Masking sound control system, sound masking system, masking sound control method, and program

Abstract

The present invention addresses the problem of alleviating both discomfort given to people by noise and discomfort given to people by masking sound, with improved balance. In a masking-sound control system (1), an optimization processing unit (1d) uses, as a control parameter, at least one of the reproduction sound pressure level (Pa(j)) of masking sound M((j)) emitted from a speaker (2(j)) and speaker position, which is the position of the speaker (2(j)). The optimization processing unit (1d) obtains, as an optimized parameter, a control parameter that minimizes a discomfort level, which is the degree of discomfort given to people by the masking sound (M(j)) and noise (N1) in a target area (Rt). A masking-sound adjustment unit (1e) adjusts the control parameter to the optimized parameter.

Description

[[Technical Field]]

[0001] The present invention relates to a masking sound control system, a sound masking system, a masking sound control method, and a program. [[Background Art]]

[0002] There is a sound masking technology that makes it difficult for people to hear noise and suppresses leakage of conversations.

[0003] For example, the masking system disclosed in Patent Document 1 aims to reduce discomfort given to people in a space while ensuring a high masking effect in the space. Accordingly, the masking system collects human speech with a microphone and acquires an audio signal sequence representing the speech. Then, the masking system extracts a plurality of audio signal sequences from different sections in the audio signal sequence, and superimposes the extracted audio signal sequences on the time axis to generate a masker audio signal and stores it in a storage medium. The masking system reproduces the masker sound stored in the storage medium, and causes a speaker to emit the reproduced sound toward one of two adjacent regions across a partition.

[0004] The masking system (sound masking system) of Patent Document 1 generates a masker sound (masking sound) that reduces discomfort (annoyance) given to people in a space. However, the masker sound generated by superimposing human speech on the time axis is an unfamiliar sound to people, and thus causes discomfort (annoyance) to people in the space with respect to the masker sound. Therefore, there is a demand for more balanced suppression of both the annoyance caused by noise to people and the annoyance caused by masking sound to people. [[Prior Art Literature]] [[Patent Literature]]

[0005] [[Patent Document 1]] Japanese Patent No. 6007481 [[Summary of the Invention]]

[0006] The purpose of this disclosure is to provide a masking sound control system, a sound masking system, a masking sound control method, and a program that can suppress both the nuisance caused by noise and the nuisance caused by masking sounds in a more balanced manner.

[0007] The masking sound control system of this disclosure controls a masking sound emitted from a speaker to mask noise within a target area. The masking sound control system comprises an optimization processing unit and a masking sound adjustment unit. The optimization processing unit is before The speaker's position is the speaker location. Place The control parameters are determined as optimization parameters, which minimize the degree of nuisance caused to people by the masking sound and the noise in the target area. The masking sound adjustment unit adjusts the control parameters to the optimization parameters. The masking sound control system of this disclosure controls a masking sound emitted from a speaker to mask noise within a target area. The masking sound control system comprises an optimization processing unit and a masking sound adjustment unit. The optimization processing unit uses at least one of the playback sound pressure level of the masking sound emitted from the speaker and the speaker position, which is the location of the speaker, as control parameters, and determines the control parameters as optimization parameters that minimize the degree of nuisance, which is the degree of nuisance caused to people by the masking sound and the noise in the target area. The masking sound adjustment unit adjusts the control parameters to the optimization parameters. The optimization processing unit determines the optimization parameters by finding the optimal solution to an optimization problem that minimizes the degree of nuisance, with the degree of nuisance as the objective function. The optimization processing unit determines the optimization parameters by finding the optimal solution to the optimization problem using the steepest descent method. The optimization processing unit sets the learning rate that determines the convergence when finding the optimal solution in the steepest descent method to a first value when the sound pressure of the masking sound increases, and sets the learning rate to a second value that is smaller than the first value when the sound pressure of the masking sound decreases. The masking sound control system of this disclosure controls a masking sound emitted from a speaker to mask noise within a target area. The masking sound control system comprises an optimization processing unit and a masking sound adjustment unit. The optimization processing unit uses at least one of the playback sound pressure level of the masking sound emitted from the speaker and the speaker position, which is the position of the speaker, as control parameters, and determines the control parameters as optimization parameters that minimize the level of nuisance, which is the degree of nuisance caused to people by the masking sound and the noise in the target area. The masking sound adjustment unit adjusts the control parameters to the optimization parameters. The level of nuisance is an area nuisance level, which is the degree of nuisance caused to people by the masking sound and the noise across the target area. The optimization processing unit determines the control parameters as optimization parameters that minimize the area nuisance level. The area nuisance level is a value based on a plurality of point nuisance levels, which are the degree of nuisance caused to people by the masking sound and the noise at each of a plurality of listening points within the target area. The aforementioned area nuisance level is the integral value obtained by integrating the degree of nuisance in the target area. The masking sound control system of this disclosure controls a masking sound emitted from a speaker to mask noise within a target area. The masking sound control system comprises an optimization processing unit and a masking sound adjustment unit. The optimization processing unit uses at least one of the playback sound pressure level of the masking sound emitted from the speaker and the speaker position, which is the position of the speaker, as control parameters, and determines the control parameters as optimization parameters that minimize the level of nuisance, which is the degree of nuisance caused to people by the masking sound and the noise in the target area. The masking sound adjustment unit adjusts the control parameters to the optimization parameters. The level of nuisance is an area nuisance level, which is the degree of nuisance caused to people by the masking sound and the noise across the target area. The optimization processing unit determines the control parameters as optimization parameters that minimize the area nuisance level. The area nuisance level is a value based on a plurality of point nuisance levels, which are the degree of nuisance caused to people by the masking sound and the noise at each of a plurality of listening points within the target area. Each of the aforementioned multiple point nuisance levels is a value based on the difference between a masking detection value, which is the detected sound pressure of the masking sound at each of the multiple listening points, and a masking target value, which is the target sound pressure of the masking sound at each of the multiple listening points. The masking target value is set to the second lower limit if the noise detection value, which indicates the sound pressure of the noise at each of the multiple listening points, is smaller than the first lower limit, and is set to the second upper limit if the noise detection value is larger than the first upper limit. If the noise detection value is greater than or equal to the first lower limit and less than or equal to the first upper limit, the masking target value is set to a larger value as the noise detection value increases, within the range greater than the second lower limit and less than the second upper limit. The masking sound control system of this disclosure controls a masking sound emitted from a speaker to mask noise within a target area. The masking sound control system comprises an optimization processing unit and a masking sound adjustment unit. The optimization processing unit uses at least one of the playback sound pressure level of the masking sound emitted from the speaker and the speaker position, which is the location of the speaker, as control parameters, and determines the control parameters as optimization parameters that minimize the level of nuisance, which is the degree of nuisance caused to people by the masking sound and the noise in the target area. The masking sound adjustment unit adjusts the control parameters to the optimization parameters. The level of nuisance is a value based on a single-point nuisance level, which is the degree of nuisance caused to people by the masking sound and the noise at one listening point within the target area. The single-point nuisance level is a value based on the difference between a masking detection value and a masking target value. The masking detection value is the detected sound pressure value of the masking sound at the single listening point. The masking target value is the target value of the sound pressure of the masking sound at one listening point. The masking target value is set to the second lower limit if the noise detection value indicating the sound pressure of the noise at one listening point is less than the first lower limit, and to the second upper limit if the noise detection value is greater than the first upper limit. If the noise detection value is greater than or equal to the first lower limit and less than or equal to the first upper limit, the masking target value is set to a larger value as the noise detection value increases, within a range greater than the second lower limit and less than the second upper limit.

[0008] The sound masking system of this disclosure comprises the masking sound control system described above and the speaker.

[0009] The masking sound control method of this disclosure controls a masking sound emitted from a speaker to mask noise within a target area. The masking sound control method includes an optimization processing step and a masking sound adjustment step. The optimization processing step is before The speaker's position is the speaker location. Place The control parameter is determined as the optimization parameter, which minimizes the degree of nuisance caused to people by the masking sound and the noise in the target area. The masking sound adjustment step adjusts the control parameter to the optimization parameter. The masking sound control method of this disclosure controls a masking sound emitted from a speaker to mask noise within a target area. The masking sound control method includes an optimization step and a masking sound adjustment step. The optimization step determines the control parameter as an optimization parameter, using at least one of the playback sound pressure level of the masking sound emitted from the speaker and the speaker position, which is the position of the speaker, as control parameters, and minimizing the degree of nuisance, which is the degree of nuisance caused to people by the masking sound and the noise in the target area. The masking sound adjustment step adjusts the control parameter to the optimization parameter. The optimization step determines the optimization parameter by finding the optimal solution to an optimization problem that minimizes the degree of nuisance, with the degree of nuisance as the objective function. The optimization step determines the optimization parameter by finding the optimal solution to the optimization problem using the steepest descent method. The optimization process step involves setting the learning rate that determines the convergence when finding the optimal solution in the steepest descent method to a first value when the sound pressure of the masking sound increases, and setting the learning rate to a second value smaller than the first value when the sound pressure of the masking sound decreases. The masking sound control method of this disclosure controls a masking sound emitted from a speaker to mask noise within a target area. The masking sound control method includes an optimization step and a masking sound adjustment step. The optimization step determines an optimization parameter that minimizes the nuisance level, which is the degree of nuisance caused to people by the masking sound and the noise, in the target area, using at least one of the playback sound pressure level of the masking sound emitted from the speaker and the speaker position, which is the position of the speaker, as control parameters. The masking sound adjustment step adjusts the control parameter to the optimization parameter. The nuisance level is an area nuisance level, which is the degree of nuisance caused to people by the masking sound and the noise across the target area. The optimization step determines an optimization parameter that minimizes the area nuisance level. The area nuisance level is a value based on a plurality of point nuisance levels, which are the degree of nuisance caused to people by the masking sound and the noise at each of a plurality of listening points within the target area. The aforementioned area nuisance level is the integral value obtained by integrating the degree of nuisance in the target area. The masking sound control method of this disclosure controls a masking sound emitted from a speaker to mask noise within a target area. The masking sound control method includes an optimization step and a masking sound adjustment step. The optimization step determines an optimization parameter that minimizes the nuisance level, which is the degree of nuisance caused to people by the masking sound and the noise, in the target area, using at least one of the playback sound pressure level of the masking sound emitted from the speaker and the speaker position, which is the position of the speaker, as control parameters. The masking sound adjustment step adjusts the control parameter to the optimization parameter. The nuisance level is an area nuisance level, which is the degree of nuisance caused to people by the masking sound and the noise across the target area. The optimization step determines an optimization parameter that minimizes the area nuisance level. The area nuisance level is a value based on a plurality of point nuisance levels, which are the degree of nuisance caused to people by the masking sound and the noise at each of a plurality of listening points within the target area. Each of the aforementioned multiple point nuisance levels is a value based on the difference between a masking detection value, which is the detected sound pressure of the masking sound at each of the multiple listening points, and a masking target value, which is the target sound pressure of the masking sound at each of the multiple listening points. The masking target value is set to the second lower limit if the noise detection value, which indicates the sound pressure of the noise at each of the multiple listening points, is smaller than the first lower limit; it is set to the second upper limit if the noise detection value is larger than the first upper limit; and if the noise detection value is greater than or equal to the first lower limit and less than or equal to the first upper limit, it is set to a value that is larger the larger the noise detection value is, within the range of greater than the second lower limit and less than the second upper limit. The masking sound control method of this disclosure controls a masking sound emitted from a speaker to mask noise within a target area. The masking sound control method includes an optimization step and a masking sound adjustment step. The optimization step determines an optimization parameter that minimizes the nuisance level, which is the degree of nuisance caused to people by the masking sound and the noise, in the target area, using at least one of the playback sound pressure level of the masking sound emitted from the speaker and the speaker position, which is the position of the speaker, as control parameters. The masking sound adjustment step adjusts the control parameter to the optimization parameter. The nuisance level is a value based on a point nuisance level, which is the degree of nuisance caused to people by the masking sound and the noise at one listening point within the target area. The point nuisance level is a value based on the difference between a masking detection value, which is the detected sound pressure of the masking sound at the one listening point, and a masking target value, which is the target sound pressure of the masking sound at the one listening point. The masking target value is set to the second lower limit if the noise detection value, which indicates the sound pressure of the noise at one listening point, is smaller than the first lower limit; if the noise detection value is larger than the first upper limit, it is set to the second upper limit; and if the noise detection value is greater than or equal to the first lower limit and less than or equal to the first upper limit, it is set to a value that is larger as the noise detection value increases, within a range greater than the second lower limit and less than the second upper limit.

