Acoustic processing method, acoustic processing system, and program

Abstract

This acoustic processing system comprises: a first processing unit 41 that executes first processing on an acoustic signal Ak representing a target sound, and thereby generates an acoustic signal Bk for cancelling the target sound; and a second processing unit 42 that executes second processing on the acoustic signal Bk to reduce nonlinear distortion of a reproduced sound from a speaker unit by utilizing a nonlinear model simulating nonlinearity of a characteristic parameter related to the speaker unit, and thereby generates an acoustic signal Ck to be supplied to the speaker unit.

Description

Acoustic processing method, acoustic processing system, and program

[0001] The present disclosure relates to a technique for reducing a target sound such as an environmental sound.

[0002] For example, noise canceling technology has been proposed that uses the results of collecting environmental sounds present around the listener to cancel the environmental sounds heard by the listener. For example, Patent Document 1 discloses a technology that generates a cancellation signal to cancel muffled sounds synchronized with the rotation of an automobile engine.

[0003] Japanese Patent Application Laid-Open No. 2018-163223

[0004] When the output delay of a speaker unit used for noise cancellation is large, there is a problem that the performance of reducing the target sound (hereinafter referred to as "noise reduction performance") decreases. For example, if a high-performance speaker unit with small waveform distortion such as nonlinear distortion is used, the output delay is suppressed, and as a result, sufficient noise reduction performance can be achieved. However, a large magnet is required to structurally reduce waveform distortion, making it difficult to reduce the weight of the speaker unit. In consideration of the above circumstances, one aspect of the present disclosure aims to achieve both improved noise reduction performance for the target sound and a lightweight speaker unit.

[0005] In order to solve the above problems, an acoustic processing method according to one aspect of the present disclosure performs a first process on a first acoustic signal representing a target sound to generate a second acoustic signal for canceling the target sound, and performs a second process on the second acoustic signal using a nonlinear model that simulates the nonlinearity of characteristic parameters related to the speaker unit to reduce nonlinear distortion in the reproduced sound from the speaker unit, thereby generating a third acoustic signal that is supplied to the speaker unit.

[0006] An acoustic processing system according to one aspect of the present disclosure includes a first processing unit that performs a first processing on a first acoustic signal representing a target sound to generate a second acoustic signal for canceling the target sound, and a second processing unit that performs a second processing on the second acoustic signal using a nonlinear model that simulates the nonlinearity of characteristic parameters related to the speaker unit to reduce nonlinear distortion in the reproduced sound from the speaker unit, thereby generating a third acoustic signal to be supplied to the speaker unit.

[0007] A program according to one aspect of the present disclosure causes a computer system to function as a first processing unit that performs a first processing on a first acoustic signal representing a target sound to generate a second acoustic signal for canceling the target sound, and a second processing unit that performs a second processing on the second acoustic signal to reduce nonlinear distortion in the sound reproduced from the speaker unit using a nonlinear model that simulates the nonlinearity of characteristic parameters related to the speaker unit, thereby generating a third acoustic signal to be supplied to the speaker unit.

[0008] 1 is a schematic diagram of an acoustic system in a first embodiment. FIG. 2 is a block diagram illustrating the configuration of an acoustic system. FIG. 3 is a cross-sectional view of a speaker unit. FIG. 4 is a block diagram illustrating the configuration of a processing unit. FIG. 5 is a block diagram illustrating the configuration of a first processing unit. FIG. 6 is a schematic diagram of a nonlinear model. FIG. 7 is an explanatory diagram regarding the nonlinearity of characteristic parameters. FIG. 8 is a flowchart of second processing. FIG. 9 is a flowchart of processing executed by an acoustic processing system. FIG. 10 is a measurement result of nonlinear distortion accompanying reproduced sound. FIG. 11 is a measurement result of a waveform of reproduced sound. FIG. 12 is a measurement result of a delay difference depending on whether or not second processing is performed. FIG. 13 is a block diagram illustrating the configuration of an acoustic system in a modified example.

[0009] A: Embodiments Figure 1 is a schematic diagram illustrating the configuration of an acoustic system 100 according to one embodiment of the present disclosure. The acoustic system 100 is an in-vehicle system installed in a vehicle 10 such as an automobile. The vehicle 10 is a mobile object equipped with, for example, an engine 11, four tires 12, and four seats 13-1 to 13-4.

[0010] 2 is a block diagram illustrating the configuration of the sound system 100. As illustrated in FIGS. 1 and 2, the sound system 100 includes four microphones 20-1 to 20-4, four speaker units 30-1 to 30-4, and a sound processing system 40. The four microphones 20-1 to 20-4 and the four speaker units 30-1 to 30-4 may be considered as elements that make up the sound processing system 40.

