TWI640983B - Method for simulating room acoustics effect - Google Patents

Abstract

An indoor acoustic effect simulation method includes: particleizing an acoustic feature into a volumeless sound particle; and when the sound particle hits an obstacle after being emitted from a sound source model, calculating an impact point position by using an indoor environment model, and Taking the surface material information of the obstacle, defining a direction selection area, randomly selecting a deflection direction from the sound direction; when the sound particle enters a sound collection area along the deflection direction, obtaining a listener model according to the deflection direction Relevant computing information, the operation of the sound characteristics of the sound particles, and the sound characteristics suitable for simulating the sound effects of the real world sound, including position, direction of travel, spectrum distribution, sound intensity, bandwidth, There is time, phase and behavior. The listener model can further calculate the reverberation time and the overall sound effect of the sound receiving area according to the tracking result of the sound particles and the loading information of the sound particles according to the indoor environment model. According to this method, a window simulation software with an indoor acoustic effect can be developed, which is easy to operate and can simulate diverse indoor environments by graphics.

Description

Indoor sound effect simulation method

The present invention relates to an acoustic effect simulation method, and more particularly to a method for simulating an indoor acoustic effect by particleizing sound.

Advances in technology have prompted people to pursue better multimedia sound effects, and the sound quality experience in various occasions has also received attention. In addition to the space that emphasizes sound effects such as concert halls or cinemas, the sound quality in lecture halls, classrooms, offices, cars, and even general home spaces has gradually gained attention. Therefore, the design of the building structure and interior space needs to consider and predict the sound quality after completion.

The quality of the sound in the space is affected by the decor and furnishings in the space. More specifically, all kinds of furniture and human body in the space will affect the final listening effect. However, existing methods for predicting the sound quality of indoor spaces, such as Sabine and Eyring formulas, cannot take into account details such as the position and spatial structure of the objects. Moreover, the relevant evaluation tools are mostly designed for large spaces such as concert halls and churches. They are limited to rectangular spaces and only consider the influence of space walls, and their purpose is only to produce reverberation sound effects, and the operation mode needs to have a certain degree. Relevant knowledge and technology.

In view of this, building structure and interior design workers need a sound quality simulation tool that is easier to operate than conventional techniques, can add more environmental impact factors to simulate a more diverse space environment, and its simulation results Can be closer to the actual sound effect.

An object of the present invention is to provide a room acoustic effect simulation method for developing an indoor sound effect window simulation software, which is simple in operation, can simulate a diverse indoor environment by graphics, and can make the simulation result closer to reality. Sound effect.

In order to achieve the above object, the present invention provides a room acoustic effect simulation method, the method comprising: particleizing an acoustic feature to form a volumeless sound particle; defining a sound source model, an indoor environment model, and a listener model, wherein The sound source model is used to emit sound particles, and a plurality of sound particles are arranged on a travel path to form a sound line. The indoor environment model provides volume, surface area and surface material information of an obstacle, and the listener model provides a a sound receiving area; when the sound particles hit the obstacle, the indoor environment model uses a microprocessor to calculate a collision point position, and uses the surface material information of the obstacle to define a direction selection area, the direction selection area includes a plurality of possible directions of deflection, and the possible deflection directions are starting from the collision point; then, a direction of deflection after the sound particle hits the collision point is randomly selected from the possible deflection directions When the sound particles travel in the direction of the deflection and enter the sound receiving area, the listener model takes the direction of the deflection According to the operation information related to the radio area, and based on the operation information, the sound characteristics currently carried by the sound particles are calculated, and the sound characteristics of the sound receiving area can be obtained, or the sound characteristics currently carried by the sound particles can be further transferred by the head correlation. The function and the equal-impedance curve are calculated to obtain human perception information. In addition, the listener model can also use the piggybacking information of the sound particles to calculate the reverberation time and overall sound effect of the radio region. The above sound particles, sound source model, indoor environment model, and listener model can all be displayed in a user interface using digital graphics.

In an embodiment, the indoor environment model provides a virtual space for adding a plurality of objects; and setting a plurality of sound parameters of the objects in the virtual space, including position, shape, size, area, orientation, and surface a material; and setting a temperature value, a humidity value, and an atmospheric pressure value of the virtual space.

In one embodiment, the information on the loading of the sound particles includes sound characteristics such as position, direction of travel, sound intensity, spectral distribution, bandwidth, time of presence, phase, and behavior. The sound source model emits a plurality of sound particles in different directions in a radial direction. In the piggybacking information of each sound particle, the setting of the sound characteristics may be the same or different, and the values of the sound characteristics may be changed with time to simulate the three-dimensional directivity of the speaker.