[0010] The program of this disclosure causes a computer system to execute the masking sound control method described above. [Brief explanation of the drawing]

[0011] [Figure 1] Figure 1 is a block diagram showing a masking sound control system according to an embodiment. [Figure 2] Figure 2 is a plan view showing the working space including the target area in the masking sound control system described above. [Figure 3] Figure 3 is a perspective view showing the working space, including the target area, in the masking sound control system described above. [Figure 4] Figure 4 is a graph showing the relationship between the noise detection value and the masking target value in the masking sound control system described above. [Figure 5] Figure 5 shows the waveforms of noise and masking sound in the masking sound control system described above. [Figure 6] Figure 6 is a flowchart showing the masking sound control method of an embodiment. [Figure 7] Figure 7 shows an image of the control parameter update process in the steepest descent method in the second modified example. [Figure 8] Figure 8 is a waveform diagram showing the tracking response of the masking sound to noise when the learning rate is fixed at a constant value. [Figure 9] Figure 9A shows the waveforms of noise and masking sound when using the steepest descent method with a learning rate μ = 0.02. Figure 9B shows the waveforms of noise and masking sound when using the steepest descent method with a learning rate μ = 0.2. [Figure 10] Figure 10 is a waveform diagram showing the response of the masking sound to noise when using the steepest descent method in the second modified example. [Figure 11]FIG. 11 is a waveform diagram showing noise and masking sound when the steepest descent method described above is used. [Figure 12] FIGS. 12A and 12B are perspective views showing a noise suppressing device according to a third modification. [Figure 13] FIG. 13 is a side view showing the arrangement of the noise suppressing device described above. [Figure 14] FIG. 14 is a perspective view showing the noise suppressing device for the personal booth described above. [Figure 15] FIG. 15 is a block diagram showing the masking sound control system described above. DESCRIPTION OF EMBODIMENTS

[0012] The following embodiments generally relate to a masking sound control system, a sound masking system, a masking sound control method, and a program. More specifically, the following embodiments relate to a masking sound control system, a sound masking system, a masking sound control method, and a program that control a masking sound emitted from a speaker for masking noise in a target area. The embodiments described below are merely examples of embodiments of the present disclosure. The present disclosure is not limited to the following embodiments, and various modifications can be made according to design and the like as long as the effects of the present disclosure can be achieved.

[0013] (1) Overview of sound masking system A sound masking system 100 according to an embodiment of the present disclosure includes a masking sound control system 1, a speaker 2(j), and a microphone 3, as shown in FIG. 1. In the present embodiment, the number of speakers 2(j) is plural, but it is sufficient that the number is at least one. In addition, although the number of microphones 3 is one in FIG. 1, a plurality of microphones may be provided.

[0014] In the sound masking system 100, speaker 2(j) is installed in the target area Rt. The masking sound control system 1 controls the masking sound M(j) emitted by speaker 2(j). The masking sound M(j) masks noise N1 within the target area Rt, making it difficult for people in the target area Rt to hear noise N1, and suppressing the leakage of conversations between people in the target area Rt.

[0015] In this embodiment, the space containing the target area Rt is assumed to be a working space R10 where multiple people perform desk work, work, meetings, and other tasks, as well as take breaks, such as an open office, coworking space, or free-address office. Such a working space R10 is generally divided into a first area R11 and a second area R12, as shown in Figure 2. Figure 3 is a perspective view showing an example of the working space R10.

[0016] Area 1, R11, is used for face-to-face or remote meetings and discussions, and also includes a break area. In other words, since communication between people takes place in Area 1, R11, the voices of people in Area 1 are likely to be generated as noise N1. On the other hand, Area 2, R12, is used for individual desk work and other tasks.

[0017] However, when people hear the voices of those around them, they tend to become distracted by what is being said, and their concentration decreases. In other words, the voices of those around them interfere with a person's thinking and reduce their intellectual productivity. Therefore, the intellectual productivity of people in the second area R12 is reduced by the noise N1, which is the voices generated in the first area R11.

[0018] Therefore, the sound masking system 100 emits a masking sound M(j) from speaker 2(j) so that noise N1 is difficult for people in the second area R12 to hear. People in the second area R12 can suppress the decrease in intellectual productivity because noise N1 is difficult to hear when masking sound M(j) is applied over it. The masking sound M(j) is preferably an ambient sound such as the sound of a babbling brook, rain, a waterfall, or wind. Ambient sounds are sounds that are difficult for people to notice and can also help people relax or improve their concentration. Alternatively, the masking sound M(j) may be white noise.

[0019] As described above, in masking technology that makes noise N1 difficult to hear by emitting a masking sound, the masking effect increases as the volume of the masking sound increases, making it difficult for people in the second area R12 to hear noise N1. However, when the volume of the masking sound is increased, there is a problem that the masking sound becomes loud and bothersome for people in the second area R12. In other words, there were people who found the masking sound annoying.

[0020] Therefore, the masking sound control system 1 of the sound masking system 100 has the following configuration, with the second area R12 as the target area Rt. The second area R12 (target area Rt) is a space where we want to suppress the decrease in intellectual productivity caused by noise N1 generated in the first area R11.

[0021] The masking sound control system 1 controls the masking sound M(j) emitted from speaker 2(j) to mask noise N1 within the second area R12 (target area Rt). The masking sound control system 1 comprises an optimization processing unit 1d and a masking sound adjustment unit 1e. The optimization processing unit 1d uses at least one of the playback sound pressure level of the masking sound M(j) emitted from speaker 2(j) and the speaker position, which is the position of speaker 2(j), as control parameters, and determines the control parameters that minimize the level of nuisance, which is the degree of nuisance caused to people by the masking sound M(j) and noise N1, in the second area R12, as optimization parameters. The masking sound adjustment unit 1e adjusts the control parameters to the optimization parameters.

[0022] The sound masking system 100 and masking sound control system 1, having the above configuration, adjust the control parameters of the masking sound M(j) emitted by speaker 2(j) to optimization parameters that minimize the level of nuisance in the second area R12 (target area Rt). Therefore, the sound masking system 100 and masking sound control system 1 can suppress both the nuisance caused by noise N1 and the nuisance caused by masking sound M(j) in a more balanced manner.

[0023] Furthermore, Area 2R12 is a relatively large space capable of accommodating multiple people (for example, several to several dozen people). In such a large Area 2R12, the sound pressure of the masking sound reaching people's ears varies depending on their location within Area 2R12. In other words, Area 2R12 contains a mix of people who can easily hear the masking sound and people who cannot. Therefore, even if the sound pressure of the masking sound is locally optimized within Area 2R12, most people within Area 2R12 still found the masking sound annoying.

[0024] Therefore, in the masking sound control system 1 of the sound masking system 100, the degree of nuisance is preferably the area nuisance level Wa, which is the degree of nuisance caused to people by the masking sound M(j) and noise N1 across the second area R12 (target area Rt). In this case, the optimization processing unit 1d determines the control parameters that minimize the area nuisance level Wa as the optimization parameters.

[0025] The sound masking system 100 and masking sound control system 1, having the above configuration, adjust the control parameters of the masking sound M(j) emitted by speaker 2(j) to optimization parameters that minimize the area nuisance level Wa in the second area R12 (target area Rt). Therefore, the sound masking system 100 and masking sound control system 1 can suppress the level of nuisance caused to people by the masking sound M(j) and noise N1 over a wide area.

[0026] (2) Details of the sound masking system The sound masking system 100 comprises a masking sound control system 1 and a speaker 2(j). The sound masking system 100 further comprises a microphone 3 and a sound source 4.

[0027] (2.1) Points to be heard In the second area R12 (target area Rt), multiple listening points Q(k) are set. In Figure 1, k = 1, 2, 3, ..., km, and there are km listening points Q(1)-Q(km) set in the second area R12.

[0028] Then, at the listening point Q(k), a composite masking sound Mt(k) arrives, which is formed by the superposition of masking sounds M(j) emitted from multiple speakers 2(j). For a person at the listening point Q(k), the composite masking sound Mt(k) is superimposed on the noise N1, making it difficult to hear the noise N1.

[0029] In this embodiment, as shown in Figures 2 and 3, mutually orthogonal X, Y, and Z axes are virtually set up to define the three-dimensional position within the second area R12. The X and Y axes extend horizontally, and the Z axis extends vertically. Let x be the coordinate defined by the X axis (X coordinate), y be the coordinate defined by the Y axis (Y coordinate), and z be the coordinate defined by the Z axis (Z coordinate). That is, the three-dimensional position within the second area R12 is represented by the coordinates x, y, z. In this case, the position of each of the multiple listening points Q(k) in the second area R12 is represented by the coordinates xk, yk, zk. That is, the multiple listening points Q(k) are located three-dimensionally within the second area R12, and listening point Q(k) can be represented as Q(xk, yk, zk).

[0030] The listening point Q(k) preferably corresponds to a location within the second area R12 where an individual concentrates on their work. For example, the X and Y coordinates of listening point Q(k) correspond to the positions of desks, chairs, etc., on the horizontal plane. The Z coordinate of listening point Q(k) is set according to the posture of the person at listening point Q(k) (sitting, standing, etc.), the expected range of the person's height, etc.

[0031] Furthermore, multiple listening points Q(k) may be arranged in a three-dimensional grid within the second area R12.

[0032] (2.2) Speakers In this embodiment, multiple speakers 2(j) are installed in the second area R12 (target area Rt). In Figure 1, j = 1, 2, 3, ..., jm, and jm speakers 2(1)-2(jm) are installed in the second area R12. The positions of each of the multiple speakers 2(j) in the second area R12 are represented by coordinates xj, yj, zj.

[0033] Speaker 2(j) is installed on the ceiling, floor, and walls of Area 2R12. Speaker 2(j) receives a masking signal (drive signal for speaker 2(j)) from the masking sound control system 1 and emits a masking sound M(j) corresponding to the masking signal. The sound pressure of the masking sound M(j) reproduced by speaker 2(j) is called the reproduced sound pressure level Pa(j) (unit: dB). The reproduced sound pressure level is also called the output sound pressure level.

[0034] Furthermore, some of the multiple speakers 2(j) may be installed in the first area R11.

[0035] (2.3) Microphone In this embodiment, microphone 3 corresponds to a sound collection device. Microphone 3 (one microphone 3 in Figure 1) is installed in the first area R11. Microphone 3 collects noise N1 generated in the first area R11 and outputs the noise signal Bn to the masking sound control system 1.

[0036] (2.4) Masking sound control system The masking sound control system 1 comprises an optimization processing unit 1d and a masking sound adjustment unit 1e. The masking sound control system 1 further comprises a noise detection unit 1a, a target sound pressure calculation unit 1b, a masking sound detection unit 1c, and a storage unit 1f.

[0037] The masking sound control system 1 preferably includes a computer system. The computer system executes a program to realize some or all of the functions of the masking sound control system 1. The computer system primarily includes a processor that operates according to the program. The processor can be of any type as long as it can realize its functions by executing a program. The processor consists of one or more electronic circuits, including a semiconductor integrated circuit (IC) or LSI (Large Scale Integration). Here, we refer to them as ICs and LSIs, but the terminology changes depending on the degree of integration; they may also be called system LSIs, VLSIs (Very Large Scale Integrations), or ULSIs (Ultra Large Scale Integrations). Field-Programmable Gate Arrays (FPGAs), which are programmed after the LSI is manufactured, or reconfigurable logic devices that allow for the reconfiguration of internal junctions or the setup of internal circuit compartments within the LSI, can also be used for the same purpose. Multiple electronic circuits may be integrated on a single chip or provided on multiple chips. Multiple chips may be aggregated in a single device or provided in multiple devices. The program is recorded on a non-temporary recording medium such as a computer-readable ROM, optical disc, or hard disk drive. The program may be pre-stored on the recording medium or supplied to the recording medium via a wide-area communication network, including the Internet.

[0038] The masking sound control system 1 may be implemented using either a single computer or multiple computer devices working in conjunction with each other. Furthermore, the masking sound control system 1 may be constructed as a cloud computing system.

[0039] (2.4.1) Storage section The memory unit 1f stores coordinate data (coordinate data) indicating the three-dimensional positions of the microphone 3, multiple listening points Q(k), and multiple speakers 2(j) (speaker positions) in the working space R10. By referring to the coordinate data, the distance and direction between the microphone 3, listening points Q(k), and speakers 2(j) can be determined. The coordinate data is created using the X, Y, and Z coordinates mentioned above.

[0040] (2.4.2) Noise detection unit The noise detection unit 1a determines the noise detection value Pn(k) (in dB) based on the noise signal Bn, which is the signal of noise N1 collected by the microphone 3. The noise detection value Pn(k) represents the sound pressure of noise N1 at each of the multiple listening points Q(k) within the second area R12 (target area Rt).