[0011] As illustrated in FIG. 1 , four microphones 20-1 to 20-4 are installed at different positions within the vehicle 10. Similarly, four speaker units 30-1 to 30-4 are also installed at different positions within the vehicle 10. Specifically, one microphone 20-k (k = 1 to 4) and one speaker unit 30-k are installed at positions corresponding to each of the four seats 13-1 to 13-4 within the vehicle 10. That is, a microphone 20-k and a speaker unit 30-k are installed for each listener seated in each seat 13-k. For example, each microphone 20-k is installed on the ceiling of the vehicle cabin at a position corresponding to the seat 13-k, and each speaker unit 30-k is installed on the inner wall surface of the door corresponding to each seat 13-k.

[0012] Each of the four microphones 20-1 to 20-4 is a sound collection device that collects ambient sounds and generates an acoustic signal Ak. The acoustic signal Ak is a sample sequence representing a waveform of a sound wave. The sound represented by the acoustic signal Ak includes environmental sounds heard by a listener seated in the seat 13-k. In this embodiment, the environmental sounds are, for example, engine noise (e.g., engine booming) generated by the rotation of the engine 11 of the vehicle 10. For convenience, an A / D converter that converts each acoustic signal Ak from analog to digital and an amplifier that amplifies the acoustic signal Ak are not shown. The acoustic signal Ak is an example of a "first acoustic signal."

[0013] The acoustic processing system 40 is a signal processing device that processes each acoustic signal Ak. For example, the acoustic processing system 40 is realized by a DSP (Digital Signal Processor) dedicated to processing the acoustic signal Ak. The acoustic processing system 40 is realized by, for example, one or more semiconductor chips.

[0014] As illustrated in FIG. 2, the sound processing system 40 includes four processing units U1 to U4 corresponding to different speaker units 30-k. Each processing unit Uk processes an audio signal Ak to generate an audio signal Ck. The audio signal Ck is a sample sequence for canceling the environmental sound contained in the audio signal Ak. Specifically, each processing unit Uk generates an audio signal Ck representing a canceling sound. The canceling sound represented by the audio signal Ck is a sound that is in an antiphase relationship with the environmental sound of the audio signal Ak. The audio signal Ck is an example of a "third audio signal."

[0015] The acoustic signal Ck is supplied to the speaker unit 30-k. The speaker unit 30-k is a sound emitting device that emits the canceling sound represented by the acoustic signal Ck. For convenience, a D / A converter that converts the acoustic signal Ck from digital to analog and an amplifier that amplifies the acoustic signal Ck are not shown in the figure.

[0016] 3 is a cross-sectional view of the speaker unit 30-k. The speaker unit 30-k includes a frame 31, a magnet 32, a voice coil 33, a diaphragm , an edge 35, and a damper .

[0017] The frame 31 is a structure that forms the exterior of the speaker unit 30-k. The magnet 32 ​​is a circular permanent magnet. The voice coil 33 is a coil that can be displaced in the axial direction within the magnetic field generated by the magnet 32. The diaphragm 34 is a truncated cone-shaped structure. The inner peripheral edge of the diaphragm 34 is fixed to the voice coil 33. The outer peripheral edge of the diaphragm 34 is connected to the frame 31 via an edge 35. The diaphragm 34 and the frame 31 are also connected via a damper 36. The edge 35 and the damper 36 are circular elastic bodies. Sound waves are radiated as the diaphragm 34 moves back and forth in the axial direction.

[0018] 4 is a block diagram illustrating the configuration of any one of the processing units Uk. Each processing unit Uk includes a first processing unit 41 and a second processing unit 42. The first processing unit 41 and the second processing unit 42 may be realized by a single semiconductor chip or by multiple separate semiconductor chips.

[0019] The first processing unit 41 generates an audio signal Bk by performing signal processing (hereinafter referred to as "first processing") on the audio signal Ak. Like the audio signal Ck, the audio signal Bk is a sample series for canceling out the environmental sound contained in the audio signal Ak. That is, the first processing unit 41 generates an audio signal Bk representing a cancellation sound. The cancellation sound represented by the audio signal Bk is a sound that is in an antiphase relationship with the environmental sound of the audio signal Ak. As explained above, the first processing is active noise canceling (ANC) processing that generates a cancellation audio signal Bk from the audio signal Ak. Note that the audio signal Bk is an example of a "second audio signal."

[0020] 5 is a block diagram illustrating the configuration of the first processing unit 41. The first processing unit 41 includes an adaptive filter 411 and a control unit 412. The adaptive filter 411 generates an acoustic signal Bk by performing filter processing on the acoustic signal Ak. The control unit 412 adaptively controls the frequency response of the adaptive filter 411 in accordance with the acoustic signal Ak. A control signal Q of the vehicle 10 may be used to control the adaptive filter 411 by the control unit 412. The control signal Q is a periodic signal that fluctuates, for example, with a period corresponding to the rotation speed of the engine 11.

