JPH07168587A - Sound environment simulation device and sound environment analysis method - Google Patents

Abstract

(57) [Summary] [Purpose] To realize a realistic sound field reproduction based on the simulation results of the acoustic characteristics of the room, and to promote efficient acoustic design and proposal-based sales. [Structure] The sound environment simulation experience device 1 includes a sound environment analysis unit 1, a sound field reproduction unit 2, and an output unit 3. Sound environment analysis unit 1
Divides the analysis space into several planes, calculates the reflected sound using the reflectance and form factor of each divided plane, and reaches when the volume radiated from a predetermined sound source reaches a predetermined sound receiving point. Calculate volume time series information. In the sound field reproduction unit 2, the impulse response calculation unit converts the reached volume time series information into impulse response data. The reproduced sound generation unit generates a reproduced sound by convolving the impulse response with the dry source based on the position information of the experience person input from the linked virtual reality device, and outputs the sound to the headphones.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a sound environment simulation experience device capable of predicting acoustic performance by computer simulation and confirming it by experience when designing a room acoustic such as a listening room or a concert hall. Furthermore, the present invention relates to a room acoustic analysis method for analyzing acoustic characteristics of a space.

[0002]

2. Description of the Related Art In recent years, AV (audio / visual)
Due to the remarkable progress in high functionality and high performance of related devices, it has become possible to widely enjoy high quality music and images in general households. In particular, it is becoming more and more popular to watch movie films and music that are as realistic as movie theaters and concert halls at home, as is called the "home theater." Therefore, in a general home, it is desired to install a room with sufficient acoustic design and sound insulation design.

However, considering the actual housing environment, it is difficult to provide a dedicated listening room. For example, it is realistic to use the living room and the listening room together. Therefore, in the future, there is an increasing demand for realizing a living room as a space with sufficient acoustic design and sound insulation design in consideration of livability and aesthetics of a living room where various furniture and electronic devices are arranged. .

Conventionally, in acoustic design, acoustic characteristic data of a room to be designed is obtained, and the specification of the room is evaluated based on the acoustic characteristic data. Then, as a method of calculating the acoustic characteristic data, an experimental method or an analytical method has been used. As an experimental method, for example, there is a method of creating a model of an actual room and directly measuring impulse response data in the model room to obtain acoustic characteristics. However, it is difficult to disseminate this method to general households from the viewpoint of work man-hours and costs.

As an analytical method, there is a method of obtaining acoustic characteristics by computer simulation.
This method can obtain output according to various conditions,
It has been rapidly put into practical use in recent years because of its capability of precise analysis, easy processing of output results, and easy feedback to design work. Room sound field analysis 3
The two-dimensional computer simulation is roughly divided into a numerical analysis method that considers the wave nature of sound and a solution method based on the geometrical acoustic theory that ignores the wave nature.

A typical numerical analysis method is a finite element method (FEM) or a boundary element method (BEM). FE
M and BEM are methods of calculating the acoustic eigenmode and sound pressure distribution by solving the Helmholtz equation, which is the governing equation of the stationary sound field. This method can analyze the diffraction phenomenon and the interference phenomenon by considering the wave nature of the sound, but has a problem that unsteady calculation such as echo time pattern calculation is difficult. In particular, in the analysis of high-frequency sounds, the number of grids required for calculation becomes enormous, the memory of the computer is insufficient, and at the same time, it takes too much calculation time, so that the analysis is practically impossible.

On the other hand, the geometrical acoustic methods such as the sound ray method and the virtual image method are calculated on the assumption that the sound ray is geometrically reflected on the wall surface. This method does not cause a problem due to the memory limitation as compared with the numerical analysis method, but in a room having a size shorter than the wavelength of sound, the geometrical reflection is not accurately performed, so that the calculation accuracy is deteriorated. Therefore, it is difficult to analyze low frequency sounds in a small space such as a home listening room. Also, if the ray tracing calculation takes a long time,
Since the amount of calculation becomes enormous, the reverberation time, which is an important element of the room acoustic characteristics, is approximated using the Sabine or Eyring equations. For this reason, there is a problem that an error occurs in the calculation when a device or a complicated space where unevenness is present is targeted for calculation.

[0008]

Furthermore, recently, a sound environment simulation experience device intended to simulate a sound environment in a room by reproducing a sound field based on acoustic characteristic data and making a person listen to the sound field is provided. What is called is realized. As a pseudo experience device in a broad sense, there is, for example, a system in which an actual model space is installed, and a speaker arranged inside the device outputs sound to allow an experience person to listen to it. However, such a system is not suitable for home use.

Another pseudo experience device is USP 4,731,84
8 is shown. This device generates an output signal by adding a reverberation effect to an input audio signal and reproduces and outputs it to a headphone. In this device, a signal branched from an audio signal is generated as a reverberation signal in which a time delay of a reflected sound is added using a delay circuit, and an input signal and this reverberation signal are combined to output a signal. However, since this device does not consider the high-order reflected sound, it may feel strange compared with the actual sound. Further, information such as the virtual sound source and the position of the experience person needs to be input by the operator as instruction data, and there is a problem that the immediacy of the reproducing operation is lacking and the target space has a geometrical constraint.

Therefore, an object of the present invention is to provide a sound environment simulation experience device capable of reproducing a sound environment in an arbitrary space without using an actual room or model. Another object of the present invention is to provide a sound environment simulating experience device capable of generating a reproduced sound in consideration of a change in acoustic characteristics when a user moves or changes a head direction. is there.

Another object of the present invention is to provide a sound environment analysis method capable of calculating a room acoustic characteristic in consideration of the influence of a high-order reflected sound.

[0012]

According to another aspect of the present invention, there is provided an acoustic environment simulating experience device for calculating an echo time pattern of a space to be analyzed by a geometric acoustic method using a condition relating to a space to be analyzed. A time pattern calculating means, an impulse response calculating means for calculating a response for each frequency band using the echo time pattern calculated by the echo time pattern calculating means, and calculating an impulse response by combining the responses; Reproduced sound generating means for generating reproduced sound data by convolving the sound source data of the sound with the impulse response calculated by the impulse response calculating means, and the reproduced sound data generated by the reproduced sound generating means are converted into voice and output. And output means.

In the sound environment simulation experience apparatus according to the invention of claim 2, in addition to the invention of claim 1, the echo time pattern calculation means further includes analysis condition data such as the shape of the space to be analyzed and the physical property data of the wall surface. Receive and set the division analysis surface by dividing the wall surface into predetermined sections,
Analysis data setting means for setting the sound source and the sound receiving point at an arbitrary position in the analysis target space, and the arrival sound of the radiated sound emitted from the sound source set at the arbitrary position in the analysis target space to the sound receiving point. Time-series data calculation means for calculating time-series data of volume, and data conversion means for performing a predetermined interpolation operation on the time-series data of volume of arrival sound calculated by the time-series data calculation means to convert into an echo time pattern It has and.

In the sound environment simulation experience device according to the invention of claim 3, in addition to the invention of claim 2, the sound field reproducing means further comprises:
Dry source input means for inputting anechoic sound source data, head related transfer function storage means for storing head related transfer function data, and impulse response calculation means for sound source data input from the dry source input means And a reproduced sound generation means for generating reproduced sound data by convolving the impulse response thus generated and the head related transfer function read from the head related transfer function storage means.

In the sound environment simulation experience device according to the invention of claim 4, in addition to the invention of claim 3, the sound field reproducing means further comprises:
A position information detecting means for detecting information of at least one of the position of the head of the experience person and the direction of the head in the analysis target space is provided, and the reproduced sound generating means has the position detected by the position information detecting means. The impulse response and the head related transfer function corresponding to the information are extracted and convoluted with the sound source data.

In addition to the invention of claim 1, the sound environment simulation experience device according to the invention of claim 5 is further obtained by a head related transfer function storing means for storing a head related transfer function and an impulse response calculating means. The impulse response and the head related transfer function read from the head related transfer function storing means are convoluted to calculate a combined response for each head direction, and a combined response generating means for anechoic sound source data. And the reproduced sound generating means for generating reproduced sound data by convolving the calculated combined response.

