Sound control method and device based on 3D space acoustic calibration, terminal equipment and storage medium

By adjusting the main beam and null beam of the speaker unit based on a 3D spatial acoustic model, the problem of traditional speaker calibration ignoring the spatial propagation path is solved, achieving a more efficient speaker calibration effect and enhancing the sound quality of the speaker system.

CN121645073APending Publication Date: 2026-03-10GUANGZHOU SHENGDA ELECTRIC CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202511839275.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-12-08
Publication Date
2026-03-10

AI Technical Summary

Technical Problem

Traditional audio calibration methods ignore the propagation path of sound in space and reflected sound, resulting in poor calibration results.

Method used

Based on a 3D spatial acoustic model, by adjusting the direction and energy amplitude weights of the main beam and null beam, the speaker unit is controlled to focus the main lobe energy in the listening area, reducing ineffective radiation paths and suppressing reflected sound energy.

Benefits of technology

It improves the effect of sound calibration, concentrates sound wave energy in the listening area, reduces ineffective radiation paths in the side lobe area, and suppresses the intensity of reflected sound energy, thereby improving the sound quality of the audio system.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN121645073A_ABST
    Figure CN121645073A_ABST
Patent Text Reader

Abstract

The invention discloses a sound equipment control method and device based on 3D space acoustic calibration, terminal equipment and a storage medium, and the method comprises the steps: constructing a Gaussian model used for solving the energy amplitude weight corresponding to each sound wave radiation direction based on a 3D space acoustic model, and determining an energy amplitude weight curve according to the Gaussian model; determining a time delay parameter according to the distance between the central point of the sound unit and the central point of the listening area so as to determine a main beam parameter; determining a null beam direction according to a direction angle of the sound unit pointing to the target reflection surface; extracting an energy amplitude weight with the same direction as the nulling beam from the energy amplitude weight curve as a nulling beam energy amplitude weight, and taking an opposite phase of the main beam in the same direction as a nulling beam phase so as to generate a nulling beam parameter; according to the main beam parameter and the null beam parameter, the sound unit is controlled, and the sound calibration precision can be improved.
Need to check novelty before this filing date? Find Prior Art

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of acoustic devices, in particular to a sound control method and device based on 3D space acoustic calibration, a terminal device and a storage medium. BACKGROUND

[0002] With the continuous improvement of people's requirements for audio experience, multi-channel sound systems are increasingly widely used in different scenarios.

[0003] However, different space environments have a significant impact on the propagation of sound. Traditional sound calibration often only measures the frequency response of the sound system for calibration, ignoring the propagation path of sound in space, the proportion of reflected sound and direct sound, etc., resulting in poor existing sound calibration effect. SUMMARY

[0004] The embodiments of the present application provide a sound control method and device based on 3D space acoustic calibration, a terminal device and a storage medium, which can adjust the main beam and null beam according to the position of the listening area, so that the main lobe energy can be focused on the listening area path, reducing the invalid radiation path and suppressing the reflected sound energy intensity.

[0005] To achieve the above-mentioned purpose, the first aspect of the embodiments of the present application provides a sound control method based on 3D space acoustic calibration, comprising: According to the size of the sound field space and the reflection coefficient of each reflecting surface, a 3D space acoustic model is generated; For each sound unit in the sound field space, the center angle is determined according to the direction angle of the sound unit center point pointing to the listening area center point in the sound field space; the direction corresponding to the center angle is taken as the reference direction, and the deviation center angle is determined according to the included angle between each sound wave radiation direction and the reference direction; the main lobe width is determined according to the opening angle of the listening area; The polar coordinates are constructed with the sound unit center point as the pole, each deviation center angle as the polar angle, and the energy amplitude weight corresponding to each deviation center angle as the polar radius; based on the polar coordinates, the center angle, the main lobe width and the preset initial energy amplitude weight, a Gaussian model for solving the energy amplitude weight corresponding to each sound wave radiation direction is constructed, and then the energy amplitude weight curve of each sound wave radiation direction is determined according to the Gaussian model; According to the distance between the sound unit center point and the listening area center point, the time delay parameter is determined, and the main beam parameter is generated according to the energy amplitude weight curve and the time delay parameter; The reflection surface with a reflection coefficient exceeding a preset reflection coefficient is taken as a target reflection surface, a null beam direction is determined according to a direction angle of the sound unit pointing to the target reflection surface, an energy amplitude weight identical to the null beam direction is extracted from an energy amplitude weight curve as a null beam energy amplitude weight, and an opposite phase of a main beam in the same direction is taken as a null beam phase; and null beam parameters are generated according to the null beam direction, the null beam energy amplitude weight and the null beam phase. The sound unit is controlled according to the main beam parameters and the null beam parameters.

[0006] Further, the Gaussian model for solving the energy amplitude weight corresponding to each sound wave radiation direction is: ; wherein, is the energy amplitude weight; is the deviation central angle; A is the initial energy amplitude weight; is the central angle; is a main lobe width control parameter, which is converted from the main lobe width.

[0007] Further, the energy amplitude weight curve of each sound wave radiation direction determined according to the Gaussian model comprises: for each sound wave radiation direction, whether the current sound wave radiation direction is within the main lobe range is judged according to the corresponding deviation central angle and the main lobe width; if yes, the corresponding energy amplitude weight is calculated according to the Gaussian model; if no, it is determined that the sound wave radiation direction is within the side lobe range, a cosine roll-off window is superimposed on the basis of the Gaussian model to obtain a side lobe attenuation enhancement model: ; wherein, is the boundary angle of the main lobe width; is a preset roll-off transition bandwidth.

[0008] Further, the initial energy amplitude weight is determined by the following method: the distances between each sound unit and the listening area are calculated, and the sound unit closest to the listening area is taken as a reference sound unit; and the distance between the reference sound unit and the listening area is taken as a reference distance; a preset initial energy weight is taken as the initial energy amplitude weight of the reference sound unit; for non-reference sound units other than the reference sound unit, the distance between the non-reference sound unit and the listening area is calculated to obtain a first distance; a distance compensation amplitude coefficient is determined according to the ratio of the first distance to the reference distance; The preset initial energy weight is compensated based on the distance compensation amplitude coefficient to generate the initial energy amplitude weight corresponding to the non-reference audio unit.

