduction characteristic is as shown in FIG. 18E, and the noise canceling of the characteristic is also performed. . That is, the noise power on the high frequency side can be reduced.

  As a result, as described with reference to FIG. 15, the noise canceling system in which the feedforward method and the adaptive filter method are sequentially connected obtains a noise canceling effect as shown in FIG. It becomes possible. That is, the high frequency noise (noise) that has a large power can be made equal to or lower than the reference value Plow.

  Further, the shape of the noise power shown in FIG. 17A and the noise power shown in FIG. However, it is also important that FIGS. 17F and 18F, which are the power after the final noise canceling, have very similar shapes. This does not mean that all noises have a similar shape, but indicates that the noise cancellation system of the adaptive filter system absorbs noise differences to some extent.

  Furthermore, the noise canceling system according to the first embodiment having the configuration shown in FIG. 15 is greatly different from that of a feedback system that uses only an internal microphone that also incorporates music components to be reproduced.

  In other words, in the adaptive filter type noise canceling system, the adaptive filter circuit 52 processes the signal supplied from the external microphone and the microphone amplifier 31 to generate a noise canceling signal, and thus does not affect the music component. .

  Therefore, the noise canceling system configured as shown in FIG. 15 does not require a feedback type noise canceling system. That is, the advantage (Advantage 4) that it is not necessary to use the equalizer in the form of the equalizer 16 used in the feedback type noise canceling system shown in FIG.

[Second Embodiment]
FIG. 19 is a block diagram for explaining a noise canceling system according to a second embodiment to which an embodiment of the present invention is applied. In the case of the noise canceling system shown in FIG. 19, the DSP / CPU unit 60 realizes the function of the adaptive filter circuit 61 and the function of the FB filter circuit 63.

  19, the DSP / CPU 60 includes an adaptive filter circuit 61, a filter control circuit 62 that controls the adaptive filter circuit 61, an FB filter circuit 63, a switch circuit 64, and a power calculation unit 65. And a combining unit 66 is provided.

  As shown in FIG. 19, the output signal from the external microphone and microphone amplifier unit 31 is supplied to the adaptive filter circuit 61 and the filter control unit 62. The output signal from the internal microphone and microphone amplifier unit 37 is supplied to the FB filter circuit 63, the input terminal of the switch circuit 64, and the power calculation unit 65.

  As will be described later, immediately after the operation, the power calculation unit 65 checks whether the power of the output signal from the internal microphone and the microphone amplifier unit 37 has been reduced by the function of the feedback type noise canceling system.

  When the output power from the internal microphone and the microphone amplifier unit 37 is decreased, the noise can be expected to be further reduced by the adaptive filter type noise canceling system. The circuit 64 is turned on. As a result, the output signal from the internal microphone and microphone amplifier unit 37 is supplied to the filter control unit 62.

  The filter control unit 62 operates when both the output signal from the external microphone and microphone amplifier unit 31 and the output signal from the internal microphone and microphone amplifier unit 37 are supplied. The filter control unit 62 controls the transfer function (coefficient) K of the adaptive filter circuit 61 based on the output signal from the external microphone and microphone amplifier unit 31 and the output signal from the internal microphone and microphone amplifier unit 37. .

  When the power calculation unit 65 confirms that the output power from the internal microphone and microphone amplifier unit 37 has decreased to a predetermined value or less, the power calculation unit 65 turns off the switch circuit 64. Thus, the adaptive filter circuit 61 holds the adaptively changed transfer function K, and the adaptive filter circuit 61 appropriately performs noise reduction processing.

Therefore, in the case of the noise canceling system shown in FIG.
(1) First, the input from the internal microphone and the microphone amplifier unit 37 is processed by a normal feedback type noise canceling system portion, and noise is reduced.
(2) The transfer function (coefficient) K of the adaptive filter 61 is updated by the filter control unit 62 using signals from the external microphone and microphone amplifier unit 31 and the internal microphone and microphone amplifier unit 37.
(3) The input to the internal microphone and microphone amplifier unit 37 is minimized (not necessarily minimal).
The noise cancellation process is performed by the procedure described above.

  Thereby, in the case of the noise canceling system shown in FIG. 19, a large reduction amount can be realized as compared with the case where noise canceling is performed only by the feedback method. (Advantage 1) can be realized.

  In the twin type noise canceling system shown in FIGS. 12 and 14, the transfer function α of the FF filter circuit is fixed, whereas in the case of the noise canceling system shown in FIG. The transfer function (coefficient) K of the circuit 61 is variable. As a result, the noise can be minimized by automatically following the noise that changes every moment that cannot be completely eliminated by the function of the FB filter circuit 63. (Advantage 2) can be realized.

  In addition, compared with a simple adaptive filter type noise canceling system, the function of the FB filter circuit 63 can be used in advance to design an auditory noise canceling characteristic to some extent. Since it is only necessary to focus on the residual signal, an auditory design is easier than a noise canceling system using only the adaptive filter method. (Advantage 3) can be realized.

  FIG. 20 is a flowchart for explaining the operation of the noise canceling system in which the feedback method and the adaptive filter method shown in FIG. 19 are combined. The processing of FIG. 20 is executed by the DSP / CPU unit 60 when the noise canceling system is powered on or when an execution instruction from the user is received through an operation unit (not shown).

  First, the DSP / CPU unit 60 operates a first noise canceling system (in this example, a feedback type noise canceling system) (step S201).

  Then, the power calculation unit 65 of the DSP / CPU unit 60 determines whether or not the output power from the internal microphone and microphone amplifier unit 37 has decreased (step S202). The power deriving method here is obtained by, for example, a method of obtaining the sum of squares while holding data for a certain section. It is possible to determine whether or not the output power has decreased according to the change in power obtained by the method.

  If it is determined in step S202 that the output power from the internal microphone and the microphone amplifier unit 37 has not decreased, it indicates that the first noise canceling system is not performing noise cancellation. In this case, since the user cannot enjoy the original purpose of noise reduction, the process shown in FIG.

  If it is determined in step S202 that the output power from the internal microphone and microphone amplifier unit 37 has decreased as originally designed, the second noise canceling system is operated (step S203). In the second embodiment, the second noise canceling system is an adaptive filter type noise canceling system.

  In this case, the power calculation unit 65 turns on the switch circuit 64, causes the filter control unit 62 to function, and starts processing to adaptively update the transfer function (coefficient) K of the adaptive filter circuit 61 (step S204).