In an embodiment, the indoor environment model calculates a sound intensity value that is attenuated by the air during the process of moving the sound particle from the sound source model to the collision point according to the location of the collision point and the current position of the sound particle.

In one embodiment, the surface material information includes surface material absorption coefficient, reflection coefficient, refractive index, and scattering coefficient of a plurality of different frequency bands. After calculating the air attenuation, the indoor environment model calculates the sound particle from the collision point by using the surface material absorption coefficient. The sound intensity value carried in an instant, and the scattering coefficient is used to define the direction selection area. The step of defining the direction selection area includes: considering that the sound particle has a part of energy acting on the law of reflection, and another part of the energy acting on the law of refraction to determine a main deflection direction starting from the collision point in the direction selection area; defining a maximum diffusion angle And determining the range of the direction selection area based on the scattering coefficient, the main deflection direction, and the maximum diffusion angle. In addition, the energy of the sound particles before the collision is subtracted from the energy used after the deflection in the main deflection direction to form a divided sound particle, and the above steps of defining the direction selection region are performed. The range of the direction selection area of the divided sound particles is determined.

In an embodiment, the listener model performs a screening process in order to avoid re-sampling the collected sound particles, and the steps include: determining whether the number of deflections of any at least two of the sound particles is equal. If the number of deflections is equal, compare the existence time and the deflection direction of the two sound particles to determine whether the two sound particles are the same sound particle and take the average value of the piggybacked information or discard one of the two sound particles.

In summary, the method of the present invention takes the sound source data into consideration based on the defined sound particles, and makes the sound source model closer to the solid sound source, taking into account air attenuation, surface material absorption and providing randomly selected sound particles. The mechanism of the deflection direction after collision causes the reflected sound particles to collide in the indoor environment, while the refracted sound particles form a new virtual virtual source outside, which can be further used to simulate the degree of influence of the outdoor environment by the indoor sound, so that the entire acoustic effect The model is more appropriate to reflect the real situation. Furthermore, the method of the present invention is suitable for use in a 3D graphics development environment, while simplifying the operation of the developed window simulation software.

The above and other technical contents, features and advantages of the present invention will be apparent from the following detailed description of the preferred embodiments. The directional terms mentioned in the following embodiments, such as upper, lower, left, right, front or rear, etc., are only used to refer to the directions of the accompanying drawings. Therefore, the directional terms are used for illustration only and are not intended to limit the invention.

FIG. 1A is a schematic conceptual diagram of a method for simulating indoor acoustic effects of the present invention. This method particleizes the acoustic characteristics to define a volumeless sound particle P, which is used to simulate the propagation characteristics and energy changes of the sound in space. It should be noted that the sound particles P defined by this method are different from the conventional phonons, and the two cannot be confused. The method also provides a sound source model 100, an indoor environment model 200, and a listener model 300 for calculating and recording energy changes caused by the sound particles P being affected by the sound source, the environment, and the listener on their travel paths. And through the microprocessor to actually achieve the computing functions of the three models 100, 200 and 300.

As shown in FIG. 1A, in an embodiment, in order to achieve the function of digitizing the graphic display, a window simulation software 10 can be designed and a database 20 required for the operation of the window simulation software 10 can be provided. The window simulation software 10 includes a system management module 11 for managing an audio particle management module 12, an object management module 13, a virtual space management module 14, and a graphics module 15. The required database 20 includes an acoustic particle parameter database 21, a sound source database 22, an object database 23, an acoustic parameter database 24, a sensory loudness compensation database 25, and a head related transfer function. The database 26 is associated with a psychoacoustic database 27. The sound particle P, the sound source model 100, the indoor environment model 200, and the listener model 300 can all be displayed in a user interface 400 in the form of a digitized graphic by the drawing module 15.

The sound source model 100 emits a plurality of sound particles P in different directions in a radial direction, wherein each sound particle P has a different piggybacking information according to electrical characteristics, speaker structure and directivity of the corresponding sound source model 100, and These piggybacking information is recorded in these sound particles P. A plurality of sound particles P sequentially emitted by the sound source model 100 and arranged in a straight line on the same traveling path can be regarded as a sound line, and therefore it is suitable to use the sound line tracking method as the basis of the algorithm of the present invention.

The indoor environment model 200 defines a direction selection region 230 in a diffuse concept with a reflection law or a law of refraction, thereby obtaining a deflection direction in which the sound particles P collide with the obstacle 210 and leave the collision point P _C . This direction selection area 230 is a collection area of all possible deflection directions starting from the collision point P _C , and FIG. 1A shows these possible deflection directions in a plurality of broken lines. These deflection directions have the possibility that the incident angle is equal to the reflection angle without obeying the law of reflection, and also has the possibility that the incident angle is equal to the angle of refraction without obeying the law of refraction. Therefore, a deflection direction can be randomly selected from the direction selection area 230 in accordance with the concept of the probability distribution.