[0041] Specifically, the noise detection unit 1a uses the coordinate data from the storage unit 1f to determine the distance between the microphone 3 and each of the multiple listening points Q(k). The noise detection unit 1a performs attenuation correction to reduce the magnitude of the noise signal Bn based on the distance between the microphone 3 and the listening points Q(k). That is, the noise detection unit 1a attenuates and corrects the noise signal Bn so that it represents the magnitude of the noise N1 that has reached each of the multiple listening points Q(k). Then, based on the attenuated and corrected noise signal Bn, the noise detection unit 1a determines a noise detection value Pn(k) that indicates the sound pressure of the noise N1 at each of the multiple listening points Q(k).

[0042] The noise detection unit 1a performs attenuation correction to reduce the magnitude of the noise signal Bn based on the distance between the microphone 3 and the listening point Q(1), and determines the noise detection value Pn(1) at the listening point Q(1). The noise detection unit 1a also performs attenuation correction to reduce the magnitude of the noise signal Bn based on the distance between the microphone 3 and the listening point Q(2), and determines the noise detection value Pn(2) at the listening point Q(2). The noise detection unit 1a similarly determines the noise detection values ​​Pn(3)-Pn(km) at the listening point Q(3)-Q(km).

[0043] (2.4.3) Target sound pressure calculation unit The target sound pressure calculation unit 1b determines the masking target value Pd(k) based on the noise detection value Pn(k) at each of the multiple listening points Q(k) (in dB). The masking target value Pd(k) is the target value of the sound pressure of the synthesized masking sound Mt(k) at each of the multiple listening points Q(k). In other words, the masking sound control system 1 determines the masking target value Pd(k) at each of the multiple listening points Q(k).

[0044] Generally, noise can be classified into significant noise and unintentional noise. People tend to find significant noise more annoying than unintentional noise. Significant noise is sounds that have meaning, such as people talking or speaking, and corresponds to noise N1 in this embodiment. If the sound pressure of significant noise audible to people is about 30 dB or higher, people will understand the content of the conversation and become aware of (pay attention to) the significant noise. Unintentional noise is sounds that do not have meaning, such as ambient noise, air conditioning noise, and traffic noise, and corresponds to masking sound in this embodiment. People are less bothered by (less aware of) unintentional noise than significant noise, but if the unintentional noise level exceeds 50 dB, people will perceive the unintentional noise itself as noisy. Therefore, it is preferable that the sound pressure of the masking sound at listening point Q(k) is 50 dB or less. Also, since the background noise in an office is less than about 40 dB, it is preferable that the sound pressure of the masking sound at listening point Q(k) is 40 dB or higher.

[0045] Furthermore, in order to make noise N1 less audible at listening point Q(k), the sound pressure of the masking sound at listening point Q(k) must be at least 10 dB higher than the sound pressure of noise N1 at listening point Q(k).

[0046] Therefore, in this embodiment, the target sound pressure calculation unit 1b determines the masking target value Pd(k) from the noise detection value Pn(k) using the following equation 1. That is, the masking target value Pd(k) at listening point Q(k) is uniquely determined according to the sound pressure of the noise N1 at listening point Q(k).

[0047]

number

[0048] As shown in Figure 4, the target sound pressure calculation unit 1b sets the masking target value Pd(k) to the second lower limit of 40dB if the noise detection value Pn(k) is less than the first lower limit of 30dB. The target sound pressure calculation unit 1b sets the masking target value Pd(k) to the second upper limit of 50dB if the noise detection value Pn(k) is greater than the first upper limit of 40dB. The target sound pressure calculation unit 1b sets the masking target value Pd(k) to Pn(k) + 10dB if the noise detection value Pn(k) is greater than or equal to the first lower limit of 30dB and less than or equal to the first upper limit of 40dB. In other words, if the noise detection value Pn(k) is 30 dB or more and 40 dB or less, the target masking value Pd(k) is set to a value that is greater than the second lower limit of 40 dB and less than the second upper limit of 50 dB, and the value increases as the noise detection value Pn(k) increases.

[0049] Furthermore, the specific values ​​for the first lower limit, second lower limit, first upper limit, and second upper limit are not limited to any particular value.

[0050] (2.4.4) Masking sound detection unit The masking sound detection unit 1c determines the masking detection value Pm(k), which is the detected sound pressure value of the synthesized masking sound Mt(k) at each of the multiple listening points Q(k).

[0051] The masking sound detection unit 1c uses the coordinate data from the storage unit 1f to determine the distance between each of the multiple speakers 2(j) and each of the multiple listening points Q(k). The masking sound detection unit 1c also receives data from the masking sound adjustment unit 1e regarding the playback sound pressure level Pa(j) of the masking sound M(j) emitted by each of the multiple speakers 2(j). Based on the distances between each of the multiple speakers 2(j) and each of the multiple listening points Q(k), and the playback sound pressure level Pa(j) of the masking sound M(j), the masking sound detection unit 1c performs attenuation correction to reduce the playback sound pressure level Pa(j), thereby determining the sound pressure of the masking sound M(j) that has reached each of the multiple listening points Q(k). Then, the masking sound detection unit 1c determines the sound pressure of the synthesized masking sound Mt(k) at listening point Q(k) as the masking detection value Pm(k) based on the sound pressure of each masking sound M(j) that has reached listening point Q(k).

[0052] In other words, the masking sound detection unit 1c can determine the masking detection value Pm(k) at each of the multiple listening points Q(k) based on the coordinates xj, yj, zj of speaker 2(j), the coordinates xk, yk, zk of listening point Q(k), and the playback sound pressure level Pa(j) of the masking sound M(j). In this case, the masking detection value Pm(k) is expressed by the following equation 2. Note that the coordinates xj, yj, zj of speaker 2(j) correspond to the speaker positions indicating the respective positions of speaker 2(j).

[0053]

number

[0054] (2.4.5) Optimization Processing Unit The optimization processing unit 1d uses at least one of the following as control parameters: the sound pressure level Pa(j) of the masking sound M(j) emitted from speaker 2(j), and the speaker position, which is the location of speaker 2(j). The optimization processing unit 1d then determines the control parameters that minimize the area nuisance level Wa as the optimization parameters. The area nuisance level Wa is the degree of nuisance caused to people by the masking sound M(j) and noise N1 across the second area R12 (target area Rt). In other words, the area nuisance level Wa is a single nuisance level evaluated for the entire area of ​​the second area R12, which can accommodate multiple people.

[0055] Furthermore, it is preferable that the optimization processing unit 1d determines the control parameters that minimize the area nuisance level Wa as optimization parameters based on the relationship between the control parameters and the area nuisance level Wa (see Equation 7).

[0056] In this embodiment, the area nuisance level Wa is preferably a value based on multiple point nuisance levels Wp(k), which are the degree of nuisance caused to people by the masking sound M(j) and noise N1 at each of the multiple listening points Q(k) within the second area R12.

[0057] Furthermore, it is preferable that each of the multiple point nuisance levels Wp(k) is a value based on the difference between the masking detection value Pm(k), which is the detected sound pressure value of the masking sound M(j) at each of the multiple listening points Q(k), and the masking target value Pd(k), which is the target sound pressure value of the masking sound M(j) at each of the multiple listening points Q(k).

[0058] The following explains the point-based nuisance level Wp(k), the area-based nuisance level Wa, and the optimization process.

[0059] (Points indicating level of annoyance) First, in this embodiment, multiple speakers 2(j) each emit a masking sound M(j). In this case, the point nuisance level Wp(k) is the degree of nuisance caused to a person at each of the multiple listening points Q(k) by the combined masking sound Mt(k) (a sound obtained by superimposing the masking sounds M(j) emitted from each of the multiple speakers 2(j)) and the noise N1. That is, the point nuisance level Wp(k) is the degree of nuisance caused by the combined masking sound Mt(k) and the noise N1, and is the local degree of nuisance at listening point Q(k) within the second area R12.

[0060] Specifically, the point nuisance level Wp(k) is determined as the error between the masking detection value Pm(k) and the masking target value Pd(k). Preferably, the point nuisance level Wp(k) is expressed as the square root of the difference obtained by subtracting the masking detection value Pm(k) from the masking target value Pd(k), as shown in Equation 3 below.

[0061]

number

[0062] Here, since the masking target value Pd(k) is expressed by equation 1 above, the point nuisance level Wp(k) is expressed by the following equation 4.

[0063]

number

[0064] The point-based nuisance level Wp(k) is the degree of nuisance a person feels at listening point Q(k) due to the synthesized masking sound Mt(k) and noise N1. The larger the error between the masking detection value Pm(k) and the masking target value Pd(k), the higher the point-based nuisance level Wp(k).

[0065] Specifically, if the sound pressure of the synthesized masking sound Mt(k) reaching listening point Q(k) is too high, the person at listening point Q(k) will perceive the synthesized masking sound Mt(k) as loud, and the point nuisance level Wp(k) will increase.

[0066] Furthermore, the point nuisance Wp(k) expressed by equations 3 and 4 above reflects not only the level of nuisance a person feels due to the synthesized masking sound Mt(k), but also the level of nuisance a person feels due to the noise N1. In other words, if the sound pressure of the synthesized masking sound Mt(k) reaching the listening point Q(k) is too low, a person at the listening point Q(k) will perceive the noise N1 as loud, and the point nuisance Wp(k) will increase.

[0067] The point nuisance level Wp(k) can be expressed as Wp(xk,yk,zk).

[0068] (Area nuisance level) The area nuisance level Wa corresponds to the integral of the degree of nuisance in the second area R12, based on the point nuisance level Wp(k) at each of the multiple listening points Q(k). For example, the area nuisance level Wa is the volume integral of the degree of nuisance in the second area R12.

[0069] For example, the area nuisance level Wa is the value obtained by dividing the point nuisance level Wp(k) by volume, as shown in Equation 5 below.

[0070]

number

[0071] As a process of integraling the point nuisance level Wp(k) by volume, the sum of the point nuisance levels Wp(k) at each of the multiple listening points Q(k) may be calculated. That is, as shown in Equation 6 below, the sum of the point nuisance levels Wp(k) at each of the multiple listening points Q(k) may be used as the area nuisance level Wa.

[0072]

number

[0073] (Relationship between control parameters and area nuisance level) From equations 1-6 above, the area nuisance level Wa is a function of the coordinates xj, yj, zj of speaker 2(j) and the reproduced sound pressure level Pa(j), and is expressed as Wa(x1…xjm, y1…yjm, z1…zjm, Pa(1)…Pa(jm)). Furthermore, the area nuisance level Wa(x1…xjm, y1…yjm, z1…zjm, Pa(1)…Pa(jm)) is expressed from equations 1-6 as follows: Equation 7 corresponds to the relationship between the control parameter and the area nuisance level Wa. In equation 7, each of x1…xjm, y1…yjm, z1…zjm, and Pa(1)…Pa(jm) can be a control parameter. That is, at least one of x1…xjm, y1…yjm, z1…zjm, and Pa(1)…Pa(jm) is a control parameter.

[0074]

number

[0075] (Optimization process) In this embodiment, the control parameters are defined as the respective playback sound pressure levels Pa(1)...Pa(jm) of the masking sounds M(1)...M(jm). The optimization parameters corresponding to the playback sound pressure levels Pa(1)...Pa(jm) are defined as the optimized sound pressure Pb(1)...Pb(jm). In this case, the coordinates x1...xjm, y1...yjm, z1...zjm, which indicate the positions of the speakers 2(1)-2(jm), are constants (fixed values) based on the coordinate data in the storage unit 1f.

[0076] The optimization processing unit 1d performs optimization processing to determine the playback sound pressure level Pa(1)...Pa(jm) that minimizes the area nuisance level Wa, and sets this as the optimized sound pressure Pb(1)...Pb(jm). The optimized sound pressure Pb(1)...Pb(jm) is the playback sound pressure level Pa(1)-Pa(jm) of the masking sound M(1)-M(jm) used to minimize the area nuisance level Wa, and serves as the control value for the playback sound pressure level Pa(1)-Pa(jm).

[0077] The optimization processing unit 1d preferably determines the optimization parameters by finding the optimal solution to an optimization problem that minimizes the area nuisance Wa, with the area nuisance Wa as the objective function. In particular, the optimization processing unit 1d preferably determines the optimization parameters by finding the optimal solution to the optimization problem using the steepest descent method.

[0078] The steepest descent method is one of the methods for solving optimization problems that seek the minimum value of an objective function. In the steepest descent method, the objective function (area nuisance level Wa) is plotted against the control parameters (here, the playback sound pressure level Pa(1)...Pa(jm)) and this surface is used as the objective function surface. Then, the steepest descent method searches for the minimum value of the objective function from the slope (gradient) of the objective function surface.