  Then, the power calculation unit 65 determines whether or not the output power from the internal microphone and the microphone amplifier unit 37 has decreased and the power has fallen below a predetermined reference value Plow (step S205).

  If it is determined in step S205 that the output power from the internal microphone and microphone amplifier unit 37 is not lower than the reference value Plow, the processing from step S205 is repeated. As a result, the transfer function (coefficient) K of the adaptive filter circuit 61 is continuously updated and waits for the output power to fall below the reference value Plow.

  If it is determined in step S205 that the output power from the internal microphone and microphone amplifier unit 37 has fallen below the reference value Plow, it means that the noise has already been reduced to the target level. In this case, the power calculator 65 turns off the switch circuit 64 (step S206), and ends the process shown in FIG.

  Thereby, the update of the transfer function K of the adaptive filter circuit 61 by the filter control unit 62 is stopped. Even if updating of the transfer function K of the adaptive filter circuit 61 is stopped, the transfer function K of the adaptive filter circuit 61 is not emptied, and the value of the transfer function K is maintained as it is, so that the noise level is reduced. Is also kept constant.

  21 and 22 are diagrams for explaining the noise reduction characteristics of the noise canceling system according to the second embodiment shown in FIG.

  First, as shown in FIG. 21 (A), it is assumed that noise power having a relatively large peak on the low frequency side and a small peak on the high frequency side is generated. In this case, in the noise canceling system having the configuration shown in FIG. 19, the noise canceling is performed by the feedback-type first noise canceling system having the characteristics shown in FIG.

  As a result, the noise power after the noise canceling can be reduced on the low frequency side as shown in FIG. 21C, but the peak on the low frequency side is still large. For this reason, the second noise canceling system of the adaptive filter system can further reduce noise (noise) with respect to the noise power shown in FIG. 21D (FIG. 21C).

  Since the adaptive filter type noise canceling system automatically adapts to a frequency component having a large noise power, the noise canceling characteristic is as shown in FIG. 21E, and the noise canceling of the characteristic is also performed. Is called.

  As a result, as described with reference to FIG. 19, the noise canceling system in which the feedback method and the adaptive filter method are sequentially connected can comprehensively obtain the noise canceling effect as shown in FIG. Is possible. That is, the low-frequency noise (noise) that has a large power can be reduced to the reference value Plow or less.

  As is clear from the noise reduction characteristics shown in FIG. 21, a noise canceling effect greater than that of the noise canceling system using the feedback method alone is obtained.

  In addition, an adaptive filter type single noise canceling system simply adapts to a frequency component having a large power, and there is a high degree of uncertainty that this may have an audibly sufficient noise canceling effect.

  However, the characteristics of the feedback-type noise canceling system are designed in advance according to the frequency band in which the auditory effect is high. In this way, a certain amount of noise canceling effect can be ensured with the feedback-type first noise canceling system. In addition, the residual component can be further reduced by an adaptive filter type noise canceling system.

  Accordingly, the noise canceling system having the configuration shown in FIG. 19 can achieve a higher noise canceling effect than the noise canceling system of the feedback method alone or the adaptive filter method alone, both physically and auditorily. .

  Further, as shown in FIG. 22A, it is assumed that noise power having a relatively large peak on the high frequency side and a small peak on the low frequency side is generated. In this case, in the noise canceling system having the configuration shown in FIG. 19, the noise canceling is performed by the feedback-type first noise canceling system having the characteristics shown in FIG.

  Thereby, as shown in FIG. 22C, the noise power after the noise canceling has a high peak in the high band, although the noise power in the low band is reduced. For this reason, the second noise canceling system of the adaptive filter system can further reduce noise (noise) with respect to the noise power shown in FIG. 22D (FIG. 22C).

  Since the adaptive filter type noise canceling system automatically adapts to a frequency component having a large noise power, the noise canceling characteristic is as shown in FIG. 22E, and the noise canceling of the characteristic is also performed. Is called. That is, the noise power on the high frequency side can be reduced.

  As a result, as described with reference to FIG. 19, the noise canceling effect in which the feedback method and the adaptive filter method are sequentially connected can obtain a noise canceling effect as shown in FIG. Is possible. That is, the high frequency noise (noise) that has a large power can be made equal to or lower than the reference value Plow.

  Further, the shape of the noise power shown in FIG. 21A is greatly different from the shape of the noise power shown in FIG. However, it is also important that FIGS. 21F and 22F, which are the power after the final noise canceling, have very similar shapes. This does not mean that all noises have a similar shape, but indicates that the noise cancellation system of the adaptive filter system absorbs noise differences to some extent.

[Third Embodiment]
FIG. 23 is a block diagram for explaining a noise canceling system according to a third embodiment to which an embodiment of the present invention is applied. In the case of the noise canceling system shown in FIG. 23, the DSP / CPU unit 70 realizes the function of the FF filter circuit 71, the function of the FB filter circuit 72, and the function of the adaptive filter circuit 73. Yes.

  23, the DSP / CPU 70 includes an FF filter circuit 71, an FB filter circuit 72, an adaptive filter circuit 73, a filter control circuit 74 for controlling the adaptive filter circuit 73, and a switch circuit 75. A power calculator 76 and a combiner 77 are provided.

  As shown in FIG. 23, the output signal from the external microphone and microphone amplifier unit 31 is supplied to the FF filter circuit 71, the adaptive filter circuit 73, and the filter control unit 74. The output signal from the internal microphone and microphone amplifier unit 37 is supplied to the input terminal of the switch circuit 75, the power calculation unit 76, and the FB filter circuit 72.

  As will be described later, immediately after the operation, the power calculation unit 76 determines whether the power of the output signal from the internal microphone and the microphone amplifier unit 37 has been reduced by the function of the noise canceling system of the feedforward method and the feedback method. To check.

  When the output power from the internal microphone and microphone amplifier unit 37 is reduced, the noise can be expected to be further reduced by the adaptive filter type noise canceling system. Circuit 75 is turned on. As a result, the output signal from the internal microphone and microphone amplifier unit 37 is supplied to the filter control unit 74.

  The filter control unit 74 operates when both the output signal from the external microphone and microphone amplifier unit 31 and the output signal from the internal microphone and microphone amplifier unit 37 are supplied. The filter control unit 74 controls the transfer function (coefficient) K of the adaptive filter circuit 73 based on the output signal from the external microphone and microphone amplifier unit 31 and the output signal from the internal microphone and microphone amplifier unit 37. .