[0021] The second processing unit 42 in FIG. 4 generates an acoustic signal Ck by performing signal processing (hereinafter referred to as "second processing") on the acoustic signal Bk. The sound waves (hereinafter referred to as "reproduced sound") emitted from the speaker unit 30-k in response to the supply of the acoustic signal Ck are accompanied by nonlinear distortion due to the nonlinearity of each element constituting the speaker unit 30-k. The second processing is signal processing for reducing the nonlinear distortion of the reproduced sound. In other words, the second processing unit 42 generates the acoustic signal Ck from the acoustic signal Bk so as to reduce the nonlinear distortion in the reproduced sound of the speaker unit 30-k.

[0022] Specifically, the second processing unit 42 reduces nonlinear distortion by using a nonlinear model M that simulates the behavior of the speaker unit 30-k. The nonlinear model M is a nonlinear mathematical model that simulates the nonlinearity of the characteristic parameters related to the speaker unit 30-k.

[0023] 6 is a schematic diagram of the nonlinear model M. The input voltage u(t) of the nonlinear model M is the voltage supplied to the voice coil 33. That is, the signal level of the acoustic signal Ck corresponds to the input voltage u(t). The input current i(t) is the current flowing through the voice coil 33.

[0024] The nonlinear model M includes the displacement x(t) of the diaphragm 34 in the axial direction, the electrical resistance Re and inductance Le(x(t)) of the voice coil 33, and the force coefficient (electromagnetic conversion coefficient) Bl(x(t)) of the voice coil 33. The force coefficient Bl(x(t)) is the product of the magnetic flux density B and the winding width l of the voice coil 33. The nonlinear model M also includes the mechanical resistance Rms, mass Mms, and compliance Cms(x(t)) of the vibration system in the speaker unit 30-k. The compliance Cms(x(t)) is the reciprocal of the spring constant Kms(x(t)) of the vibration system.

[0025] 7 is a diagram illustrating the nonlinearity of the characteristic parameters related to the speaker unit 30-k. As illustrated in FIG. 7, the inductance Le(x(t)) changes nonlinearly according to the displacement x(t) of the diaphragm 34. The force coefficient Bl(x(t)) and the compliance Cms(x(t)) (= 1 / Kms(x(t))) also change nonlinearly according to the displacement x(t) of the diaphragm 34.

[0026] As explained above, the inductance Le(x(t)), force coefficient Bl(x(t)), and spring constant Kms(x(t)) in the nonlinear model M are characteristic parameters that change nonlinearly in response to the displacement x(t) of the diaphragm 34. The nonlinear relationship ( FIG. 7 ) between each characteristic parameter (Le(x(t)), Bl(x(t)), Kms(x(t))) and the displacement x(t) is measured in advance using an actual speaker unit 30-k. Note that the relationship between each characteristic parameter and the displacement x(t) may also be estimated by various simulations.

[0027] The nonlinear model M applied to the second processing differs for each processing unit Uk. Specifically, the nonlinear model M used in the second processing of the processing unit Uk simulates the behavior of the speaker unit 30-k. Because the operating characteristics differ for each speaker unit 30-k, the nonlinear model M also differs for each speaker unit 30-k. In other words, the nonlinear model M corresponding to each speaker unit 30-k is used for the second processing related to that speaker unit 30-k. Specifically, the numerical values ​​of the various parameters (constant parameters and characteristic parameters) included in the nonlinear model M differ for each speaker unit 30-k.

[0028] For example, the numerical values ​​of constant parameters such as electrical resistance Re, mechanical resistance Rms, and mass Mms in the nonlinear model M differ for each speaker unit 30-k. Furthermore, the numerical values ​​of characteristic parameters such as inductance Le(x(t)), force coefficient Bl(x(t)), and spring constant Kms(x(t)) in the nonlinear model M also differ for each speaker unit 30-k. For example, the characteristics of change in each characteristic parameter with respect to the displacement x(t) of the diaphragm 34 (i.e., the relationship between the displacement x(t) and the characteristic parameter) differ for each speaker unit 30-k.

[0029] Of the four speaker units 30-1 to 30-4, focusing on speaker unit 30-k1 (k1 = 1 to 4) and speaker unit 30-k2 (k2 = 1 to 4, k2 ≠ k1) for the sake of convenience, the second processing related to speaker unit 30-k1 uses a nonlinear model M corresponding to speaker unit 30-k1, and the second processing related to speaker unit 30-k2 uses a nonlinear model M corresponding to speaker unit 30-k2. The nonlinear model M used in the second processing related to speaker unit 30-k1 and the nonlinear model M used in the second processing related to speaker unit 30-k2 have different parameter values. Note that speaker unit 30-k1 is an example of a "first speaker unit," and speaker unit 30-k2 is an example of a "second speaker unit."