[0009] Furthermore, the 3D spatial acoustic model is divided into three-dimensional meshes, and several mesh points are generated within the 3D spatial acoustic model. Based on the actual sound pressure level of each grid point at each frequency point, generate the actual frequency response curve corresponding to each grid point; Determine the scene frequency response curve corresponding to the current scene type; The target grid point is determined based on the user's location. The actual frequency response curve of the target grid point is compared with the scene frequency response curve to determine the frequency response deviation. The frequency response of the audio unit is calibrated based on the frequency response deviation.

[0010] Furthermore, a microphone is also provided within the acoustic space; The determination of the actual sound pressure level includes: Based on the reference sound pressure level of each speaker unit and the position coordinates of the microphone in the 3D spatial acoustic model, the direct sound pressure level of each speaker unit at the microphone is calculated, and then the total direct sound pressure level of all speaker units at the microphone is calculated. The total reflected sound pressure level at the microphone is calculated based on the direct total sound pressure level and the measured total sound pressure level collected by the microphone; the reflected sound pressure level of each speaker unit at the microphone is determined based on the proportional relationship of the reference sound pressure levels of each speaker unit. The reflected sound pressure level and direct sound pressure level of each speaker unit at the microphone are used to determine the reflected sound contribution coefficient of each speaker unit at the microphone. For each grid point, the direct sound pressure level contribution of each speaker unit to the grid point is calculated based on the distance between the grid point and the speaker unit and the reference sound pressure level of the speaker unit. The contribution of each speaker unit to the reflected sound pressure level of the grid point is determined based on the contribution of each speaker unit to the direct sound pressure level of the grid point, the contribution coefficient of each speaker unit to the reflected sound at the microphone, and the comprehensive reflection coefficient of the reflecting surface corresponding to the grid point. The actual sound pressure level of the grid point is determined based on the direct sound pressure level contribution and reflected sound pressure level contribution of each audio unit to the grid point.

[0011] Furthermore, the overall reflection coefficient of the reflecting surface corresponding to the grid point is determined in the following way: Calculate the distance of each grid point to each reflecting surface; The reflection weight of each reflecting surface is determined based on the distance between the grid points and each reflecting surface. The overall reflection coefficient is obtained by weighted fusion based on the reflection weight and corresponding reflection coefficient of each reflective surface.

[0012] A second aspect of this application provides a sound control device based on 3D spatial acoustic calibration, comprising: The system includes a 3D spatial acoustic model construction module, an angle determination module, an energy amplitude weight determination module, a main beam parameter determination module, a null beam parameter determination module, and an audio control module. The 3D spatial acoustic model building module is used to generate a 3D spatial acoustic model based on the size of the sound field space and the reflection coefficient of each reflecting surface. The angle determination module is used to determine the center angle for each speaker unit in the sound field space based on the direction angle from the center point of the speaker unit to the center point of the listening area in the sound field space; to determine the deviation from the center angle based on the angle between each sound wave radiation direction and the reference direction, using the direction corresponding to the center angle as the reference direction; and to determine the main lobe width based on the angle of the listening area. The energy amplitude weight determination module is used to construct polar coordinates with the center point of the audio unit as the pole, each deviation angle as the polar angle, and the energy amplitude weight corresponding to each deviation angle as the polar radius. Based on the polar coordinates, the center angle, the main lobe width, and the preset initial energy amplitude weight, a Gaussian model is constructed to solve the energy amplitude weight corresponding to each sound wave radiation direction. Then, the energy amplitude weight curve of each sound wave radiation direction is determined according to the Gaussian model. The main beam parameter determination module is used to determine the time delay parameter based on the distance between the center point of the audio unit and the center point of the listening area, and to generate the main beam parameter based on the energy amplitude weighting curve and the time delay parameter. The null beam parameter determination module is used to identify reflective surfaces with a reflection coefficient exceeding a preset reflection coefficient as target reflective surfaces, determine the null beam direction based on the directional angle of the audio unit pointing towards the target reflective surface, extract the energy amplitude weights with the same direction as the null beam from the energy amplitude weight curve as the null beam energy amplitude weights, and use the opposite phase of the main beams with the same direction as the null beam phases; and generate null beam parameters based on the null beam direction, null beam energy amplitude weights, and null beam phases. The audio control module is used to control the audio unit according to the main beam parameters and null beam parameters.

[0013] A third aspect of this application provides a terminal device including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the above-described sound control method based on 3D spatial acoustic calibration.

[0014] A fourth aspect of this application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described sound control method based on 3D spatial acoustic calibration.

[0015] The following benefits can be obtained by implementing the present invention: Compared to existing technologies, this invention constructs a Gaussian model in polar coordinates based on the location of the listening area to solve the energy amplitude weights corresponding to each sound wave radiation direction, thereby determining the energy amplitude weights of the main beam in each sound wave radiation direction. The Gaussian model concentrates the sound wave energy in the listening area, reducing the energy amplitude of ineffective radiation paths in the side lobe region. In addition, for strong reflective surfaces, a null-dip beam with consistent energy amplitude but opposite phase is constructed to suppress the intensity of reflected sound energy, thereby adjusting the ratio of direct sound to reflected sound and further improving the sound calibration effect. Attached Figure Description

[0016] Figure 1 This is a flowchart illustrating an embodiment of the present invention that provides a sound control method based on 3D spatial acoustic calibration; Figure 2 This is a schematic diagram of the structure of a sound control device based on 3D spatial acoustic calibration, according to an embodiment of the present invention. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] To resolve the above issues, please refer to [link / reference]. Figure 1 An embodiment of the present invention provides a sound control method based on 3D spatial acoustic calibration, comprising: S1. Generate a 3D spatial acoustic model based on the dimensions of the sound field space and the reflection coefficients of each reflecting surface.