According to the sixth aspect of the sound environment simulating device, the sound field reproducing means for generating reproduced sound data by convoluting the impulse response data into the anechoic sound source data, and the sound field reproducing means. And output means for converting the reproduced sound data into sound and outputting the sound. In the sound environment simulation experience device according to the invention of claim 7, in addition to the invention of claim 6, the sound field reproducing means further comprises dry source input means for inputting anechoic sound source data, and head transmission. Head-related transfer function storing means for storing function data, position information detecting means for detecting information on at least one of the position of the head and the orientation of the head of the experiencer in the analysis target space, and position information detecting means And a reproduced sound generating means for generating reproduced sound data by extracting an impulse response and a head related transfer function corresponding to the position information detected by the sound source data and convolving the convolution with the sound source data.

A sound environment analyzing apparatus according to the invention of claim 8 is
Room information input means for inputting room information such as shape and size of analysis space to be analyzed, sound source information input means for inputting sound source information such as sound source position and sound pressure, number of sound receiving points, position, etc. Sound receiving point information inputting means for inputting sound receiving point information, analysis plane dividing means for dividing a surface forming an analysis space into a predetermined number of surfaces to be analyzed, and sound reflected from a sound source arranged in the analysis space. The time series of the reached sound volume at the sound receiving point by grouping the sound volume at which the line arrives at the predetermined sound receiving point in the analysis space directly or while reflecting on the dividing surface with a predetermined time interval for each arrival time. Sound field reproduction is performed by averaging, for each group, the time-series information calculation means for obtaining information, and the reached sound volume grouped at a predetermined sound receiving point with the number of sound rays reaching within the time interval. Sound field reproduction that converts into data for possible sound field reproduction And a data conversion unit.

A sound environment analyzing apparatus according to the invention of claim 9 is
Room information input means for inputting room information such as shape and size of analysis space to be analyzed, sound source information input means for inputting sound source information such as sound source position and sound pressure, number of sound receiving points, position, etc. Sound receiving point information inputting means for inputting sound receiving point information, analysis plane dividing means for dividing a surface forming an analysis space into a predetermined number of surfaces to be analyzed, and sound reflected from a sound source arranged in the analysis space. The time series of the reached sound volume at the sound receiving point by grouping the sound volume at which the line arrives at the predetermined sound receiving point in the analysis space directly or while reflecting on the dividing surface with a predetermined time interval for each arrival time. Sound field reproduction is performed by averaging, for each group, the time-series information calculation means for obtaining information, and the reached sound volume grouped at a predetermined sound receiving point with the number of sound rays reaching within the time interval. Sound field reproduction that converts into data for possible sound field reproduction A data conversion unit, and a sound field reproduction data interpolation means for performing interpolation of the volume of each group were averaged with respect to the time axis.

A sound environment analysis method according to a tenth aspect of the present invention divides a surface constituting an analysis space to be analyzed into several surfaces, radiates a sound ray from a predetermined sound source in the analysis space, and divides the surface. The volume that reaches each surface is calculated, the time until the sound radiated from a predetermined sound source reaches that surface is calculated, and the reflected volume on that surface is calculated based on the incident volume. In the sound environment analysis method that obtains time-series information of the sound reaching the predetermined sound receiving point by performing the same calculation using this reflected sound volume as a new sound source, when calculating the reflected sound volume at each divided surface, After the number of reflections due to is exceeded a predetermined number of times, incident sounds are accumulated at a predetermined time interval, and then the accumulated sounds are reflected at once.

The sound environment analysis method according to the invention of claim 11 is characterized in that, in contrast to the invention of claim 10, when the reflected sound volume on the divided surface is below a predetermined limit value, the sound is not reflected from the surface. And The sound environment analysis method according to the invention of claim 12 is different from the invention of claim 10 or 11 in that the predetermined time interval is before a predetermined set time calculated based on the predicted value of the reverberation time of the analysis space. And later is different, and the later time interval is characterized by being longer than the previous time interval.

The sound environment analysis method according to the invention of claim 13 is the same as that of claim 11 or claim 12, when all the reflected sound volumes are below the limit value, It is characterized in that some or all of them are added and added to the time-series information of the sound at the sound receiving point, and the time-series information is integrated in order from the earliest arrival time to the sound receiving point to the earliest arrival time.

According to a fourteenth aspect of the present invention, there is provided a sound environment analysis method, which includes analysis conditions such as a shape of a space to be analyzed, physical property data of a wall surface, a divided analysis surface in which the wall surface is divided into predetermined sections, a sound source and a sound receiving unit. The arrival sound receives the time-series data of the arrival sound volume at which the sound emitted from the sound source reaches the sound receiving point set at an arbitrary position in the analysis space. A time-series data calculation step of calculating for each direction incident on the sound point and for each frequency band of the sound source,
An echo time pattern calculation step of calculating an echo time pattern by performing a predetermined interpolation operation on the time series data of the volume of the arrival sound.

The sound environment analysis method according to the invention of claim 15 is the sound environment analysis method according to claim 14, wherein the time-series data calculation step is the volume of the direct sound in which the sound ray emitted from the sound source directly reaches the sound receiving point. A data calculation step, and a sound volume data calculation step for a lower-order reflected sound in which a sound ray radiated from a sound source is reflected by a divided analysis surface at a predetermined number of times or less and a reflected sound reaches other divided analysis surfaces and a sound receiving point. , The step of calculating the volume data of the higher-order reflected sound in which the reflected sound of which the number of times the sound ray radiated from the sound source is reflected by the divided analysis surface reaches the sound receiving point is a predetermined number or more, Calculating time series data by grouping at predetermined time intervals.

In the sound environment analysis method according to the sixteenth aspect of the present invention, in addition to the fifteenth aspect of the invention, the division analysis is performed in which the volume data calculation step for the high-order reflected sound is calculated in the volume data calculation step for the low-order reflected sound. It is possible to repeat the process of integrating the volume data on the surface at predetermined time intervals, setting a new sound source, and calculating the volume reached to the divided analysis surface where the new radiated sound emitted from this sound source arrives. Characterize.

In the sound environment analysis method according to the seventeenth aspect of the present invention, the echo time pattern for each frequency band is calculated using the geometrical acoustic method at the sound receiving point set at an arbitrary position in the analysis target space. And the step of calculating the impulse response by calculating and synthesizing the response for each frequency band by convoluting the bandpass filter for each frequency band in the echo time pattern, and detecting the position information of the experience person. And a step of receiving the dry source of the sound source and convolving the dry source with impulse response and head related transfer function corresponding to the detected position information, and generating the reproduced sound data into voice. And a step of converting and outputting.

In the sound environment analyzing method according to the eighteenth aspect of the present invention, the step of detecting position information in the seventeenth aspect of the invention detects the moving position and the head direction of the experiencer as the position information. It is characterized by doing. The sound environment analysis method according to the invention of claim 19 is the same as the invention of claim 17,
The step of detecting the position information is characterized by detecting the orientation of the head of the experiencer as the position information.

[0028]

According to the first aspect of the invention, the echo time pattern calculating means calculates the echo time pattern of the analysis target space by using the geometrical acoustic method using the condition relating to the analysis target space. The impulse response calculation means calculates the response for each frequency band using the calculated echo time pattern, and further calculates the impulse response by combining the responses. The reproduced sound generation unit convolves the calculated impulse response with the anechoic sound source data to generate reproduced sound data. The generated reproduction sound data is converted into voice and output as voice from the output means.

Further, in the invention of claim 2, the analysis data setting means of the echo time pattern calculation means receives the analysis condition data such as the shape of the space to be analyzed and the physical property data of the wall surface, and divides the wall surface into predetermined sections. Thus, the divided analysis plane is set, and the sound source and the sound receiving point are set at arbitrary positions in the analysis target space. The time-series data calculation means calculates time-series data of the volume of the sound reaching the sound receiving point of the radiated sound emitted from the sound source set at an arbitrary position in the analysis target space. Then, the data conversion means performs a predetermined interpolation operation on the time-series data of the volume of the arrival sound to convert it into an echo time pattern.

In the sound field reproducing means of the invention of claim 3,
The dry source input means inputs anechoic sound source data. Then, the reproduced sound generation unit generates reproduced sound data by convolving the input sound source data with the impulse response and the head related transfer function read from the head related transfer function storage unit. In the sound field reproducing means of the invention of claim 4, the position information detecting means detects at least one of the position of the head and the direction of the head of the experiencer in the analysis target space. Then, the reproduced sound generating means extracts the impulse response and the head related transfer function corresponding to the position information detected by the position information detecting means, and convolves them into the sound source data.