In order to establish the overall impulse response of the sound source to the listener, it is necessary to collect the sound particle loading information that the listener will get at a particular location. However, since the sound particle P is a spatial point where there is no volume, it is impossible to ensure that all the sound particles P that the listener should collect at a specific position can pass the specific position. Therefore, the listener model 300 sets a sound receiving area 310 with the specific position as a reference point. Whenever the sound particles P pass through the sound receiving area 310, the listener model 300 reads the current piggybacking information of the sound particles P for processing.

FIG. 1B is a schematic diagram of a single sound particle P transfer path and energy change in FIG. 1A. The sound particle P of one embodiment of the present invention is a pulse energy packet having a propagation characteristic that is linearly advanced and impinges on the obstacle 210. For example, when the sound particle P hits the obstacle 210, part of the energy will change the direction of travel, that is, reflection, in a rebound manner; and the rest of the energy will change the direction of travel, that is, refraction, in a penetrating manner. The piggybacking information of each of the sound particles P includes sound characteristics such as position, traveling direction, sound intensity, spectrum distribution, bandwidth, presence time, phase, and behavior, and these sound characteristics can be stored in the sound particle parameter database 21. The "behavior" in the acoustic characteristics refers to the relationship that the coefficient of random scattering varies with frequency. In the following description, the sound intensity is also simply referred to as "sound strength".

As shown in FIG. 1B, the sound particle P initially carries a pulse signal δ . The sound source model 100 performs Fourier transform on the pulse signal δ to obtain an initial spectral distribution H _p0 . The initial spectral distribution H _p0 through the sound source model 100 is adjusted into a first spectral distribution of H _p1, then through the indoor environment model 200 is adjusted to a second spectral distribution H _p2, and finally through the listener model 300 is adjusted so as to be listened to spectrum Distribution H _p . This process calculates the change after the initial spectral distribution H _{p0 is} affected by the sound source, space, and listener's head. H _p obtained spectral distribution of the pulse response h _p in time domain via the inverse Fourier transform on the basis of the pulse response h _p domain at this time can be calculated human single particles P is supplied to the sound the listener perceived information.

2 is an embodiment of a sound source model 100. The sound source model 100 includes a uniform radiation vector generator 120, a sound particle initialization module 140, a sound source directivity processing module 160, and a speaker database 180. The uniform radiation vector generator 120 imparts a different direction to each of the sound particles P, so that the sound source model 100 can emit a plurality of sound particles P in different directions in a spherical shape. The sound particle initialization module 140 initializes the piggybacking information of each sound particle P. Next, the sound source directivity processing module 160 takes a three-dimensional directivity data D _{source of} a speaker from the speaker database 180, and calculates an initial sound intensity I _p0 of the sound particles P in a certain initial direction D _p0 . The initial sound intensity I _{p0 of the} sound particle P is adjusted to a first sound intensity value I _p1 . In this way, the first sound intensity values I _p1 mounted on the sound particles P in different directions are different, thereby simulating the unevenness of the sound intensity emitted by the physical speakers in each direction. In addition, the sound source model 100 can take the frequency response data H _source in the sound source database 22, such as the speaker database 180, and calculate it with the initial spectrum distribution H _{p0 of the} sound particles P to obtain a first spectrum distribution H _p1 , and The first spectral distribution H _p1 and the first sound intensity value I _{p1 are} mounted on the acoustic particles P to be emitted.

FIG. 3A is an embodiment of an indoor environment model 200. The indoor environment model 200 includes a particle collision information module 220, an air attenuation processing module 240, a surface absorption processing module 260, a double band synthesis module 270, and a scattering processing module 280. The modules inside the dashed box 250 in FIG. 3A are used to perform a spatial reflection and refraction iterative operation flow as shown in FIG. 3B for a single sound particle P.

First, the mounting information of the sound particles P obtained by the particle collision information module 220 after being processed by the sound source model 100 is an iterative initial value (step S210), and is obtained according to the current position P _{p0 of the} sound particle P and the traveling direction D _p0 . The collision point P _C with the obstacle 210 is positioned, and the surface material information of the obstacle 210 is obtained (step S220). The obstacle 210 of the present embodiment is a space wall; however, in other embodiments the obstacle 210 may also be a still life or an animal in a virtual space. The surface material information includes surface material absorption coefficient α _fc , scattering coefficient S , reflection coefficient, and refractive index in a plurality of different frequency bands. In addition, the virtual space management module 14 is included in the embodiment, and can be used to record the environmental parameters, volume, area, shape, and coordinates of each boundary of the virtual space, and set a temperature value and a humidity value of the virtual space. With an atmospheric pressure value. The plurality of virtual space information described above may affect the degree of attenuation and the direction of travel of the plurality of sound particles P as they propagate in the virtual space.

