JP6466251B2 - Sound field reproduction system - Google Patents

Description

  The present invention relates to a sound field reproduction system for controlling a sound field at one or a plurality of listening positions included in a vehicle interior space.

  In recent years, “super-realistic systems” have attracted attention. The ultra-realistic system is a system that obtains, transmits, and reproduces the "sense of the senses" as physically as possible in order to create the feeling of being at the location. The ultra-realistic system can be applied not only to entertainment such as movies and broadcasts, but also to various fields such as medical care, education, and art.

  In order to realize a highly realistic sound system, various sound systems have been studied, including sound field reproduction technology. Sound field reproduction technology is a physical reproduction of the desired sound field in consideration of the wave nature of the sound, and a higher sense of presence is obtained with respect to existing acoustic systems, and its development is expected. . Major sound field reproduction techniques include wavefront synthesis, boundary sound field control, and higher-order ambisonics.

  Wavefront synthesis (WFS) is a technique for presenting sound coming from a specific direction by reproducing sound pressure on a plane based on Huygens' principle. The wavefront synthesis method can be easily used as compared with other sound field reproduction technologies, and can reproduce all sound fields in the region surrounded by the speaker array. However, in the wavefront synthesis method, since the boundary surface is divided and controlled, the direction of sound that can be expressed is limited. In addition, a microphone array is required to collect the sound field used in the wavefront synthesis method, and there is a problem that the installation of the array is not easy.

  Boundary sound field control (BoSC) is a technique that applies Kirchhoff-Helmholtz integral equations and inverse system theory. Reproduce the sound field by measuring the sound pressure and sound pressure gradient on the boundary surface of the area of the sound field that you want to reproduce, and using the inverse system to reproduce the sound pressure and sound pressure gradient in another space Can do. The boundary sound field control cannot reproduce a wide range of sound fields like the wavefront synthesis method, but can reproduce sound fields with higher accuracy. However, since the inverse system theory is applied, there is a one-to-one relationship between the sound collection system and the reproduction system for boundary sound field control. Therefore, the sound field information that can be used in the boundary sound field control sound collection and reproduction system is difficult to use in another boundary sound field control sound collection and reproduction system. For the same reason, it cannot be used for an existing stereo sound source, 5.1ch surround sound source, or the like.

  Higher order ambisonics (HOA) is a technique for representing a sound field at a boundary surface using a spherical harmonic function and reproducing the sound field based on the Kirchhoff-Helmholtz integral equation (see, for example, Patent Document 1). In higher-order ambisonics, the collected sound field information is expressed by spherical harmonic coefficients, so the collected sound data can be applied to any reproduction system. Therefore, the collected sound data can be applied to an existing reproduction system such as stereo reproduction or 5.1ch surround. It is also possible to apply existing sound source data to an ambisonics playback system. Higher-order ambisonics is expected to develop as a practical high-sense system because of its wide range of applicable areas.

JP, 2014-161122, A

  By the way, the conventional higher-order ambisonics has a problem that it can generate only one region (control region) in which the sound field can be reproduced with high accuracy. For example, when considering the interior space of a vehicle, it is necessary to consider not only the driver but also other passengers as users who listen to music, etc., so there are a plurality of control areas corresponding to each user. Although necessary, conventional high-order ambisonics can only provide a realistic sound field to one user.

  The present invention was created in view of such points, and its purpose is to reproduce a sound field that can reproduce a sound field with high accuracy in a control region corresponding to each of one or a plurality of users. To provide a system.

In order to solve the above-described problems, the sound field reproduction system of the present invention is compatible with a plurality of speakers and a plurality of speakers, weights the audio signal, and sets a gain corresponding to each weight. In order to realize a control region corresponding to the listening position for each combination using a predetermined wavefront, assuming a listening position for each combination of a plurality of gain adjusting means and one to a plurality of users the weights corresponding to each of the required plurality of speakers, and the speaker weight storage means for storing for each of the combinations, the combination determining means for determining the content of the combination according to the actual listening mode by a number of people and the listening position of the user Then, the respective weights of the plurality of speakers corresponding to the content determined by the combination determining unit are read from the speaker weight storing unit, and a plurality of Gain setting means for setting each gain of the obtaining adjustment means. In addition, it is desirable that the combination determination unit determines the contents of the combination so that the number of users and the number of control areas realized in an acoustic space with a plurality of speakers are equal. The predetermined wavefront described above is preferably a plane wave.