[0019] In a preferred embodiment, the dimensions of the sound field space and the reflection coefficients of the reflecting surfaces can be calibrated in advance by manual measurement. First, a digital sound field space model is constructed based on the actual dimensions of the sound field space. Then, the reflection coefficients of each reflecting surface in the sound field space are marked within the sound field space model to generate the aforementioned 3D spatial acoustic model.

[0020] In other alternative embodiments, the above-mentioned 3D spatial acoustic model can also be automatically generated. Specifically, microphones and several speaker units are arranged in the acoustic space. The speaker units can be distributed on the four walls of the sound field space, and the microphones are sequentially placed at at least three different locations within the sound field space. Next, the main control device sends a synchronization trigger command to each speaker unit, and each speaker unit simultaneously emits an MLS wideband detection signal, covering the entire frequency band from 20Hz to 20kHz, with a transmission duration of 5 seconds (to ensure that the reflected sound can be completely captured); the microphone begins to collect data at the same time as the signal transmission starts, with a collection duration of 10 seconds (covering the signal transmission period and the subsequent reflection sound attenuation period). The collected data is transmitted back to the main control device in real time in the format of "timestamp-signal amplitude" and stored in a high-speed cache (data transmission rate ≥100Mbps to avoid data loss). Next, the main control device analyzes the acquired multi-channel signals. Through joint analysis in the time and frequency domains, it achieves accurate separation of direct sound and reflected sound, providing data for subsequent size calculations and reflective surface positioning. Specifically, the direct sound, as the earliest signal to reach the microphone, directly reflects the distance relationship between the speaker and the microphone. Based on the microphone data, the three-dimensional dimensions of the room can be deduced: First, envelope detection (using Hilbert transform) is performed on the acquired signals of the microphone channel. The moment when the signal amplitude first reaches 10% of the peak value is identified as T0 as the arrival time of the direct sound. Based on T0, the original signal corresponding to the envelope from T0 to T0+ΔT1 (ΔT1 = the duration of the direct sound, about 2-3ms, defined as the signal amplitude dropping to 30% of the peak value) is extracted as the direct sound segment, and the rest is taken as the reflected sound segment, thus achieving the separation of direct sound and reflected sound. Next, based on the direct sound, the distance from each speaker to the microphone unit is calculated according to "distance = speed of sound × time". Microphones are fixed at three or more different reference points in the room (e.g., center, front, rear, spaced at least 1m apart). At each point, the direct sound signal from all speaker units is collected following the above process, obtaining multiple sets of propagation time data between the speakers and microphones (each set contains data from M speakers × 1 microphone, where M is the number of speakers). The distance from each speaker to different microphone points is calculated. Combined with the known relative coordinates of the speakers, a system of spatial geometric equations is established for each microphone point. The equations are solved using the least squares method to obtain the room's length (front-to-back direction), width (left-to-right direction), and height (up-down direction), ultimately yielding the dimensions of the sound field space. The equations for multiple microphone points are solved simultaneously to obtain the absolute coordinates of each microphone point. Combining the spatial distribution of the three points with the speaker unit coordinates, the room boundary plane is fitted, and the room's length, width, and height are finally calculated. After determining the dimensions of the sound field space, a corresponding digital sound field space model is generated.

[0021] Next, the reflection coefficient of each reflecting surface is determined. For the direct sound and reflected sound along the same propagation path, the amplitude ratio at the same frequency point is calculated, i.e., reflection coefficient α = reflected sound amplitude / direct sound amplitude. Based on the reflection coefficient, the reflecting surfaces are divided into strong reflecting surfaces (α > 0.7), medium reflecting surfaces (0.3 ≤ α ≤ 0.7), and weak reflecting surfaces (α < 0.3). The reflection coefficient of each reflecting surface is marked in the sound field spatial model, forming the above-mentioned 3D spatial acoustic model.

[0022] In a preferred embodiment, after the 3D spatial acoustic model is generated, the 3D spatial acoustic model is displayed on the interface of the main control device.

[0023] S2. For each speaker unit in the sound field space, determine the center angle based on the direction angle from the center point of the speaker unit to the center point of the listening area in the sound field space; take the direction corresponding to the center angle as the reference direction, determine the deviation from the center angle based on the angle between each sound wave radiation direction and the reference direction; determine the main lobe width based on the angle of the listening area. Optionally, the aforementioned listening area can be a fixed area pre-defined within the sound field space. Alternatively, it can be an area dynamically created by the user based on the generated 3D spatial acoustic model and according to actual needs. For example, the user can draw the desired listening area on the interface of the presented spatial acoustic model. Upon detecting the user's operation of drawing a listening area, the main control device sets the corresponding area in the spatial acoustic model as the listening area.

[0024] For each speaker unit in the sound field space, the geometric center point of the speaker unit is taken as the center point of the speaker unit, and the geometric center point of the listening area is taken as the center point of the listening area. The direction of the line connecting the center point of the speaker unit and the center point of the listening area is taken as the reference direction, the angle of the reference direction is determined as the central angle, and the angle between each sound wave radiation direction and the reference direction is determined as the deviation from the central angle. The angle formed by the center point of the speaker unit to the two sides of the listening area is taken as the width of the main lobe.

[0025] S3. Construct polar coordinates with the center point of the speaker unit as the pole, each deviation angle from the center as the polar angle, and the energy amplitude weight corresponding to each deviation angle as the polar radius. Based on the polar coordinates, the center angle, the main lobe width, and the preset initial energy amplitude weight, construct a Gaussian model to solve the energy amplitude weight corresponding to each sound wave radiation direction. Then, determine the energy amplitude weight curve of each sound wave radiation direction according to the Gaussian model. In a preferred embodiment, the Gaussian model used to solve for the energy amplitude weights corresponding to each sound wave radiation direction is: ; in, Energy amplitude weighting; The deviation from the center angle; A is the initial energy amplitude weight; The central angle; This is the main lobe width control parameter, which is calculated from the main lobe width.