[0079] Specifically, the optimization processing unit 1d partially differentiates the objective function, area nuisance Wa, with respect to each of the control parameters (Pa(1)...Pa(jm)). First, the optimization processing unit 1d sets initial values ​​for each of the control parameters (Pa(1)...Pa(jm)) and finds the gradient of the objective function surface at each initial value. Then, the optimization processing unit 1d updates each of the control parameters (Pa(1)...Pa(jm)) with a value obtained by multiplying the found gradient by the learning rate μ (step size parameter). The optimization processing unit 1d then finds the gradient of the objective function surface at the updated control parameters (Pa(1)...Pa(jm)). The optimization processing unit 1d repeatedly updates these control parameters (Pa(1)...Pa(jm)) and calculates the gradient of the objective function surface, thereby optimizing each of the control parameters (Pa(1)...Pa(jm)) and finding the control parameters (Pa(1)...Pa(jm)) that minimize the area nuisance level Wa, which are then used as the optimization parameters (Pb(1)...Pb(jm)).

[0080] Equation 8 below is an example of the optimization process described above, and shows the process of updating the playback sound pressure level Pa(1). Pa(1)new is the updated control parameter, and Pa(1)old is the control parameter before the update. μ is the learning rate (step size parameter).

[0081]

number

[0082] As described above, the optimization processing unit 1d performs the optimization process using the gradient of the objective function, area nuisance level Wa (the partial derivative of the control parameter). In this case, if the partial derivative is approximated by finite difference approximation, as shown in Equation 9 below, it is not necessary to determine the value of the area nuisance level Wa itself. However, it is necessary to define Equation 7 above, which is the relationship between the control parameter and the area nuisance level Wa.

[0083]

number

[0084] As described above, the optimization processing unit 1d determines the optimized sound pressure Pb(1)...Pb(jm) as an optimization parameter for each of the multiple speakers 2(j) to adjust the masking sound M(j) so as to minimize the area nuisance Wa(r).

[0085] (2.4.6) Masking sound adjustment section The masking sound adjustment unit 1e adjusts the playback sound pressure level Pa(j) of the masking sound M(j) emitted by speaker 2(j) to the optimized sound pressure Pb(j).

[0086] Specifically, sound source 4 stores masking sound data. Masking sound adjustment unit 1e reads the masking sound data from sound source 4 and generates a masking signal (drive signal for speaker 2(j)) based on the masking sound data. Then, masking sound adjustment unit 1e outputs the masking signal to each of the multiple speakers 2(j), causing each of the multiple speakers 2(j) to output the masking sound M(j). At this time, masking sound adjustment unit 1e adjusts the playback sound pressure level Pa(j) of the masking sound M(j) emitted by each of the multiple speakers 2(j) to the optimized sound pressure Pb(j).

[0087] Figure 5 shows the waveform Gn1 of noise N1 (noise detection value Pn(k)) and the waveform Gm1 of masking sound M(j) in this embodiment. As shown in Figure 5, the masking sound control system 1 optimizes the playback sound pressure level Pa(j) of the masking sound M(j) according to the sound pressure of noise N1. In other words, the masking sound control system 1 responds in real time to the generation conditions of noise N1 and optimally controls the playback sound pressure level Pa(j) of the masking sound M(j).

[0088] (2.4.7) Advantages The masking sound control system 1, having the configuration described above, controls the playback sound pressure level Pa(j) of the masking sound M(j) emitted from speaker 2(j) to an optimized sound pressure Pb(j) in order to minimize the area nuisance level Wa. As a result, the masking sound control system 1 can suppress the degree of nuisance caused to people by the masking sound M(j) and noise N1 over a wide area. Furthermore, the masking sound control system 1 can optimally control the playback sound pressure level Pa(j) of the masking sound M(j) in real time in response to the generation conditions of noise N1 by determining the masking target value Pd(k) according to the noise N1 collected by microphone 3.

[0089] Furthermore, the optimization processing unit 1d determines the control parameters that minimize the area nuisance level Wa as optimization parameters based on Equation 7, which is the relationship between the control parameters and the area nuisance level Wa. As a result, the masking sound control system 1 can accurately determine the optimization parameters that minimize the area nuisance level Wa.

[0090] Furthermore, the optimization processing unit 1d uses the area nuisance level Wa as its objective function and finds the optimal solution to an optimization problem that minimizes the area nuisance level Wa, thereby determining the optimization parameters. As a result, the masking sound control system 1 can accurately determine the optimization parameters that minimize the area nuisance level Wa.

[0091] Furthermore, the optimization processing unit 1d determines the optimization parameters by finding the optimal solution to the optimization problem using the steepest descent method. As a result, the masking sound control system 1 can accurately determine the optimization parameters that minimize the area nuisance level Wa.

[0092] Furthermore, the area nuisance level Wa is a value based on multiple point nuisance levels Wp(k), which represent the degree of nuisance caused to people by the masking sound M(j) and noise N1 at each of the multiple listening points Q(k) within the second area R12 (target area Rt). As a result, the masking sound control system 1 can suppress the degree of nuisance caused to people by the masking sound M(j) and noise N1 over a wide area based on the spatial area nuisance level Wa.

[0093] Furthermore, the area nuisance level Wa is the integral of the degree of nuisance in the second area R12 (target area Rt). As a result, the masking sound control system 1 can suppress the degree of nuisance caused by the masking sound M(j) and noise N1 to people over a wide area based on the spatial area nuisance level Wa.

[0094] Furthermore, the area nuisance level Wa is a value obtained by dividing the degree of nuisance in the second area R12 (target area Rt) by volume. As a result, the masking sound control system 1 can suppress the degree of nuisance caused to people by the masking sound M(j) and noise N1 over a wide area based on the area nuisance level Wa in three-dimensional space.

[0095] Furthermore, the area nuisance level Wa is the sum of the nuisance levels Wp(k) of multiple points. As a result, the masking sound control system 1 can suppress the degree of nuisance caused to people by the masking sound M(j) and noise N1 over a wide area based on the spatial area nuisance level Wa obtained by simple calculation.

[0096] Furthermore, each of the multiple point nuisance levels Wp(k) is a value based on the difference between the masking detection value Pm(k) at each of the multiple listening points Q(k) and the masking target value Pd(k) at each of the multiple listening points Q(k). As a result, the masking sound control system 1 can suppress the degree of nuisance caused by the masking sound M(j) and noise N1 to people over a wide area, based on the area nuisance level Wa obtained from the highly accurate point nuisance levels Wp(k).

[0097] Furthermore, the masking target value Pd(k) is set according to the graph shown in Figure 4. As a result, the masking sound control system 1 can accurately reflect the degree of nuisance a person feels due to the masking sound M(j) and noise N1.

[0098] Furthermore, the optimization processing unit 1d determines the optimization parameters for each of the multiple speakers 2(j) that minimize the area nuisance level Wa. Then, the masking sound adjustment unit 1e adjusts the control parameters of the multiple speakers 2(j) to the multiple optimization parameters. As a result, even when using multiple speakers 2(j), the masking sound control system 1 can suppress the level of nuisance caused by the masking sound M(j) and noise N1 to people over a wide area.

[0099] (3) Masking sound control method The masking sound control system 1 described above includes a computer system. The program of this embodiment causes the computer system to execute the masking sound control method shown in the flowchart of Figure 6. The masking sound control method includes an optimization processing step S1 and a masking sound adjustment step S2.

[0100] In optimization processing step S1, the optimization processing unit 1d uses at least one of the following as control parameters: the sound pressure level Pa(j) of the masking sound M(j) emitted from speaker 2(j), and the coordinates xj, yj, zj (speaker position) indicating the position of speaker 2(j). The optimization processing unit 1d then determines the control parameters that minimize the level of nuisance (e.g., area nuisance level Wa) as the optimization parameters. The level of nuisance is the degree to which the masking sound M(j) and noise N1 cause nuisance to people in the second area R12 (target area Rt).

[0101] In masking sound adjustment step S2, the masking sound adjustment unit 1e adjusts the control parameters of the masking sound M(j) emitted by speaker 2(j) to the optimized parameters.

[0102] The masking sound control method described above can suppress both the nuisance caused by noise N1 and the nuisance caused by masking sound M(j) in a more balanced manner.

[0103] (4) First variation The control parameters may also be the coordinates x1…xjm, y1…yjm, z1…zjm (speaker position) of speaker 2(j). In this case, the masking sound adjustment unit 1e is equipped with a drive mechanism that makes the position of each of the multiple speakers 2(j) variable. The drive mechanism has a motor, cylinder, or slider, etc., to move the multiple speakers 2(j) in three or two dimensions. The optimization processing unit 1d then determines the coordinates x1…xjm, y1…yjm, z1…zjm of each of the multiple speakers 2(j) that minimize the area nuisance Wa, as the optimization coordinates (optimization parameters). The masking sound adjustment unit 1e moves each of the multiple speakers 2(j) so that each of their coordinates matches their respective optimization coordinates.

[0104] Furthermore, the control parameters may be both the playback sound pressure levels Pa(1)...Pa(jm) of the masking sounds M(1)...M(jm) and the coordinates x1...xjm, y1...yjm, z1...zjm (speaker position) of speaker 2(j).

[0105] Multiple listening points Q(k) are located two-dimensionally in the target area Rt (second area R12), and listening points Q(k) may be represented as Q(xk,yk). In other words, multiple listening points Q(k) may be arranged two-dimensionally in the target area Rt. In this case, the area nuisance level Wa is the value obtained by integraling the degree of nuisance in the target area Rt over its area. As a result, the masking sound control system 1 can suppress the degree of nuisance caused to people by the masking sound M(j) and noise N1 over a wide area based on the area nuisance level Wa in a two-dimensional space.

[0106] Furthermore, the area nuisance level Wa is not limited to the integral of the degree of nuisance in the target area Rt. The area nuisance level Wa is sufficient as long as it reflects the degree of nuisance in the target area Rt.

[0107] The optimization processing unit 1d may use the conjugate gradient method, Newton's method, or a quasi-Newton's method to find the optimal solution to the optimization problem.

[0108] The masking sound control system 1 may have a microphone installed at each of the multiple listening points Q(k). In this case, the masking sound detection unit 1c determines the masking detection value Pm(k) at each of the multiple listening points Q(k) based on the output of the microphones installed at each of the multiple listening points Q(k). The noise detection unit 1a also determines the noise detection value Pn(k) at each of the multiple listening points Q(k) based on the output of the microphones installed at each of the multiple listening points Q(k).

[0109] In the above embodiment, the noise detection unit 1a performs attenuation correction to reduce the magnitude of the noise signal Bn using the coordinate data stored in the memory unit 1f. The masking sound detection unit 1c also performs attenuation correction to reduce the reproduced sound pressure level Pa(j) using the coordinate data stored in the memory unit 1f. However, the noise detection unit 1a and the masking sound detection unit 1c may use coordinate data read from a position sensor such as a beacon instead of the coordinate data from the memory unit 1f. Position sensors are provided on each of the microphone 3 and speaker 2(j), and the coordinate data indicating the three-dimensional position of each of the microphone 3 and speaker 2(j) is transmitted to the masking sound control system 1. Furthermore, position sensors may also be provided on each of the listening points Q(k), and the coordinate data indicating the three-dimensional position of the listening point Q(k) may be transmitted to the masking sound control system 1. In this configuration, even if the positions of the microphone 3, speaker 2(j), and listening point Q(k) are changed, the changed positions can be easily reflected.

[0110] Furthermore, it is preferable that the noise detection unit 1a and the masking sound detection unit 1c perform attenuation correction that takes into account the reverberation characteristics of the space, such as the room constant of the working space R10 or the target area Rt, as attenuation correction. The reverberation characteristics data is stored in the storage unit 1f. With this configuration, attenuation correction can be performed based on the actual reverberation characteristics of the space, thus improving the accuracy of attenuation correction.

[0111] The memory unit 1f may store data for the transfer function between each of the multiple speakers 2(j) and each of the multiple listening points Q(k), and data for the transfer function between the microphone 3 and each of the multiple listening points Q(k). In this case, the noise detection unit 1a can determine the noise detection value Pn(k) from the noise signal Bn based on the transfer function between the microphone 3 and each of the multiple listening points Q(k). The masking sound detection unit 1c can also determine the masking detection value Pm(k) based on the transfer function between each of the multiple speakers 2(j) and each of the multiple listening points Q(k).

[0112] The masking sound control system 1, sound masking system 100, masking sound control method, and program of this disclosure may also be used in a personal booth (individual booth) that secures an individual workspace, in addition to a working space where multiple people work.