  Since it is possible to set the minimum required listening position according to the actual listening mode by the user, the sound field can be reproduced with high accuracy in the control area corresponding to each of one or a plurality of users. it can. Also, when changing the combination of listening positions, it is only necessary to read out the necessary speaker weights and set the gain, so that the process for reproducing the sound field can be simplified.

  Further, it is desirable that the weight stored in the above-described speaker weight storage means is determined using a sound field transfer function between the plurality of speakers and one or a plurality of control regions. Using a transfer function of a sound field between each speaker and each control region, for example, applying a higher-order ambisonics method that expresses a sound field using a spherical harmonic function to one or more control regions. By determining the weight of each speaker, it is possible to reproduce the sound field in each control region with high accuracy.

  The weight stored in the speaker weight storage means described above is preferably determined by the speaker weight determination means using the transfer function measured by the transfer function measurement means. By applying the higher-order ambisonics method, which expresses the sound field using spherical harmonics, to one or more control regions at the same time and determining the weight of each speaker, A transfer function that takes into account the influence of reflection or the like that occurs in the space can be obtained, and the sound field can be reproduced with higher accuracy in one or a plurality of control regions.

  Further, it is desirable that the above-described transfer function measuring means identifies the transfer function by setting the tap coefficient of the adaptive filter using the LMS algorithm. In particular, the transfer function measuring means described above includes a measurement signal source that generates a measurement signal, an adaptive filter to which the measurement signal is input, a plurality of speakers that output the measurement signal to an acoustic space including a plurality of control regions, A microphone that is installed in a plurality of control areas, picks up sounds output from a plurality of speakers, a calculation unit that calculates a difference between the output signal of the microphone and the output signal of the adaptive filter, and outputs an error signal; A measurement signal and an error signal are input, and it is desirable to include an LMS algorithm processing unit for setting tap coefficients of the adaptive filter by using the LMS algorithm, and to identify the characteristics of the adaptive filter as a transfer function. This makes it possible to accurately measure the transfer function.

  The speaker weight determination means described above is such that when the high-order ambisonics conversion function matrix is B, the speaker weight matrix is P, and the reconversion function matrix in which each element is set by the transfer function is C, these It is desirable to determine the weights of a plurality of speakers based on the relationship between each matrix. Further, a relational expression of CP = B is established among the conversion function matrix B, the speaker weighting matrix P, and the reconversion function matrix C described above, and the speaker weight determining means determines whether the plurality of speakers are based on the relational expression. It is desirable to determine the weight. As a result, it is possible to easily determine the weights of a plurality of speakers using the transformation function matrix B and the retransformation function matrix C having known values.

  In addition, it is preferable that an operation unit operated by the user is further provided, and the combination determination unit determines the content of the combination in accordance with a user instruction using the operation unit. Thereby, it is possible to easily acquire the actual listening mode by the user without causing the configuration to be complicated.

  The apparatus further comprises listening position detecting means for detecting listening positions of one or a plurality of users included in a predetermined detection range, and the combination determining means is configured to perform a combination based on the listening position detected by the listening position detecting means. It is desirable to determine the contents. Thereby, it is possible to easily acquire the actual listening mode by the user without performing any operation by the user himself.

It is explanatory drawing which shows the spherical harmonic expression of the transfer function C at the time of paying attention to the sound field of the plane wave seen from the origin. It is explanatory drawing of the transfer function from each speaker to the control area | region set other than the origin. It is explanatory drawing of the transfer function from each speaker to the control area | region set other than the origin. It is a figure which shows the diameter of the control area | region of the reproduced sound field. It is a figure which shows the structure of a sound field reproduction apparatus. It is a flowchart which shows the operation | movement procedure which determines the weight P of a speaker using the sound field reproduction apparatus shown in FIG. It is a figure which shows the structure of the sound field control apparatus of one Embodiment to which this invention is applied. It is explanatory drawing which shows an example of the combination of a listening position in the case of applying the some control area reproduced by this embodiment to actual vehicle interior space. It is a figure which shows the structure of the sound field reproduction apparatus which determines the weight P of a speaker for every arbitrary combination K which consists of at least 1 in several listening positions according to actual acoustic space. It is a figure which shows the detailed structure of the transfer function measurement part shown in FIG. It is a flowchart which shows the operation | movement procedure which determines the weight P of a speaker using the sound field reproduction apparatus shown in FIG. 9 and FIG. It is explanatory drawing which performs C characteristic (transfer characteristic) identification. It is a figure which shows the structure of the sound field control apparatus of a modification.