[0026] To concentrate sound wave energy amplitude within the main listening area, achieving direct sound beam focusing and improving sound energy density and clarity in the listening area, this invention uses a Gaussian function to design an amplitude weighting curve, ensuring that the main lobe of the speaker's energy is precisely directed towards the listening area. The center angle of the Gaussian function is set as the direction angle from which the speaker points to the center of the listening area (i.e., the aforementioned center angle); the main lobe width is set as the angle corresponding to the listening area (e.g., if the listening area is within 30° of the speaker, then the main lobe width is set to 30°). In specific implementation, the aforementioned Gaussian model can be used; where σ is the main lobe width control parameter, calculated from the main lobe width, ensuring that the main lobe energy is concentrated within the listening area; A is the initial amplitude coefficient (default setting is 1, corresponding to 0dB gain). Preferably, based on the conversion of main lobe width = 2.355σ (e.g., a 30° main lobe corresponds to σ≈12.74), in this preferred embodiment, the main lobe width = 2.355σ is 1 / 2 the angle range corresponding to the peak amplitude. In the general standards of acoustic and radio frequency engineering, amplitude attenuation of 1 / 2 corresponds to power attenuation of 1 / 4, i.e., -3dB. At this threshold, the main lobe (energy concentration area) and side lobes (energy attenuation area) can be accurately distinguished, which is close to the actual listening experience and equipment performance. In the above Gaussian model, A is the peak amplitude of the main lobe. When the amplitude drops to A / 2, substituting it into the Gaussian model, we can obtain... =ln2≈0.693 → ≈±1.177σ, main lobe width = 2×1.177σ≈2.354σ, rounded to 2.355σ.

[0027] Based on the above embodiments, by inputting the offset angle corresponding to each sound wave radiation direction into the Gaussian model, the energy amplitude weight of each sound wave radiation direction can be obtained, ultimately generating an energy amplitude weight curve. The Gaussian model ensures that the closer each sound wave radiation direction is to the listening area, the greater the energy amplitude weight, thus accurately directing the main lobe of the sound system towards the listening area.

[0028] In another preferred embodiment, determining the energy amplitude weighting curves for each sound wave radiation direction based on the Gaussian model includes: For each sound wave radiation direction, based on the corresponding offset angle and the main lobe width, it is determined whether the current sound wave radiation direction is within the main lobe range; if so, the corresponding energy amplitude weight is calculated according to the Gaussian model; if not, it is determined that the sound wave radiation direction is within the side lobe range, and a cosine roll-off window is superimposed on the Gaussian model to obtain the side lobe attenuation enhancement model: ; in, The boundary angle of the main lobe width; This is the preset roll-off transition bandwidth.

[0029] In this embodiment, different models are used to calculate the energy amplitude weights for the sound wave radiation direction of the main lobe region and the sound wave radiation direction of the side lobe region, so as to enhance the side lobe attenuation and improve the sound calibration effect.

[0030] Specifically, when the sound wave radiation direction is within the main lobe region (i.e., not exceeding the main lobe width), the corresponding energy amplitude weighting curve is calculated according to the aforementioned Gaussian model. When the sound wave radiation direction is within the side lobe region, a cosine roll-off window is superimposed on the current Gaussian model to generate the aforementioned side lobe attenuation enhancement model. The preset roll-off transition bandwidth can be schematically set to 5°, thereby causing the acoustic energy in the side lobe region to decay rapidly.

[0031] In a preferred embodiment, the initial energy amplitude weight is determined in the following manner: Calculate the distance between each speaker unit and the listening area, and designate the speaker unit closest to the listening area as the reference speaker unit; use the distance between the reference speaker unit and the listening area as the reference distance; use a preset initial energy weight as the initial energy amplitude weight of the reference speaker unit; for non-reference speaker units other than the reference speaker unit, calculate the distance between the non-reference speaker unit and the listening area to obtain a first distance; determine the distance compensation amplitude coefficient based on the ratio of the first distance to the reference distance; compensate the preset initial energy weight based on the distance compensation amplitude coefficient to generate the initial energy amplitude weight corresponding to the non-reference speaker unit.

[0032] In this embodiment, the sound calibration effect is further improved by using a distance gain compensation algorithm. Based on the principle that sound energy attenuation is inversely proportional to the square of the distance, the straight-line distance from each channel speaker unit to the center of the listening area is first extracted using a 3D spatial acoustic model. The nearest channel is used as the reference, and the nearest speaker unit is used as the reference speaker unit. The distance between the reference speaker unit and the listening area is used as the reference distance. For the reference speaker unit, a preset initial energy weight, for example, 1, is used as the initial energy amplitude weight of the reference speaker unit. Since sound energy in a free field is inversely proportional to the square of the distance, and its amplitude is also inversely proportional to the distance, the distance compensation amplitude coefficient for non-reference audio units can be calculated using the following formula: ; in, Let n be the distance compensation amplitude coefficient for the nth speaker unit. The above reference distance, Let n be the distance between the nth speaker unit and the center point of the listening area.

[0033] Next, the preset initial energy weights are compensated and fused using the following formula: .

[0034] Will Substituting the original Gaussian model, we can obtain the Gaussian model corresponding to the non-reference speaker unit: ; This embodiment enables distance-based gain compensation, achieving energy equalization of gain over long distances. This ensures that the main beam of the speaker units at different distances reaches the listening area with equal energy. It focuses the sound wave energy of a single speaker unit into the listening area and achieves balanced sound wave energy across different speaker units at the listening area, significantly improving sound calibration.

[0035] S4. Determine the time delay parameters based on the distance between the center point of the speaker unit and the center point of the listening area, and generate the main beam parameters based on the energy amplitude weighting curve and the time delay parameters. Specifically, after determining the energy amplitude weighting curve corresponding to each speaker unit according to the above steps, the direct sound propagation time τ is calculated for each speaker unit and the center point of the listening area based on distance = speed of sound × time (e.g., the distance from the left front speaker unit to the center of the listening area is 3.2m, and the speed of sound is 340m / s, then τ = 3.2 / 340 ≈ 9.4ms), thus obtaining the above time delay parameters; finally, based on the time delay parameters and the energy amplitude weighting curve, the main beam parameters of each speaker unit can be obtained.