Next, as shown in FIG. 3B, in order to determine whether the energy of the sound particles P is attenuated (step S232), the sound particle management module 12 continues to determine whether the energy of the sound particles P has been during the flight of the sound particles P during the flight path and the collision time. It is attenuated to zero, and it is judged whether or not it stops traveling (step S234); if it is attenuated to zero or stopped, the tracking of the sound particle P is ended (step S290).

If the energy of the sound particle P does not decay to zero and stops, the air attenuation processing module 240 calculates the travel distance and the travel time according to the position of the collision point P _C and the current position P _{p0 of the} sound particle P, and then Calculating the total sound intensity value after the sound particle P travels from the sound source model 100 to the impact collision point P _C , the total sound intensity value after being attenuated by the air, the sound intensity value of each frequency band, and the existence time, according to the previous loading information of the sound particle P The first spectrum distribution H _p1 in the adjustment is adjusted (step S240).

In practice, the sound intensity value after being attenuated by air can be used. To find, where IL ( l ) is the sound intensity, the unit is dB; l represents the travel distance; a is the attenuation coefficient (Attenuation Coefficient), which is defined as the sound pressure level attenuated per kilometer of the sound wave travel, in dB/ Km. The calculation method of the attenuation coefficient a takes into account the influence of nitrogen molecules and oxygen molecules in the air, and is obtained according to the atmospheric pressure, temperature, relative humidity and the frequency of the sound signal in the environment. The attenuation coefficient function a ( f ) according to the frequency is calculated using the set temperature and humidity values during the simulation. When the sound particle P travels by 1 m from the starting point in space, its spectral distribution H ( f ) changes as calculated by the following equation (1). (dB) (1)

Next, the surface absorption processing module 260 calculates the sound intensity values of the respective bands at the moment of leaving the collision point P _C after the impact of the sound particles P using the surface material absorption coefficient α _fc of the plurality of different frequency bands of the surface material, and then passes through the octave band synthesis module. 270 synthesize the total sound intensity values of the sound intensity values I ₆₃ , I ₁₂₅ , I ₂₅₀ , I ₅₀₀ , I _1k , I _2k , I _4k , I _{8k of} each frequency band, and accordingly adjust the spectrum distribution adjusted in step S240 again to obtain The sound particle P leaves the spectrum distribution at the moment of the collision point P _C and updates the piggybacking information of the sound particle P (step S260). After the energy change is calculated in steps S240 and S260, the traveling direction is calculated (step S280). If the sound particle P has the next collision, then jump back to step S220 to calculate the air attenuation between the current collision and the next collision again, and the surface absorption of the next collision point until the last collision, and then calculate the final collision point. The air attenuation between the sound receiving areas 310. In this manner, the indoor environment model 200 obtains the second sound intensity value I _p2 and the second spectrum distribution H _p2 by adjusting the spectrum distribution a plurality of times.

In practice, the spectrum information can be recorded with the octave intensity I _fc , so it must be converted into the spectral distribution H _Room to obtain the spatial impulse response. The first sound intensity value I _p1 transmitted by the sound model 100 is decomposed into eight band sound intensity values I ₆₃ , I ₁₂₅ , I ₂₅₀ , I ₅₀₀ , I _1k , I _2k , I _4k in the indoor environment model 200 . I _8k . The relationship between the first sound intensity value I _p1 and each band sound intensity value I _fc is as shown in the following formula (2). (2)

Next, we use the Octave Filter to decompose the first spectral distribution H _p1 from the sound model 100 into the spectral distribution H _fc of each frequency band. The relationship between the first spectral distribution H _p1 and the spectral distribution H _{fc of} each frequency band is as follows: (3)

Finally, the sound intensity I _fc on each frequency band obtained after the iterative operation is taken as a weight, and the second spectrum distribution carried by the sound particles P after air attenuation (step S240) and surface absorption (step S260) is calculated by the following equation (4). H _p2 : (4)

As step S280, the traveling direction in order to calculate the scattering process module 280 using the current traveling direction D _p0 and the surface normal N can be calculated to give a main deflecting direction D _SR according to the law of reflection. Then, using the deflection processing method shown in FIG. 3C, a main deflection direction _Dp2 and a scattering coefficient S are used to calculate a possible deflection direction _Dp2 as a new traveling direction. If the sound particle P has the next collision and the end condition is not reached, the process _{returns to} step S220 to calculate the deflection direction D _p2N after the next collision again, otherwise the tracking of the sound particle P is ended.