  A sound field reproduction system according to an embodiment to which the present invention is applied will be described below with reference to the drawings.

(1) Spherical harmonic function The sound field of a plane wave seen from the origin is

It becomes. here,

Is a position vector indicating the position of interest,

Is the wave vector of plane waves, Jn (kr) is the first kind Bessel function,

Is a spherical harmonic function. Also,

Is

It becomes. Also,

The sound field of plane waves seen from

It becomes. Therefore,

It becomes. here,

Is

It becomes.

  next,

The sound field of the speaker in

It becomes. here,

Is

It becomes.

Seen from

The sound field of the speaker in

It becomes. here,

Is as follows.

FIG. 1 is an explanatory diagram showing a spherical harmonic expression of the transfer function C when attention is paid to a plane wave sound field viewed from the origin. In FIG. 1, M speakers L _{1 to} L _M are arranged on the circumference around the origin, and this is a high-order ambisonics (HOA) conversion function at the origin.

Is expressed by the above-described equation (2). Also, the transfer function from each speaker to the origin

Is expressed by the above-described equation (9).

(2) the sound field area s ₁ Basic Concept any plane waves of the plurality regions sound reproduction, s _2, ···, a s _N, a plurality of speakers L _1, L _2, ···, a L _M Use to reproduce. Wave vector of plane wave

The

When, in order to reproduce the plane wave, such as the following, it is necessary to obtain the weight P _a speaker for all areas (alpha) and spherical harmonics (n, m).

Therefore,

Is a normal high-order ambisonics (HOA) conversion function. Here, assuming that order (n) and degree (m) of the spherical harmonic functions are given, the conversion function of each reproduction region is as follows.

To summarize the above formula:

Therefore, the multi-region reproduction formula is as follows.

When the speaker weight is P, the higher order ambisonics (HOA) transformation matrix is B, the retransformation matrix is C,

become that way. Here, each matrix is as follows.

Therefore, the speaker weights for reproducing the sound field are as follows.

The sound field can be reproduced by outputting this weight by a speaker.

  In such a technique, the transfer function of the sound field from the speaker to the control area is expressed by spherical harmonic expansion, the transfer function of each control area is obtained, and the weight of the speaker that reproduces each sound field from the transfer function is obtained. The sound field of a plurality of control areas is reproduced.

  2 and 3 are explanatory diagrams of a transfer function from each speaker to a control region set other than the origin. Thus, the value of each element of the inverse transformation matrix C expressed by the equation (24) can be determined by obtaining the transfer function up to each of the plurality of control regions set other than the origin. Thereby, each element on the right side of the equation (25) becomes known, and the weight of each speaker can be determined.

  FIG. 4 is a diagram showing the diameter of the reproduced sound field control region. In the example shown in FIG. 4, it can be seen that plane waves that have arrived along the X direction indicated by the horizontal axis are generated, and there are four locations with few errors. Specifically, a control region having a diameter d is reproduced at the following position.

x = −0.5 m, y = 0.5 m: d = 24 cm
x = −0.5 m, y = −0.5 m: d = 22 cm
x = 0.5 m, y = 0.5 m: d = 26 cm
x = 0.5 m, y = −0.5 m: d = 32 cm
The example shown in FIG. 4 shows a case where four control areas are realized, but the diameter d of these control areas varies depending on the position and number of each control area. In particular, regarding the number of control areas, the smaller the diameter d, the larger the diameter d, and the larger the number, the smaller the diameter d. Therefore, if the number of users is N and the number of control regions realized in the actual acoustic space is M, M is set equal to N in order to set the diameter d of each control region as large as possible. From the viewpoint of securing a large diameter d, setting a large number of control areas beyond the number of users should be avoided.