[0036] S5. Take the reflective surface with a reflection coefficient exceeding the preset reflection coefficient as the target reflective surface, and determine the null beam direction based on the directional angle of the audio unit pointing to the target reflective surface; extract the energy amplitude weight with the same direction as the null beam from the energy amplitude weight curve, and take it as the null beam energy amplitude weight; take the opposite phase of the main beam with the same direction as the null beam phase; generate null beam parameters based on the null beam direction, null beam energy amplitude weight and null beam phase. To suppress reflected sound and adjust the ratio of reflected sound to direct sound, this embodiment of the invention generates a null-beam in the direction of the strong reflective surface, reducing the radiation of sound in that direction, controlling the generation of reflected sound from the source, while ensuring that the spatial sense of immersion brought by early reflected sound is not affected.

[0037] Specifically, in this invention, a strong reflective surface with a reflection coefficient greater than 0.7 can be used as the target reflective surface. The directional angle of each audio unit pointing towards the strong reflective surface is calculated to obtain the null beam direction. Then, a phase cancellation method is used to design the null weight. The energy amplitude weight of the main beam with the same direction as the null beam is extracted from the energy amplitude weight curve to obtain the null beam energy amplitude weight. The opposite phase of this main beam is used as the null beam phase. This achieves the application of a weight to the audio unit in the direction of the strong reflective surface that is equal in amplitude and opposite in phase to the main beam in the same direction. For example, if the amplitude weight of a main beam in this direction is 0.8 and the phase is 0°, then the null weight is set to 0.8 and the phase is 180°. The null beam generated in the above manner can undergo destructive interference with the main beam in the corresponding direction to achieve mutual energy cancellation, ultimately resulting in a significant attenuation of the sound pressure level in the direction of the strong reflective surface, suppressing reflected sound.

[0038] In a preferred embodiment, after determining the target reflective surface, the method further includes: generating a null beam steering vector based on the azimuth angle of the center of the speaker unit pointing towards the target reflective surface and the corresponding mode length; generating a main beam steering vector based on the azimuth angle of the center of the speaker unit pointing towards the center point of the listening area in the sound field space and the corresponding mode length; determining whether the angle between the null beam steering vector and the main beam steering vector is greater than 120°, and if not, not performing subsequent operations.

[0039] In this embodiment, the direction of the guide vector represents the energy focusing direction of the beam. The main beam guide vector corresponds to the target sound source direction, and the null beam guide vector in the interference direction needs to form a large angle with it. Essentially, this ensures that the weighting coefficient of the beam in the interference direction is out of phase with the coefficient of the main beam, thus canceling out their energy. After obtaining the target reflecting surface, the angle between the null beam guide vector and the main beam guide vector is checked to see if it is greater than 120°. When the angle between the two guide vectors is ≥120°, their inner product is ≤-0.5. At this time, the beam gain in the interference direction will be significantly reduced. If the angle is <120°, the absolute value of the inner product decreases, the interference suppression effect weakens, and may even cause the main beam sidelobes to rise due to energy superposition, thus compromising the sound field calibration accuracy. Therefore, the above scheme can further improve the accuracy of sound field calibration.

[0040] S6. Control the audio unit according to the main beam parameters and null beam parameters; Specifically, the main control device sends the main beam parameters and null beam parameters to each influence control unit. Each influence control unit first calculates the time delay difference between the reference speaker unit and the other non-reference speakers based on the main beam parameter delay, using the speaker unit closest to the center of the listening area as the reference speaker unit. This time delay difference is used as the start-up time difference between the non-reference speakers and the reference speaker unit. The start-up time point of each speaker unit is determined based on the start-up time difference, so that the direct sound from all speakers arrives in the listening area simultaneously. For example, assuming the left channel speaker unit is 5ms away from the listening area and its delay is 6ms, the difference between the two is 1ms. Then, 1ms after the right channel speaker unit is started, the left channel speaker unit is started, so that the direct sound from both the left and right channels arrives in the listening area simultaneously.

[0041] Furthermore, based on the energy amplitude weights of each sound wave radiation direction in the main beam parameters, the energy amplitude of the main signal in each sound wave radiation direction is determined. At the same time, based on the null beam parameters, the energy amplitude and phase of the null correction signal in the corresponding direction are determined. This allows each speaker unit to synthesize the corresponding sound signal according to the corresponding parameters, thereby controlling the speaker unit and improving the sound calibration effect.

[0042] In a preferred embodiment, after controlling the audio unit according to the main beam parameters and the null beam parameters, the method further includes: The 3D spatial acoustic model is divided into three-dimensional meshes, generating several mesh points within the model. Based on the actual sound pressure level at each frequency point, an actual frequency response curve is generated for each mesh point. The scene frequency response curve corresponding to the current scene is determined based on the current scene type. A target mesh point is determined based on the user's location, and its actual frequency response curve is compared with the scene frequency response curve to determine the frequency response deviation. The speaker unit's frequency response is calibrated based on the frequency response deviation.

[0043] In an optional embodiment, microphones can be placed at each grid point to collect the actual sound pressure level, thereby generating the actual frequency response curve for each grid point. Then, based on the current scene type, the scene frequency response curve corresponding to the scene is extracted from a preset scene frequency response database. In actual applications, there are various scenarios, such as conference scenarios and cinema scenarios. Different scenarios have different frequency responses. For example, for a conference scenario, which focuses on speech intelligibility, it is necessary to enhance the mid-frequency band (800Hz-3kHz) gain and suppress low-frequency (<200Hz) and high-frequency (>8kHz) noise. Its scene frequency response curve needs to meet the following requirements: a smooth transition (±1dB) from 200Hz to 800Hz, a 1-2dB boost from 800Hz to 3kHz, a smooth attenuation (-0.5dB per octave) from 3kHz to 8kHz, and a 2-3dB attenuation from 20Hz to 200Hz. For cinema scenarios, the pursuit of balanced sound across the entire frequency range and impactful sound effects requires strengthening low frequencies (50Hz-200Hz) and mid-high frequencies (3kHz-8kHz) to ensure a full sound field. The scene frequency response curve must meet the following requirements: 2-3dB boost from 20Hz to 50Hz (enhancing the impact of low frequencies), smoothness from 50Hz to 3kHz (±1dB), 1dB boost from 3kHz to 8kHz (enhancing detail), 1dB attenuation from 8kHz to 20kHz (suppressing air noise), and fluctuations across the entire frequency range ≤±2dB. To make the sound effects more closely resemble actual application scenarios, this invention pre-stores the scene frequency response curves corresponding to each scene in a scene frequency response database, and then the corresponding scene frequency response curve can be extracted based on the current scene type. After extracting the scene frequency response curve, corresponding grid points are matched based on the user's current location to obtain the aforementioned target grid points. Then, the actual frequency response curve of the target grid points is compared with the scene frequency response curve to determine the frequency response deviation. Finally, the speaker unit's frequency response is calibrated based on the frequency response deviation. This achieves frequency response calibration based on the scene and user location, further improving the accuracy of sound calibration.