Incidentally, the step S280 of calculating the traveling direction may be simultaneously processed in parallel with the steps S240 and S260 for calculating the energy change, and the traveling direction is not necessarily calculated after the energy change is calculated.

As shown in FIG. 3C, considering that part of the energy of the sound particle P acts on the law of reflection, and another part of the energy acts on the law of refraction, the scattering processing module 280 can use the law of reflection to determine a main deflection direction starting from the collision point P _C . D _SR ; further determining the direction selection area 230 according to the scattering coefficient S and a maximum diffusion angle θ _DR ; the maximum diffusion angle θ _DR represents the main deflection direction D _SR and the marginal direction D _B1 or D _B2 of the direction selection area 230 Angle. In addition, the scatter processing module 280 also divides the energy carried by the pre-collision sound particles P by the energy used to reflect it in the main deflection direction D _SR to be divided into a refracted sound particle, and uses the law of refraction to determine the refracted sound. The main deflection direction D _SR ' of the particles, the above-described step of defining the direction selection region 230 is performed to define a range of direction selection regions 230' that refract the sound particles. The direction selection area 230' also includes a maximum diffusion angle θ _DR ', which represents the angle between the main deflection direction D _SR ' and the marginal direction D _B1 ' or D _B2 '. The range of this direction selection area 230' is applicable to the case where the sound particles P penetrate the obstacle 210. The method of the present invention does not need to limit its order when defining the two-direction selection areas 230 and 230'.

The above-described main deflection direction D _SR can be expressed by the specular reflection angle θ _SR ; the main deflection direction D _SR ' can be expressed by the refraction angle θ _SR '. The range of the specular reflection angle θ _SR (or the refraction angle θ _SR ') is . The range of the scattering coefficient S is . The range of the maximum diffusion angle θ _DR (or θ _DR ') is . Thus, a possible reflection direction angle θ = θ _SR + S * θ _DR can be used to define the range of the direction selection region 230 of the reflection; the same operation can also be used to calculate the possible refraction direction angle θ ' to define the refraction Direction selection area 230' range.

It is worth mentioning that the information of the reflected sound particles and the refracted sound particles will inherit the sound particles P before the split. That is, the energy of the reflected sound particle may be the energy of the sound particle P before the collision multiplied by the reflection coefficient; similarly, the energy of the refracted sound particle may be the energy of the sound particle P before the collision multiplied by the refractive index, but the reflected sound particle The sum of the energy carried by the refracting sound particles will be smaller than the energy of the sound particles P before the collision.

In the present embodiment, since the diffuse distribution of the ideal diffusing surface exhibits a cosine distribution pattern, the probability distribution pattern of all possible deflecting directions in the 230 or 230' in the direction selecting region is set to a cosine distribution or a bell shape. The curve distribution, according to which a deflection direction is selected from the 230 or 230' in the direction selection area as the operation data required for the listener model 300. In other embodiments, the probability distribution pattern may also be, for example, a normal distribution, a Gamma distribution, a Beta distribution, a Poisson distribution, or a Binomial distribution, and is used to randomly obtain a traveling direction after the sound particles are updated within a specified range.

As shown in FIG. 4A, the listener model 300 includes a related psychoacoustic model such as a head related impulse response database 320 and a loudness weighting module 340, and these psychoacoustic models are stored in the psychoacoustic database 27. In addition, the psychoacoustic model can also record about Auditory Masking, Missing Fundamental, Interaural Time Difference, Interaural Level Difference, and binaural masking ( Binaural Masking Level Difference and the like to establish a listener model 300. The head related impulse response database 320 used in this embodiment includes, but is not limited to, a Head Related Impulse Response provided by the Center for Image Processing and Integrated Computing (CIPIC). , HRIR) database. The loudness weighting module 340 is an equal-tone curve drawn according to the standard ISO226 of the international standard organization, and then converts the equal-cord curve into a loudness level curve corresponding to the sound pressure level as an independent variable, and the loudness level curve can be A loudness weight function with different sound pressure levels corresponding to different frequencies is obtained.