[0113] In particular, the masking sound control system 1, sound masking system 100, masking sound control method, and program of this disclosure can be applied to a semi-closed personal booth having a structure in which at least one of the top and side surfaces of the enclosure is open.

[0114] In the following modified examples, components similar to those in the masking sound control system 1, sound masking system 100, masking sound control method, and program of this embodiment are denoted by the same reference numerals and their descriptions are omitted.

[0115] (5) Second variation Figure 7 shows an image of the control parameter update process in the steepest descent method. Figure 7 is a graph with the control parameter on the horizontal axis and the objective function on the vertical axis. Specifically, the control parameter is the sound pressure level Pa(j) of the masking sound M(j), and the objective function is the area nuisance level Wa. Curve L1 in Figure 7 shows the relationship between the sound pressure level Pa(j) of the masking sound M(j) and the area nuisance level Wa. The slope of curve L1 represents the gradient of the objective function surface.

[0116] In the steepest descent method, the playback sound pressure level Pa(j) is sequentially updated from the initial value V0 to values ​​V1, V2, ..., Vn, so as to minimize the area nuisance Wa. The update interval H of the playback sound pressure level Pa(j) is determined by the learning rate μ. If the slope of curve L1 (the gradient of the objective function surface) is Δd, then H is expressed as H = μ × Δd. In other words, the update interval H is the value obtained by multiplying the slope Δd of curve L1 by the learning rate μ. This learning rate μ is a hyperparameter that determines the convergence when finding the optimal solution in the steepest descent method, and determines the tracking ability of the playback sound pressure level Pa(j) to the noise N1.

[0117] Figure 8 shows the tracking responsiveness of the masking sound M(j) to noise N1 in an optimization process using the steepest descent method with the learning rate μ fixed to a constant value. In Figure 8, the waveform Gm10 of the masking sound M(j) tracks the waveform Gn10 of noise N1 with a constant time constant determined by the fixed learning rate μ.

[0118] Figure 9A shows the waveform Gn11 of noise N1 and the waveform Gm11 of masking sound M(j) when the steepest descent method is used with a learning rate μ = 0.02. In this case, the tracking response of masking sound M(j) to noise N1 is low, and at the rising edge of noise N1, masking sound M(j) is smaller than noise N1 (see C11 in Figure 9A), resulting in a reduced masking effect by masking sound M(j).

[0119] Figure 9B shows the waveforms Gn12 of noise N1 and Gm12 of the masking sound M(j) when using the steepest descent method with a learning rate μ=0.2. In this case, the masking sound M(j) responds quickly to the rising edge of noise N1 (see C12 in Figure 9B) and rises in sync with noise N1. However, because the tracking responsiveness of the masking sound M(j) to noise N1 is too high, the sound pressure fluctuations of the masking sound M(j) become drastic, resulting in an unnatural and unpleasant masking sound M(j).

[0120] As described above, there is a trade-off between the tracking responsiveness of the masking sound M(j) to the noise N1, which is determined by the learning rate μ of the optimization process, and the naturalness of the masking sound M(j).

[0121] Therefore, it is preferable that the optimization processing unit 1d sets the learning rate μ to a first value μ1 when the reproduced sound pressure level Pa(j), which corresponds to the sound pressure of the masking sound M(j), increases, and sets the learning rate μ to a second value μ2, which is smaller than the first value μ1, when the reproduced sound pressure level Pa(j) decreases. In other words, the optimization processing unit 1d makes the learning rate μ (first value μ1) when the reproduced sound pressure level Pa(j) increases greater than the learning rate μ (second value μ2) when the reproduced sound pressure level Pa(j) decreases. In this way, by setting the learning rate μ to a first value μ1 when the reproduced sound pressure level Pa(j) increases, the optimization processing unit 1d increases the tracking responsiveness of the masking sound M(j) to the noise N1 when the noise N1 increases, thereby increasing the masking effect of the masking sound M(j). Furthermore, the optimization processing unit 1d sets the learning rate μ to a second value μ2 when the playback sound pressure level Pa(j) decreases, thereby suppressing fluctuations in the playback sound pressure level Pa(j) of the masking sound M(j) when the noise N1 decreases, and making the impression of the masking sound M(j) more natural. As a result, a natural tracking response of the masking sound M(j) can be achieved in the optimization process.

[0122] Figure 10 shows the tracking responsiveness of the masking sound M(j) to noise N1 in the steepest descent method, where the learning rate μ is switched to a first value μ1 or a second value μ2 in response to the increase or decrease in the reproduced sound pressure level Pa(j). In Figure 10, the waveform Gm20 of the masking sound M(j) tracks the fluctuations of the waveform Gn20 of noise N1 with high responsiveness during the rising edge of noise N1, and gradually attenuates during the falling edge of noise N1.

[0123] Figure 11 shows the waveforms Gn21 of noise N1 and Gm21 of masking sound M(j) in the steepest descent method, where the learning rate μ is switched between the first value μ1 and the second value μ2, with the first value μ1 = 0.2 and the second value μ2 = 0.02. In this case, the masking sound M(j) responds quickly to the rising edge of noise N1 and rises in sync with noise N1 (see C21 in Figure 11). Furthermore, the sound pressure fluctuation of the masking sound M(j) is suppressed, and the impression of the masking sound M(j) is more natural.

[0124] In the above explanation, the control parameter was described as the playback sound pressure level Pa(j) of the masking sound M(j), but the control parameter may also be the speaker position, which is the position of speaker 2(j). Even when the control parameter is the speaker position, the masking sound M(j) responds quickly to the rise of noise N1 and rises in accordance with noise N1. Furthermore, the impression of the masking sound M(j) becomes more natural.

[0125] (6) Third variation Figures 12A and 12B show the external appearance of the noise suppression device SF. The noise suppression device SF constitutes the sound masking system 100A. As shown in Figure 13, the noise suppression device SF is placed between the personal booth PB located in the first area R11 and the second area R12 where individual desk work and other tasks are performed.

[0126] As shown in Figure 14, the personal booth PB is equipped with a front wall 91, a left wall 92, and a right wall 93, and a personal space 90 is formed within the personal booth PB surrounded by the front wall 91, the left wall 92, and the right wall 93. The personal space 90 is enclosed on three sides by the front wall 91, the left wall 92, and the right wall 93, and the rear of the personal booth PB is open, forming an opening (entrance / exit) 94 for entering and exiting the personal space 90. People enter the personal space 90 through the opening 94 and exit the personal space 90 through the opening 94.

[0127] Since remote meetings and phone calls take place in personal space 90, the voices of people in personal space 90 may leak out of personal space 90, mainly through the opening 94, and be transmitted to the second area R12 as noise N1.

[0128] Therefore, a noise suppression device SF is placed opposite the opening 94 of the personal booth PB, and the second area R12 is designated as the target area Rt, making the noise N1 less audible to people within the target area Rt.

[0129] The noise suppression device SF comprises a masking sound control system 1A, a speaker 2(1), a microphone 3, a base 81, and a plate member 82. The speaker 2(1) emits a masking sound M(1). The base 81 is to which the speaker 2(1) is mounted. The plate member 82 extends upward from above the base 81. The microphone 3 collects noise N1. The surface of the plate member 82 has a front surface 821 (first surface) and a rear surface 822 (second surface) that face each other. The microphone 3 collects noise N1 generated on the side of the front surface 821 of the plate member 82. The speaker 2(1) emits a masking sound M(1) toward the side of the rear surface 822 of the plate member 82.

[0130] Specifically, the base 81 is a rectangular box-shaped structure made of resin, metal, or wood, and is placed on the floor surface F1. Casters with wheels are attached to the four corners of the underside of the base 81, making it movable on the floor surface F1. A rectangular frame-shaped support 83 is attached to the top surface of the base 81. The support 83 is erected on the top surface of the base 81, and two microphones 3 are attached to the front of the support 83. The microphones 3 collect sound from the front. A plate member 82 is attached to the rear surface of the support 83. The plate member 82 is rectangular in shape, with the front surface 821 of the plate member 82 corresponding to the first surface and the rear surface 822 of the plate member 82 corresponding to the second surface. The number of microphones 3 may be one or three or more.

[0131] A bar-type speaker 2(1) with a long enclosure is rotatably mounted on the upper surface of the base portion 81. The pivot axis of the speaker 2(1) extends along the longitudinal direction of the speaker 2(1). In other words, the speaker 2(1) is rotatably mounted on the base portion 81, and the installation angle of the speaker 2(1) relative to the base portion 81 is variable. Specifically, a rectangular opening is formed on the upper surface of the base portion 81, and the speaker 2(1) is placed in this opening and rotatably supported by a rotation mechanism having a rotation axis or the like. When the rotation angle of the speaker 2(1) changes, the output direction of the masking sound M(1) emitted by the speaker 2(1) changes. In other words, the speaker 2(1) is rotatably mounted on the base portion 81, and the installation angle of the speaker 2(1) relative to the base portion 81 is variable. The rotation angle of speaker 2(1) can be adjusted, for example, within a range of 0 degrees to 90 degrees. As a result, the masking effect of the masking sound M(1) emitted from speaker 2(1) can be finely adjusted.

[0132] The masking sound control system 1A is housed inside the base portion 81.

[0133] Furthermore, it is preferable that the noise suppression device SF is positioned to close the opening 94 of the personal booth PB. Specifically, by closing the opening 94, the plate member 82 functions as at least one of a partition and a sound insulation material. For example, the plate member 82 functions as a partition separating the inside of the personal space 90 of the personal booth PB from the outside of the personal space 90. Alternatively, the plate member 82 functions as a sound insulation material that insulates between the inside of the personal space 90 and the outside of the personal space 90. The plate member 82 may have both the function of a partition and the function of a sound insulation material. In other words, the noise suppression device SF has added value as at least one of a partition and a sound insulation material.

[0134] When the noise suppression device SF is positioned to block the opening 94, the front surface 821 of the plate member 82 faces the personal space 90, and the microphone 3 facing the personal space 90 collects the voices of people present in the personal space 90 as noise N1. In addition, the rear surface 822 of the plate member 82 faces the target area Rt, and the speaker 2(1) facing the target area Rt emits a masking sound M(1) toward the target area Rt.

[0135] In the masking sound control system 1A of such a noise suppression device SF, one listening point Q(1) is set in the target area Rt (second area R12) (see Figure 13). The masking sound control system 1A uses the playback sound pressure level Pa(1) of the masking sound M(1) as a control parameter. The masking sound control system 1A determines the control parameter that minimizes the level of nuisance, which is the degree of nuisance caused to people by the masking sound M(1) and noise N1 in the target area Rt, as an optimization parameter. The masking sound control system 1A then adjusts the control parameter to the optimization parameter. In this modified example, the level of nuisance is a value based on a single-point nuisance value Wp(1), which is the degree of nuisance caused to people by the masking sound M(1) and noise N1 at one listening point Q(1) in the target area Rt.

[0136] The noise suppression device SF described above has a relatively heavy speaker 2(1) placed on a base portion 81 that constitutes the lower part of the noise suppression device SF. As a result, the center of gravity is located at the bottom of the noise suppression device SF, so the noise suppression device SF can be stably installed on the floor surface F1.

[0137] Furthermore, by attaching the speaker 2(1) to the base portion 81, it is easy to enlarge the speaker 2(1). As a result, the sound quality of the masking sound M(1) that reproduces ambient sounds and human voices can be improved.

[0138] The following describes the optimal control of the masking sound M(1) by the masking sound control system 1A.

[0139] As shown in Figure 15, the masking sound control system 1A comprises a noise detection unit 1a, a target sound pressure calculation unit 1b, a masking sound detection unit 1c, an optimization processing unit 1g, a masking sound adjustment unit 1e, and a storage unit 1f. The noise detection unit 1a, target sound pressure calculation unit 1b, masking sound detection unit 1c, masking sound adjustment unit 1e, and storage unit 1f in the masking sound control system 1A differ from the masking sound control system 1 described above in that they target one listening point Q(1) and one speaker 2(1), but basically perform processing almost the same as the masking sound control system 1, targeting one listening point Q(1) and one speaker 2(1). The optimization processing unit 1g in the masking sound control system 1A differs from the optimization processing unit 1d of the masking sound control system 1 described above in that it finds optimization parameters that minimize the point nuisance Wp(1) rather than optimization parameters that minimize the area nuisance Wa.

[0140] The memory unit 1f stores coordinate data (coordinate data) indicating the three-dimensional positions of the two microphones 3, one listening point Q(1), and one speaker 2(1) (speaker position) in the working space R10. In this modified example, the three-dimensional positions of the two microphones 3 and the one speaker 2(1) in the coordinate data correspond to the respective positions of the two microphones 3 and the one speaker 2(1) of the noise suppression device SF, which is installed in predetermined positions to block the opening 94.