[0044] In other alternative embodiments, it is not necessary to set the microphone at each grid point sequentially. Instead, the microphone can be fixed at one grid point to calculate the actual sound pressure level at the other grid points, thereby improving data acquisition efficiency.

[0045] Specifically, based on the reference sound pressure level of each speaker unit and the position coordinates of the microphone in the 3D spatial acoustic model, the direct sound pressure level of each speaker unit at the microphone is calculated, and then the total direct sound pressure level of all speaker units at the microphone is calculated. The total reflected sound pressure level at the microphone is calculated based on the direct total sound pressure level and the measured total sound pressure level collected by the microphone; the reflected sound pressure level of each speaker unit at the microphone is determined based on the proportional relationship of the reference sound pressure levels of each speaker unit. The reflected sound pressure level and direct sound pressure level of each speaker unit at the microphone are used to determine the reflected sound contribution coefficient of each speaker unit at the microphone. For each grid point, the direct sound pressure level contribution of each speaker unit to the grid point is calculated based on the distance between the grid point and the speaker unit and the reference sound pressure level of the speaker unit. The contribution of each speaker unit to the reflected sound pressure level of the grid point is determined based on the contribution of each speaker unit to the direct sound pressure level of the grid point, the contribution coefficient of each speaker unit to the reflected sound at the microphone, and the comprehensive reflection coefficient of the reflecting surface corresponding to the grid point. The actual sound pressure level of the grid point is determined based on the direct sound pressure level contribution and reflected sound pressure level contribution of each audio unit to the grid point.

[0046] To illustrate, the direct sound pressure level at the microphone for each speaker unit is calculated using the following formula: ; in, Let be the direct sound pressure level of the nth speaker unit at the i-th frequency point at the microphone. For the nth speaker unit, the sound pressure level at the i-th frequency point at the reference distance is the aforementioned reference sound pressure level; The distance from the nth speaker unit to the microphone is calculated based on the microphone's position coordinates in the 3D spatial acoustic model and the nth speaker unit's position coordinates in the 3D spatial acoustic model. The reference distance in this invention is 1m.

[0047] After calculating the direct sound pressure level of each speaker unit at the microphone, the total direct sound pressure level of all speaker units at the microphone is calculated using the following formula; ;in, For the i-th frequency point, the total sound pressure level of all audio units at the microphone is directly reached.

[0048] Next, based on the measured total sound pressure level collected by the microphone at its location and the direct total sound pressure level of all speaker units at the microphone, the total reflected sound pressure level is calculated using the following formula: ;in, For the i-th frequency point, the total reflected sound pressure level of all audio units at the microphone; The measured total sound pressure level at the i-th frequency point, collected by the microphone at its location.

[0049] After obtaining the total reflected sound pressure level, the total reflected sound pressure level is allocated according to the reference sound pressure level ratio of each speaker unit to obtain the reflected sound pressure level of a single speaker unit at the microphone: ;in, Let i be the frequency point, and n be the reflected sound pressure level of the speaker unit at the microphone. Let N be the reference sound pressure level of the Kth speaker unit at the i-th frequency point, and N be the total number of speaker units.

[0050] Next, the contribution coefficient of the reflected sound from the microphone by the speaker unit is determined using the following formula; ;in, The contribution coefficient of reflected sound; Then, the contribution of each speaker unit to the direct sound pressure level at the grid point is calculated using the following formula; = ;in, Let P be the direct sound pressure level contribution of each speaker unit at grid point P at the i-th frequency. Let be the distance from the nth speaker unit to grid point P.

[0051] Then, the contribution of the speaker unit to the reflected sound pressure level at the grid points is calculated using the following formula; = ;in, This is the overall reflection coefficient of the reflecting surface corresponding to the grid point. This represents the contribution of the reflected sound pressure level of the nth speaker unit at grid point P at the i-th frequency.

[0052] Next, the contribution of a single speaker unit to the total sound pressure level at the grid points is calculated using the following formula: ;in, , where is the total sound pressure level contribution of the nth speaker unit at grid point P at the i-th frequency point.

[0053] Finally, the actual sound pressure level at grid point P after energy superposition by the multi-speaker unit is calculated using the following formula: ;in, This represents the actual sound pressure level at grid point P.

[0054] In a preferred embodiment, the overall reflection coefficient of the reflecting surface corresponding to the grid point is determined in the following way: Calculate the distance of the grid points to each reflecting surface; determine the reflection weight of each reflecting surface based on the distance between the grid points and each reflecting surface; and perform weighted fusion based on the reflection weight of each reflecting surface and the corresponding reflection coefficient to obtain the comprehensive reflection coefficient.

[0055] Specifically, based on the distance between the reflecting surface and the grid points and the reflection coefficient of the reflecting surface itself, the reflection coefficients of multiple reflecting surfaces can be fused using the inverse distance weighting method to obtain the comprehensive reflection coefficient of the grid points, as shown in the following formula: ;in, Let be the reflection coefficient of the j-th reflecting surface; Let J be the reflection weight of the j-th reflecting surface, and J be the total number of reflecting surfaces. = ; in, Let p be the distance between grid point p and the j-th reflecting surface. The square of this distance; Let p be the distance between grid point p and the h-th reflecting surface. The square of this distance.