[0030] According to the above configuration, the characteristics unique to each speaker unit 30-k are reflected in each nonlinear model M. Therefore, compared to a configuration in which the nonlinear model M used in the second processing is common to the four speaker units 30-1 to 30-4, the nonlinear distortion caused by the characteristics unique to each speaker unit 30-k can be reduced with high accuracy by the second processing.

[0031] In the nonlinear model M described above, the following formulas (1) and (2) hold true. The symbol v(t) in formulas (1) and (2) represents the velocity of the diaphragm 34 (=∂x / ∂t), and the symbol a(t) represents the acceleration a(t) of the diaphragm 34 (=∂ 2 x / ∂ 2 t).

[0032] 8 is a flowchart of the second process executed by the second processing unit 42. The second process is repeated for each sample of the acoustic signal Bk. The second process includes a linear process S31 and a non-linear process S32.

[0033] The linear processing S31 is a signal processing for calculating a displacement parameter related to the displacement x(t) of the diaphragm 34. The displacement parameter is the displacement x(t), velocity v(t) (=∂x / ∂t), and acceleration a(t) (=∂ 2 x / ∂ 2 t).

[0034] A linear model is used for the linear processing S31. The linear model is a mathematical model in which the nonlinear characteristic parameters (Le(x(t)), Bl(x(t)), Kms(x(t))) in the nonlinear model M are replaced with constants that are independent of the displacement x(t). Specifically, the linear model is expressed by the following equations (3) and (4). That is, Equation (3) is a relational expression obtained by replacing the inductance Le(x(t)) of Equation (1) with the constant Le and the force coefficient Bl(x(t)) with the constant Bl. Similarly, Equation (4) is a relational expression obtained by replacing the force coefficient Bl(x(t)) of Equation (2) with the constant Bl and the spring constant Kms(x(t)) with the constant Kms.

[0035] The second processing unit 42 calculates ideal displacement parameters of the speaker unit 30-k for the acoustic signal Bk by analyzing the simultaneous differential equations expressed by Equations (3) and (4) (i.e., linear simulation). Specifically, the signal level of the acoustic signal Bk is applied as an input voltage u(t) to each linearized equation to calculate displacement parameters (x(t), v(t), a(t)) when the acoustic signal Bk is supplied to the speaker unit 30-k, which is assumed to be a linear system. The acoustic signal Bk corresponds to a target signal representing the reproduced sound to be radiated by the speaker unit 30-k. The target signal is a signal representing an ideal sound wave without accompanying nonlinear distortion.

[0036] Note that a known method such as a general state space model may be arbitrarily employed to solve the simultaneous differential equations in the linear processing S31. For example, an analysis using a state space model is disclosed in, for example, Huang, X. Feng, S. Chen, and Y. Shen, "Analysis of total harmonic distortion of miniature loudspeakers used in mobile phones considering nonlinear acoustic damping," The Journal of the Acoustical Society of America, vol. 149, no. 3, pp. 1579-1588, Mar. 2021, doi: 10.1121 / 10.0003644.

[0037] The nonlinear processing S32 in Fig. 8 is signal processing that generates an acoustic signal Ck using a nonlinear model M expressed by Equation (1) and Equation (2). Specifically, the second processing unit 42 generates an acoustic signal Ck of an input voltage u[n] to be supplied to the speaker unit 30-k to displace the diaphragm 34 according to the displacement parameters calculated by the linear processing S31. The input voltage u[n] corresponds to the signal level of the acoustic signal Ck. Note that the symbol n refers to the number of one sample in a discrete signal. In the following description, each variable is represented using the number n.

[0038] From the above-mentioned formula (2), the following formula (5) representing the input current i[n] is derived.

[0039] The second processing unit 42 calculates the input current i[n] by applying the displacement parameters (x(t), v(t), a(t)) calculated by the linear processing S31 to Equation (5). Note that the force coefficient Bl(x[n]) in Equation (5) is set to a value corresponding to the displacement x(t) calculated by the linear processing S31 based on the nonlinear relationship ( FIG. 7 ) between the force coefficient Bl(x(t)) and the displacement x(t) measured in advance. Similarly, the spring constant Kms(x[n]) in Equation (5) is set to a value corresponding to the displacement x(t) calculated by the linear processing S31 based on the nonlinear relationship ( FIG. 7 ) between the spring constant Kms(x[n]) and the displacement x(t) measured in advance.