Taking the sound particles P that enter the sound receiving area 310 after the collision in the deflecting direction D _p2 as an example, the listener model 300 takes out the deflection direction D _p2 in the current carrying information P _inf of each sound particle P, and then de novo The partial correlation impulse response database 320 obtains a head related impulse response h _head corresponding to the deflection direction D _p2 . The head related impulse response h _{head is} Fourier transformed to obtain a Head Related Transfer Function H _head , and these head related transfer functions H _head can be stored in the head related transfer function database 26. The head-related transfer function H _head and the spectral distribution H _p2 in the current piggybacking information P _inf of the sound particles P are calculated to obtain the spectral distribution H _p of the sound particles P entering the human ear from different directions. This spectral distribution H _p can be used to further calculate the volume of subjective perception of the human ear or to simulate human perception information such as the acoustic effects brought about by real world sounds. In order to obtain these human perception information, the spectral distribution H _p needs to be weighted by the loudness weighting module 340 to obtain the loudness compensated spectral distribution H _pw , and the data required for the weighting operation can be taken from the perceived loudness compensation database 25 . in.

After obtaining the spectral distribution H _{p of} each sound particle P affected by the sound source, environment and listener, it is inverse-Fourier transform to obtain the impulse response h _p in the time domain. Since there is a time difference between each sound particle P starting from the sound source and being collected by the listener, each sound particle P impulse response h _{p must} be delayed, and finally a total impulse response is synthesized.

As shown in FIG. 4B, in order to avoid repeated sampling, the listener model 300 provides a time threshold and an angle threshold for performing a plurality of reflected sound particles P ₁ , P ₂ , P ₃ collected by the sound receiving area 310. Screening process. The step of the screening process includes: determining whether the number of times of deflection of any two reflected sound particles P ₁ , P ₂ is equal; if the number of times of deflection is equal, the time difference between the two reflected sound particles P ₁ , P ₂ and the time threshold Comparing, and comparing the deviation angle difference between the two reflected sound particles P ₁ , P ₂ with the angle threshold; when the difference between the two reflected sound particles P ₁ , P ₂ is less than the time threshold, and the deviation angle difference When it is less than the angle threshold, the average of the information carried by the two reflected sound particles P ₁ , P ₂ is taken, or only one of them is retained, and the other reflected sound particle P ₁ or P _{2 is} discarded. The above description is made by taking the reflected sound particles P ₁ , P ₂ , P ₃ as an example, and the same screening procedure is also applicable to the refracted sound particles P1 ′, P2 ′ collected by the sound receiving region 310 ′ as shown in FIG. 4B . P3'.

According to the indoor acoustic effect simulation method of the present invention, for example, the game engine Unity 3D can be used as a development environment, and a window simulation software 10 having a 3D drawing module 15 and a user interface 400 as shown in FIG. 1A can be constructed, which can be used for cross The platform simulates the indoor acoustic effect. The window emulation software 10 is suitable for execution on a personal computer, a laptop or a handheld device, and is convenient for the user to operate in various situations.

The sound source model 100 of the present invention takes into account the physical characteristics of the speaker including, for example, three-dimensional directivity and the frequency response of the speaker. Accordingly, the developed window simulation software 10 can provide a user to place a virtual sound source anywhere in the virtual space to simulate a single sound source environment, and can also add multiple virtual sound sources to simulate such as a home theater and a car. Multiple sound source environments such as internal space. In addition to the location and number of sound sources, the direction of the virtual sound source, the volume of the sound source and the physical characteristics of the speakers can also be set by the user. In addition, the sound source model 100 can be evaluated and simulated on the subwoofer source after a specific adjustment.

The indoor environment model 200 of the present invention considers characteristics such as air attenuation when the sound travels in space and surface material absorption and surface material scattering at the time of collision. These characteristics imply environmental parameters such as temperature and relative humidity, as well as acoustic characteristics such as position, direction of travel, spectral distribution, sound intensity, bandwidth, duration, phase, and behavior of sound waves in space. Accordingly, the developed window simulation software 10 can provide the user to freely plan the shape and size of the space, and add furniture or characters to any place in the space according to the scene to be simulated. The user can drag the selected space decoration, furniture or character to any position in the virtual space, and use the functions of rotation to adjust the orientation of the object. The object management module 13 can be automatically associated with the acoustic parameter database 24 according to the requirements of the object database 23, and the sound parameters corresponding to different object surfaces can be given according to different frequencies. The acoustic parameter database 24 includes a plurality of acoustic parameters of the objects in the virtual space, such as position, shape, size, area, orientation, and surface material. The spatial shape planned by the window simulation software 10 is not limited to a regular shape having a clear corner relationship. For various surfaces such as space walls, furniture or people, users can set a variety of different surface materials, each surface material can adjust its sound characteristics, such as the degree of reflection of sound, the degree of scattering of sound and the degree of sound transmission Wait. The reverberation time of the virtual space is calculated according to the degree of energy attenuation of all the sound particles corresponding to the virtual space at each frequency. The energy decay curve can be derived from the overall impulse response. The manner in which the overall impulse response, the energy decay curve, and the reverberation time are calculated from the sound intensities of the respective frequencies of all the sound particles as a function of time is not intended to limit the present invention. In addition, the indoor environment model 200 of the present invention is suitable for calculating the reverberation time in the space by using the formula proposed by Sabine or Eyring to provide a space designer as a reference for evaluating the sound quality; The gears and specifications provided by the speaker manufacturer are used to produce the sound effects expected to be heard in a particular space.