[0056] Based on the above method embodiments, the present invention provides corresponding device embodiments. One embodiment of the present invention provides a sound control device based on 3D spatial acoustic calibration, including: a 3D spatial acoustic model construction module, an angle determination module, an energy amplitude weight determination module, a main beam parameter determination module, a null beam parameter determination module, and a sound control module. The 3D spatial acoustic model building module is used to generate a 3D spatial acoustic model based on the size of the sound field space and the reflection coefficient of each reflecting surface. The angle determination module is used to determine the center angle for each speaker unit in the sound field space based on the direction angle from the center point of the speaker unit to the center point of the listening area in the sound field space; to determine the deviation from the center angle based on the angle between each sound wave radiation direction and the reference direction, using the direction corresponding to the center angle as the reference direction; and to determine the main lobe width based on the angle of the listening area. The energy amplitude weight determination module is used to construct polar coordinates with the center point of the audio unit as the pole, each deviation angle as the polar angle, and the energy amplitude weight corresponding to each deviation angle as the polar radius. Based on the polar coordinates, the center angle, the main lobe width, and the preset initial energy amplitude weight, a Gaussian model is constructed to solve the energy amplitude weight corresponding to each sound wave radiation direction. Then, the energy amplitude weight curve of each sound wave radiation direction is determined according to the Gaussian model. The main beam parameter determination module is used to determine the time delay parameter based on the distance between the center point of the audio unit and the center point of the listening area, and to generate the main beam parameter based on the energy amplitude weighting curve and the time delay parameter. The null beam parameter determination module is used to identify reflective surfaces with a reflection coefficient exceeding a preset reflection coefficient as target reflective surfaces, determine the null beam direction based on the directional angle of the audio unit pointing towards the target reflective surface, extract the energy amplitude weights with the same direction as the null beam from the energy amplitude weight curve as the null beam energy amplitude weights, and use the opposite phase of the main beams with the same direction as the null beam phases; and generate null beam parameters based on the null beam direction, null beam energy amplitude weights, and null beam phases. The audio control module is used to control the audio unit according to the main beam parameters and null beam parameters.

[0057] Based on the above method embodiments, the present invention provides corresponding terminal device embodiments.

[0058] One embodiment of this application provides a terminal device, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the above-described audio control method based on 3D spatial acoustic calibration.

[0059] Based on the above-described method embodiments, the present invention provides corresponding computer-readable storage medium embodiments.

[0060] One embodiment of this application provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the above-described sound control method based on 3D spatial acoustic calibration.

[0061] Computer devices can be computing devices such as smartphones, tablets, desktop computers, and cloud servers. These devices may include, but are not limited to, processors and memory. Those skilled in the art will understand that the figures are merely examples of computer devices and do not constitute a limitation on the computer device. The device may include more or fewer components than illustrated, or a combination of certain components, or different components; for example, it may also include input / output devices, network access devices, etc.

[0062] The processor referred to can be a Central Processing Unit (CPU), but it can also be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor.

[0063] In some embodiments, the memory can be an internal storage unit of a computer device, such as a hard drive or RAM. In other embodiments, the memory can be an external storage device of the computer device, such as a plug-in hard drive, Smart Media Card (SMC), Secure Digital (SD) card, or Flash Card. Furthermore, the memory can include both internal and external storage units of the computer device. The memory is used to store operating systems, applications, bootloaders, data, and other programs, such as program code for computer programs. The memory can also be used to temporarily store data that has been output or will be output.

[0064] This application provides a computer program product that, when run on a computer device, enables the computer device to execute the steps described in the various method embodiments above.

[0065] In the several embodiments provided in this application, it will be understood that each block in the flowchart or block diagram may represent a module, segment, or portion of code, which contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than those shown in the figures. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved.

[0066] If a function is implemented as a software module and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device to execute all or part of the steps of the methods in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as a USB flash drive, a portable hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0067] The above are preferred embodiments of the present invention. It should be noted that, for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A sound control method based on 3D space sound school qualification, characterized in that, The method comprises the following steps: According to the size of the sound field space and the reflection coefficient of each reflecting surface, a 3D space acoustic model is generated; For each sound unit in the sound field space, the central angle is determined according to the direction angle of the sound unit center point pointing to the center point of the listening area in the sound field space; the reference direction is determined according to the direction corresponding to the central angle; the deviation central angle is determined according to the included angle between each sound wave radiation direction and the reference direction; the main lobe width is determined according to the opening angle of the listening area; The polar coordinates are constructed with the sound unit center point as the pole, each deviation central angle as the polar angle, and the energy amplitude weight corresponding to each deviation central angle as the polar radius; the Gaussian model for solving the energy amplitude weight corresponding to each sound wave radiation direction is constructed based on the polar coordinates, the central angle, the main lobe width, and the preset initial energy amplitude weight; then the energy amplitude weight curve of each sound wave radiation direction is determined according to the Gaussian model; The time delay parameter is determined according to the distance between the sound unit center point and the listening area center point; the main beam parameter is generated according to the energy amplitude weight curve and the time delay parameter; The reflecting surface with a reflection coefficient exceeding a preset reflection coefficient is taken as a target reflecting surface; the null beam direction is determined according to the direction angle of the sound unit pointing to the target reflecting surface; the same energy amplitude weight as the null beam direction is extracted from the energy amplitude weight curve as the null beam energy amplitude weight; and the opposite phase of the main beam in the same direction is taken as the null beam phase; The null beam parameter is generated according to the null beam direction, the null beam energy amplitude weight, and the null beam phase; The sound unit is controlled according to the main beam parameter and the null beam parameter.

2. The sound control method based on 3D spatial sound calibration according to claim 1, wherein, The Gaussian model for solving the energy amplitude weight corresponding to each sound wave radiation direction is as follows: ; wherein, is an energy amplitude weight; is an off-center angle; A is an initial energy amplitude weight; is a center angle; is a main lobe width control parameter, converted from a main lobe width.