(3) Specific Configuration of Multi-field Sound Field Reproduction FIG. 5 is a diagram illustrating an example of the configuration of a sound field reproduction device. The sound field reproduction device 100 shown in FIG. 5 performs processing for determining the speaker weight P based on the above-described equation (25), and includes a control region position setting unit 110, an operation unit 112, a transfer function setting unit 120, An HOA function setting unit 130 and a speaker weight determination unit 140 are provided.

  The control area position setting unit 110 sets the positions of a plurality of control areas. The operation unit 112 is configured using a numeric keypad, various knobs, a touch panel, or the like operated by the user, and receives input of position data and the like input by the user. The control area position setting unit 110 described above sets the positions of a plurality of control areas based on a user instruction (position data) using the operation unit 112.

  The transfer function setting unit 120 expresses the transfer function of the sound field between the plurality of speakers and the plurality of control regions by spherical harmonic expansion, and uses the positions of the plurality of control regions set by the control region position setting unit 110. Then, the transfer function shown by the above-described equations (11) to (13) is set (calculated).

  The HOA function setting unit 130 sets a high-order ambisonics conversion function. Specifically, the value of each element of the high-order ambisonics transformation function matrix expressed by the above-described equation (23) is held.

  The speaker weight determination unit 140 uses a transfer function set by the transfer function setting unit 120 to input a plurality of speaker weights necessary for reproducing a plurality of control areas by plane waves to the plurality of speakers. A weight for weighting the audio signal is determined.

  Specifically, when the transformation function matrix of higher-order ambisonics is B, the weight matrix of the speaker is P, and the reconversion function matrix in which each element is set by the transfer function is C, The indicated relational expression holds. The speaker weight determination unit 140 actually determines the weight matrix P of a plurality of speakers using the above-described equation (25) based on the equation (21).

  FIG. 6 is a flowchart showing an operation procedure for determining the speaker weight P using the sound field reproduction device 100 shown in FIG.

  First, the control region position setting unit 110 sets the positions of a plurality of control regions (step 100).

  Next, the transfer function setting unit 120 sets a transfer function using the positions of the plurality of control regions set by the control region position setting unit 110 (step 102).

  Next, the speaker weight determination unit 140 uses the transfer function set by the transfer function setting unit 120 to input a plurality of speaker weights necessary for reproducing a plurality of control areas. The weight for weighting the audio signal is determined (step 104).

  By the way, the sound field reproduction system of the present invention reproduces the sound field of one or a plurality of control regions corresponding to each combination by assuming a combination of positions that can be the listening position of the user in advance. FIG. 7 is a diagram showing a configuration of a sound field control apparatus according to an embodiment to which the present invention is applied.

  A sound field control device 500 shown in FIG. 7 is connected to the output side of the audio device 600, and includes M amplifiers 510-1, 510-2,..., 510-M, a parameter storage unit 520, parameter setting. A unit 526, a combination determination unit 522, an operation unit 524, and M speakers 530-1, 530-2,..., 530-M.

  Each of the amplifiers 510-1 and the like can change the gain, and amplifies (or attenuates) the audio signal input from the audio device 600 with a predetermined gain and outputs the amplified signal. The output audio signal is input to the speaker 530-1 or the like corresponding to one to one.

  The parameter storage unit 520 assumes a listening position by one or a plurality of users, and uses a predetermined wavefront, and a plurality of control areas necessary for realizing one or a plurality of control areas corresponding to the listening position on a one-to-one basis. A weight corresponding to each of the speakers 530-1 and the like is stored for each arbitrary combination K including at least one of a plurality of listening positions.

  The combination determination unit 522 determines the listening position combination K in accordance with the actual listening mode by the user. Specifically, the sound field control device 500 of the present embodiment includes an operation unit 524 operated by a user, and the combination determination unit 522 responds to a user instruction using the operation unit 524. A listening position combination K is determined.

  The amplifier 510-1 and the like described above are gain adjusting means, the parameter storage unit 520 is speaker weight storage means, the combination determination unit 522 is combination determination means, the parameter setting unit 526 is gain setting means, and the operation unit 524 is operation means. Correspond to each.