The listener model 300 of the present invention considers psychoacoustic imparting to the listener such as a head related transfer function such as a head related transfer function, and sets the sound receiving area 310 or 310' to simulate the sound effect heard by the listening point at different positions. The user can add a listener to the space according to his needs and adjust the position and azimuth of the listener.

Compared to the prior art, the method of the present invention allows for more variations in the simulated spatial environment, and its simulation results for acoustic effects are closer to actual conditions than the simpler operation.

The above is only the preferred embodiment of the present invention, and the scope of the invention is not limited thereto, that is, the simple equivalent changes and modifications made by the scope of the invention and the description of the invention are All remain within the scope of the invention patent. In addition, any of the objects or advantages or features of the present invention are not required to be achieved by any embodiment or application of the invention. In addition, the abstract sections and headings are only used to assist in the search of patent documents and are not intended to limit the scope of the invention.

10‧‧‧Windows simulation software

11‧‧‧System Management Module

12‧‧‧Sound Particle Management Module

13‧‧‧Object Management Module

14‧‧‧Virtual Space Management Module

15‧‧‧Drawing module

20‧‧‧Database

21‧‧‧Sound Particle Parameter Database

22‧‧‧Source database

23‧‧‧ Object database

24‧‧‧Acoustic parameter database

25‧‧‧Feeling loudness compensation database

26‧‧‧ head related transfer function database

27‧‧‧Psychoacoustic database

100‧‧‧ source model

120‧‧‧Uniform Radiation Vector Generator

140‧‧‧Sound Particle Initialization Module

160‧‧‧Source directional processing module

180‧‧‧Speaker Database

200‧‧‧ indoor environment model

210‧‧‧ obstacles

220‧‧‧Particle collision information module

230, 230’‧‧ Directional selection area

240‧‧‧Air attenuation processing module

250‧‧‧dotted box

260‧‧‧Surface Absorption Processing Module

270‧‧‧ octave band synthesis module

280‧‧‧scatter processing module

S210, S220, S232, S234, S240, S260, S280, S290‧‧‧Steps of spatial reflection and refraction iterative operation flow

300‧‧‧Lister model

310, 310’‧‧‧ Radio area

320‧‧‧ head related impulse response database

340‧‧‧ Loudness weighting module

400‧‧‧User interface

D _B1 , D _B2 ‧‧‧ (reflection) direction selection direction of the marginal direction

D _B1 ', D _B2 '‧‧‧ (refraction) direction selection direction of the marginal direction

D _p0 ‧‧‧ initial direction of the sound particles

D _p2N , D _p2 , D _p2 '‧‧‧ _Deflection direction after collision

D _source ‧‧‧3D directional data of the speaker

D _SR ‧‧‧ (reflection) main deflection direction

D _SR '‧‧‧ (refraction) main deflection direction

H _head ‧‧‧Head related transfer function

H _p0 ‧‧‧ initial spectral distribution of sound particles

H _p1 ‧‧‧First spectrum distribution of sound particles affected by sound source

H _p2 ‧‧‧Second spectrum distribution of sound particles affected by sound source and environment

H _p ‧‧‧Split distribution of sound particles affected by sound source, environment and listeners

H _pw ‧‧‧Spread spectrum distribution after loudness compensation

H _Room ‧‧‧Spectral spectral distribution based on octave intensity

H _source ‧‧‧Frequency response data in the loudspeaker database

h _p ‧‧‧ Impulse response after spectral distribution conversion

I _p0 ‧‧‧Initial sound intensity of sound particles

I _p1 ‧‧‧The first sound intensity value of the sound particle after being affected by the sound source

I _p2 ‧‧‧Second sound intensity values of sound particles affected by sound source and environment

I ₆₃ , I ₁₂₅ , I ₂₅₀ , I ₅₀₀ , I _1k , I _2k , I _4k , I _8k ‧‧‧ Sound intensity values for each band

N ‧‧‧ surface normal vector

P‧‧‧Sound Particles

P ₁ , P ₂ , P ₃ ‧ ‧ reflected sound particles

P ₁ ', P ₂ ', P ₃ '‧‧‧ refracted sound particles

P _C ‧‧‧ collision point

P _inf ‧‧‧Information on the sound particles

θ ‧‧‧ possible reflection direction angle

θ '‧‧‧ possible refraction direction angle

θ _DR ‧‧‧ (reflection) maximum diffusion angle

θ _DR '‧‧‧ (refraction) maximum diffusion angle

θ _SR ‧‧‧ specular angle

θ _SR '‧‧‧Reflection angle

δ ‧‧‧The sound particles originally carried the pulse signal

FIG. 1A is a schematic conceptual diagram of a method for simulating indoor acoustic effects according to the present invention.