In the fifth aspect of the present invention, the combined response generation means convolves the impulse response obtained by the impulse response calculation means and the head related transfer function read from the head related transfer function storage means for each head direction. Calculate the composite response. Then, the reproduced sound generation means generates reproduced sound data by convolving the synthetic response with the anechoic sound source data.

In the sixth aspect of the invention, the sound field reproducing means generates reproduced sound data by convoluting the impulse response data with the anechoic sound source data. The reproduced sound data generated by the sound field reproducing means is converted into voice and output from the output means. In the sound field reproducing means of the invention of claim 7, the dry source input means inputs anechoic sound source data. The position information detecting means detects information on at least one of the position of the head and the orientation of the head of the experiencer in the analysis target space. The reproduced sound generating means generates reproduced sound data by extracting the impulse response and the head related transfer function corresponding to the detected position information and convolving the convoluted sound source data.

In the invention of claim 8, the room information input means inputs room information such as the shape and size of the analysis space to be analyzed. The sound source information input means inputs sound source information such as a sound source position and sound pressure. The sound receiving point information input means inputs sound receiving point information such as the number and position of sound receiving points. The analysis surface dividing surface means divides the surfaces forming the analysis space into a predetermined number of surfaces to be analyzed. The time-series information calculating means determines the volume at which the sound ray reflected from the sound source arranged in the analysis space reaches the predetermined sound receiving point in the analysis space directly or while reflecting on the dividing surface at a predetermined time interval for each arrival time. By performing the grouping with, the time series information of the reached volume at the sound receiving point is obtained. Then, the sound field reproduction data converting means averages the grouped reached sound volumes at a predetermined sound receiving point for each group with the number of sound rays reaching within the time interval to reproduce the sound field. Is converted to data for sound field reproduction that enables

In the ninth aspect of the present invention, the room information input means inputs room information such as the shape and size of the analysis space to be analyzed. The sound source information input means inputs sound source information such as a sound source position and sound pressure. The sound receiving point information input means inputs sound receiving point information such as the number and position of sound receiving points. The analysis surface dividing means divides the surfaces forming the analysis space into a predetermined number of surfaces to be analyzed. The time-series information calculating means determines the volume at which the sound ray reflected from the sound source arranged in the analysis space reaches the predetermined sound receiving point in the analysis space directly or while reflecting on the dividing surface at a predetermined time interval for each arrival time. By performing the grouping with, the time series information of the reached volume at the sound receiving point is obtained. The sound field reproduction data conversion means can reproduce the sound field by averaging the reached sound volume grouped at a predetermined sound receiving point for each group with the number of sound rays reaching within the time interval. And convert it to sound field playback data. The sound field reproduction data interpolation means interpolates the averaged volume of each group with respect to the time axis.

According to the tenth aspect of the present invention, the surface forming the analysis space to be analyzed is divided into several surfaces, and a sound ray is emitted from a predetermined sound source in the analysis space to reach each of the divided surfaces. The volume of the sound emitted from a given sound source is calculated, the time until the sound radiated from a predetermined sound source reaches the surface is calculated, and the reflected sound volume on that surface is calculated based on the incident sound volume. As a sound source, in the sound environment analysis method that obtains time-series information of the sound reaching a predetermined sound receiving point by performing the same calculation, when calculating the reflected sound volume on each divided surface, the reflection by the divided surface is performed a predetermined number of times. After exceeding, the incident sound is accumulated at a predetermined time interval,
After that, the accumulated sounds are reflected at once.

In the eleventh aspect of the present invention, it is assumed that when the reflected sound volume on the divided surface becomes equal to or less than a predetermined limit value, the reflection from that surface is not performed. In the invention of claim 12, the predetermined time interval is different before and after the predetermined set time calculated based on the predicted value of the reverberation time of the analysis space, and the subsequent time interval is greater than the previous time interval. long.

In the thirteenth aspect of the present invention, when all the reflected sound volumes are equal to or less than the limit value, a part or all of the sound energy remaining on each divided surface is added, and the time series information of the sound at the sound receiving point is added. The time series information is integrated in order from the earliest arrival time to the sound receiving point to the earliest arrival time. Claim 1
In the invention of 4, the analysis conditions such as the shape of the space to be analyzed, the physical property data of the wall surface, etc., the divided analysis surface dividing the wall surface into predetermined sections, the sound source and the sound receiving point are set. next,
The time-series data of the volume of the arrival sound at which the emission sound emitted from the sound source reaches the sound receiving point set at an arbitrary position in the analysis space, for each direction in which the arrival sound enters the sound receiving point, and It is calculated for each frequency band of the sound source. Furthermore, the echo time pattern is calculated by performing a predetermined interpolation operation on the time series data of the volume of the arrival sound.

In the fifteenth aspect of the invention, the time-series data calculating step first calculates volume data of a direct sound in which the sound ray emitted from the sound source directly reaches the sound receiving point. Next, the volume data of the low-order reflected sound in which the reflected sound whose number of times the sound ray radiated from the sound source is reflected on the divided analysis surface reaches the predetermined number of times or less reaches the other divided analysis surface and the sound receiving point is calculated. Furthermore, the volume data of the higher-order reflected sound in which the reflected sound whose sound ray radiated from the sound source is reflected by the divided analysis surface a predetermined number of times or more reaches the sound receiving point is calculated. Further, the sound volume data at the sound receiving point is grouped at predetermined time intervals to calculate time series data.

In the sixteenth aspect of the invention, in the volume data calculation step for the high-order reflected sound, the volume data on the divided analysis plane calculated by the volume data calculation for the low-order reflected sound are integrated at predetermined time intervals and newly added. A sound source is set, and a process of calculating the arrival sound volume to the divided analysis plane to which a new radiated sound radiated from this sound source arrives is repeatedly performed. Claim 1
In the seventh aspect of the invention, first, at the sound receiving point set at an arbitrary position in the analysis target space, the echo time pattern for each frequency band is calculated using the geometric acoustic method.
Next, the impulse response is calculated by convoluting the bandpass filter for each frequency band in the echo time pattern and calculating and synthesizing the response for each frequency band. Further, the position information of the experience person is detected. Further, it receives a dry source of the sound source and convolves the impulse response and the head related transfer function corresponding to the detected position information with the dry source to generate reproduced sound data. Then, the reproduced sound data is converted into voice and output.

According to the eighteenth aspect of the invention, as the position information, information on the moving position of the experiencer and the orientation of the head is detected. In the nineteenth aspect of the invention, the orientation of the head of the experiencer is detected as the position information.

[0041]

Embodiments of the present invention will be described in detail below with reference to the drawings. FIG. 1 is a flowchart showing the processing of the sound environment analysis method according to the first embodiment of the present invention. In FIG. 1, the process A is an initial reflected sound analysis process, and the process B is a higher-order reflected sound analysis process.

First, the step A will be described. Referring to FIG. 1, a solid surface such as a wall that firstly inputs a space to be analyzed is divided into several patches (planes) 4 as shown in FIG. Step S1
1), a number is given to each surface 4 (in FIG. 2, for example, i surface, j surface, k surface). Next, in FIG. 2, the sound source S
The sound factor radiated from 1 arrives at each divided surface 4 (for example, the view coefficient Fsi for the i-plane), and the view coefficient at which each sound reaches each surface 4 other than the sound receiving point (for example, from the i-plane) The view factor for the j-th surface is Fij), and the sound receiving point R from each surface 4
A form factor reaching 2 (for example, a form factor reaching the sound receiving point R2 from the j-th surface is FjR) is calculated (step S12 in FIG. 1).

The method of calculating these form factors is based on the radiant heat ray tracing method used for radiation calculation in a thermal environment or a light environment. First, a predetermined number of sound rays are radiated from the sound source S1, and a surface 4 to which each sound ray reaches is searched from among the divided surfaces 4 (for example, i surface), and the volume reaching the surface 4 is calculated. Then, the ratio of the sound reaching the surface 4 from the sound source S1 (form factor Fsi) is obtained. By repeating this calculation, the view factor for all the divided surfaces 4 is obtained from the sound source S1. next,
Using the arrival sound volume of each divided surface 4 (for example, i surface) other than the sound source as a sound source, the sound volume reaching the other divided surface 4 other than the i surface (for example, j surface) is calculated as described above, and a new sound source is obtained. The form factor Fij from the i-plane to the j-plane is obtained. This calculation is repeated for all the divided surfaces 4 to obtain all the view factors between the surfaces. Finally, the sound receiving point R from each divided surface 4 (for example, k surface)
Similarly, the view factor FkR for 2 is obtained, and this calculation is repeated over all the division planes 4 to obtain all view factors necessary for the calculation of the echo time pattern.