[0141] The noise detection unit 1a determines the noise detection value Pn(1) (in dB) based on the noise signal Bn, which is the signal of noise N1 collected by the two microphones 3. The noise detection value Pn(1) represents the sound pressure of noise N1 at one listening point Q(1) within the second area R12 (target area Rt). Specifically, the noise detection unit 1a performs attenuation correction to reduce the magnitude of the noise signal Bn based on the distance between the microphones 3 and the listening point Q(1), and determines the noise detection value Pn(1) at the listening point Q(1). The noise detection value Pn(1) is, for example, the average value of the noise detection values ​​based on the respective sound collection results of the two microphones 3.

[0142] The target sound pressure calculation unit 1b determines the masking target value Pd(1) based on the noise detection value Pn(1) at listening point Q(1) (in dB). The masking target value Pd(1) is the target sound pressure value of the masking sound M(1) at listening point Q(1). In other words, the masking sound control system 1A determines the masking target value Pd(1) at listening point Q(1). The target sound pressure calculation unit 1b determines the masking target value Pd(1) from the noise detection value Pn(1) using the above-described equation 1. That is, the masking target value Pd(1) at listening point Q(1) is uniquely determined according to the sound pressure of the noise N1 at listening point Q(1).

[0143] Specifically, as shown in Figure 4, the target sound pressure calculation unit 1b sets the masking target value Pd(1) to the second lower limit of 40 dB if the noise detection value Pn(1) is less than the first lower limit of 30 dB. The target sound pressure calculation unit 1b sets the masking target value Pd(1) to the second upper limit of 50 dB if the noise detection value Pn(1) is greater than the first upper limit of 40 dB. The target sound pressure calculation unit 1b sets the masking target value Pd(1) to Pn(1) + 10 dB if the noise detection value Pn(1) is greater than or equal to the first lower limit of 30 dB and less than or equal to the first upper limit of 40 dB. In other words, if the noise detection value Pn(1) is 30 dB or more and 40 dB or less, the target sound pressure calculation unit 1b sets the masking target value Pd(1) to a value that is greater than the second lower limit of 40 dB and less than the second upper limit of 50 dB, with the value increasing as the noise detection value Pn(1) increases.

[0144] The masking sound detection unit 1c determines the masking detection value Pm(1), which is the detected sound pressure of the masking sound M(1) at the listening point Q(1). Based on the distance between speaker 2(1) and listening point Q(1), and the reproduced sound pressure level Pa(1) of the masking sound M(1), the masking sound detection unit 1c performs attenuation correction to reduce the reproduced sound pressure level Pa(1), thereby determining the sound pressure of the masking sound M(1) that has reached listening point Q(1). Then, based on the sound pressure of the masking sound M(1) that has reached listening point Q(1), the masking sound detection unit 1c determines the sound pressure of the masking sound M(1) at listening point Q(1) as the masking detection value Pm(1).

[0145] The optimization processing unit 1g calculates the point nuisance level Wp(1) based on the difference between the masking detection value Pm(1) and the masking target value Pd(1). The point nuisance level Wp(1) is expressed by equations 3 and 4 described above.

[0146] The optimization processing unit 1g determines the optimal sound pressure Pb(1) for the playback sound pressure level Pa(1) of the masking sound M(1) that minimizes the point nuisance Wp(1) expressed in Equation 4 above. Here, the optimized sound pressure Pb(1) is the optimization parameter. Equation 10 below is an example of the optimization process by which the optimization processing unit 1g determines the optimization parameter, and shows the update process of the playback sound pressure level Pa(1). Pa(1)new is the control parameter after the update, and Pa(1)old is the control parameter before the update. μ is the learning rate (step size parameter). The coordinates x1, y1, z1 indicate the position of speaker 2(1).

[0147]

number

[0148] The masking sound adjustment unit 1e adjusts the playback sound pressure level Pa(1) of the masking sound M(1) emitted by the speaker 2(1) to the optimized sound pressure Pb(1).

[0149] The masking sound control system 1A, having the above configuration, controls the playback sound pressure level Pa(1) of the masking sound M(1) emitted from speaker 2(1) to an optimized sound pressure Pb(1) in order to minimize the point nuisance Wp(1). As a result, the masking sound control system 1A can suppress both the nuisance caused by noise N1 and the nuisance caused by the masking sound M(1) in a more balanced manner.

[0150] Furthermore, the masking sound control system 1A determines a masking target value Pd(1) according to the noise N1 collected by the microphone 3, and can optimally control the playback sound pressure level Pa(1) of the masking sound M(1) in real time in response to the generation conditions of the noise N1.

[0151] Furthermore, the optimization processing unit 1g uses the point nuisance level Wp(1) as its objective function and finds the optimal solution to an optimization problem that minimizes the point nuisance level Wp(1), thereby determining the optimization parameters. As a result, the masking sound control system 1A can accurately determine the optimization parameters that minimize the point nuisance level Wp(1).

[0152] Furthermore, the optimization processing unit 1g determines the optimization parameters by finding the optimal solution to the optimization problem using the steepest descent method. As a result, the masking sound control system 1A can accurately determine the optimization parameters that minimize the point nuisance level Wp(1).

[0153] In the above explanation, the control parameter was described as the playback sound pressure level Pa(1) of the masking sound M(1), but the control parameter may also be the speaker position, which is the position of speaker 2(1). The same effect can be obtained even if the control parameter is the speaker position.

[0154] Furthermore, as shown in Figure 12A, it is preferable that the noise suppression device SF further comprises a portable battery 84. If the noise suppression device SF is connected to a commercial power source, the noise suppression device SF is powered by the commercial power source. The portable battery 84 is equipped with a secondary battery, and if the noise suppression device SF is connected to a commercial power source, the secondary battery is charged by the commercial power source. If the noise suppression device SF is not connected to a commercial power source, the portable battery 84 powers the noise suppression device SF with the discharge power of the secondary battery.

[0155] Furthermore, it is preferable that the noise suppression device SF is further equipped with an uninterruptible power supply. The uninterruptible power supply is used to send a shutdown command to the computer system of the masking sound control system 1A and to safely shut down the system when the power supply from the commercial power source and the portable battery 84 is interrupted.

[0156] Furthermore, during the installation and maintenance of the noise suppression device SF, user interface equipment such as a keyboard and monitor are connected to the computer system of the masking sound control system 1A to perform software installation and other related tasks.

[0157] Furthermore, the noise suppression device SF preferably includes an optimal control mode, a fixed mode, and an off mode as operating modes. The optimal control mode is an operating mode that optimally controls the playback sound pressure level Pa(1) of the masking sound M(1) as described above. The fixed mode is an operating mode that fixes the playback sound pressure level Pa(1) of the masking sound M(1) to a constant value. The off mode is an operating mode that sets the playback sound pressure level Pa(1) of the masking sound M(1) to zero and stops the output of the masking sound M(1).

[0158] Furthermore, the noise suppression device SF may be configured to output at least one additional sound effect from speaker 2(1) in addition to the masking sound M(1) consisting of ambient sounds. In this case, the noise suppression device SF may mix the sound effect with the masking sound M(1) and output both the masking sound M(1) and the sound effect from speaker 2(1), or it may output only one of the masking sound M(1) or the sound effect from speaker 2(1). For example, if the masking sound M(1) is the sound of waves, the sound effect could be the cry of a seagull, the rustling of trees, etc.

[0159] Furthermore, if a combination of ambient sounds and performance sounds is considered a sound source set, it is preferable that the noise suppression device SF pre-stores data for multiple sound source sets in a storage medium such as the storage unit 1f. In this case, the noise suppression device SF reads one of the multiple sound source sets from memory in response to user operations, and uses the ambient sounds and performance sounds of the read sound source set.

[0160] Furthermore, speaker 2(1) may be mounted on the base 81 so as to be slidable in the vertical direction. In this case, the vertical position of speaker 2(1) relative to the base 81 is variable. When the vertical position of speaker 2(1) changes, the height at which the masking sound M(1) emitted from speaker 2(1) is output changes. As a result, the masking effect of the masking sound M(1) emitted from speaker 2(1) can be finely adjusted. The sliding mechanism of speaker 2(1) may be a structure in which speaker 2(1) is slid manually by a person, or a structure in which speaker 2(1) is slid by a driving force such as a motor or cylinder.

[0161] Furthermore, it is preferable that the microphone 3 be configured to slide freely in the vertical direction. For example, the vertical position of the microphone 3 is variable on the front surface of the plate member 82 or support 83. The closer the microphone 3 is to the mouth of a person (noise source) in the personal space 90, the more accurately it can collect noise N1. Therefore, by making the vertical position of the microphone 3 variable, the microphone 3 can be moved closer to the noise source. As a result, the microphone 3 can collect noise N1 with high accuracy, and the masking effect of the masking sound M(1) is improved. The sliding mechanism of the microphone 3 may be a structure in which the microphone 3 is slid manually by a person, or a structure in which the microphone 3 is slid by a driving force such as a motor or cylinder.

[0162] Furthermore, the optimization processing unit 1d of the noise suppression device SF may set the learning rate μ to a first value μ1 when the reproduced sound pressure level Pa(1) of the masking sound M(1) increases, and to a second value μ2 which is smaller than the first value μ1 when the reproduced sound pressure level Pa(1) decreases.

[0163] Furthermore, the noise suppression device SF may be equipped with multiple speakers 2(j). In this case, the noise suppression device SF determines the control parameters of each of the multiple speakers 2(j) that minimize the point nuisance Wp(1) as optimization parameters.

[0164] Furthermore, the noise suppression device SF may be equipped with multiple speakers 2(j) and set multiple listening points Q(k). In this case, the noise suppression device SF, similar to the masking sound control system 1, determines the control parameters of each of the multiple speakers 2(j) that minimize the area nuisance Wa as optimization parameters.

[0165] Furthermore, the configurations provided by the sound masking system 100 and the masking sound control system 1 may be appropriately applied to the noise suppression device SF (sound masking system 100A) and the masking sound control system 1A.

[0166] As shown in Figures 13 and 14, a person inside the personal space 90 of the personal booth PB may conduct remote meetings, make phone calls, etc., facing the front wall 91 of the personal booth PB. In this case, the noise suppression device SF is located behind the person inside the personal space 90. Alternatively, as shown in Figure 15, a person inside the personal space 90 may conduct remote meetings, make phone calls, etc., facing the opening 94 of the personal booth PB. In this case, the noise suppression device SF is located in front of the person inside the personal space 90.

[0167] (7) Summary A masking sound control system (1, 1A) according to the first embodiment controls a masking sound (M(j)) emitted from a speaker (2(j)) to mask noise (N1) within a target area (Rt). The masking sound control system (1) comprises an optimization processing unit (1d, 1g) and a masking sound adjustment unit (1e). The optimization processing units (1d, 1g) control at least one of the playback sound pressure level (Pa(j)) of the masking sound (M(j)) emitted from the speaker (2(j)) and the speaker position (xj, yj, zj), which is the position of the speaker (2(j)), as a control parameter (x1...xjm, y1...yjm, z1...zjm, Pa(1)...Pa(jm)). The optimization processing unit (1d) determines the control parameters (Pb(1)...Pb(jm), etc.) that minimize the level of nuisance (Wa, Wp(k)), which is the degree of nuisance caused to people by the masking sound (M(j)) and noise (N1) in the target area (Rt), as optimization parameters. The masking sound adjustment unit (1e) adjusts the control parameters to the optimization parameters.

[0168] The masking sound control system (1, 1A) described above can suppress both the nuisance caused by noise (N1) and the nuisance caused by the masking sound (M(j)) in a more balanced manner.

[0169] In the second embodiment of the masking sound control system (1, 1A), in the first embodiment, it is preferable that the optimization processing unit (1d, 1g) determines the control parameters that minimize the level of annoyance (Wa, Wp(k)) as optimization parameters based on the relationship equation (Equation 7) between the control parameters and the level of annoyance (Wa, Wp(k)).

[0170] The masking sound control system described above (1, 1A) can accurately determine the optimization parameters that minimize the level of nuisance (Wa, Wp(k)).

[0171] In the third embodiment of the masking sound control system (1, 1A) according to the embodiment, in the first or second embodiment, the optimization processing unit (1d, 1g) preferably determines the optimization parameters by finding the optimal solution to an optimization problem that minimizes the nuisance level (Wa, Wp(k)), with the nuisance level (Wa, Wp(k)) as the objective function.

[0172] The masking sound control system described above (1, 1A) can accurately determine the optimization parameters that minimize the level of nuisance (Wa, Wp(k)).