3. The sound control method based on 3D spatial sound calibration according to claim 2, wherein, The energy amplitude weight curve of each sound wave radiation direction determined according to the Gaussian model comprises the following steps: For each sound wave radiation direction, it is judged whether the current sound wave radiation direction is within the main lobe range according to the corresponding deviation central angle and the main lobe width; If yes, the corresponding energy amplitude weight is calculated according to the Gaussian model; If no, it is determined that the sound wave radiation direction is within the sidelobe range; the sidelobe attenuation enhancement model is obtained by superimposing a cosine roll-off window on the basis of the Gaussian model: ; wherein, is a boundary angle of the main lobe width; is a predetermined roll-off transition bandwidth.

4. The sound control method based on 3D spatial sound calibration according to claim 3, wherein, The initial energy amplitude weight is determined by the following method: The distance between each sound unit and the listening area is calculated; the sound unit closest to the listening area is taken as the reference sound unit; and the distance between the reference sound unit and the listening area is taken as the reference distance; The preset initial energy weight is taken as the initial energy amplitude weight of the reference sound unit; For non-reference sound units other than the reference sound unit, the distance between the non-reference sound unit and the listening area is calculated to obtain a first distance; The distance compensation amplitude coefficient is determined according to the ratio of the first distance to the reference distance; The preset initial energy weight is compensated according to the distance compensation amplitude coefficient to generate the initial energy amplitude weight corresponding to the non-reference sound unit.

5. The sound control method based on 3D spatial sound calibration according to claim 1, wherein, After the sound unit is controlled according to the main beam parameter and the null beam parameter, the following steps are further included: The 3D space acoustic model is divided into a three-dimensional grid to generate a plurality of grid points in the 3D space acoustic model; According to the actual sound pressure level of each grid point at each frequency point, an actual frequency response curve corresponding to each grid point is generated; According to the current scene type, a scene frequency response curve corresponding to the current scene is determined; According to the position of the user, a target grid point corresponding to the position of the user is determined, the actual frequency response curve of the target grid point is compared with the scene frequency response curve, and a frequency response deviation is determined; The frequency response of the sound unit is calibrated according to the frequency response deviation.

6. The sound control method based on 3D spatial sound calibration according to claim 5, wherein, The acoustic space is also provided with a microphone; The determination of the actual sound pressure level comprises: According to the reference sound pressure level of each sound unit and the position coordinates of the microphone in the 3D space acoustic model, the direct sound pressure level of each sound unit at the microphone is calculated, and then the total direct sound pressure level of all sound units at the microphone is calculated; According to the total direct sound pressure level and the measured total sound pressure level collected by the microphone, the total reflected sound pressure level at the microphone is calculated; according to the proportional relationship of the reference sound pressure levels of each sound unit, the reflected sound pressure level of each sound unit at the microphone is determined; According to the reflected sound pressure level and the direct sound pressure level of each sound unit at the microphone, the reflected sound contribution coefficient of each sound unit at the microphone is determined; For each grid point, according to the distance between the grid point and the sound unit, the reference sound pressure level of the sound unit, the direct sound pressure level contribution of each sound unit to the grid point is calculated; According to the direct sound pressure level contribution of each sound unit to the grid point, the reflected sound contribution coefficient of each sound unit at the microphone and the comprehensive reflection coefficient of the reflection surface corresponding to the grid point, the reflected sound pressure level contribution of each sound unit to the grid point is determined; According to the direct sound pressure level contribution and the reflected sound pressure level contribution of each sound unit to the grid point, the actual sound pressure level of the grid point is determined.

7. The sound control method based on 3D spatial sound calibration according to claim 6, wherein, The comprehensive reflection coefficient of the reflection surface corresponding to the grid point is determined by the following method: The distance of the grid point from each reflection surface is calculated; According to the distance between the grid point and each reflection surface, the reflection weight of each reflection surface is determined; According to the reflection weight of each reflection surface and the corresponding reflection coefficient, the comprehensive reflection coefficient is obtained by weighted fusion.

8. A sound control device based on 3D spatial sound learning, characterized by, It comprises: A 3D space acoustic model construction module, an angle determination module, an energy amplitude weight determination module, a main beam parameter determination module, a null beam parameter determination module and a sound control module; The 3D space acoustic model construction module is used to generate a 3D space acoustic model according to the size of the sound field space and the reflection coefficients of each reflection surface; The angle determination module is used to determine the center angle of each sound unit in the sound field space according to the direction angle of the sound unit center point pointing to the center point of the listening area in the sound field space; taking the direction corresponding to the center angle as the reference direction, the deviation from the center angle is determined according to the included angle between each sound wave radiation direction and the reference direction; the main lobe width is determined according to the opening angle of the listening area; The energy amplitude weight determination module is configured to construct a polar coordinate with a center point of the acoustic unit as a polar point, with each deviation center angle as a polar angle, and with an energy amplitude weight corresponding to each deviation center angle as a polar radius; construct a Gaussian model for solving the energy amplitude weight corresponding to each sound wave radiation direction based on the polar coordinate, the center angle, the main lobe width, and a preset initial energy amplitude weight; and then determine an energy amplitude weight curve of each sound wave radiation direction according to the Gaussian model. The main beam parameter determination module is configured to determine a time delay parameter according to a distance between the center point of the acoustic unit and the center point of the listening area, and generate the main beam parameter according to the energy amplitude weight curve and the time delay parameter. The null beam parameter determination module is configured to take a reflection surface with a reflection coefficient exceeding a preset reflection coefficient as a target reflection surface, determine a null beam direction according to a direction angle of the acoustic unit pointing to the target reflection surface, extract an energy amplitude weight in the same direction from the energy amplitude weight curve as a null beam energy amplitude weight, and take an opposite phase of the main beam in the same direction as a null beam phase. The null beam parameter determination module is configured to generate the null beam parameter according to the null beam direction, the null beam energy amplitude weight, and the null beam phase. The acoustic control module is configured to control the acoustic unit according to the main beam parameter and the null beam parameter.

9. A terminal device, comprising: The computer program is stored in the memory and configured to be executed by the processor, and the processor implements the acoustic control method based on the 3D space sound calibration according to any one of claims 1 to 7 when executing the computer program.

10. A computer-readable storage medium, characterized in that, The computer program is stored in the memory and configured to be executed by the processor, and the processor implements the acoustic control method based on the 3D space sound calibration according to any one of claims 1 to 7 when executing the computer program.