Next, the step of calculating the echo time pattern is started. First, the volume of the direct sound that reaches the sound receiving point R2 directly from the sound source S1 without hitting any divided surface 4 is obtained (step S13 in FIG. 1). This is calculated based on the inverse square law of the distance between the sound source and the sound receiving point (r in FIG. 2). At the same time, the arrival time is obtained by dividing the distance between the sound source and the sound receiving point by the speed of sound.

Subsequently, the volume and arrival time of the reflected sound of the sound radiated from the sound source S1 reaching the sound receiving point R2 through some divided surfaces are calculated (step S14 in FIG. 1). These reflected sounds are obtained by calculating each form factor obtained in the previous step S12. That is, here, first, the primary reflected sound that reaches the sound receiving point R2 from the sound source S1 through one divided surface 4 is calculated. For example, in FIG. 2, in the case of a sound from the sound source S1 via the i-side 4 to the sound receiving point R2, the form factor is Fsi × FiR, and this form factor and the sound absorption coefficient αi of the i-side 4 are radiated from the sound source S1. The volume P1 that reaches the sound receiving point R2 is obtained by applying the entire volume. Similarly to the direct sound, the arrival time of the primary reflected sound is calculated by obtaining the distances between the sound source and the i-plane, and the distances between the i-plane and the sound receiving point, and dividing the sum of these by the sound speed. After repeating these processes for all the divided surfaces 4, the calculation of the secondary reflected sound that reaches the sound receiving point R2 from the sound source S1 via the two divided surfaces 4 is performed in the same manner as the primary reflected sound. Calculate by the method. For example, in FIG. 2, in the case of a sound that reaches the sound receiving point R2 from the sound source S1 via the i-side 4 and the j-side 4, the view factor is Fsi × Fij × FjR, and this and the i-side 4 and the j-side 4 respectively. By multiplying the sound absorption coefficients αi and αj of the sound volume to the total sound volume emitted from the sound source S1, the sound volume P2 reaching the sound receiving point R2 is obtained. The arrival time is obtained by adding up the distances between the sound source and the i surface, between the i surface and the j surface, and between the j surface and the sound receiving point, and dividing by the speed of sound. k-plane 4
Similarly, the volume P3 and the arrival time of the third-order reflected sound passing through are also calculated.

FIG. 3 is a schematic diagram showing the high-order reflected sound analysis step of B. As shown in FIG. 3, after the low-order reflected sound analysis is performed in the step A (steps S11 to S14 in FIG. 1), the sound energy reaching each split surface corresponds to the time when the split surface is reached. Then, each arrival time is divided into m groups at a preset time Δt interval, and the sound energies divided into the same group are added together. Where △
The smaller t is, the more accurate calculation can be performed. afterwards,
The sound energy collected for each Δt is used as a new sound source in the next reflection, and the sound energy of a certain group is emitted from all n patches at the time corresponding to the group and calculated in the same manner as the low-order reflected sound calculation. , The sound reaching the predetermined sound receiving point and the sound energy reaching the next division surface are calculated (step S15 in FIG. 1).

The above process is performed for all m groups, and the calculation is terminated as soon as the volume reaching all the divided surfaces becomes equal to or lower than a preset specified value (step S16 in FIG. 1). With this method, conventionally, the reflected sound ray increased exponentially every time the sound ray was reflected on the wall surface, but the reflected sound ray is n (total number of divisions) × m (all groups) even if reflection is repeated. The number is constant, so that it is possible to calculate the response up to higher-order reflected sounds.

FIG. 4 is a flow chart showing the processing of the sound environment analysis method of the second embodiment according to the present invention. In FIG. 4, first, the reverberation time is calculated from the volume of the analysis space, the surface area, and the wall average sound absorption coefficient using the Eyring's prediction formula, and the preset time is set to the calculated reverberation time in order to improve the calculation accuracy. The added time is newly set as the set time (step S21 in FIG. 4). The process A is an initial reflected sound calculation process (steps S22 to S25 in FIG. 4), and the process B is the same process as that of the first embodiment until the set time described above is reached ( Step S2 in FIG.
6), the process C is performed after the calculation time exceeds the set time (see FIG.
It is a high-order reflected sound analysis step of step B transition to step S27). Since the steps A and B are the same process steps as the first embodiment according to the present invention, the description of the operation will be omitted.

Next, in the step C, the same calculation as in the step B is performed by using the time interval Δt ′ that is longer than the sound energy grouping time Δt used in the step B (FIG. 4).
Step S28). After that, the calculation is finished as soon as the volume reaching all the divided surfaces becomes equal to or lower than the preset specified value (step S29 in FIG. 4). In this method, the accuracy of calculation is improved by making the analysis time interval Δt as small as possible in the process B, which greatly affects the slope of the time-series information of the sound reaching the sound receiving point, with the set time as a delimiter. And the slope of the time series information is not so affected, but in the C process which is indispensable for obtaining the absolute value of the time series information,
By advancing the processing using Δt ′ which is larger than t, it is possible to improve the calculation speed while maintaining excellent calculation accuracy, and in Δt, higher-order reflection which was impossible due to the memory of the computer. It is possible to calculate up to the sound.

FIG. 5 shows the sound at a predetermined sound receiving point obtained in the first or second embodiment of the present invention in the third embodiment of the present invention (the flowchart showing the flow of processing is omitted). It shows how to add the residual sound energy of each divided surface to the time series information. In FIG. 5, by the first or second method of the present invention, response calculation is performed from time ta to time tb at which sound energy reaches a predetermined sound receiving point, and time series information at the sound receiving point is obtained ( (Thin line in FIG. 5). Next, after tb, the sound energy remaining on each divided surface is added regardless of the arrival time (the portion surrounded by the diagonal lines in FIG. 5). After that, the total sound energy is added to the time series information from ta to tb (Fig. 5).
, And the echo time pattern at the sound receiving point is obtained. Finally, the reverberation characteristic can be obtained by integrating the echo time pattern from tb to ta. With this method, reverberation time with extremely high accuracy can be obtained as compared with the reverberation time conventionally calculated using Erying's prediction formula or the like.

FIG. 6 shows an embodiment according to the present invention.
It is a schematic diagram of the space model used for calculation. As shown in the figure, an omnidirectional sound source 1 and a sound receiving point R2 are installed in a cuboid room (analysis space) 30 of about 5.4 m × 3.6 m × 2.25 m and having a size of 12 tatami mats. In addition to windows and doors, a sound absorbing material and a reflecting material are attached to each wall surface of the room 30, and the average sound absorbing rate is about 0.26 from the ratio of the area occupied by each member in the entire room and the sound absorbing rate. Becomes When the reverberation time is calculated from the average sound absorption coefficient and the volume and surface area using the Erying equation, it is about 280 msec. The measured reverberation time is about 340 msec, and the predicted value has an error of about 18% compared to the measured value. This model is divided into about 400 planes, and the time interval Δt until the set time is 0.1 mse.
c, the subsequent time interval Δt ′ is set to 5 msec, and the time between these divisions is 30% of the value obtained by the Erying equation.
The response was calculated with an additional 360 msec.

As a result, 100th reflection, 500 msec
It is possible to calculate the response up to, and the reverberation characteristic as shown in FIG. 7 was obtained from the echo time pattern obtained by this calculation. Compared with the energy of the direct sound that directly reaches the sound receiving point from the sound source, the time until it is attenuated by -60 dB, that is, the reverberation time is about 310 msec from FIG. 7, and the error from the actual measurement value is about 9%, which is very accurate. Good value was obtained.

As described above, it is possible to calculate the response up to the higher-order reflected sound, which is impossible with the conventional method, and it is possible to calculate the echo time pattern at a predetermined sound receiving point for a long time. Further, after the high-order reflected sound analysis, the residual energy is added to the echo time pattern and integrated, whereby a highly accurate reverberation characteristic can be obtained. Furthermore,
By using the predicted value obtained from the Eyring's equation or the like for controlling the analysis time interval width, it is possible to speed up the calculation while maintaining excellent calculation accuracy. As a result, highly accurate reverberation time prediction is possible. Therefore, it is very useful for designing a listening room or the like that requires high-performance acoustic characteristics over a wide range of sounds including the low range.