  When the power calculation unit 76 confirms that the output power from the internal microphone and microphone amplifier unit 37 has decreased to a predetermined value or less, the power calculation unit 76 turns off the switch circuit 75. Thus, the adaptive filter circuit 73 holds the adaptively changed transfer function K, and the adaptive filter circuit 73 appropriately performs noise reduction processing.

Therefore, in the case of the noise canceling system shown in FIG.
(1) First, the input from the external microphone and microphone amplifier unit 31 is processed by a normal feedforward type noise canceling system, and then the input from the internal microphone and microphone amplifier unit 37 is a normal feedback method. The noise canceling system is used to reduce noise.
(2) The transfer function (coefficient) K of the adaptive filter 73 is updated by the filter control unit 74 using signals from the external microphone and microphone amplifier unit 31 and the internal microphone and microphone amplifier unit 37.
(3) The input to the internal microphone and microphone amplifier unit 37 is minimized (not necessarily minimal).
The noise cancellation process is performed by the procedure described above.

  Thereby, in the case of the noise canceling system shown in FIG. 23, a large reduction amount can be realized as compared with the case where the noise canceling is performed by the twin method using the feed forward method and the feedback method. (Advantage 1) can be realized.

  Further, in the twin type noise canceling system shown in FIGS. 12 and 14, the transfer function α of the FF filter circuit and the transfer function β of the FB filter circuit are fixed, whereas the noise canceling shown in FIG. In the case of the system, the transfer function K of the adaptive filter circuit 73 is variable. As a result, the noise can be minimized by automatically following the noise that changes every moment that cannot be eliminated by the functions of the FF filter circuit 71 and the FB filter circuit 72. (Advantage 2) can be realized.

  In addition, compared with a simple adaptive filter type noise canceling system, the function of the FB filter circuit 71 and the FB filter circuit 72 can be designed in advance with some expectation of auditory noise canceling characteristics. Since the adaptive filter portion only needs to focus on the residual signal, the auditory design is easier than a noise canceling system using only the adaptive filter method. (Advantage 3) can be realized.

  FIG. 24 is a flowchart for explaining the operation of the noise canceling system that combines the feedforward method, the feedback method, and the adaptive filter method shown in FIG. The processing in FIG. 24 is executed in the DSP / CPU unit 70 when the noise canceling system is powered on or when an execution instruction from the user is received through an operation unit (not shown).

  First, the DSP / CPU unit 70 operates the first noise canceling system (in this example, a feedforward type noise canceling system and a feedback type noise canceling system) (step S301).

  Then, the power calculation unit 76 of the DSP / CPU unit 70 determines whether or not the output power from the internal microphone and microphone amplifier unit 37 has decreased (step S302). The power deriving method here is obtained by, for example, a method of obtaining the sum of squares while holding data for a certain section. It is possible to determine whether or not the output power has decreased according to the change in power obtained by the method.

  If it is determined in step S302 that the output power from the internal microphone and microphone amplifier unit 37 has not decreased, this indicates that noise cancellation is not successful in the first noise canceling system. In this case, since the user cannot enjoy the benefit of noise reduction, which is the original purpose, the process shown in FIG. 24 is terminated.

  If it is determined in step S302 that the output power from the internal microphone and microphone amplifier unit 37 has decreased as originally designed, the second noise canceling system is operated (step S303). In the third embodiment, the second noise canceling system is an adaptive filter type noise canceling system.

  In this case, the power calculation unit 76 turns on the switch circuit 75, causes the filter control unit 74 to function, and starts processing for adaptively updating the transfer function (coefficient) K of the adaptive filter circuit 73 (step S304).

  Then, the power calculation unit 76 determines whether or not the output power from the internal microphone and the microphone amplifier unit 37 has decreased and the power has fallen below a predetermined reference value Plow (step S305).

  When it is determined in step S305 that the output power from the internal microphone and microphone amplifier unit 37 is not lower than the reference value Plow, the processing from step S305 is repeated. As a result, the transfer function K of the adaptive filter circuit 73 is continuously updated and waits for the output power to fall below the reference value Plow.

  If it is determined in step S305 that the output power from the internal microphone and microphone amplifier unit 37 has fallen below the reference value Plow, this means that the noise has already been reduced to the target level. In this case, the power calculator 76 turns off the switch circuit 75 (step S306), and ends the process shown in FIG.

  Thereby, the update of the transfer function K of the adaptive filter circuit 73 by the filter control unit 74 is stopped. Even if updating of the transfer function K of the adaptive filter circuit 73 is stopped, the transfer function K of the adaptive filter circuit 73 is not emptied, and the value of the transfer function K is maintained as it is, so that the noise level is reduced. Is also kept constant.

  25 and 26 are diagrams for explaining the noise reduction characteristics of the noise canceling system according to the third embodiment shown in FIG.

  First, as shown in FIG. 25A, it is assumed that noise power having a relatively large peak on the low frequency side and a small peak on the high frequency side is generated. In this case, in the noise canceling system having the configuration shown in FIG. 23, the noise canceling is performed by the first noise canceling system of the feedback method and the feedforward method having the characteristics shown in FIG. .

  As a result, the noise power after the noise canceling can be reduced as shown in FIG. 25C, but the low frequency side peak is still large. For this reason, the noise reduction shown in FIG. 25D (FIG. 25C) can be further reduced by the adaptive filter type second noise canceling system.

  Since the adaptive filter type noise canceling system automatically adapts to a frequency component having a large noise power, the noise canceling characteristic is as shown in FIG. 25E, and the noise canceling of the characteristic is also performed. Is called.

  As a result, as described with reference to FIG. 23, in addition to the feedback method and the feedforward method, the noise canceling system in which the adaptive filter method is sequentially connected as shown in FIG. A noise canceling effect can be obtained. That is, the noise having a large power (noise) can be reduced to the reference value Plow or less.

  As is clear from the noise reduction characteristics shown in FIG. 25, this state clearly obtains a larger noise canceling effect than the twin-type noise canceling system B having both the feedforward method and the feedback method.

  In addition, an adaptive filter type single noise canceling system simply adapts to a frequency component having a large power, and there is a high degree of uncertainty that this may have an audibly sufficient noise canceling effect.