  FIG. 8 is an explanatory diagram illustrating an example of a combination of listening positions when a plurality of control regions reproduced in the present embodiment are applied to an actual vehicle interior space. The vehicle interior space S1 shown in FIG. 8 is provided with a plurality of seats D (for example, 6 seats), which is a minimum of one driver and a maximum of one driver and five other passengers. Combinations of listening positions corresponding to up to six people are possible.

  The user selects one or a plurality of listening positions among the six seats D shown in FIG. In this selection, the seating positions of one or more users who actually listen to the audio sound are designated. For example, the operation unit 524 is operated by the user in a state where a seat diagram schematically representing the six seats D shown in FIG. 8 is displayed on a display device (not shown), and one or a plurality of seats are operated. D is selected. Note that it is not always necessary to use a schematic seat map, and it is only necessary to specify the content of the combination of “front row / middle row / back row” and “left / right”. The combination determination unit 522 determines the listening position combination K according to the operation result of the operation unit 524.

The parameter setting unit 526 reads the weights of the plurality of speakers 530-1 and the like corresponding to the combination K determined by the combination determination unit 524 from the parameter storage unit 520, and sets the gains of the amplifiers 510-1 and the like. To do. Specifically, corresponding to the combination K, each of the elements P ₁ , P ₂ ,..., P _M of the speaker weight matrix P determined by using the above-described equation (25) is used as an amplifier. 510-1, 510-2,..., 510-M.

  As described above, in the sound field reproduction system of this embodiment, the higher-order ambisonics method for expressing the sound field using a spherical harmonic function is simultaneously applied to one or a plurality of control regions, and the weight of each speaker is determined. It is possible to reproduce the sound field with high accuracy in one or a plurality of control areas.

  In addition, since the user operates the operation unit 524 to specify the positions of the plurality of control areas, the sound field at the plurality of positions specified by the user can be realized with high accuracy.

  Also, by using the above-described equation (25), it is possible to easily determine the weights of a plurality of speakers using the conversion function matrix B and the reconversion function matrix C having known values.

  FIG. 9 shows a sound field in which the speaker weight P is determined for each arbitrary combination K consisting of at least one of a plurality of listening positions in accordance with the actual acoustic space (for example, the vehicle interior space shown in FIG. 8). It is a figure which shows the structure of a reproduction apparatus. The sound field reproduction device 200 shown in FIG. 9 performs a process of determining the speaker weight P based on the above-described equation (25) for each of a plurality of combinations K of listening positions. A function measurement unit 220, a HOA function setting unit 130, and a speaker weight determination unit 140 are provided. The HOA function setting unit 130 and the speaker weight determination unit 140 are basically the same as those shown in FIG. 5, and the combination setting unit 210 and the transfer function measurement unit 220 will be described in detail below.

  The combination setting unit 210 designates one of a plurality of combinations K of listening positions, and sets one or a plurality of control areas corresponding to the designated combination K.

  The transfer function measurement unit 220 measures the transfer function of the sound field between the plurality of speakers 530-1 and the like and one or more control regions set by the combination setting unit 210. The transfer function measurement unit 220 identifies the transfer function by setting the tap coefficient of the adaptive filter using the LMS algorithm.

  The transfer function measuring unit 220 described above corresponds to the transfer function measuring unit, and the speaker weight determining unit 140 corresponds to the speaker weight determining unit.

  FIG. 10 is a diagram showing a detailed configuration of the transfer function measuring unit 220 shown in FIG. As shown in FIG. 10, the transfer function measurement unit 220 includes a measurement signal source 221, switches 222 and 226, a C characteristic identification unit 224, N microphones 225-1, 225-2,..., M speakers. 530-1, 530-2,.

  The measurement signal source 221 generates a measurement signal (for example, a white noise signal) for measuring a transfer function. The switch 222 selectively inputs the measurement signal output from the measurement signal source 221 to any of the M speakers 530-1 or the like. The C characteristic identification unit 224 has one or more of each of the M speakers 530-1, etc. and one of the N microphones 225-1, etc. (one-to-one at the listening position included in the combination K of interest). The transfer function of the acoustic space with the corresponding microphone) is identified. A specific configuration and operation for identifying the transfer function will be described later. The switch 226 selects an output corresponding to the listening position included in the combination K of interest from among the outputs of the N microphones 225-1 and the like, and inputs the selected output to the C characteristic identification unit 224.