[0040] Furthermore, the following equation (6) representing the input voltage u[t] can be derived from the above equation (1). In equation (6), time t is replaced with the number n of each sample, just like in equation (5). The symbol i'[n] in equation (6) means the time derivative of the input current i[n].

[0041] The second processing unit 42 calculates the input voltage u[n] by applying the displacement parameters (x(t), v(t), a(t)) calculated by the linear processing S31 and the input current i[n] calculated by the formula (5) to the formula (6). Note that the inductance Le(x[n]) in the formula (6) is set to a value corresponding to the displacement x(t) calculated by the linear processing S31 based on the nonlinear relationship ( FIG. 7 ) between the inductance Le(x(t)) and the displacement x(t) measured in advance.

[0042] The second processing unit 42 generates a time series of the input voltage u[n] by repeating the second processing exemplified above. The time series of the input voltage u[n] is supplied to the speaker unit 30-k as the acoustic signal Ck. Therefore, nonlinear distortion in the reproduced sound of the speaker unit 30-k is reduced.

[0043] As described above, in this embodiment, the displacement parameters (x(t), v(t), a(t)) of the diaphragm 34 are calculated by linear processing S31 using a linear model, and the acoustic signal Ck is generated by applying the displacement parameters to nonlinear processing S32 using a nonlinear model M. Therefore, nonlinear distortion can be reduced with high accuracy by simple processing.

[0044] 9 is a flowchart of the processing executed by the sound processing system 40. For example, the processing of FIG. 9 is executed for each sample of the sound signal Ak. Furthermore, the processing of FIG. 9 is executed in parallel in each of the four sound processing systems 40 of the sound system 100.

[0045] When the processing starts, the first processing unit 41 acquires the acoustic signal Ak (S1). The first processing unit 41 generates the acoustic signal Bk by executing a first processing on the acoustic signal Ak (S2). The second processing unit 42 generates the acoustic signal Ck by executing a second processing on the acoustic signal Bk (S3). As illustrated in FIG. 8 , the second processing includes a linear processing step S31 and a nonlinear processing step S32. The second processing unit 42 supplies the acoustic signal Ck to the speaker unit 30-k (S4).

[0046] As described above, for each of the four speaker units 30-1 to 30-4, the sound signal Bk is generated by the first processing and the sound signal Ck is generated by the second processing. Therefore, the environmental sounds heard by the listener can be reduced at multiple points corresponding to the different speaker units 30-k.

[0047] As described above, in this embodiment, the acoustic signal Bk for canceling the environmental sound is generated by the first processing, and then the second processing using the nonlinear model M is executed on the acoustic signal Bk, thereby reducing nonlinear distortion in the sound reproduced from the speaker unit 30-k. As a result of the reduction in nonlinear distortion, the output delay is reduced. The output delay is the delay between the pickup of the environmental sound by the microphone 20-k and the emission of the canceling sound by the speaker unit 30-k (for example, a processing delay by the sound processing system 40).

[0048] The smaller the output delay, the better the environmental sound silencing performance achieved by the first processing. Therefore, even when a lightweight speaker unit 30-k with a relatively large output delay is used, the environmental sound silencing performance can be improved. In other words, according to this embodiment, it is possible to achieve both improved environmental sound silencing performance and a lightweight speaker unit 30-k.

[0049] FIG. 10 shows the measurement results of nonlinear distortion (THD: Total Harmonic Distortion) associated with the reproduced sound. FIG. 10 shows both the nonlinear distortion in this embodiment, in which the second processing is performed on the acoustic signal Bk, and the nonlinear distortion in a configuration in which the second processing is not performed (hereinafter referred to as the "comparison"). According to this embodiment, it can be seen that the execution of the second processing on the acoustic signal Bk reduces nonlinear distortion compared to the comparison. The effect of reducing nonlinear distortion is particularly noticeable in the frequency range below 100 Hz, where environmental sounds (engine noise) exist.

[0050] Fig. 11 shows the measurement results of the waveform of the reproduced sound when the acoustic signal Ak is a sine wave. Fig. 11 shows both the measurement results when the acoustic signal Ak is a sine wave of 47 Hz and the measurement results when the acoustic signal Ak is a sine wave of 59 Hz. The horizontal axis in Fig. 11 represents the sample number n. Each measurement result in Fig. 11 shows both the waveform of the reproduced sound in this embodiment and the waveform of the reproduced sound in a comparative example. It can be seen from Fig. 11 that the second processing reduces nonlinear distortion, and as a result, the delay of the reproduced sound is also reduced.

