A method and system for testing spatial sound effects of a sound system

Abstract

The application discloses a kind of spatial sound effect test method and system of acoustic system, the method includes the following steps: S1, in the preset stage, the multimodal perception data of test environment is synchronously collected;S2, generate three-dimensional sound pressure distribution heat map, and build initial sound field model based on this;S3, generate correction sound field model, and vibration data collected by the three-axis accelerometer is synchronously recorded;S4, generate sound image positioning trajectory chart, and calculate the deviation angle of actual sound image position and theoretical position;S5, determine the sound quality score of each test stage.The application constructs the closed-loop test framework of physical environment perception-sound field real-time modeling-human-computer interaction analysis-fault linkage tracing, provides the full-process test solution covering hardware state, acoustic characteristics and human factor perception for immersive audio system, significantly improves the spatial sound effect reliability of digital playback device.

Description

TECHNICAL FIELD

[0001] The present application belongs to the field of digital technology playing equipment, more particularly, it relates to a kind of spatial sound effect test method of sound system. BACKGROUND

[0002] The current mainstream sound system has been fully digitalized, and immersive panorama sound (such as Dolby Atmos, DTS:X) and object audio (Object Audio) technology are represented, and its core features include: using multi-channel independent power amplifier to drive distributed loudspeaker array, supporting three-dimensional space sound image positioning; transmitting digital audio stream through HDMI eARC / optical fiber, cooperating with panorama sound decoder to generate dynamic metadata; introducing room correction algorithm (such as Audys frequency response stability score y) for preliminary sound field optimization;

[0003] Such equipment significantly improves the sound field space sense, but the test of acoustic performance still relies on traditional methods, which has the following shortcomings:

[0004] Static test limitation: based on single-point sound pressure meter measurement frequency response curve (such as REW software), only the static sound pressure distribution of fixed listening position can be reflected;

[0005] Separation type analysis defect: loudspeaker displacement monitoring (laser range finder), vibration detection (accelerometer) and acoustic test are independent, and the data cannot be coupled in real time;

[0006] Human factor evaluation is missing: the influence of listener head movement on sound image positioning depends on subjective listening evaluation, and there is no quantitative model;

[0007] Based on the above, we propose a kind of spatial sound effect test method and system of sound system, which solves the above problems. SUMMARY

[0008] The purpose of the present application is to solve the problems existing in the prior art, and a kind of spatial sound effect test method and system of sound system are proposed, which constructs a closed-loop test framework of physical environment perception-sound field real-time modeling-human-computer interaction analysis-fault linkage tracing, provides a full-process test solution covering hardware state, acoustic characteristics and human factor perception for immersive audio system, and significantly improves the spatial sound effect reliability of digital playing equipment.

[0009] To achieve the above purpose, the present application provides the following technical scheme:

[0010] A kind of spatial sound effect test method of sound system, comprising the following steps:

[0011] S1, in the preset stage, the multi-modal perception data of test environment is synchronously collected;

[0012] S2. Perform sound pressure level mapping processing on the sound field distribution data in the initial calibration stage to generate a three-dimensional sound pressure distribution heat map, and construct an initial sound field model based on this.

[0013] S3. Based on the speaker position data calibrated in real time by the laser rangefinder, the initial sound field model is dynamically corrected to generate a corrected sound field model, and the vibration data collected by the triaxial accelerometer is recorded simultaneously.

[0014] S4. The frequency band energy of the audio signal in the dynamic test phase is analyzed by the spectrum analyzer, the energy distribution characteristics of each frequency band are extracted and the abnormal frequency band offset is marked. At the same time, the head rotation trajectory data of the listener captured by the high-speed camera is combined with the corrected sound field model to generate a sound image positioning trajectory map and calculate the deviation angle between the actual sound image position and the theoretical position.

[0015] S5. Based on the three-dimensional sound pressure distribution heat map, the corrected sound field model, the abnormal frequency band offset, the vibration data, and the sound image positioning trajectory map, determine the sound effect quality score for each test stage.

[0016] Preferably, in step S1, the preset stage includes an initial calibration stage, a dynamic testing stage, and a final verification stage; the initial calibration stage plays a benchmark test signal, the dynamic testing stage plays a spatial sound effect test signal, and the final verification stage plays pink noise.

[0017] The multimodal sensing data includes: sound field distribution data collected by the microphone array, speaker position data acquired by the laser rangefinder, vibration data recorded by the triaxial accelerometer, frequency response curve data captured by the spectral analyzer, and listener head movement data captured by the high-speed camera.

[0018] Preferably, the microphone array adopts a spherical distributed topology structure, and the sound field distribution data it collects includes time-frequency domain sound pressure vector information. Furthermore, during the dynamic testing phase, the laser rangefinder is triggered to recalibrate the speaker position once every preset time interval.