  By the way, the arrangement of the M speakers 530-1 and the like and the N microphones 225-1 and the like in this embodiment is actually performed using an acoustic space. For example, in the example shown in FIG. 8, six microphones 225-1 and the like are installed at the head position of each user when six users are seated in six seats D installed in the vehicle interior space. Then, M speakers 510-1 and the like are installed along the wall surface forming the vehicle interior space so as to surround these seats. When the combination K is designated by the combination setting unit 210, the combination K (or information for specifying the listening position corresponding to the combination) is input to the switch 226. The switch 226 performs a selection operation in which only the output of the microphone 225-1 or the like installed in the seat corresponding to the listening position included in the combination K is input to the C characteristic identification unit 224.

  FIG. 11 is a flowchart showing an operation procedure for determining the speaker weight P using the sound field reproduction device 200 shown in FIGS. 9 and 10.

  First, the combination setting unit 210 selects one from a plurality of combinations K of listening positions (step 200). Next, the combination setting unit 210 (or another control unit (not shown, and the C characteristic identification unit 224 may perform the same operation)) includes one control region included in the selected combination K. One microphone 225-1 etc. corresponding to is selected (step 202), and one speaker 530-1 etc. is selected (step 204).

  Next, a measurement signal is output from the measurement signal source 221 (step 206). This measurement signal is output from the speaker 530-1 selected in step 204.

  Next, the C characteristic identification unit 224 picks up the measurement signal using the microphone 225-1 selected in Step 202 (Step 208), and identifies the C characteristic (transfer function) between the selected speaker and the microphone. (Step 210).

  Next, the combination setting unit 210 determines whether there is another speaker that has not been selected (step 212). In some cases, an affirmative determination is made, and the process returns to step 204 to repeat the same identification operation for other speakers. Further, when the identification operation for all the speakers is completed, a negative determination is made in the determination in step 212.

  Next, the combination setting unit 210 determines whether there is another control area (other microphone) that has not been selected (step 214). In some cases, an affirmative determination is made, and the process returns to step 202 to repeat the same identification operation for other control regions. Further, when the identification operation for all the control regions is completed, a negative determination is made in the determination of step 214.

  Next, the speaker weight determination unit 140 reproduces one or a plurality of control regions corresponding to the combination K using a transfer function measured for a combination of all speakers and one or a plurality of microphones corresponding to the combination K. Therefore, the weights of the M speakers necessary to weight the audio signals input to the M speakers are determined (step 216).

  Next, the combination setting unit 210 determines whether there is another combination K that has not been selected (step 218). If there is an unselected combination K, an affirmative determination is made, and the process returns to step 200 and the processing after the selection operation of another combination K is repeated. When selection for all combinations K is completed, a negative determination is made in the determination in step 218, and a series of processes for determining the weights of speaker weights for all combinations K is completed.

  Next, a specific configuration and operation for identifying the C characteristic (transfer characteristic) will be described. FIG. 12 is an explanatory diagram for performing C characteristic (transfer characteristic) identification. In the explanatory diagram shown in FIG. 12, for example, the case where the C characteristic between the speaker 530-1 and the microphone 225-1 is identified is shown.

  As shown in FIG. 12, the C characteristic identification unit 224 includes an adaptive filter 224A, a calculation unit 224B, and an LMS algorithm processing unit 224C, and identifies the characteristic of the adaptive filter 224A as a transfer function.

  The adaptive filter 224A is an FIR type filter to which the measurement signal output from the measurement signal source 221 is input. The arithmetic unit 224B calculates the difference between the output signal of the microphone 530-1 and the like and the output signal of the adaptive filter 224A, and outputs an error signal ε.

  The LMS algorithm processing unit 224C receives the measurement signal and the error signal, and sets the tap coefficient W (n) of the adaptive filter 224A by using the LMS algorithm as the adaptive algorithm.

  As described above, in the sound field reproduction device 200 according to the present embodiment, the weight of each speaker is determined by simultaneously applying a higher-order ambisonics method that expresses a sound field using a spherical harmonic function to a plurality of control regions. When measuring the transfer function, it is possible to obtain a transfer function that takes into account the effects of reflection, etc. that occur in the acoustic space, and to reproduce the sound field in a plurality of control areas with higher accuracy. Become.