  However, the characteristics of the noise canceling system of the feed-forward method and the feedback method are designed in advance according to the frequency band that is audibly effective. Thus, a certain noise canceling effect can be ensured by the first noise canceling system of the feedforward method and the feedback method. In addition, the residual component can be further reduced by an adaptive filter type noise canceling system.

  Therefore, the noise canceling system having the configuration shown in FIG. 23 can achieve a higher noise canceling effect than the noise canceling system of the twin method alone or the adaptive filter method alone, both physically and auditorily. .

  Further, as shown in FIG. 26A, it is assumed that noise power having a relatively large peak on the high frequency side and a small peak on the low frequency side is generated. In this case, in the noise canceling system having the configuration shown in FIG. 23, noise canceling is performed by the feed-forward and first feedback noise canceling systems having the characteristics shown in FIG. .

  As a result, the noise power after the noise canceling has a large peak on the high frequency side although the noise power is reduced as shown in FIG. For this reason, the second noise canceling system of the adaptive filter system can further reduce the noise (noise) with respect to the noise power shown in FIG. 26D (FIG. 26C).

  Since the adaptive filter type noise canceling system automatically adapts to a frequency component having a large noise power, the noise canceling characteristic is as shown in FIG. 26E, and the noise canceling of the characteristic is also performed. Is called. That is, the noise power on the high frequency side can be reduced.

  Thus, as described with reference to FIG. 23, in addition to the feedforward method and the feedback method, the noise canceling system in which the adaptive filter method is sequentially connected as shown in FIG. A noise canceling effect can be obtained. That is, the high frequency noise (noise) that has a large power can be made equal to or lower than the reference value Plow.

  Further, the shape of the noise power shown in FIG. 25A is greatly different from the shape of the noise power shown in FIG. However, it is also important that FIGS. 25F and 26F, which are the power after the final noise canceling, have very similar shapes. This does not mean that all noises have a similar shape, but indicates that the noise cancellation system of the adaptive filter system absorbs noise differences to some extent.

[Fourth Embodiment]
27 and 28 are block diagrams for explaining a noise canceling system according to a fourth embodiment to which an embodiment of the present invention is applied. As shown in FIG. 27, the noise canceling system according to the fourth embodiment is similar to the noise canceling system according to the first to third embodiments described above except for the DSP / CPU unit 80. Composed.

  The DSP / CPU unit 80 of the noise canceling system according to the fourth embodiment is provided with two adaptive filter circuits in addition to the FF filter circuit and the FB filter circuit.

  28, the DSP / CPU 80 includes an FF filter circuit 81, an FB filter circuit 82, an adaptive filter circuit unit 83, an adaptive filter circuit unit 84, a switch circuit 85, and a power calculation unit. 86 and a combining unit 87 are provided.

  The adaptive filter circuit unit 83 includes weighting filters 831 and 836, an adaptive filter circuit 832, a filter control circuit 833 that controls the adaptive filter circuit 832, a switch circuit 834, and a power calculation unit 835.

  Similarly, the adaptive filter circuit unit 84 includes weighting filters 841 and 846, an adaptive filter circuit 842, a filter control circuit 843 that controls the adaptive filter circuit 842, a switch circuit 844, and a power calculation unit 845. .

  The weighting filters 831 and 836 of the adaptive filter circuit unit 83 and the weighting filters 841 and 846 of the adaptive filter circuit unit 84 are used for setting different movable ranges for the adaptive filter circuit unit 83 and the adaptive filter circuit unit 84, respectively. It is.

  As shown in FIG. 28, the output signal from the microphone amplifier 312 of the external microphone and microphone amplifier unit 31 is supplied to the FF filter circuit 81.

  An output signal from the microphone amplifier 312 of the external microphone and microphone amplifier unit 31 is supplied to the filter control unit 833 through the weighting filter 831 of the adaptive filter circuit unit 83 and also to the adaptive filter circuit 832.

  Similarly, an output signal from the external microphone and microphone amplifier 312 of the microphone amplifier unit 31 is supplied to the filter control unit 843 through the weighting filter 841 of the adaptive filter circuit unit 84 and also to the adaptive filter circuit 842.

  As described above, the output signal from the microphone amplifier 312 of the external microphone and microphone amplifier unit 31 is supplied to the FF filter circuit 81, the adaptive filter circuit unit 83, and the adaptive filter circuit unit 84.

  The output signal from the microphone amplifier 272 of the internal microphone and microphone amplifier 37 is supplied to the FB filter circuit 82, the input terminal of the switch circuit 85, and the power calculator 86.

  The output terminal of the switch circuit 85 is connected to the switch circuit 834 of the adaptive filter circuit unit 83 and the power calculation unit 835 via the weighting filter 836 of the adaptive filter circuit unit 83. The output terminal of the switch circuit 834 is connected to the filter control unit 833.

  Similarly, the output terminal of the switch circuit 85 is connected to the switch circuit 844 of the adaptive filter circuit unit 84 and the power calculation unit 845 via the weighting filter 846 of the adaptive filter circuit unit 84. The output terminal of the switch circuit 844 is connected to the filter control unit 843.

  As will be described later, immediately after the operation, the power calculation unit 86 determines whether the power of the output signal from the internal microphone and the microphone amplifier unit 37 has been reduced by the function of the noise canceling system of the feedforward method and the feedback method. To check.

  When the output power from the internal microphone and the microphone amplifier unit 37 is reduced, the noise can be expected to be further reduced by the adaptive filter type noise canceling system. Circuit 85 is turned on.

  As a result, the output signal from the internal microphone and microphone amplifier unit 37 is supplied to the input terminal of the switch circuit 834 and the power calculation unit 835 through the weighting filter 836 of the adaptive filter circuit unit 83. Similarly, the output signal from the internal microphone and microphone amplifier unit 37 is supplied to the input terminal of the switch circuit 844 and the power calculation unit 845 through the weighting filter 846 of the adaptive filter circuit unit 84.

  In this case, similarly to the power calculation unit 86, the power calculation units 835 and 845 turn on the switch circuits 834 and 844 when the output power from the internal microphone and the microphone amplifier unit 37 is reduced. As a result, output signals from the internal microphone and microphone amplifier unit 37 are supplied to the filter control units 833 and 843.

  The filter control units 833 and 843 operate when both the output signal from the external microphone and microphone amplifier unit 31 and the output signal from the internal microphone and microphone amplifier unit 37 are supplied through the weighting filters, respectively.