[0051] FIG. 12 shows the measurement results of the delay difference between when the second processing is performed (present embodiment) and when it is not performed (comparison example). The difference in the rise time of the reproduced sound in this embodiment compared to the rise time of the reproduced sound in the comparison example is shown as the delay difference on the vertical axis of FIG. 12. The delay difference is expressed as the number of samples. The smaller the delay difference value within the negative number range, the more the output delay of the reproduced sound is improved due to the reduction of nonlinear distortion. It can also be seen from FIG. 12 that according to this embodiment, the delay of the reproduced sound is significantly reduced compared to the comparison example. Furthermore, the effect of delay reduction is particularly noticeable in the frequency range below 100 Hz where environmental sound (engine noise) is present. Therefore, environmental sound can be effectively reduced by the first processing.

[0052] B: Modifications Specific modifications that can be added to the above-mentioned embodiments are exemplified below. Two or more embodiments arbitrarily selected from the following examples may be combined as appropriate within the scope of not mutually contradicting each other.

[0053] (1) In the above embodiment, engine noise is exemplified as an example of the environmental sound to be canceled by the first processing. However, environmental sound is not limited to this example. For example, road noise caused by contact between the road surface and each tire 12 of the vehicle 10 is also conceivable as an environmental sound to be canceled by the first processing. That is, the first processing may be, for example, road noise canceling (RNC) processing that generates an acoustic signal Bk for canceling road noise from an acoustic signal Ak. In the first processing for canceling road noise, an acceleration signal in addition to the acoustic signal Ak may be used to control the frequency response of the adaptive filter 411. The acceleration signal is a signal detected by an acceleration sensor installed in the vehicle 10. The intensity of the road noise correlates with the acceleration of the vehicle 10.

[0054] As can be understood from the above examples, the first processing is comprehensively expressed as processing for canceling environmental sounds such as engine noise or road noise, and the specific configuration or procedure for the first processing is arbitrary.

[0055] (2) In the above embodiment, environmental sounds such as engine noise and road noise are exemplified as targets of cancellation, but the targets of cancellation by the first process are not limited to environmental sounds. Any type of target sound, not limited to environmental sounds, may be considered as targets of cancellation by the first process.

[0056] (3) In the above embodiment, an example has been given in which an acoustic signal Bk is generated from an acoustic signal Ak generated by one microphone 20-k, but multiple acoustic signals Ak generated by different microphones 20-k may be used to generate the acoustic signal Bk in the first processing. For example, the first processing unit 41 of the processing unit Uk may generate the acoustic signal Bk by performing the first processing on four systems of acoustic signals A1 to A4 generated by different microphones 20-k.

[0057] (4) In the above embodiment, the numerical values ​​of the various parameters (constant parameters and characteristic parameters) included in the nonlinear model M are different for each speaker unit 30-k, but the nonlinear model M may be common to multiple speaker units 30-k. For example, the numerical values ​​of the parameters in the nonlinear model M may be common to multiple speaker units 30-k that have common operating characteristics.

[0058] (5) In each of the above embodiments, an audio system 100 including multiple pairs of microphones 20-k and speaker units 30-k has been exemplified. However, an audio system 100 including only one pair of microphones 20-k and speaker units 30-k is also contemplated.

[0059] (6) In the above-described embodiments, the sound processing system 40 is realized by a DSP dedicated to processing the sound signal Ak, but the sound processing system 40 may also be realized by using a general-purpose computer system, as shown in Fig. 13. The sound processing system 40 in Fig. 13 includes a control device 51 and a storage device 52.

[0060] The control device 51 is composed of one or more processors that control each element of the sound processing system 40. Specifically, the control device 51 is composed of one or more types of processors, such as a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), or an SPU (Sound Processing Unit).

[0061] The storage device 52 is one or more memories that store programs executed by the control device 51 and data used by the control device 51. The storage device 52 is configured with a known recording medium such as a magnetic recording medium or a semiconductor recording medium. The control device 51 executes the programs stored in the storage device 52 to function as the first processing unit 41 and the second processing unit 42 for generating the acoustic signal Ck from the acoustic signal Ak.

[0062] In the configuration of FIG. 13 , the functions of the acoustic processing system 40 are realized through cooperation between one or more processors constituting the control device 51 and a program stored in the storage device 52. The program according to the present disclosure may be provided in a form stored on a computer-readable recording medium and installed on a computer. The recording medium may be, for example, a non-transitory recording medium, such as an optical recording medium (optical disk) such as a CD-ROM, but may also include any known type of recording medium, such as a semiconductor recording medium or a magnetic recording medium. Note that a non-transitory recording medium includes any recording medium other than a transient, propagating signal, and does not exclude volatile recording media. Furthermore, in a configuration in which a distribution device distributes a program via a communication network, the storage medium that stores the program in the distribution device corresponds to the non-transitory recording medium described above.