[0173] In the fourth embodiment of the masking sound control system (1, 1A) according to the embodiment, in the third embodiment, it is preferable that the optimization processing unit (1d, 1g) obtains the optimization parameters by finding the optimal solution to the optimization problem using the steepest descent method.

[0174] The masking sound control system described above (1, 1A) can accurately determine the optimization parameters that minimize the level of nuisance (Wa, Wp(k)).

[0175] In the fifth embodiment of the masking sound control system (1, 1A) according to the embodiment, in the fourth embodiment, it is preferable that the optimization processing unit (1d, 1g) sets the learning rate (μ) that determines the convergence when finding the optimal solution in the steepest descent method to a first value (μ1) when the sound pressure (Pa(j)) of the masking sound (M(j)) increases, and sets the learning rate (μ) to a second value (μ2) which is smaller than the first value (μ1) when the sound pressure (Pa(j)) of the masking sound (M(j)) decreases.

[0176] The masking sound control system (1, 1A) described above can improve the tracking responsiveness of the masking sound (M(j)) to the noise (N1), and can make the impression of the masking sound (M(j)) more natural.

[0177] In the sixth embodiment of the masking sound control system (1), in any one of the first to fifth embodiments, the level of nuisance is preferably an area nuisance level (Wa), which is the degree of nuisance caused to people by the masking sound (M(j)) and noise (N1) over the target area (Rt). The optimization processing unit (1d) determines the control parameters that minimize the area nuisance level (Wa) as optimization parameters.

[0178] The masking sound control system (1) described above can suppress the degree of nuisance caused to people by the masking sound (M(j)) and noise (N1) over a wide range.

[0179] In the masking sound control system (1) of the seventh embodiment, in the sixth embodiment, the area nuisance level (Wa) is preferably a value based on a plurality of point nuisance levels (Wp(k)) which are the degree of nuisance caused to people by the masking sound (M(j)) and noise (N1) at each of a plurality of listening points (Q(k)) within the target area (Rt).

[0180] The masking sound control system (1) described above can suppress the degree of nuisance caused to people by masking sound (M(j)) and noise (N1) over a wide area based on the spatial area nuisance level (Wa).

[0181] In the masking sound control system (1) of the eighth embodiment, in the seventh embodiment, the area nuisance level (Wa) is preferably the integral of the degree of nuisance in the target area (Rt).

[0182] The masking sound control system (1) described above can suppress the degree of nuisance caused to people by masking sound (M(j)) and noise (N1) over a wide area based on the spatial area nuisance level (Wa).

[0183] In the masking sound control system (1) of the ninth embodiment, in the eighth embodiment, the area nuisance level (Wa) is preferably a value obtained by dividing the degree of nuisance in the target area (Rt) by volume.

[0184] The masking sound control system (1) described above can suppress the degree of nuisance caused to people by masking sound (M(j)) and noise (N1) over a wide area based on the area nuisance level (Wa) in a three-dimensional space.

[0185] In the masking sound control system (1) of the tenth embodiment, in any one of the seventh to ninth embodiments, the area nuisance level (Wa) is preferably the sum of a plurality of point nuisance levels (Wp(k)).

[0186] The masking sound control system (1) described above can suppress the degree of nuisance caused to people by masking sound (M(j)) and noise (N1) over a wide area, based on the spatial area nuisance level (Wa) obtained by simple calculations.

[0187] In the masking sound control system (1) of the 11th embodiment, it is preferable that in any one of the 7th to 10th embodiments, each of the multiple point nuisance levels (Wp(k)) is a value based on the difference between a masking detection value (Pm(k)), which is the detected sound pressure value of the masking sound (M(j)) at each of the multiple listening points (Q(k)), and a masking target value (Pd(k)), which is the target sound pressure value of the masking sound (M(j)) at each of the multiple listening points (Q(k)).

[0188] The masking sound control system (1) described above can suppress the level of nuisance caused by masking sound (M(j)) and noise (N1) to people over a wide area, based on an area nuisance level (Wa) determined from a highly accurate point nuisance level (Wp(k)).

[0189] In the masking sound control system (1) of the 12th embodiment, in the 11th embodiment, the masking target value (Pd(k)) is preferably set to the second lower limit if the noise detection value (Pn(k)) indicating the sound pressure of the noise (N1) at each of the plurality of listening points (Q(k)) is smaller than the first lower limit. The masking target value (Pd(k)) is preferably set to the second upper limit if the noise detection value (Pn(k)) is larger than the first upper limit. The masking target value (Pd(k)) is preferably set to a larger value as the noise detection value (Pn(k)) increases, within the range of being greater than the second lower limit and smaller than the second upper limit, provided that the noise detection value (Pn(k)) is greater than or equal to the first lower limit and less than or equal to the first upper limit.

[0190] The masking sound control system (1) described above can accurately reflect the degree of annoyance a person feels from a masking sound (M(j)).

[0191] In the masking sound control system (1A) of the 13th embodiment, in any one of the first to sixth embodiments, the level of nuisance is preferably a value based on a single-point nuisance level (Wp(k)), which is the degree of nuisance caused to a person by the masking sound (M(j)) and noise (N1) at a single listening point (Q(k)) within the target area (Rt).

[0192] The masking sound control system (1A) described above can suppress the degree of nuisance caused to people by the masking sound (M(j)) and noise (N1) based on the point nuisance level (Wp(k)).

[0193] In the masking sound control system (1A) of the 14th embodiment, in the 13th embodiment, the point nuisance level (Wp(k)) is preferably a value based on the difference between a masking detection value (Pm(k)), which is the detected sound pressure of the masking sound (M(j)) at one listening point (Q(k)), and a masking target value (Pd(k)), which is the target sound pressure of the masking sound (M(j)) at one listening point (Q(k)).

[0194] The masking sound control system (1A) described above can determine a highly accurate point nuisance level (Wp(k)).

[0195] In the masking sound control system (1A) of the 15th embodiment, in the 14th embodiment, the masking target value (Pd(k)) is preferably set to the second lower limit if the noise detection value (Pn(k)) indicating the sound pressure of the noise (N1) at one listening point (Q(k)) is less than the first lower limit. The masking target value (Pd(k)) is preferably set to the second upper limit if the noise detection value (Pn(k)) is greater than the first upper limit. The masking target value (Pd(k)) is preferably set to a value greater than the second lower limit and less than the second upper limit, provided that the noise detection value (Pn(k)) is greater than or equal to the first lower limit and less than or equal to the first upper limit, and the larger the noise detection value (Pn(k)), the larger the value.

[0196] The masking sound control system (1A) described above can accurately reflect the degree of annoyance a person feels from a masking sound (M(j)).

[0197] In the masking sound control system (1, 1A) of the 16th embodiment, the number of speakers (2(j)) is multiple in any one of the first to 15 embodiments. The optimization processing unit (1d, 1g) preferably determines the optimization parameters that minimize the level of nuisance (Wa, Wp(k)) for each of the multiple speakers (2(j)). The masking sound adjustment unit (1e) preferably adjusts the control parameters of the multiple speakers (2(j)) to the multiple optimization parameters.

[0198] The masking sound control system (1, 1A) described above can suppress both the nuisance caused by noise (N1) and the nuisance caused by masking sound (M(j)) in a more balanced manner, even when using multiple speakers (2(j)).

[0199] In the masking sound control system (1, 1A) of the 17th embodiment, it is preferable that the masking sound (M(j)) is an ambient sound in any one of the first to 16th embodiments.

[0200] The masking sound control system described above (1, 1A) makes the masking sound (M(j)) less noticeable to people, and can also help people relax or improve their concentration.

[0201] A sound masking system (100, 100A) according to the 18th embodiment comprises a masking sound control system (1, 1A) according to any one of the first to 17 embodiments, and a speaker (2(j)).

[0202] The sound masking systems described above (100, 100A) can suppress both the nuisance caused by noise (N1) and the nuisance caused by the masking sound (M(j)) in a more balanced manner.

[0203] A sound masking system (100A) according to the 19th embodiment preferably further comprises a base (81), a plate member (82), and a microphone (3) in the 18th embodiment. A speaker (2(j)) can be attached to the base (81). The plate member (82) extends upward from above the base (81). The microphone (3) collects noise (N1). The surface of the plate member (82) has a first surface (821) and a second surface (822) that face each other. The microphone (3) collects noise (N1) generated on the side of the first surface (821) of the plate member (82). The speaker (2(j)) emits a masking sound (M(j)) toward the side of the second surface (822) of the plate member (82).

[0204] The aforementioned sound masking system (100A) has its center of gravity located at the bottom, allowing it to be stably installed on the floor surface (F1).

[0205] In the 20th embodiment of the sound masking system (100A), in the 19th embodiment, it is preferable that the speaker (2(j)) is rotatably mounted on the base portion (81).

[0206] The aforementioned sound masking system (100A) allows for fine-tuning of the masking effect using the masking sound (M(j)).

[0207] In the 21st embodiment of the sound masking system (100A), in the 19th embodiment, it is preferable that the speaker (2(j)) is mounted on the base portion (81) so as to be slidable in the vertical direction.

[0208] The aforementioned sound masking system (100A) allows for fine-tuning of the masking effect using the masking sound (M(j)).

[0209] In the 22nd embodiment of the sound masking system (100A), in any one of the 19th to 21st embodiments, it is preferable that the microphone (3) is configured to slide freely in the vertical direction.

[0210] The aforementioned sound masking system (100A) can accurately collect noise (N1) and improve the masking effect of the masking sound (M(j)).

[0211] In the 23rd embodiment of the sound masking system (100A) according to the embodiment, in any one of the 19th to 22nd embodiments, the plate member (82) preferably functions as at least one of a partition and a sound insulation material.

[0212] The sound masking system (100A) described above has added value as at least one of a partition and a sound insulation material.

[0213] A masking sound control method according to a 24th embodiment controls a masking sound (M(j)) emitted from a speaker (2(j)) to mask noise (N1) within a target area (Rt). The masking sound control method includes an optimization processing step (S1) and a masking sound adjustment step (S2). The optimization processing step (S1) sets at least one of the playback sound pressure level (Pa(j)) of the masking sound (M(j)) emitted from the speaker (2(j)) and the speaker position (xj, yj, zj), which is the position of the speaker (2(j)), as a control parameter (x1...xjm, y1...yjm, z1...zjm, Pa(1)...Pa(jm)). The optimization step (S1) determines the control parameters (Pb(1)...Pb(jm), etc.) that minimize the level of nuisance (Wa, Wp(1)), which is the degree of nuisance caused to people by the masking sound (M(j)) and noise (N1) in the target area (Rt), as optimization parameters. The masking sound adjustment step (S2) adjusts the control parameters to the optimization parameters.

[0214] The masking sound control method described above can suppress both the nuisance caused by noise (N1) and the nuisance caused by the masking sound (M(j)) in a more balanced manner.

[0215] The program of the 25th embodiment causes a computer system to execute the masking sound control method of the 24th embodiment.

[0216] The program described above can suppress both the nuisance caused by noise (N1) and the nuisance caused by masking sound (M(j)) in a more balanced manner. [Explanation of symbols]

[0217] 100, 100A Sound Masking System SF Noise Suppression Device (Sound Masking System) 1. 1A Masking Sound Control System 1d, 1g Optimization Processing Unit 1e Masking sound adjustment section 2(j) speaker 3. Microphone (sound collection device) Rt Target Area N1 Noise M(j) Masking sound Q(k) Listening points Bn Noise Signal Wp(k) Point Annoyance Level Wa Area Nuisance Level Pa(j) Playback sound pressure level (sound pressure of masking sound) Pb(j) Optimized Sound Pressure Pn(k) Noise detection value Pd(k) Masking Target Value Pm(k) Masking detection value S1 Optimization process step S2 Masking Sound Adjustment Step x1…xjm, y1…yjm, z1…zjm Speaker position (control parameter) Pa(1)...Pa(jm) Playback sound pressure level (control parameter) Pb(1)...Pb(jm) Optimization parameters μ learning rate μ1 1st value μ2 2nd value 81 Base 82 Plate members 821 Front (first side) 822 Back side (2nd side)

Claims

1. A masking sound control system that controls the masking sound emitted from a speaker to mask noise within a target area, An optimization processing unit that uses the speaker position, which is the position of the speaker, as a control parameter, and determines the control parameter that minimizes the level of nuisance, which is the degree of nuisance caused to people by the masking sound and the noise in the target area, as an optimization parameter, The system includes a masking sound adjustment unit that adjusts the control parameters to the optimization parameters. Masking sound control system.

2. The optimization processing unit determines the control parameter that minimizes the level of annoyance as the optimization parameter, based on the relationship between the control parameter and the level of annoyance. A masking sound control system according to claim 1.