[0019] Preferably, in step S2, the generation of the three-dimensional sound pressure distribution heat map includes: performing spatial interpolation processing on the sound field distribution data to complete the sound pressure values ​​of unmeasured points, and using an acoustic mapping algorithm to convert the sound pressure level into a heat map with spatial coordinate attributes.

[0020] Preferably, in step S3, the dynamic correction includes: reconstructing the sound wave propagation path function based on the real-time drift of the speaker position data, and marking abnormal mechanical vibrations above a threshold in the vibration data as sound field distortion compensation factors and inputting them into the corrected sound field model;

[0021] The vibration data is synchronously bound to the surface of each speaker unit housing via a triaxial accelerometer, and the recorded vibration spectrum is used to identify abnormal areas of sound field interference caused by device resonance.

[0022] Preferably, in S4, the frequency band energy analysis includes: dividing the frequency response curve captured by the spectral analyzer into critical frequency band groups, detecting the frequency response flatness by calculating the energy variance between groups, and marking the abnormal frequency band offset in the corresponding frequency band when the variance exceeds the limit;

[0023] The markers for the abnormal frequency band offsets are associated with the timestamps of the dynamic testing phase and are spatially-frequency coupled with the abnormal sound pressure regions in the corrected sound field model.

[0024] Preferably, the generation of the acoustic image localization trajectory map includes: calculating the spatial pointing vector of the binaural auditory axis based on head motion data captured by a high-speed camera, mapping the virtual sound source position in the modified sound field model, and generating an acoustic image movement path topology map.

[0025] A spatial sound effect testing system for an audio system, the system being used to implement the above-described method, comprising:

[0026] The multimodal sensing subsystem, configured in the test environment, includes: a microphone array module for acquiring sound field distribution data during the initial calibration, dynamic testing, and final verification phases; a laser ranging module for acquiring speaker position data in real time; a vibration sensing module, including a triaxial accelerometer, for recording speaker vibration data; a spectral analysis module for capturing frequency response curve data of audio signals; and a motion capture module, including a high-speed camera, for capturing listener head movement data.

[0027] The sound field modeling subsystem, connected to the multimodal sensing subsystem, includes: a heat map generation unit, which performs sound pressure level mapping processing on the sound field distribution data in the initial calibration stage to generate a three-dimensional sound pressure distribution heat map; and a dynamic correction unit, which corrects the initial sound field model based on the real-time position data of the laser ranging module and outputs the corrected sound field model.

[0028] The analysis and evaluation subsystem includes: a spectrum feature analysis unit, which extracts the energy distribution characteristics of each frequency band based on the frequency response curve data from the spectrum analysis module and marks the offset of abnormal frequency bands; a sound image trajectory generation unit, which combines head motion data from the motion capture module with the corrected sound field model to generate a sound image positioning trajectory map and calculates the deviation angle between the actual sound image position and the theoretical position; and a scoring decision unit, which generates sound quality scores for each stage based on the three-dimensional sound pressure distribution heat map, the corrected sound field model, the offset of abnormal frequency bands, vibration data, and the sound image positioning trajectory map, and evaluates the spatial sound performance by measuring the rate of change of scores during the dynamic testing stage and the final verification stage.

[0029] Preferably, the laser ranging module triggers position recalibration at a preset period, and the position drift output by the module is input into the dynamic correction unit to reconstruct the sound wave propagation path function. The abnormal mechanical vibration data of the vibration sensing module is marked as a sound field distortion compensation factor and stored in the corrected sound field model.

[0030] The spectral feature analysis unit performs frequency band energy analysis: it divides the frequency response curve into critical frequency band groups, detects the frequency response flatness through the energy variance between groups, associates the timestamp of the dynamic test stage with the abnormal frequency band offset, and performs spatial-frequency domain coupling with the sound pressure anomaly region in the corrected sound field model.

[0031] Technical effects and advantages of the present invention: The spatial sound effect testing method and system for an audio system provided by the present invention have the following effects compared with the prior art:

[0032] Real-time accurate modeling of dynamic sound field is achieved. Through the closed-loop linkage of laser ranging and sound field correction unit, the sound wave propagation path is automatically reconstructed when the speaker is displaced or vibrates, eliminating the sound pressure distribution distortion caused by temperature drift / mechanical resonance, and ensuring that the three-dimensional sound field model continuously matches the physical environment.

[0033] Establish a cross-dimensional fault tracing mechanism, couple and analyze vibration spectrum data, abnormal frequency band offset and acoustic image positioning deviation angle, accurately pinpoint the causal relationship between mechanical resonance and acoustic distortion, and provide quantifiable diagnostic basis for loudspeaker hardware faults.