In the above embodiment, the incident sound is accumulated at a predetermined time interval, and the number of reflections until the accumulated sound is reflected at one time, that is, the low-order reflected sound is calculated. I tried to perform up to the third order, but not limited to this,
The number of times may be 0, 1, or 2 as long as the analysis speed can be secured, and may be 4 or the like as long as the calculation capability allows.

Further, in the above embodiment, the sound energy is used as the prescribed value which is the final judgment for the termination of the calculation of the reflected sound. You may use. Further, in the above embodiment, the number of divisions of the surface of the analysis space in the specific example is about 400, but the number of divisions is not limited to this, and may be changed according to the size of the analysis space, other conditions, and the like. It goes without saying that it is good.

Further, in the above embodiment, the analysis process is realized by software using a computer, but instead of this, hardware having the same function may be used. Next, a fourth embodiment of the present invention will be described. The simulated sound environment experience device according to the fourth exemplary embodiment of the present invention calculates acoustic characteristics using a virtual space that models a real room or the like, and reproduces realistic three-dimensional sound using the acoustic characteristics. It is a device for outputting. Then, the experiencer can experience the state of listening to the sound in the actual room by listening to the sound reproduced by the sound environment simulation experience device.

As shown in FIG. 8, the sound environment simulation experience device comprises a sound environment analysis unit 1, a sound field reproduction unit 2, and an output unit 3. The sound environment analysis unit 1 analyzes the acoustic characteristics of the target space by simulation and calculates an echo time pattern. The sound field reproducing unit 2 calculates impulse response data using the echo time pattern, and reproduces the non-reverberant sound of the sound to be experienced into a continuous sound signal in real time using the impulse response data.

The output unit 3 uses headphones or VR (Vi
rtual Reality) device, which converts the continuous sound signal generated by the sound field reproducing unit 2 into a sound and outputs the sound. The general operation of the sound environment simulation experience device having the above configuration will be described below. The device can reproduce changes in how the sound is heard when a person stands, sits, walks, or turns his face indoors. And
Specifically, in the sound environment analysis unit 1, each position in the room,
The echo time pattern for each direction and each frequency of the sound source is obtained. Then, the sound field reproducing unit 2 calculates the impulse response using the echo time pattern. Further, the source sound is processed using the impulse response to generate a reproduced sound signal having a stereoscopic effect, and the reproduced sound signal is output to the output unit 3.

The structure and operation of each unit will be described below. First, the detailed configuration of the sound environment analysis unit 1 is shown in FIG. The sound environment analysis unit 1 includes a room information input unit 11, a sound source information input unit 12, a sound receiving point information input unit 13, an analysis plane division unit 14, an echo time pattern calculation unit 15, and a memory 18. , And the echo time pattern calculator 15
Includes a time-series information calculation unit 16 and a data conversion unit 17.

Operation of the sound environment analysis unit 1 FIGS. 9 to 1
This will be described with reference to FIG. This sound environment analysis unit 1 is (1)
For each arbitrary position in the analysis space, (2) For each incident direction of the sound incident on the sound receiving point set in the analysis space,
(3) An echo time pattern is calculated for each sound source having a plurality of predetermined frequency bands.

First, the room information input unit 11 inputs room model information such as the shape, size, and sound absorption coefficient of the space to be analyzed (step S100). Next, the position information of the sound receiving point (human head) in the room is input from the sound receiving point input unit 13 (step S110). Furthermore, the incident direction of the sound to the sound receiving point is input from the sound receiving point input unit 13 (step S120).

Further, sound source information such as the frequency and volume of the sound source is input from the sound source information input section 12 (step S13).
0). Further, the analysis surface dividing unit 14 divides a solid surface such as a wall forming the analysis target space into several surfaces, and adds reference numbers to the respective divided surfaces. The divided surface is set in a rectangular area whose one side is shorter than the wavelength of the sound generated from the sound source. In addition, the sound absorption coefficient of each wall is set using the sound absorption coefficient of the wall input from the room information input unit 11. The sound receiving point is set to, for example, a dodecahedron that approximates the shape of the human head (step S140).

An example of the analysis space model set by the above processing is shown in FIG. In a certain analysis processing example, the division plane is set by dividing one wall surface into about 400 pieces. Further, when the frequency of the sound source is 500 Hz, the length of the short side of the dividing surface is set to 45 cm. The sound absorption coefficient αi is assigned to the dividing surface i, and the sound absorption coefficient αj is assigned to the dividing surface j.
Is added.

In addition, the sound absorption coefficient data of the wall material is generally obtained for each octave band in many cases. Therefore, the frequency of the sound source is set and input for each 1/8 octave band, for example. When the analysis space model is set, the calculation process of the echo time pattern is started (step S200).

First, the time-series information calculation unit 16 calculates a form factor for each divided surface (step S210). The view factor is a combination of a sound source and each split surface on which the sound emitted from the sound source directly arrives, a combination of one split surface and another split surface on which the sound reflected from the split surface arrives, the split surface and the split surface. It is calculated for the combination with the sound receiving point where the sound reflected from the surface reaches. These form factor calculation methods are used for radiation calculation in a thermal environment, lighting calculation, etc.,
It is based on the radiant heat ray tracing method. (1) Form factor F for the combination of the sound source S and the division plane i
SC First, suppose a plurality of sound rays, for example, 100,000 sound rays that are uniformly radiated from the sound source S in all directions. Then, a division plane reached by each sound ray is searched for along the direction in which each sound ray is emitted. In the analysis space model 10 shown in FIG. 12, the sound ray SRsk reaches the dividing plane k, and the sound ray SRsi1,
SRsi2 reaches the division plane i.

Next, the volume P1 of the arrival sound reaching the division plane from the sound source is calculated, and the ratio P1 / Ps of the volume P1 to the volume Ps of the sound source is calculated. This volume ratio is obtained as the form factor FSC for the divided surface of the sound source S. This calculation is performed for all the divided planes, and the form factor for all the divided planes is obtained from the sound source S. (2) Form factor FCC between each division plane Using the sound reaching the division plane from the sound source as a new sound source, the arrival volume of the sound reaching another division plane is calculated using the same method as in (1) above. , Find the view factor Fcc between the two divided surfaces. (3) Form factor F for the combination of the divided surface and the sound receiving point R
CR Further, the form factor FCR for each sound receiving point is similarly obtained from each divided surface.

By these calculations, all form factors necessary for calculating the reached volume time series information at the sound receiving point can be obtained. When the form factor is obtained, time-series data of the volume of the sound reaching the sound receiving point from the sound source is calculated. The sound emitted from the sound source is classified into a direct sound that directly reaches the sound receiving point and a reflected sound that reaches the sound receiving point after being reflected by the dividing surface.

First, the reached sound volume of the direct sound is obtained.
In the analysis space model 10, the sound emitted along the sound ray SRSR corresponds to the direct sound. The volume Pr at the sound receiving point R is obtained based on the inverse square law using the distance between the sound source S and the sound receiving point R. At the same time, the arrival time TR depends on the sound source S
It is obtained by dividing the distance between the sound receiving point R and the sound receiving point R (step S220).

Then, the volume of the reflected sound is calculated (step S230). In the calculation of the volume of the reflected sound for each sound ray, the number of reflected sound rays to be calculated is multiplied by the number of division planes each time the number of reflections increases, and the calculation increases as the number of geometrical increases increases. Limited by memory capacity for storing results. Therefore, here, the reflected sound of the low order and the reflected sound of the high order are distinguished from each other, and the volume is calculated by different methods. In the analysis space model 10, the number of reflections of 3 times or less is defined as a low next time, and the number of reflections of 4 times or more is defined as a high next time.

First, the sound volume reached when the sound radiated from the sound source S reaches the sound receiving point R after being reflected by the three or less divided surfaces is calculated. These reached volumes are calculated using the respective form factors obtained in the previous step S210. (Primary reflected sound) In the analysis space model 10, the primary reflected sound (SRs) from the sound source S to the sound receiving point R via the divided surface k.
k, SRkr) is calculated by using the form factor FsC1 between the sound source S and the dividing surface k and the form factor F between the dividing surface k and the sound receiving point R.
C1R and the reflectance αk given to the dividing surface k are used. Then, the volume PR2 at the sound receiving point R is obtained by the following equation.