  The filter control units 833 and 843 transfer the transfer functions (coefficients) of the adaptive filter circuits 832 and 842 based on the output signal from the external microphone and microphone amplifier unit 31 and the output signal from the internal microphone and microphone amplifier unit 37. K1 and K2 are controlled.

  When each of the power calculation units 835 and 845 confirms that the output power from the internal microphone and the microphone amplifier unit 37 has decreased to a predetermined value or less in each movable range (movable band), the switch circuit 834 and 844 are turned off.

  As a result, the adaptively changing transfer functions K1 and K2 are held in the adaptive filter circuits 832 and 842, and the adaptive filter circuits 832 and 842 appropriately perform noise reduction processing.

  Of course, it may be confirmed that, of the power calculation units 835 and 845, the output power from the internal microphone and the microphone amplifier unit 37 is not more than a predetermined value, and the other cannot be confirmed. In this case, the operation of the other adaptive filter circuit unit is continuously performed, and the adjustment of the adaptive filter circuit is continuously performed.

  When the power calculation unit 86 confirms that the output power from the internal microphone and the microphone amplifier unit 37 is equal to or less than a predetermined value in the entire movable range (all bands), the switch circuit 85 is also turned off. The transfer functions K1 and K2 of the two adaptive filter circuit units realized by the DSP / CPU unit 80 are fixed.

Therefore, in the case of the noise canceling system shown in FIGS.
(1) First, the input from the external microphone and microphone amplifier unit 31 is processed by a normal feedforward type noise canceling system, and noise is reduced at a cancellation point. Next, the input from the internal microphone and the microphone amplifier unit 37 is processed by a normal feedback type noise canceling system, and noise is reduced.
(2) The transfer functions (coefficients) K1 and K2 of the adaptive filters 832 and 842 are updated by the filter control units 833 and 834 using the signals from the external microphone and microphone amplifier unit 31 and the internal microphone and microphone amplifier unit 37.
(3) As described above, weighting W1 is performed on the input signal to the filter control unit 833. The input signal to the filter control unit 843 is weighted W2. Thereby, the movable ranges in the adaptive filter circuit units 83 and 84 are made different.
(4) The input to the internal microphone and microphone amplifier unit 37 is minimized (not necessarily minimal).
The noise cancellation process is performed by the procedure described above.

  Thus, in the case of the noise canceling system shown in FIG. 27 and FIG. 28, the system having both the feedforward method and the feedback method described with reference to FIG. 23 and one adaptive filter method noise canceling system. As compared with the above, a large reduction amount of noise can be realized. (Advantage 1) can be realized.

  In the case of the noise canceling system shown in FIG. 27, two adaptive filter circuit units are provided. As a result, even if there are two peaks in the noise (noise) that cannot be eliminated by the feed-forward and feedback-type noise canceling systems, the noise can be automatically tracked and minimized. (Advantage 2) can be realized.

  Further, even when compared with a simple adaptive filter type noise canceling system, it is possible to design with a certain degree of auditory noise canceling characteristics in advance using a feedforward type and feedback type noise canceling system. For this reason, the adaptive filter circuit unit only needs to focus on the residual signal, so that it is easier to design audibly than a noise canceling system using only the adaptive filter method. (Advantage 3) can be realized.

  FIG. 29 is a flowchart for explaining the operation of the noise canceling system that combines the feedforward method, the feedback method, and the two adaptive filter methods shown in FIGS. 27 and 28. The processing in FIG. 29 is executed in the DSP / CPU unit 80 when the noise canceling system is powered on or when an execution instruction from the user is received through an operation unit (not shown).

  First, the DSP / CPU unit 80 operates a first noise canceling system (in this example, a feedforward type noise canceling system and a feedback type noise canceling system) (step S401).

  Then, the power calculation unit 86 of the DSP / CPU unit 80 determines whether or not the output power from the internal microphone and microphone amplifier unit 37 has decreased (step S402). The power deriving method here is obtained by, for example, a method of obtaining the sum of squares while holding data for a certain section. It is possible to determine whether or not the output power has decreased according to the change in power obtained by the method.

  If it is determined in step S402 that the output power from the internal microphone and microphone amplifier unit 37 has not decreased, this indicates that noise cancellation is not successful in the first noise canceling system. In this case, since the user cannot enjoy the original purpose of reducing noise (noise), the processing shown in FIG. 29 is terminated.

  If it is determined in step S402 that the output power from the internal microphone and microphone amplifier unit 37 has decreased as originally designed, the switch circuit 85 (SW0) is turned on to perform the second noise canceling. The system is operated (step S403). In the second embodiment, the second noise canceling system is a noise canceling system of two adaptive filter systems including adaptive filter circuit units 83 and 84.

  In the adaptive filter circuit unit 83, the power calculation unit 835 turns on the switch circuit 834 (SW1), causes the filter control unit 833 to function, and adaptively sets the transfer function (coefficient) K1 of the adaptive filter circuit 832. The update process is started (step S404).

  Then, the power calculation unit 835 determines whether or not the output power from the internal microphone and the microphone amplifier unit 37 has decreased and the power has fallen below a predetermined reference value Plow (step S405).

  If it is determined in step S405 that the output power from the internal microphone and microphone amplifier unit 37 is not lower than the reference value Plow, the processing from step S405 is repeated. As a result, the transfer function K1 of the adaptive filter circuit 832 is continuously updated and waits for the output power to fall below the reference value Plow.

  If it is determined in step S405 that the output power from the internal microphone and microphone amplifier unit 37 has fallen below the reference value Plow, this means that the noise has already been reduced to the target level. In this case, the power calculation unit 835 turns off the switch circuit 834 (SW1) (step S406), and ends the operation of the adaptive filter circuit unit 83.

  On the other hand, in the adaptive filter circuit unit 84 as well, the power calculation unit 845 turns on the switch circuit 844 (SW2), causes the filter control unit 843 to function, and applies the transfer function (coefficient) K1 of the adaptive filter circuit 842. The update process is started (step S407).

  Then, the power calculation unit 845 determines whether the output power from the internal microphone and the microphone amplifier unit 37 has decreased and the power has fallen below a predetermined reference value Plow (step S408).

  If it is determined in step S408 that the output power from the internal microphone and microphone amplifier unit 37 is not lower than the reference value Plow, the processing from step S408 is repeated. As a result, the transfer function K2 of the adaptive filter circuit 842 is continuously updated and waits for the output power to fall below the reference value Plow.