[0063] C: Supplementary Notes From the above-described exemplary embodiments, the following configurations can be understood, for example.

[0064] An acoustic processing method according to one aspect (aspect 1) of the present disclosure generates a second acoustic signal for canceling out a target sound by performing a first process on a first acoustic signal representing the target sound, and generates a third acoustic signal to be supplied to the speaker unit by performing a second process on the second acoustic signal using a nonlinear model that simulates the nonlinearity of characteristic parameters related to the speaker unit to reduce nonlinear distortion in the reproduced sound from the speaker unit.

[0065] According to the above configuration, a second acoustic signal for canceling the target sound is generated by the first processing, and then a second processing using a nonlinear model is performed on the second acoustic signal, thereby reducing nonlinear distortion in the sound reproduced from the speaker unit. As a result of the reduction in nonlinear distortion, the output delay is reduced. The smaller the output delay, the better the target sound silencing performance achieved by the first processing. Therefore, the target sound silencing performance can be improved even when a lightweight speaker unit with a relatively large output delay is used. In other words, according to the present disclosure, it is possible to achieve both improved target sound silencing performance and a lightweight speaker unit.

[0066] The "target sound" is a sound to be canceled by the acoustic processing method of the present disclosure. For example, environmental sounds present around the listener are an example of the "target sound." For example, assuming application to a moving body such as an automobile, engine noise generated by engine rotation, road noise caused by contact between the tires and the road surface, and other low-frequency sounds of 100 Hz or less are exemplified as the "target sound" (environmental sound).

[0067] The "first processing" is signal processing (e.g., noise canceling processing) that generates a second acoustic signal for canceling the target sound. The second acoustic signal is, for example, a signal that represents a sound wave that is in the opposite phase to the target sound. Ideally, the second acoustic signal is a signal that can completely cancel the target sound, but the "second acoustic signal" also includes a signal that cannot completely cancel the target sound but can reduce it.

[0068] "Characteristic parameters" are various parameters related to the characteristics of a speaker unit. Examples of characteristic parameters include the inductance (Le(x(t))) of the voice coil in the speaker unit, the force coefficient (Bl(x(t))) which is the product of the magnetic flux density and the winding width, and the spring constant (Kms(x(t))) of the vibration system. The "nonlinearity" of a characteristic parameter refers to the relationship in which the characteristic parameter changes nonlinearly depending on the displacement of the diaphragm in the speaker unit.

[0069] A "nonlinear model" is a mathematical model that simulates a nonlinear relationship between a displacement parameter (for example, displacement, velocity, or acceleration) relating to the displacement of a diaphragm in a speaker unit and a characteristic parameter of the speaker unit.

[0070] In a specific example (Aspect 2) of Aspect 1, the second processing includes linear processing that uses a linear model that simulates the behavior of the speaker unit to calculate a displacement parameter related to the displacement of the diaphragm of the speaker unit from the second acoustic signal, and nonlinear processing that uses the nonlinear model to generate the third acoustic signal that represents a signal level to be supplied to the speaker unit to displace the diaphragm according to the displacement parameter. In the above aspect, the displacement parameter of the diaphragm is calculated by linear processing using the linear model, and the third acoustic signal is generated by applying the displacement parameter to nonlinear processing using the nonlinear model. Therefore, nonlinear distortion can be reduced with high accuracy through simple processing. Note that the "linear model" is a mathematical model in which each characteristic parameter of the speaker unit is set to a constant that is independent of the displacement of the diaphragm.

[0071] In a specific example (Aspect 3) of Aspect 1 or Aspect 2, the speaker unit is installed in a mobile object having an engine and tires, and the target sound includes engine noise generated by the rotation of the engine or road noise caused by contact between the tires and the road surface. In the above aspect, a third acoustic signal is generated and supplied to the speaker unit installed in the mobile object. Major components of noise, such as engine noise (e.g., engine booming) or road noise, are predominantly contained in a low-frequency range, for example, below 100 Hz. The second processing using a nonlinear model effectively reduces nonlinear distortion, particularly in the low-frequency range, resulting in reduced output delay in the low-frequency range. As a result of reducing output delay in the low-frequency range as described above, the first processing can generate a second acoustic signal that can effectively cancel out target sounds specific to the mobile object, such as engine noise (e.g., engine booming) or road noise.

[0072] In a specific example (Aspect 4) of any of Aspects 1 to 3, the first processing generates the second acoustic signal, and the second processing generates the third acoustic signal, for each of a plurality of speaker units. In the above aspects, the first processing and the second processing are executed for each of a plurality of speaker units. Therefore, the target sound heard by the listener can be reduced at a plurality of points corresponding to different speaker units. The above aspect is particularly suitable for an environment where a plurality of listeners are located at different positions within a space (for example, the interior space of a moving body such as a car).