PR2 = FsC1 × FC1R × (1-αk) × Ps Further, similar to the direct sound, the distances between the sound source S and the dividing surface k, and between the dividing surface k surface and the sound receiving point R are obtained, and these distances are calculated. The arrival time T2 of the primary reflected sound is calculated by dividing the sum by the speed of sound. The above calculation is performed for all primary reflected sounds. (Secondary reflected sound) In the analysis space model 10, the volume Pr2 at the sound receiving point R of the second reflected sound that reaches the sound receiving point R from the sound source S via the two divided surfaces i and j is as follows. i
And Fc between the division plane i and j, and the form factor F between the division plane j and the sound receiving point R.
C2R and the reflectances αi, α given to the division planes i, k
It is calculated by the following equation using k.

Pr2 = Fsc × FCC2 × FC2R × (1-αi) × (1−αk) × Ps Further, the arrival time T3 of the secondary reflected sound is determined by the sound source S and the division plane i.
, The division plane i and the division plane j, and the division plane j and the sound receiving point R are added together and divided by the sound velocity. (Third Reflected Sound) Furthermore, for the third reflected sound, the volume and arrival time of the sound receiving point are obtained using the same method as the above method.

Next, for the reflected sound calculation after the fourth-order reflected sound, the high-order reflected sound calculation method is applied. Below, FIG.
Will be described with reference to. (High-Next Reflected Sound Calculation) First, at the time point when the previous third-order reflected sound calculation is completed, in FIG.
Time series data of the arrival sound as shown in (a) is calculated. Therefore, the time series data of the reached sound volume is grouped at a predetermined time interval Δt. As an example, this Δt is set to 0.
Specified in 1 millisecond.

Next, the grouped volume levels are summed up for each Δt to generate added volume time series data as shown in FIG. 14 (b). Then, for the next reflection calculation, the division plane having this added volume time series data is set as a new sound source. In this new sound source, the n sound rays that actually reach during Δt are replaced by one sound ray. The replaced sound ray has a volume equal to the sum of the sound volumes of the n sound rays.

Furthermore, the sound rays radiated from each divided surface to the other divided surface and the sound receiving point are set, and the arrival of the other divided surface and the sound receiving point is performed by the same method as in the case of the low-order reflection. Calculate volume and arrival time. Such a calculation is repeated until the volume held by all the divided surfaces becomes a specified value, for example, 1/10000 or less of the total emission volume Ps from the sound source (step S250).

In the meantime, at the sound receiving point R, a large amount of sound arrives from the sound source S directly or through a plurality of divided surfaces.
This state is schematically shown in FIG. The sound rays that have arrived are grouped using a predetermined time interval (for example, Δt used in the high-order reflected sound calculation) in the same manner as in the high-order reflected sound calculation on the divided surface. Then, when a plurality of sound rays arrives within each group time Δt, these numbers are counted and the sound volumes of the plurality of sound rays are added (step S260).

In step S250, when the remaining volume of each divided surface becomes equal to or less than the predetermined value, all the remaining volume of each divided surface is added, and the time series information of the volume at the sound receiving point obtained in the above step is added. To (step S260). Through the above process, time series data of the reached sound volume at the sound receiving point as shown in FIG. 15B is calculated.

Next, the data conversion unit 17 performs the reached sound volume averaging processing and the data interpolation processing to generate an echo time pattern from the time series data of the sound volume grouped. First, the data conversion unit 17 averages the grouped volume data shown in FIG. 15B to calculate the averaged volume data shown in FIG. 15C. The averaging is performed by dividing the sound volume reached by the sound receiving points that are grouped by the number of sound rays that have reached within the time. For example, it is received within the time of the third group in FIG. 15B (since Δt = 0.1 msec, the third group is between 0.2 and 0.3 msec). Letting Pm be the total volume reaching the sound point, and N (= 3) the number of sound rays that have reached this time, one sound ray data SRr having the volume Pm before conversion is the number of sound rays N that has reached. By dividing, the average sound ray data SRa having the volume of Pm / N is calculated. (Step S27
0).

Next, the data conversion section 17 performs interpolation processing of the reached sound volume. As one example of the interpolation method, as shown in FIG. 15C, first, the average sound ray data obtained in the above step is assumed at the midpoint of a certain group time interval.
Then, the ends of this average sound ray data are connected by a straight line to perform a linear approximation. Then, using this approximation curve, N sound ray data that have arrived within each time interval are obtained. By this interpolation operation, a linearly approximated echo time pattern as shown in FIG. 15C is created (step S28).
0). Note that this interpolation is not limited to linear interpolation. For example, the interpolation may be performed using curve approximation, or the interpolation processing may be performed by performing weighting considering the arrival time of each sound ray grouped.

The time interval Δt used for grouping is
Is most suitable to use a value corresponding to the sampling frequency of an audio device such as a CD. Specifically, in the case of a CD, Δt = 22.7 microseconds corresponding to 44.1 KHz. The echo time pattern calculated by the above processing is stored in the memory 18.

Next, the configurations of the sound field reproducing section 2 and the output section 3 will be described. As shown in FIG. 16, the sound field reproducing unit 2 includes an impulse response calculation unit 21 that reads the echo time pattern obtained by the sound environment analysis unit 1 from the memory 18 and calculates an impulse response, and an output unit 3. The position information input unit 22 that reads the position information of the experience person in the room detected by the position information detection unit 31 of the VR device, the head-related transfer function storage unit 23 that stores the head-related transfer function obtained in advance, and the audition A dry source input section 24 for inputting a dry source (sound recorded in an anechoic chamber) of
Impulse response data and HRTF data for dry source data are converted to DSP (Digital Signal Processo).
r) and a reproduction sound generation unit 25 that generates a reproduction signal of continuous sound by performing a convolution operation.

The impulse response calculation unit 21 is further provided in FIG.
7, each position in the room, each incident direction,
An echo time pattern input unit 41 for reading echo time pattern data for each band and a convolution operation unit 42 for performing a convolution operation of a bandpass filter for each band.
An impulse response calculation unit 43 that calculates an impulse response by combining responses of respective bands, an impulse response output unit 44 that outputs an impulse response for each position / incident direction, and a memory 4 that stores impulse response data.
5 and.

The output unit 3 is also provided with an audio output device 32 such as a headphone that outputs the reproduced sound to the ears of the experience person based on the reproduced signal of the continuous sound output from the reproduced sound generating unit 25. The position information detection unit 31 is, for example, VPL.
Research Inc. and POLHEMUS
Inc. RB2 system (Reality Build
t for 2 :: tradename) EyePhone system. The EyePhone measures the position and orientation of the head using a sensor attached to the top of the user's head, and outputs the measured position information to the sound field reproducing unit 2 side in real time.

Next, the operations of the sound field reproducing section 2 and the output section 3 will be described. First, the impulse response calculation unit 21
The impulse response is calculated using the echo time pattern calculated by the echo time pattern calculation unit 15. The impulse response data is calculated for each position in the room and for the head direction (sound incident direction). Referring to FIG.
First, the echo time pattern data is sequentially read from the memory 18 (step S300). As described above, the echo time pattern is regulated by the acoustic physical property data in the room and is output for each 1/8 octave band, for example. Also, the echo time pattern is for each position in the room,
It is read for each incident direction and each band.

Next, the response for each frequency band is calculated by convolving the bandpass filter with the echo time pattern for each band. Furthermore, the responses obtained for each band are added. By this calculation, impulse response data for each position in the room and for each incident direction is calculated (step S320). The calculated impulse response data is stored in the memory 45, for example.

When the impulse response data is obtained, the process for reproducing the sound field is started. With reference to FIG. 19, the sound field reproducing unit 2 outputs the position information in the room and the head direction information detected by the position information detecting unit 31 of the VR device to the position information input unit 22.
Is read in real time from (step S400). The impulse response input unit 41 reads, from the impulse response calculation unit 3 (memory 45), impulse response data corresponding to the position information (room position, incident direction) read from the position information input unit 42 (step S410).

Further, the head related transfer function data is read from the head related transfer function storage section 43 for the head direction of the human at a predetermined position (step S420). here,
The head-related transfer function represents a state in which the sound incident from the surroundings toward the head changes depending on conditions such as the shape of the human head and the position of the ear, and is calculated in advance and stored in the head-related transfer function storage unit. It is stored in 43.