  If it is determined in step S408 that the output power from the internal microphone and microphone amplifier unit 37 has fallen below the reference value Plow, this means that the noise has already been reduced to the target level. In this case, the power calculation unit 845 turns off the switch circuit 844 (SW2) (step S409) and ends the operation of the adaptive filter circuit unit 84.

  As described above, when the transfer functions K1 and K2 of the adaptive filter circuits 832 and 842 of the adaptive filter circuits 83 and 84 can be fixed, the processing shown in FIG. 29 ends. Note that even if the updating of the transfer functions K1 and K2 of the adaptive filter circuits 83 and 84 is stopped, the transfer functions K1 and K2 of the adaptive filter circuits 83 and 84 are not emptied, and the values of the transfer functions K1 and K2 are Since it is held as it is, the reduction of the noise level is also kept constant.

  As described above, the adaptive filter circuit unit 83 and the adaptive filter circuit unit 84 use the weighting filters 831, 836, 841, and 846 having different weights. For this reason, the switch circuits 834 (SW1) and 844 (SW2) of the adaptive filter circuits 83 and 84 are not simultaneously turned on / off.

  Depending on the update status of the transfer functions K1 and K2 of the adaptive filter circuits 832 and 842 of the adaptive filter circuit units 84 and 84, only one of the switch circuits may be turned on / off.

  Specifically, even when the turn-on timing is the same, the switch circuit 834 and the switch circuit 844 may have different turn-off timings. However, when the power is first turned on, the switch circuits 834 and 844 may be turned on at the same time according to the result of confirmation by the power calculator 86.

  In addition, since the power calculation unit 835 of the adaptive filter circuit unit 83 and the power calculation unit 845 of the adaptive filter circuit unit 84 function independently, the timing for turning on each of the switch circuits 834 and 844 is made different. Of course it is also possible.

  Also, the power reference value PLow used in the power calculation unit 835 of the adaptive filter circuit unit 83 and the power reference value PLow used in the power calculation unit 845 of the adaptive filter circuit unit 84 can be set individually.

  Also, the switch circuit 85 located on the internal microphone and microphone amplifier unit 37 side of the switch circuits 834 and 844 is turned on / off by the function of the power calculation unit 86 instead of the on / off of the switch circuits 834 and 844. It is also possible to configure so as to control the start and end of coefficient update of the adaptive filter circuit units 83 and 84.

  As described above, it is possible to take various modes as to how the two adaptive filter circuit units 83 and 84 are controlled.

  FIG. 30 is a diagram for explaining the noise reduction characteristics of the noise canceling system according to the fourth embodiment shown in FIGS. 27 and 28.

  First, as shown in FIG. 30A, it is assumed that noise power having a relatively large peak on the low frequency side and a small peak on the high frequency side is generated. In this case, in the noise canceling system having the configuration shown in FIGS. 27 and 28, noise canceling is performed by the feedback type first feed-forward type noise canceling system having the characteristics shown in FIG. To be implemented.

  As a result, the noise power after the noise canceling can be reduced as shown in FIG. 30C, but the low frequency side peak is still large. For this reason, the second noise canceling system of the adaptive filter system can further reduce the noise (noise) with respect to the noise power shown in FIG. 30D (FIG. 30C).

  In such a case, when there is one adaptive filter circuit unit, the adaptive filter automatically adapts to the low-frequency peak, and the high-frequency peak remains unchanged.

  However, in the case of the noise canceling system of the fourth embodiment, the movable ranges of the adaptive filters 832 and 842 are set by multiplying the signals supplied to the filter control units 833 and 843 by weights W1 and W2. Restrict.

  Thereby, as shown in FIG. 30E, for example, the characteristic of the adaptive filter circuit unit 83 cancels the low-frequency noise as indicated by AF1. Further, for example, the characteristic of the adaptive filter circuit unit 84 cancels high-frequency noise as indicated by AF2.

  As a result, the adaptive filter circuits 832 and 842 minimize the peak portion within the set (restricted) band, so that it is possible to minimize both the high frequency side peak and the low frequency side peak. Become. Therefore, as shown in FIG. 30F, both the low frequency side noise and the high frequency side noise are appropriately canceled, and a high noise canceling effect can be realized.

  Each of the adaptive filter circuits 832 and 842 can adaptively change the transfer functions K1 and K2 as described above. For this reason, as shown by the fine dotted waveform in FIG. 30E, a noise cancellation signal corresponding to the generated noise can be generated and the noise can be canceled appropriately.

  In the noise canceling system according to the fourth embodiment, the case where two adaptive filter circuit units are used has been described. However, it is not limited to this. For example, it is of course possible to provide three adaptive filter circuit units so as to cancel noise in each of the low, middle and high frequencies.

  Furthermore, it goes without saying that the configuration in which the number of adaptive filter circuit units is increased to four or five is limited to the capability of the DSP / CPU unit.

  However, not all the noise canceling system parts configured by the feedforward method and the feedback method are replaced by the adaptive filter method.

  The noise canceling system combining the feedforward method, the feedback method, and the adaptive filter method according to the present invention has been conceived since it is difficult to design the weighting filter W only by the adaptive filter method.

  As an auditory effect, a feedforward type or feedback type noise canceling system that has already been proven in the market is used, and an adaptive filter type noise canceling system is used for the residual component.

  In other words, an adaptive filter type noise canceling system is introduced in addition to a feedback type, feedforward type, or a twin type noise canceling system that combines these. In addition, the optimum noise canceling performance is automatically realized for each noise.

  This facilitates the design of an adaptive filter type noise canceling system, and a higher noise canceling effect can be obtained than a noise canceling system using only the feedforward method or only the feedback method. Moreover, a higher noise canceling effect can be obtained than a twin-type noise canceling system having both a feedforward method and a feedback method.

  In the fourth embodiment described above, the two adaptive filter circuit units 83 and 84 are provided in addition to the FF filter circuit 81 and the FB filter circuit 82. However, the present invention is not limited to this.

  For example, in the noise canceling system according to the first embodiment shown in FIG. 15, it is of course possible to provide two or more adaptive filter circuit units. Further, in the noise canceling system of the second embodiment shown in FIG. 19, it is possible to adopt a configuration in which two or more adaptive filter circuit units are provided.