FIG. 1B is a schematic diagram of a single sound particle transfer path and energy change.

2 is a schematic diagram of a sound source model according to an embodiment of the present invention.

3A is a schematic diagram of an indoor environment model according to an embodiment of the present invention.

3B is a flow chart of spatial reflection and refraction iterative operations in accordance with an embodiment of the present invention.

3C is a schematic diagram of a direction selection area according to an embodiment of the present invention.

4A is a schematic diagram of a listener model in accordance with an embodiment of the present invention.

4B is a schematic diagram of repeated sampling of sound particles according to an embodiment of the present invention.

no

no

Claims (7)

A method for simulating an indoor acoustic effect, comprising: providing a sound particle having no volume and having information for loading, wherein the sound particle is a pulse energy packet having a propagation characteristic that is linearly advanced and impinges on an obstacle, and the sound particle The piggybacking information includes a plurality of acoustic characteristics including position, direction of travel, sound intensity, spectral distribution, bandwidth, presence time, phase, and behavior; defining a sound source model, an indoor environment model, and a listener model The sound source model is configured to emit the sound particle, the indoor environment model provides volume, surface area and surface material information of the obstacle, and the listener model provides a sound receiving area; when the sound particle hits the obstacle The indoor environment model uses a microprocessor to calculate a collision point position, and uses the surface material information of the obstacle to define a direction selection area, where the direction selection area includes a plurality of possible deflection directions, The possible deflection directions are all based on the collision point; from the possible deflection directions Selecting a direction of deflection after the sound particle hits the collision point; when the sound particle travels along the deflection direction and enters the sound receiving area, the listener model obtains according to the deflection direction of the sound particle The operation information related to the sound receiving area is calculated according to the operation information, and the sound characteristics of the sound particle are calculated to obtain the sound characteristics of the sound receiving area; and the sound particle, the sound source model, the indoor environment model, and the sound The listener model is digitally graphical and displayed in a user interface. The indoor acoustic effect simulation method according to the first aspect of the invention, wherein the indoor environment model provides a virtual space, the method further comprises: adding a plurality of objects in the virtual space; setting a plurality of objects in the virtual space Acoustic parameters, including position, shape, size, area, orientation, and surface material; and setting a temperature value, a humidity value, and an atmospheric pressure value of the virtual space. The indoor acoustic effect simulation method as described in claim 1, further comprising: the sound source model radiating a plurality of sound particles in different directions in a spherical shape, wherein each of the sound particles is loaded with information, the sound is The settings of the features are all different, and the values of the acoustic features change over time. The indoor acoustic effect simulation method according to claim 3, further comprising: calculating, according to the location of the collision point and the current position of the sound particle, the indoor environment model to calculate the sound particle traveling from the sound source model to the collision point The value of the sound intensity after being attenuated by the air. The indoor acoustic effect simulation method as described in claim 4, wherein the surface material information includes a surface material absorption coefficient, a reflection coefficient, a refractive index, and a scattering coefficient of a plurality of different frequency bands, and the indoor environment is calculated after the air attenuation is calculated. The model uses the surface material absorption coefficient to calculate the sound intensity value that the sound particle is mounted at the moment of leaving the collision point. The indoor acoustic effect simulation method according to claim 5, wherein the possible deflection directions have the possibility of not obeying the law of reflection and the law of refraction, and the indoor environment model uses the scattering coefficient to define the direction selection. The step includes: considering that the sound particle has a part of energy acting on the law of reflection, and another part of the energy acting on the law of refraction to determine a main deflection direction of the direction selection area starting from the collision point; defining a maximum a diffusion angle; determining a range of the direction selection region according to the scattering coefficient, the main deflection direction, and the maximum diffusion angle; and subtracting the energy of the sound particle before the collision with the major deflection direction The energy used is formed to form a segmented sound particle; and the step of defining the direction selection region is performed to determine a range of the one direction selection region of the segmented sound particle. The indoor acoustic effect simulation method according to claim 3, wherein the listener model performs a screening process on the collected sound particles, the step comprising: determining at least two of the sound particles Whether the number of times of deflection of the particles is equal; if the number of times of deflection is equal, comparing the existence time and the direction of deflection of the two sound particles to determine whether the two sound particles are the same sound particle and take the average of the mounted information or discard the One of the two sound particles.