  As described above, in the sound field reproduction system according to the present embodiment, it is possible to set a minimum listening position according to the actual listening form by the user, and therefore it corresponds to each of one or a plurality of users. The sound field can be reproduced with high accuracy in the control region. Also, when changing the listening position combination K, only the necessary speaker weight P is read and the gain is set, so that the process for reproducing the sound field can be simplified.

  In addition, the weight P stored in the parameter storage unit 520 is determined by the speaker weight determination unit 140 using the transfer function measured by the transfer function measurement unit 220. When the weight P of each speaker is determined by simultaneously applying a higher-order ambisonics method for expressing a sound field using a spherical harmonic function to one or a plurality of control regions, A transfer function that takes into account the influence of reflection or the like that occurs in the acoustic space can be obtained, and the sound field can be reproduced with higher accuracy in one or a plurality of control regions.

  The transfer function measurement unit 220 identifies the transfer function by setting the tap coefficient of the adaptive filter using the LMS algorithm. In particular, the transfer function measurement unit 220 includes a measurement signal source 221 that generates a measurement signal, an adaptive filter 224A to which the measurement signal is input, and a plurality of speakers 530 that output the measurement signal to an acoustic space including a plurality of control regions. -1 etc., which are installed in a plurality of control areas and pick up sound output from a plurality of speakers 530-1 etc., output signals from the microphone 225-1 etc. and the adaptive filter 224A The calculation unit 224B that calculates the difference between the output signals and outputs the error signal, the measurement signal and the error signal are input, and the LMS algorithm processing that sets the tap coefficient of the adaptive filter 224A by using the LMS algorithm 224C, and the characteristic of the adaptive filter 224A is identified as a transfer function. This makes it possible to accurately measure the transfer function.

  Further, the speaker weight determination unit 140 has a higher-order ambisonics conversion function matrix as B, a speaker weight matrix as P, and a reconversion function matrix in which each element is set by a transfer function as C. The weights of a plurality of speakers are determined based on the relationship between the matrices. Further, a relational expression of CP = B is established among the conversion function matrix B, the speaker weight matrix P, and the reconversion function matrix C described above, and the speaker weight determination unit 140 is configured based on the above relational expressions. The speaker weight P is determined. This makes it possible to easily determine the weights of the plurality of speakers 530-1, etc., using the conversion function matrix B and the reconversion function matrix C having known values.

  In addition, an operation unit 524 operated by the user is provided, and the combination determination unit 522 determines the content of the combination in accordance with a user instruction using the operation unit 524. Thereby, it is possible to easily acquire the actual listening mode by the user without causing the configuration to be complicated.

  In addition, this invention is not limited to the said embodiment, A various deformation | transformation implementation is possible within the range of the summary of this invention. For example, in the above-described embodiment, the user operates the operation unit 524 to instruct the listening position combination K, but may be automatically detected.

  FIG. 13 is a diagram showing a configuration of a modified sound field control device 500A. The sound field control device 500A shown in FIG. 13 differs from the sound field control device 500 shown in FIG. 7 in that the operation unit 524 is replaced with a listening position detection unit 524A. The listening position detector 524A corresponds to the listening position detector. The listening position detection unit 524A includes, for example, a camera that images a vehicle interior space and an image processing unit that recognizes a human head position included in the captured image. The listening position detection unit 524A can detect in which seat the user is seated and whether the user is an adult or a child without performing a special operation by the user. In addition, since it is necessary to store the transfer function measured beforehand corresponding to each listening position in the parameter storage unit 520, the detected head position of the adult or child is not used, but close to these. It is necessary to prepare a representative value in advance and select the closest representative value. The combination determination unit 522 determines the content of the combination based on the listening position detected by the listening position detection unit 524A. Thereby, it is possible to easily acquire the actual listening mode by the user without performing any operation by the user himself.

  In the above-described embodiment, the vehicle interior space has been described as a specific example including a plurality of control areas. However, the acoustic space to which the present invention is applied may be, for example, a room for a home theater.

  In the above-described embodiment, a case where a plurality of control areas are reproduced by plane waves has been described. However, the present invention can also be applied to a case where a plurality of control areas are reproduced using a predetermined wavefront other than plane waves. .