[Summary of effects of embodiment]
As is apparent from the above description of the embodiment, the invention according to this application can achieve the following effects.
(1) Replacing a part of the noise canceling system with an adaptive filter system, or additionally using an adaptive filter system, can give the system variability.
(2) Due to the variability of the adaptive filter system, it is possible to automatically follow the noise that changes every moment.
(3) The degree of freedom in design is improved as compared with the case where the feedforward method / feedback method / adaptive filter method are each carried out independently.
(4) Compared with the case where the feedforward method / feedback method / adaptive filter method are each carried out independently, a large amount of reduction can be realized in terms of noise reduction.
The effect that can be said.

Furthermore, in relation to the effect (3) above,
(A) The noise canceling characteristic considering the auditory characteristic which is difficult only by the adaptive filter method can be secured by the noise canceling system of the feedforward method / feedback method.
(B) The adaptive filter type noise canceling system takes charge of the noise that cannot be completely eliminated by the feedforward type / feedback type noise canceling system, and the noise can be appropriately reduced.
(C) The shift of the noise canceling characteristic due to the component variation of the feedforward type / feedback type noise canceling system can be corrected by the adaptive filter type noise canceling system.
Can be realized.

In relation to (C) above,
(A) Although the component accuracy directly affects the noise canceling characteristics, the influence can be reduced by using the adaptive filter method.
(B) Conventionally, the noise canceling characteristic is manually adjusted using a variable resistor or the like, but this can be automatically corrected.
(C) As a result, it can contribute to cost reduction.
Such effects can be achieved.

  As described above, various effects (advantages) can be obtained as compared with the case where the feed forward method / feedback method / adaptive filter method is used alone or in the twin method in which the feed forward method and the feedback method are combined. To be.

[About the feasibility of the method invention]
In the first to fourth embodiments described above, the FF filter circuit, the FB filter circuit, the adaptive filter circuit, the filter control unit, the switch circuit, and the power realized by the DSP / CPU units 50, 60, 70, and 80 are used. Each of the processes performed in each part of the calculation part corresponds to the process of each corresponding step in the method according to the present invention.

  More specifically, the processes shown in the flowcharts of FIGS. 16, 20, 24, and 29 are obtained by applying the method according to the present invention. Therefore, the method of the present invention can also be realized.

[About the feasibility of the invention of the program]
In the first to fourth embodiments described above, the FF filter circuit, the FB filter circuit, the adaptive filter circuit, the filter control unit, the switch circuit, and the power realized by the DSP / CPU units 50, 60, 70, and 80 are used. Each unit of the calculation unit can realize its function by a program executed by the corresponding DSP / CPU unit 50, 60, 70, 80.

  More specifically, the processing shown in the flowcharts of FIGS. 16, 20, 24, and 29 is a program to which the present invention is applied. Therefore, the program of the present invention can also be realized.

[Others]
In the above-described embodiment, the input sound S can be received as an external source. However, it is not always necessary to have a function of receiving an external source. That is, it is not necessary to listen to an external source such as music, and it is possible to configure as a noise reduction system that can reduce only noise.

  In the above-described embodiment, the case where the present invention is applied to a headphone system has been described as an example in order to simplify the explanation, but it is not necessary that all systems are mounted in the headphone body. For example, processing mechanisms such as an FB filter circuit, an FF filter circuit, an adaptive filter circuit unit, and a power amplifier can be divided as a box outside, or can be configured in combination with other devices. Here, the other devices may be various types of hardware capable of reproducing voice / music signals, such as portable audio players, telephone devices, and network voice communication devices.

  In particular, by applying the present invention to a mobile phone terminal and a headset connected thereto, for example, it is possible to reduce noise and make a good call even in a very noisy environment on the go. It becomes possible to do. In this case, the configuration on the headset side can be simplified by providing the FF filter circuit, the FB filter circuit, the adaptive filter circuit unit, the drive circuit, and the like on the mobile phone terminal side. Of course, all the configurations can be provided on the headset side so that the supply of audio from the mobile phone terminal can be received.

It is a figure for demonstrating the noise canceling system of a feedback system. It is a figure which shows the characteristic of the noise canceling system of a feedback system. It is a figure for demonstrating the noise canceling system of a feedforward system. It is a figure which shows the characteristic of the noise canceling system of a feedforward system. It is a figure for demonstrating the noise canceling system of an adaptive filter system. It is a figure which shows the characteristic of the noise canceling system of an adaptive filter system. It is a figure for demonstrating the calculation formula which shows the characteristic of the noise cancellation system of a feedback system. It is a figure for demonstrating the calculation formula which shows the characteristic of the noise canceling system of a feedforward system. It is a figure for demonstrating the calculation formula which shows the characteristic of the noise canceling system of an adaptive filter system. It is a block diagram for demonstrating the structural example of the noise canceling system of the adaptive filter system which considered weighting. It is a figure for demonstrating the characteristic of the noise canceling system of the adaptive filter system using the weighting shown in FIG. It is a figure for demonstrating the noise canceling system system which overlap | superposed the feedback system and the feedforward system. It is a figure for demonstrating the characteristic of the noise cancellation system of the twin system shown in FIG. It is a block diagram for demonstrating an example of the twin-type noise canceling system. It is a block diagram for demonstrating the noise canceling system of 1st Embodiment of this invention. It is a flowchart for demonstrating operation | movement of the noise canceling system of 1st Embodiment shown in FIG. It is a figure for demonstrating the noise reduction characteristic of the noise canceling system of 1st Embodiment shown in FIG. It is a figure for demonstrating the noise reduction characteristic of the noise canceling system of 1st Embodiment shown in FIG. It is a block diagram for demonstrating the noise canceling system of 2nd Embodiment of this invention. It is a flowchart for demonstrating operation | movement of the noise canceling system of 2nd Embodiment shown in FIG. It is a figure for demonstrating the noise reduction characteristic of the noise canceling system of 2nd Embodiment shown in FIG. It is a figure for demonstrating the noise reduction characteristic of the noise canceling system of 2nd Embodiment shown in FIG. It is a block diagram for demonstrating the noise canceling system of 3rd Embodiment of this invention. It is a flowchart for demonstrating operation | movement of the noise canceling system of 3rd Embodiment shown in FIG. It is a figure for demonstrating the noise reduction characteristic of the noise canceling system of 3rd Embodiment shown in FIG. It is a figure for demonstrating the noise reduction characteristic of the noise canceling system of 3rd Embodiment shown in FIG. It is a block diagram for demonstrating the noise canceling system of 4th Embodiment of this invention. It is a block diagram for demonstrating the DSP / CPU part 80 of the noise canceling system of the 4th Embodiment of this invention. It is a flowchart for demonstrating operation | movement of the noise canceling system of 4th Embodiment shown in FIG. 27, FIG. It is a figure for demonstrating the noise reduction characteristic of the noise canceling system of 4th Embodiment shown in FIG. 27, FIG.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 31 ... External microphone and microphone amplifier part, 311 ... External microphone, 312 ... Microphone amplifier, 34 ... Composition part, 35 ... Power amplifier part, 36 ... Driver, 361 ... Drive circuit, 362 ... Speaker, 37 ... Internal microphone and microphone amplifier 371 ... Internal microphone, 372 ... Microphone amplifier, 30 ... Equalizer, 50 ... DSP / CPU unit, 51 ... FF filter circuit, 52 ... Adaptive filter circuit, 53 ... File control unit, 54 ... Switch circuit, 55 ... Power calculation , 56 ... Synthesizer, 60 ... DSP / CPU, 61 ... Adaptive filter circuit, 62 ... Filter controller, 63 ... Switch circuit, 64 ... Power calculator, 65 ... FB filter circuit, 66 ... Synthesizer, 70 ... DSP / CPU section 71 ... FF filter circuit 72 ... FB filter circuit 73 ... Adaptive filter circuit 74 ... F Filter control unit 75 ... Switch circuit 76 ... Power calculation unit 77 ... Synthesis unit 80 ... DSP / CPU unit 81 ... FF filter circuit 82 ... FB filter circuit 83, 84 ... Adaptive filter circuit unit 831, 836 ... Weighting filter, 841, 846 ... Weighting filter, 832, 842 ... Adaptive filter circuit, 833, 843 ... Filter control unit, 834, 844 ... Switch circuit, 835, 845 ... Power calculation unit, 85 ... Switch circuit, 86 ... Power calculator