3. The optimization processing unit determines the optimization parameters by using the level of annoyance as the objective function and finding the optimal solution to an optimization problem that minimizes the level of annoyance. A masking sound control system according to claim 1.

4. The optimization processing unit determines the optimization parameters by finding the optimal solution to the optimization problem using the steepest descent method. The masking sound control system according to claim 3.

5. A masking sound control system for controlling a masking sound emitted from a speaker in order to mask noise within a target area, An optimization processing unit determines, as optimization parameters, the control parameters that minimize the level of nuisance caused to people by the masking sound and the noise in the target area, using at least one of the sound pressure level of the masking sound emitted from the speaker and the speaker position, respectively. The system includes a masking sound adjustment unit that adjusts the control parameters to the optimization parameters, The optimization processing unit determines the optimization parameters by finding the optimal solution to an optimization problem that minimizes the level of annoyance, using the level of annoyance as the objective function. The optimization processing unit determines the optimization parameters by finding the optimal solution to the optimization problem using the steepest descent method. The optimization processing unit described above is When the sound pressure of the masking sound increases, the learning rate that determines the convergence when finding the optimal solution in the steepest descent method is set to the first value. When the sound pressure of the masking sound decreases, the learning rate is set to a second value which is smaller than the first value. Masking sound control system.

6. The aforementioned level of nuisance is an area nuisance level, which is the degree of nuisance caused to people by the masking sound and the noise over the target area. The optimization processing unit determines the control parameters that minimize the level of nuisance in the area as optimization parameters. A masking sound control system according to claim 1.

7. The aforementioned area nuisance level is a value based on multiple point nuisance levels, which represent the degree of nuisance caused to people by the masking sound and the noise at each of the multiple listening points within the target area. The masking sound control system according to claim 6.

8. A masking sound control system for controlling a masking sound emitted from a speaker in order to mask noise within a target area, An optimization processing unit determines, as optimization parameters, the control parameters that minimize the level of nuisance caused to people by the masking sound and the noise in the target area, using at least one of the sound pressure level of the masking sound emitted from the speaker and the speaker position, respectively. The system includes a masking sound adjustment unit that adjusts the control parameters to the optimization parameters, The aforementioned level of nuisance is an area nuisance level, which is the degree of nuisance caused to people by the masking sound and the noise over the target area. The optimization processing unit determines the control parameters that minimize the level of nuisance in the area as optimization parameters, The aforementioned area nuisance level is a value based on multiple point nuisance levels, which represent the degree of nuisance caused to people by the masking sound and the noise at each of the multiple listening points within the target area. The aforementioned area nuisance level is the value obtained by integrating the degree of nuisance in the target area. Masking sound control system.

9. The aforementioned area nuisance level is the value obtained by dividing the degree of nuisance in the target area by volume. The masking sound control system according to claim 8.

10. The aforementioned area nuisance level is the sum of the nuisance levels of the multiple points. The masking sound control system according to claim 7.

11. Each of the aforementioned multiple point nuisance levels is a value based on the difference between a masking detection value, which is the detected sound pressure of the masking sound at each of the multiple listening points, and a masking target value, which is the target sound pressure of the masking sound at each of the multiple listening points. The masking sound control system according to claim 7.

12. A masking sound control system for controlling a masking sound emitted from a speaker in order to mask noise within a target area, An optimization processing unit determines, as optimization parameters, the control parameters that minimize the level of nuisance caused to people by the masking sound and the noise in the target area, using at least one of the sound pressure level of the masking sound emitted from the speaker and the speaker position, respectively. The system includes a masking sound adjustment unit that adjusts the control parameters to the optimization parameters, The aforementioned level of nuisance is an area nuisance level, which is the degree of nuisance caused to people by the masking sound and the noise over the target area. The optimization processing unit determines the control parameters that minimize the level of nuisance in the area as optimization parameters, The aforementioned area nuisance level is a value based on multiple point nuisance levels, which represent the degree of nuisance caused to people by the masking sound and the noise at each of the multiple listening points within the target area. Each of the aforementioned multiple point nuisance levels is a value based on the difference between a masking detection value, which is the detected sound pressure value of the masking sound at each of the multiple listening points, and a masking target value, which is the target sound pressure value of the masking sound at each of the multiple listening points. The aforementioned masking target value is, If the noise detection value indicating the sound pressure of the noise at each of the plurality of listening points is less than the first lower limit, it is set to the second lower limit. If the noise detection value is greater than the first upper limit, it is set to the second upper limit. If the noise detection value is greater than or equal to the first lower limit and less than or equal to the first upper limit, the value will be set to a larger value as the noise detection value increases, within a range greater than the second lower limit and less than the second upper limit. Masking sound control system.

13. The aforementioned level of nuisance is a value based on a single-point nuisance level, which is the degree of nuisance caused to a person by the masking sound and the noise at one listening point within the target area. A masking sound control system according to claim 1.

14. The aforementioned point nuisance level is a value based on the difference between the masking detection value, which is the detected sound pressure of the masking sound at one listening point, and the masking target value, which is the target sound pressure of the masking sound at one listening point. A masking sound control system according to claim 13.

15. A masking sound control system for controlling a masking sound emitted from a speaker in order to mask noise within a target area, An optimization processing unit determines, as optimization parameters, the control parameters that minimize the level of nuisance caused to people by the masking sound and the noise in the target area, using at least one of the sound pressure level of the masking sound emitted from the speaker and the speaker position, respectively. The system includes a masking sound adjustment unit that adjusts the control parameters to the optimization parameters, The aforementioned level of nuisance is a value based on a single-point nuisance level, which is the degree of nuisance caused to a person by the masking sound and the noise at one listening point within the target area. The aforementioned point nuisance level is a value based on the difference between the masking detection value, which is the detected sound pressure value of the masking sound at one listening point, and the masking target value, which is the target sound pressure value of the masking sound at one listening point. The aforementioned masking target value is, If the noise detection value indicating the sound pressure of the noise at one of the listening points is less than the first lower limit, it is set to the second lower limit. If the noise detection value is greater than the first upper limit, it is set to the second upper limit. If the noise detection value is greater than or equal to the first lower limit and less than or equal to the first upper limit, the value will be set to a larger value as the noise detection value increases, within a range greater than the second lower limit and less than the second upper limit. Masking sound control system.

16. The number of the aforementioned speakers is multiple. The optimization processing unit determines the optimization parameters that minimize the level of nuisance for each of the multiple speakers, thereby determining a plurality of optimization parameters corresponding to the multiple speakers. The masking sound adjustment unit adjusts the control parameters of the plurality of speakers to the plurality of optimization parameters. A masking sound control system according to any one of claims 1 to 15.

17. The aforementioned masking sound is an ambient sound. A masking sound control system according to any one of claims 1 to 15.

18. A masking sound control system comprising any one of claims 1 to 15, The speaker comprises Sound masking system.

19. The base portion on which the speaker is attached, A plate member extending upward from above the base portion, A microphone for collecting the aforementioned noise, and further comprising, The surface of the plate member has a first surface and a second surface that face each other. The microphone collects the noise generated on the first surface side of the plate member. The speaker emits the masking sound toward the second surface side of the plate member. The sound masking system according to claim 18.

20. The speaker is rotatably mounted on the base. The sound masking system according to claim 19.

21. The speaker is mounted on the base so as to be able to slide vertically. The sound masking system according to claim 19.

22. The microphone is configured to slide freely in the vertical direction. The sound masking system according to claim 19.

23. The aforementioned plate member functions as at least one of a partition and a sound insulation material. The sound masking system according to claim 19.

24. A masking sound control method for controlling a masking sound emitted from a speaker to mask noise within a target area, An optimization process step in which the speaker position, which is the position of the speaker, is used as a control parameter, and the control parameter that minimizes the degree of nuisance, which is the degree of nuisance caused to people by the masking sound and the noise in the target area, is determined as an optimization parameter; Includes a masking sound adjustment step of adjusting the control parameter to the optimization parameter. A method for controlling masking sounds.

25. A masking sound control method for controlling a masking sound emitted from a speaker in order to mask noise within a target area, An optimization process step in which at least one of the playback sound pressure level of the masking sound emitted from the speaker and the speaker position, which is the location of the speaker, is used as a control parameter to determine the control parameter that minimizes the degree of nuisance, which is the degree of nuisance caused to people by the masking sound and the noise in the target area, as an optimization parameter; The step includes adjusting the control parameters to the optimization parameters, The optimization processing step involves determining the optimization parameters by finding the optimal solution to an optimization problem that minimizes the level of annoyance, with the level of annoyance as the objective function. The optimization process step involves determining the optimization parameters by finding the optimal solution to the optimization problem using the steepest descent method. The optimization process step is: When the sound pressure of the masking sound increases, the learning rate that determines the convergence when finding the optimal solution in the steepest descent method is set to the first value. When the sound pressure of the masking sound decreases, the learning rate is set to a second value which is smaller than the first value. A method for controlling masking sounds.

26. A masking sound control method for controlling a masking sound emitted from a speaker in order to mask noise within a target area, An optimization process step in which at least one of the playback sound pressure level of the masking sound emitted from the speaker and the speaker position, which is the location of the speaker, is used as a control parameter to determine the control parameter that minimizes the degree of nuisance, which is the degree of nuisance caused to people by the masking sound and the noise in the target area, as an optimization parameter; The step includes adjusting the control parameters to the optimization parameters, The aforementioned level of nuisance is an area nuisance level, which is the degree of nuisance caused to people by the masking sound and the noise over the target area. The optimization processing step determines the control parameters that minimize the level of nuisance in the area as optimization parameters, The aforementioned area nuisance level is a value based on multiple point nuisance levels, which represent the degree of nuisance caused to people by the masking sound and the noise at each of the multiple listening points within the target area. The aforementioned area nuisance level is the value obtained by integrating the degree of nuisance in the target area. A method for controlling masking sounds.

27. ​​A masking sound control method for controlling a masking sound emitted from a speaker in order to mask noise within a target area, An optimization process step in which at least one of the playback sound pressure level of the masking sound emitted from the speaker and the speaker position, which is the location of the speaker, is used as a control parameter to determine the control parameter that minimizes the degree of nuisance, which is the degree of nuisance caused to people by the masking sound and the noise in the target area, as an optimization parameter; The step includes adjusting the control parameters to the optimization parameters, The aforementioned level of nuisance is an area nuisance level, which is the degree of nuisance caused to people by the masking sound and the noise over the target area. The optimization processing step determines the control parameters that minimize the level of nuisance in the area as optimization parameters, The aforementioned area nuisance level is a value based on multiple point nuisance levels, which represent the degree of nuisance caused to people by the masking sound and the noise at each of the multiple listening points within the target area. Each of the aforementioned multiple point nuisance levels is a value based on the difference between a masking detection value, which is the detected sound pressure value of the masking sound at each of the multiple listening points, and a masking target value, which is the target sound pressure value of the masking sound at each of the multiple listening points. The aforementioned masking target value is, If the noise detection value indicating the sound pressure of the noise at each of the plurality of listening points is less than the first lower limit, it is set to the second lower limit. If the noise detection value is greater than the first upper limit, it is set to the second upper limit. If the noise detection value is greater than or equal to the first lower limit and less than or equal to the first upper limit, the value will be set to a larger value as the noise detection value increases, within a range greater than the second lower limit and less than the second upper limit. A method for controlling masking sounds.

28. A masking sound control method for controlling a masking sound emitted from a speaker in order to mask noise within a target area, An optimization process step in which at least one of the playback sound pressure level of the masking sound emitted from the speaker and the speaker position, which is the location of the speaker, is used as a control parameter to determine the control parameter that minimizes the degree of nuisance, which is the degree of nuisance caused to people by the masking sound and the noise in the target area, as an optimization parameter; The step includes adjusting the control parameters to the optimization parameters, The aforementioned level of nuisance is a value based on a single-point nuisance level, which is the degree of nuisance caused to a person by the masking sound and the noise at one listening point within the target area. The aforementioned point nuisance level is a value based on the difference between the masking detection value, which is the detected sound pressure value of the masking sound at one listening point, and the masking target value, which is the target sound pressure value of the masking sound at one listening point. The aforementioned masking target value is, If the noise detection value indicating the sound pressure of the noise at one of the listening points is less than the first lower limit, it is set to the second lower limit. If the noise detection value is greater than the first upper limit, it is set to the second upper limit. If the noise detection value is greater than or equal to the first lower limit and less than or equal to the first upper limit, the value will be set to a larger value as the noise detection value increases, within a range greater than the second lower limit and less than the second upper limit. A method for controlling masking sounds.

29. Causes a computer system to execute the masking sound control method described in any one of claims 24 to 28. program.

Images