The reproduced sound generator 45 receives a dry source such as music to be reproduced from the dry source input unit 24,
Using the information such as the position of the experiencer and the orientation of the head provided from the position information input unit 22, the impulse response data and the head related transfer function data corresponding to these pieces of information are serially connected to the DSP.
Is used to perform a convolution operation (step S430). A continuous sound signal is generated by this calculation.

The generated signal is output to the audio output device 52 in real time (step S440). Through the above operation, the experience person can hear the stereophonic sound as if he / she were actually in the room by using the sound output device 52. Further, the configuration of the sound environment simulation experience device according to the fifth embodiment is shown in FIG. In the sound environment simulation experience device according to the fifth embodiment, the sound environment analysis unit 1 includes an echo time pattern calculation unit 15 and an impulse response calculation unit 21.
Then, the impulse response data of the analysis target space is calculated in advance for each position in the room using a dedicated computing server and stored in the memory.

Also, the sound field reproducing section 2 reads the corresponding impulse response data and head related transfer function from the memory based on the position information of the experiencer given in real time, and reproduces the stereophonic data from the dry source in real time. To do. Further, the configuration of the sound environment simulation experience device according to the sixth embodiment is shown in FIG. The combined response data calculation unit 26 in this device convolves the head related transfer function corresponding to each head direction, that is, the head swing direction, and the impulse response data corresponding to the incident direction in which sound enters the head side. , The combined response data in which the impulse response data and the head related transfer function are combined is calculated in advance for each sound incident method.

Then, the reproduced sound generator 24 folds the combined response data for each head direction into a dry source in real time on the DSP. With this configuration, the impulse response data and the head related transfer function data for each position and incident direction in the room are inversely proportional to the number of head directions to be handled, as compared with the case where the convolution calculation is performed by the DSP in series. Thus, the number of DSPs used can be reduced.

Further, the following various modifications can be applied to the above embodiment. The simulated sound environment experience devices according to the fourth, fifth, and sixth embodiments are configured to reproduce changes in sound when an experienceer stands, sits, walks around, or changes the direction of the face. . However, it may be configured so that at least one of these states can be reproduced. For example, (1) When the experienced person does not move The impulse response data may be calculated only for the sound receiving point fixed at one point in the room. Further, the position information detection unit 31 and the position information input unit 22 for detecting the position information of the experience person are unnecessary.

(2) When the experienced person changes only the face orientation while sitting or standing The impulse response data may be calculated only for the sound receiving point fixed at one point in the room, and the position The information detection unit 31 only needs to detect the rotation angle information of the face. (3) When an experienced person moves while standing Since the ear height of a human walking through can be specified,
The data at each position in the room can be basically obtained as two-dimensional data.

Further, regarding the present invention, it is possible to operate even if there are a plurality of sound sources, and in principle, impulse response data for each incident direction to the sound receiving point is added even if the sound sources move. Only. Further, although the apparatus according to the above-described embodiment is targeted for the indoor space, it is possible to calculate the echo time pattern by computer simulation even in the open space. Therefore, it is possible to apply the sound environment simulation experience device of the present invention to an open space by using such an echo time pattern calculation method.

As described above, according to the present invention, it is possible to calculate the multiple reflections of the sound in the target space, and it is possible to accurately predict the reverberation time from the time series information of the calculated volume. Further, the calculation of the echo time pattern to the conversion to the impulse response data can be performed for any position and the neck direction of the person. This gives V
It became possible to reproduce a realistic sound field using the R device.

As a result, the efficiency of design using the sound environment simulation experience device of the present invention or the support of the proposal-type sales can be effectively performed.

[0097]

According to the sound environment simulation experience device according to the present invention, the impulse response for each incident direction to the sound receiving point at any position in the room is transmitted to the head for each incident direction. Along with the function, by convolving the dry source of the sound source based on the position information in the room and the head direction information at the sound receiving point, sound field reproduction in any direction at any position in the room is performed, and indoor walkthrough And a playback sound field synchronized with the swinging motion is obtained, so when the head is shaken at the indoor walkthrough and at any position in the room,
It is possible to promote the linear efficiency of a stereoscopic sound field that is more realistic and more realistic, and to propose a business with a realistic simulated experience.

Further, in the inventions of claims 8 to 19, when the reflected sound volume at each divided surface is calculated, after the reflection by the divided surface exceeds a predetermined number, the incident sound is given a predetermined value. Since the sound is accumulated at time intervals and then the accumulated sound is reflected at once, it is possible to calculate up to higher-order reflected sounds and improve the accuracy of sound environment analysis.

[Brief description of drawings]

FIG. 1 is a flowchart showing a processing flow of a sound environment analysis method according to a first embodiment of the present invention.

FIG. 2 is a reflected sound ray diagram showing a sound ray radiated from a sound source to reach a sound receiving point.

FIG. 3 is a schematic diagram showing high-order reflected sound calculation in the first embodiment of the present invention.

FIG. 4 is a flowchart showing a processing flow of a sound environment analysis method according to a second embodiment of the present invention.

FIG. 5 is a diagram showing a process of adding remaining sound energy of a division surface to time series information according to a third embodiment of the present invention.

FIG. 6 is a schematic diagram of an analytical space model used for calculation of a sound environment according to the first to third embodiments of the present invention.

FIG. 7 is a diagram showing calculation results of reverberation characteristics in the example of the present invention.

FIG. 8 is a block diagram showing a schematic overall configuration of a sound environment simulation experience device according to a fourth embodiment of the present invention.

FIG. 9 is a block diagram showing a configuration of a sound environment analysis unit according to the first embodiment of the present invention.

10 is a flowchart showing an operation of the sound environment analysis unit shown in FIG.

11 is a sub-flowchart showing an operation of calculating an echo time pattern in the flowchart of FIG.

FIG. 12 is a schematic diagram showing an example of an analysis space model used in the echo time pattern calculation unit 15.

FIG. 13 is a schematic diagram showing an example of an analysis space model used in the echo time pattern calculation unit 15.

FIG. 14 is a time-series data diagram showing an example of time-series data of the reached sound volume on the division surface of the analysis space model,
(A) shows the actual reached volume, and (b) shows the grouped reached volume.

FIG. 15 is a time-series data diagram conceptually showing time-series data of arrival sound volume at a sound receiving point in an analysis space model, where (a) shows actual arrival sound volume, and (b) shows a group. 11C shows the converted arrival sound volume, and (c) shows time-series data (echo time pattern) of the interpolated arrival sound volume.

16 is a block diagram showing a detailed configuration of a sound field reproducing unit 2 and an output unit 3 of the sound environment simulation experience device shown in FIG.

17 is a block diagram showing a detailed configuration of an impulse response calculation unit 21 of the sound field reproduction unit 2 shown in FIG.

18 is a flowchart showing the operation of the impulse response calculation unit 21 shown in FIG.

19 is a flowchart showing an operation of the sound field reproducing unit 2 shown in FIG.

FIG. 20 is a block diagram showing the structure of a sound environment simulation experience device according to a fifth embodiment of the present invention.

FIG. 21 is a block diagram showing the structure of a sound environment simulation experience device according to a sixth embodiment of the present invention.

[Explanation of symbols]

 1 sound environment analysis unit 2 sound field reproduction unit 3 output unit 11 room information input unit 12 sound source information input unit 13 sound receiving point information input unit 14 analysis surface division unit 15 echo time pattern calculation unit 16 time series information calculation unit 17 data conversion Part 21 Impulse response calculation part 22 Position information input part 23 Head related transfer function storage part 24 Dry source input part 25 Playback sound generation part 31 Position information detection part 32 Voice output device 41 Echo time pattern input part 42 Convolution operation part 43 Impulse response Calculation unit 44 Impulse response output unit

 ─────────────────────────────────────────────────── ─── Continuation of front page (72) Inventor Hisashi Kodama 1006 Kadoma, Kadoma City, Osaka Prefecture Matsushita Electric Industrial Co., Ltd.