  As described above, according to the present invention, it is possible to set the minimum necessary listening position in accordance with the actual listening mode by the user. Therefore, the control area corresponding to each of one or a plurality of users. The sound field can be reproduced with high accuracy. Also, when changing the combination of listening positions, it is only necessary to read out the necessary speaker weights and set the gain, so that the process for reproducing the sound field can be simplified.

100, 200 Sound field reproduction device 110 Control region position setting unit 112 Operation unit 120 Transfer function setting unit 130 HOA function setting unit 140 Speaker weight determination unit 210 Combination setting unit 220 Transfer function measurement unit 221 Measurement signal source 222 Switch 224 C characteristic identification Unit 225-1, 255-2,... Microphone 500 sound field control device 510-1, 510-2,... Amplifier 520 parameter storage unit 522 combination determination unit 524 operation unit 524 A listening position detection unit 526 parameter setting unit 530-1, 530-2,... Speaker 600 Audio device

Claims (11)

Multiple speakers,
A plurality of gain adjusting means that correspond to each of the plurality of speakers, weight the audio signal, and set a gain corresponding to each weight;
Assuming a listening position for each combination of one to a plurality of users, each of the plurality of speakers necessary for realizing a control area corresponding to the listening position for each combination using a predetermined wavefront the corresponding the weight, and the speaker weight storage means for storing for each of the combinations,
Combination determining means for determining the content of the combination in accordance with the actual listening mode according to the number of users and the listening position ;
Gain setting means for reading the weights of the plurality of speakers corresponding to the determination content by the combination determination means from the speaker weight storage means and setting the gains of the plurality of gain adjustment means;
Sound field reproduction system characterized by comprising. In claim 1,
The sound field reproduction system, wherein the predetermined wavefront is a plane wave. In claim 1 or 2,
The sound field reproduction system characterized in that the weight stored in the speaker weight storage means is determined using a transfer function of a sound field between the plurality of speakers and the one or more control regions. . In claim 3,
The sound field reproduction system according to claim 1, wherein the weight stored in the speaker weight storage unit is determined by a speaker weight determination unit using the transfer function measured by the transfer function measurement unit. In claim 4,
The sound field reproduction system characterized in that the transfer function measurement means identifies the transfer function by setting tap coefficients of an adaptive filter using an LMS algorithm. In claim 4 or 5,
The transfer function measuring means includes
A measurement signal source for generating a measurement signal; and
An adaptive filter to which the measurement signal is input;
The plurality of speakers for outputting the measurement signal to an acoustic space including the plurality of control regions;
A microphone that is installed in the plurality of control areas, and that collects sound output from the plurality of speakers;
An arithmetic unit that calculates a difference between an output signal of the microphone and an output signal of the adaptive filter and outputs an error signal;
The measurement signal and the error signal are input, and an LMS algorithm processing unit that sets a tap coefficient of the adaptive filter by using an LMS algorithm;
The sound field reproduction system comprising: identifying the characteristic of the adaptive filter as the transfer function. In any one of Claims 4-6,
The speaker weight determining means is configured such that a higher-order ambisonics conversion function matrix is B, a speaker weight matrix is P, and a reconversion function matrix in which each element is set by the transfer function is C. A sound field reproduction system, wherein weights of the plurality of speakers are determined based on a relationship between matrices. In claim 7,
A relational expression of CP = B is established between the conversion function matrix B, the speaker weight matrix P, and the reconversion function matrix C.
The sound field reproduction system characterized in that the speaker weight determination means determines the weights of the plurality of speakers based on the relational expression. In any one of Claims 1-8,
It further comprises operating means operated by the user,
The sound field reproduction system according to claim 1, wherein the combination determining unit determines the content of the combination in accordance with a user instruction using the operation unit. In any one of Claims 1-8,
A listening position detecting means for detecting listening positions of one or more users included in a predetermined detection range;
The sound field reproduction system characterized in that the combination determining means determines the content of the combination based on the listening position detected by the listening position detecting means.   In any one of Claims 1-10,
  The combination determination means determines the content of the combination so that the number of users and the number of control areas realized in an acoustic space by the plurality of speakers are equal. .

Images