Claims (15)

First signal processing means for forming a noise cancellation signal according to a fixed transfer function from a signal collected by an external sound collecting means provided outside a housing attached to a user's ear; One or both of a second signal processing means for forming a noise cancellation signal according to a transfer function used in a fixed manner from a signal collected by an internal sound collecting means provided inside the housing;
Means for forming a noise cancellation signal from the signal collected by the external sound collecting means, and a third signal processing means capable of adaptively changing the transfer function;
Control means for controlling to change the transfer function of the third signal processing means based on the signal from the external sound collecting means and the signal from the internal sound collecting means;
A noise canceling unit comprising: a combining unit configured to combine a noise canceling signal from one or both of the first signal processing unit and the second signal processing unit and a noise canceling signal from the third signal processing unit. Ring system. The noise canceling system according to claim 1,
A noise canceling system comprising the first signal processing means out of the first signal processing means and the second signal processing means. The noise canceling system according to claim 1,
A noise canceling system comprising the second signal processing means out of the first signal processing means and the second signal processing means. The noise canceling system according to claim 1,
A noise canceling system comprising both the first signal processing means and the second signal processing means. A noise canceling system according to claim 1, claim 2, claim 3 or claim 4,
A plurality of each of the third signal processing means and the control means;
By providing weighting means for weighting the supplied signal for each of the plurality of control means provided,
A noise canceling system comprising a plurality of the third signal processing means having different movable ranges. The first signal processing means forms a noise cancellation signal from the signal collected by the external sound collecting means provided outside the housing attached to the user's ear according to a fixed transfer function. A first signal processing step, and a signal collected by an internal sound collecting means provided inside a housing attached to the user's ear, in accordance with a transfer function used in a fixed manner. Performing one or both of a second signal processing step in which the signal processing means forms a noise cancellation signal;
A means for forming a noise cancellation signal from the signal collected by the external sound collecting means, wherein the transfer function is adaptively based on the signal from the external sound collecting means and the signal from the internal sound collecting means. A third signal processing step of forming a noise cancellation signal by the third signal processing means to be changed;
A synthesis step of synthesizing the noise cancellation signal formed in one or both of the first signal processing step and the second signal processing step and the noise cancellation signal formed in the third signal processing step; A noise canceling signal forming method. The noise cancellation signal forming method according to claim 6,
A noise cancellation signal forming method for executing the first signal processing step of the first signal processing step and the second signal processing step. The noise cancellation signal forming method according to claim 6,
A noise cancellation signal forming method for executing the second signal processing step of the first signal processing step and the second signal processing step. The noise cancellation signal forming method according to claim 6,
A noise cancellation signal forming method for executing both the first signal processing step and the second signal processing step. A noise canceling method according to claim 6, claim 7, claim 8 or claim 9,
Each of the third signal processing means and the control means is provided in plural,
For each of the plurality of control means provided, a weighting means for weighting a signal supplied so as to vary the movable range is provided,
In the third signal processing step, a plurality of noise cancellation signals are formed by the plurality of third signal processing means having different movable ranges. A first signal processing step of forming a noise cancellation signal according to a fixedly used transfer function from a signal collected by an external sound collecting means provided outside a housing attached to a user's ear; Second signal processing step of forming a noise cancellation signal according to a fixed transfer function from a signal collected by an internal sound collecting means provided inside a housing attached to a user's ear And one or both of
A transfer function is adaptively changed based on the signal from the external sound collecting means and the signal from the internal sound collecting means, and a noise cancellation signal is formed from the signal collected by the external sound collecting means. 3 signal processing steps;
Performing a synthesis step of synthesizing the noise cancellation signal formed in one or both of the first signal processing step and the second signal processing step and the noise cancellation signal formed in the third signal processing step. A computer-readable noise canceling signal forming program executed by a computer mounted in a noise canceling system. A noise cancellation signal forming program according to claim 11,
A noise cancellation signal forming program for executing the first signal processing step out of the first signal processing step and the second signal processing step. A noise cancellation signal forming program according to claim 11,
A noise cancellation signal forming program for executing the second signal processing step out of the first signal processing step and the second signal processing step. A noise cancellation program according to claim 11,
A noise cancellation signal forming program for executing both the first signal processing step and the second signal processing step. A noise cancellation signal forming program according to claim 11, claim 12, claim 13 or claim 14,
A plurality of the third signal processing steps are provided, and a weighting step for weighting the signal from the external sound collecting means and the signal from the internal sound collecting means is provided for each of the plurality of third signal processing steps. And
A noise cancellation signal forming method for executing a plurality of the third signal processing steps having different movable ranges.

Images