Claims (19)

[Claims] 1. An echo time pattern calculation means for calculating an echo time pattern of the analysis target space by a geometric acoustic method using a condition relating to the analysis target space, and the echo calculated by the echo time pattern calculation means. A response for each frequency band is calculated using a time pattern, and an impulse response calculation means for calculating an impulse response by combining the responses, and an impulse response calculation means for anechoic sound source data are calculated by the impulse response calculation means. It is characterized by further comprising: reproduced sound generating means for convolving the impulse response to generate reproduced sound data; and output means for converting the reproduced sound data generated by the reproduced sound generating means into voice and outputting the sound. , Sound environment simulation experience device. 2. The echo time pattern calculation means,
Further, the analysis condition data such as the shape of the analysis target space and the physical property data of the wall surface is received, and the divided analysis surface is set by dividing the wall surface into predetermined sections and the sound source and the sound receiving point are set in the analysis target space. And an analysis data setting means to be set at an arbitrary position, and when calculating time-series data of the volume of a sound reaching a sound receiving point of a radiated sound emitted from a sound source set at an arbitrary position in the analysis target space. And a data conversion unit for performing a predetermined interpolation operation on the time-series data of the arrival sound volume calculated by the time-series data calculation unit to convert the time-series data into an echo time pattern. The sound environment simulation experience device according to claim 1. 3. The sound field reproducing means further includes dry source input means for inputting sound source data of anechoic sound, head related transfer function storing means for storing head related transfer function data, and the dry source. Reproduction sound data is generated by convolving the sound source data input from the input means with the impulse response calculated by the impulse response calculation means and the head related transfer function read from the head related transfer function storage means. The simulated sound experience device according to claim 2, further comprising a reproduced sound generating unit. 4. The sound field reproducing means further comprises position information detecting means for detecting information of at least one of the position of the head and the direction of the head of the experiencer in the analysis target space. 4. The reproduced sound generating means extracts the impulse response and the head related transfer function corresponding to the position information detected by the position information detecting means and convolves the extracted sound source data with the sound source data. Sound environment simulation experience device. 5. The simulated acoustic environment device further comprises head-related transfer function storage means for storing a head-related transfer function, impulse response obtained by the impulse response calculation means, and head-related transfer function storage means. A synthetic response generating means for calculating a synthetic response for each head direction by convolving the read head related transfer function; and convolving the synthetic response calculated by the synthetic response generating means with anechoic sound source data. The simulated sound experience device according to claim 1, further comprising: the reproduced sound generating means for generating reproduced sound data by the method. 6. A sound field reproducing means for generating reproduced sound data by convoluting impulse response data with anechoic sound source data, and converting the reproduced sound data generated by the sound field reproducing means into voice. A simulated sound environment experience device, comprising: an output unit for outputting. 7. The sound field reproducing means further includes a dry source input means for inputting anechoic sound source data, a head related transfer function storing means for storing head related transfer function data, and the analysis target. Position information detecting means for detecting at least one of the position of the head and the orientation of the head of the experiencer in space, and an impulse response and head transmission corresponding to the position information detected by the position information detecting means. 7. The simulated sound experience device according to claim 6, further comprising: reproduced sound generating means for generating reproduced sound data by extracting a function and convoluting the sound source data with the sound source data. 8. Room information input means for inputting room information such as shape and size of an analysis space to be analyzed, sound source information input means for inputting sound source information such as sound source position and sound pressure, and sound receiving point Sound receiving point information input means for inputting sound receiving point information such as the number and position of data, analysis plane dividing means for dividing the planes constituting the analysis space into a predetermined number of planes to be analyzed, and By grouping the sound volume reflected by the sound source to reach the predetermined sound receiving point of the analysis space directly or while reflecting on the divided surface with a predetermined time interval for each arrival time. A time-series information calculating means for obtaining time-series information of the reached sound volume at the sound receiving point, and grouped reached sound volume at the predetermined sound receiving point, each having a number of sound lines reaching within the time interval. Sound field reproduction by averaging for each group Characterized in that a sound field reproduction data converting means for converting the sound field reproduction data to ability, sound environment analysis device. 9. Room information input means for inputting room information such as shape and size of an analysis space to be analyzed, sound source information input means for inputting sound source information such as position of sound source, sound pressure, and sound receiving point Sound receiving point information input means for inputting sound receiving point information such as the number and position of data, analysis plane dividing means for dividing the planes constituting the analysis space into a predetermined number of planes to be analyzed, and By grouping the sound volume reflected by the sound source to reach the predetermined sound receiving point of the analysis space directly or while reflecting on the divided surface with a predetermined time interval for each arrival time. A time-series information calculating means for obtaining time-series information of the reached sound volume at the sound receiving point, and grouped reached sound volume at the predetermined sound receiving point, each having a number of sound lines reaching within the time interval. Sound field reproduction by averaging for each group Sound field reproduction data conversion means for converting the sound field reproduction data to a function, and sound field reproduction data interpolation means for interpolating the averaged volume of each group with respect to the time axis. A characteristic sound environment analysis device. 10. A surface constituting an analysis space to be analyzed is divided into several surfaces, a sound ray is radiated from a predetermined sound source in the analysis space, and the volume reaching each divided surface is calculated. , The time until the sound radiated from the predetermined sound source reaches the surface is obtained, and the reflected sound volume on the surface is calculated based on the incident sound volume. Further, this reflected sound volume is used as a new sound source, In the sound environment analysis method for obtaining the time series information of the sound reaching the predetermined sound receiving point by calculating, when calculating the reflected sound volume on each of the divided surfaces, the reflection by the divided surface exceeds a predetermined number of times. After that, the sound environment analysis method is characterized in that incident sounds are accumulated at a predetermined time interval, and then the accumulated sounds are reflected at once. 11. The sound environment analysis method according to claim 10, wherein when the reflected sound volume on the divided surface becomes equal to or less than a predetermined limit value, the reflected sound is not reflected from the surface. 12. The predetermined time interval is different before and after the predetermined set time calculated based on the predicted value of the reverberation time of the analysis space, and the subsequent time interval is longer than the previous time interval. The sound environment analysis method according to claim 10 or 11, characterized in that. 13. When all the reflected sound volumes are equal to or lower than the limit value, a part or all of the sound energy remaining on each of the divided surfaces is added and added to the time series information of the sound at the sound receiving point. 13. The sound environment analysis method according to claim 11 or 12, wherein the time series information is integrated in order from the earliest arrival time to the sound receiving point to the earliest arrival time. 14. An analysis condition such as a shape of a space to be analyzed, physical property data of a wall surface, a divided analysis surface in which the wall surface is divided into predetermined sections, a condition setting step for setting a sound source and a sound receiving point, Radiation sound emitted from a sound source is time-series data of the volume of arrival sound reaching the sound receiving point set at an arbitrary position in the analysis target space, and the arrival sound is incident on the sound receiving point. A time series data calculation step for each time and for each frequency band of the sound source, and an echo time pattern calculation step for calculating an echo time pattern by performing a predetermined interpolation operation on the time series data of the volume of the reaching sound. A method for analyzing a sound environment, which comprises: 15. The time-series data calculation step includes a step of calculating volume data of a direct sound in which a sound ray emitted from the sound source directly reaches a sound receiving point, and a sound ray emitted from the sound source is subjected to the division analysis. A volume data calculation step of a lower-order reflected sound in which a reflected sound whose number of times of reflection on a surface is equal to or less than a predetermined number reaches the other divided analysis surface and the sound receiving point; A step of calculating volume data of higher-order reflected sound in which a reflected sound having a number of times of reflection on a surface reaches a predetermined number or more, and the sound volume of the sound receiving point is grouped at predetermined time intervals. 15. The sound environment analysis method according to claim 14, further comprising a step of calculating time series data. 16. The volume data calculating step for the high-order reflected sound calculates a new volume by integrating the volume data on the divided analysis plane calculated in the volume data calculating step for the low-order reflected sound at predetermined time intervals. 16. The sound environment analysis method according to claim 15, wherein a sound source is set, and a process of calculating an arrival sound volume to a divided analysis plane to which a new radiated sound emitted from the sound source reaches is repeatedly performed. 17. A step of calculating an echo time pattern for each frequency band using a geometric acoustic method at a sound receiving point set at an arbitrary position in the analysis target space, and each frequency in the echo time pattern. Calculating a response for each frequency band by convolving a band pass filter for each band and synthesizing an impulse response; detecting the position information of the experience person; receiving the dry source of the sound source; Generating reproduced sound data by convolving the dry source with the impulse response and the head related transfer function corresponding to the detected position information; and converting the reproduced sound data into voice and outputting the sound. A sound environment analysis method comprising: 18. The step of detecting the position information comprises:
18. The sound environment analysis method according to claim 17, wherein information on the moving position of the experience person and the orientation of the head is detected as the position information. 19. The step of detecting the position information comprises:
The sound environment analysis method according to claim 17, wherein the orientation of the head of the experiencer is detected as the position information.