[0073] In a specific example (Aspect 5) of Aspect 4, the second processing for a first speaker unit among the plurality of speaker units uses a nonlinear model corresponding to the first speaker unit, and the second processing for a second speaker unit other than the first speaker unit among the plurality of speaker units uses a nonlinear model corresponding to the second speaker unit. According to the above aspect, compared to an embodiment in which the nonlinear model used for the second processing is common to a plurality of speaker units, nonlinear distortion caused by characteristics unique to each speaker unit can be reduced with high accuracy by the second processing. Because output delay is reduced by reducing nonlinear distortion, the target sound can be reduced with high accuracy using each speaker unit.

[0074] An acoustic processing system according to one aspect (aspect 6) of the present disclosure includes a first processing unit that performs a first processing on a first acoustic signal representing a target sound to generate a second acoustic signal for canceling the target sound, and a second processing unit that performs a second processing on the second acoustic signal using a nonlinear model that simulates the nonlinearity of characteristic parameters related to the speaker unit to reduce nonlinear distortion in the reproduced sound from the speaker unit, thereby generating a third acoustic signal to be supplied to the speaker unit.

[0075] A program according to one aspect (aspect 7) of the present disclosure causes a computer system to function as a first processing unit that performs a first processing on a first acoustic signal representing a target sound to generate a second acoustic signal for canceling the target sound, and a second processing unit that performs a second processing on the second acoustic signal to reduce nonlinear distortion in the sound reproduced from the speaker unit using a nonlinear model that simulates the nonlinearity of characteristic parameters related to the speaker unit, thereby generating a third acoustic signal to be supplied to the speaker unit.

[0076] 100...acoustic system, 10...vehicle, 11...engine, 12...tire, 13-k (13-1 to 13-4)...seat, 20-k (20-1 to 20-4)...microphone, 30-k (30-1 to 30-4)...speaker unit, 31...frame, 32...magnet, 33...voice coil, 34...diaphragm, 35...edge, 36...damper, 40...acoustic processing system, 41...first processing unit, 42...second processing unit, 51...control device, 52...storage device.

Claims

1. An acoustic processing method implemented by a computer system, which comprises: executing a first processing on a first acoustic signal representing a target sound to generate a second acoustic signal for canceling the target sound; and executing a second processing on the second acoustic signal using a nonlinear model that simulates the nonlinearity of characteristic parameters related to a speaker unit to reduce nonlinear distortion of the sound reproduced from the speaker unit, thereby generating a third acoustic signal to be supplied to the speaker unit.

2. The acoustic processing method of claim 1, wherein the second processing includes: a linear processing for calculating a displacement parameter relating to the displacement of a diaphragm of the speaker unit from the second acoustic signal using a linear model that simulates the behavior of the speaker unit; and a nonlinear processing for generating the third acoustic signal, which represents a signal level to be supplied to the speaker unit in order to displace the diaphragm in accordance with the displacement parameter, using the nonlinear model.

3. The acoustic processing method according to claim 1 or 2, wherein the speaker unit is installed on a moving object having an engine and tires, and the target sound includes engine noise generated by the rotation of the engine, or road noise caused by contact between the road surface and the tires.

4. The acoustic processing method according to claim 1, further comprising the steps of: generating the second acoustic signal by the first processing; and generating the third acoustic signal by the second processing, for each of a plurality of speaker units.

5. The acoustic processing method of claim 4, wherein the second processing relating to a first speaker unit among the plurality of speaker units utilizes a nonlinear model corresponding to the first speaker unit, and the second processing relating to a second speaker unit other than the first speaker unit among the plurality of speaker units utilizes a nonlinear model corresponding to the second speaker unit.

6. An acoustic processing system comprising: a first processing unit that performs a first processing on a first acoustic signal representing a target sound to generate a second acoustic signal for canceling the target sound; and a second processing unit that performs a second processing on the second acoustic signal to reduce nonlinear distortion of the sound reproduced from the speaker unit by utilizing a nonlinear model that simulates the nonlinearity of characteristic parameters related to the speaker unit, to generate a third acoustic signal to be supplied to the speaker unit.

7. A program that causes a computer system to function as a first processing unit that performs a first processing on a first acoustic signal representing a target sound to generate a second acoustic signal for canceling the target sound, and a second processing unit that performs a second processing on the second acoustic signal to reduce nonlinear distortion of the sound reproduced from the speaker unit by utilizing a nonlinear model that simulates the nonlinearity of characteristic parameters related to the speaker unit, thereby generating a third acoustic signal to be supplied to the speaker unit.

Images