[0034] The study quantifies the impact of human-factor interaction on spatial sound effects, drives the generation of sound image trajectories based on head rotation data from motion capture, objectively characterizes the degree of degradation in positioning perception under the dynamic behavior of the audience, and fills the gap in the lack of human-factor data in traditional testing.

[0035] Generate a comprehensive sound quality evaluation system that integrates multiple indicators such as sound pressure level, frequency response stability, vibration suppression, and positioning accuracy. Output a comprehensive score that reflects both the static characteristics and dynamic response of the sound field, avoiding the biased evaluation caused by single-point testing.

[0036] Optimize the calibration efficiency of digital playback devices by verifying abnormal frequency band offset and injecting sound field distortion compensation factors to improve the adaptability of the room calibration algorithm of the surround sound decoder and reduce the number of manual adjustments. Attached Figure Description

[0037] Figure 1 This is a flowchart of the spatial sound effect testing method for the audio system of the present invention;

[0038] Figure 2 This invention provides a logic diagram for determining the sound quality scores at each testing stage. Detailed Implementation

[0039] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without inventive effort are within the scope of protection of this invention.

[0040] This invention provides, for example Figure 1 The method and system for testing spatial sound effects in an audio system are presented. They construct a closed-loop testing framework encompassing "physical environment perception—real-time sound field modeling—human-computer interaction analysis—fault linkage and tracing." This framework upgrades from static detection to dynamic tracking, resolving the issue of inaccurate sound field models caused by speaker drift in digital playback devices. It also upgrades from single-dimensional separation measurement to multimodal coupling analysis, overcoming the traditional diagnostic bottleneck of fragmented vibration-audio-location data. Furthermore, it upgrades from subjective questionnaires to objective quantitative evaluation, incorporating human behavior into the spatial sound effect performance evaluation system for the first time. Ultimately, it provides a comprehensive testing solution for immersive audio systems, covering hardware status, acoustic characteristics, and human perception, significantly improving the spatial sound effect reliability of digital playback devices.

[0041] The spatial sound effect testing method for audio systems achieves multimodal collaborative perception: integrating five types of heterogeneous data—sound field distribution, device position, mechanical vibration, spectral response, and audience behavior—to realize full-dimensional coupled analysis of complex acoustic scenarios; a dynamic model correction mechanism: using a laser rangefinder to calibrate the spatial displacement of speakers in real time, achieving dynamic compensation for hardware position drift in sound field modeling for the first time; human-factor interaction verification: by fusing head trajectory with the corrected sound field model, subjective sound image positioning deviation is transformed into a quantifiable indicator; a phased quantitative evaluation chain: based on a three-stage step-by-step test, a closed-loop quality evaluation system of "initial modeling → dynamic correction → final verification" is constructed, using the rate of change indicator to dynamically reflect the system's robustness.

[0042] Includes the following steps:

[0043] S1. Synchronously collect multimodal sensing data of the test environment in the preset stage; wherein, the preset stage includes the initial calibration stage, the dynamic test stage and the final verification stage; the initial calibration stage plays the benchmark test signal, the dynamic test stage plays the spatial sound effect test signal, and the final verification stage plays pink noise.

[0044] Multimodal sensing data includes: sound field distribution data collected by microphone array, speaker position data acquired by laser rangefinder, vibration data recorded by triaxial accelerometer, frequency response curve data captured by spectral analyzer, and listener head movement data captured by high-speed camera;

[0045] In S1, the microphone array adopts a spherical distributed topology structure. The sound field distribution data it collects includes time-frequency domain sound pressure vector information. During the dynamic testing phase, the laser rangefinder is triggered to recalibrate the speaker position once every preset time interval.

[0046] S2. Perform sound pressure level mapping processing on the sound field distribution data in the initial calibration stage to generate a three-dimensional sound pressure distribution heat map, and construct an initial sound field model based on this. In S2, the generation of the three-dimensional sound pressure distribution heat map includes: performing spatial interpolation processing on the sound field distribution data to complete the sound pressure values ​​of unmeasured points, and using an acoustic mapping algorithm to convert the sound pressure level into a heat map with spatial coordinate attributes.

[0047] When generating a three-dimensional sound pressure distribution heatmap in S2, calculate the sound pressure equalization index (EQ):

[0048] in, Let be the sound pressure gradient value of the i-th grid in the heat map. P is the arithmetic mean of the sound pressure gradients across all grids in the heatmap. max To test the maximum sound pressure level in the space, P min The minimum sound pressure level in the test space is N, which is the total number of spatial grids in the heat map. This formula accurately characterizes the uniformity of the sound field by quantifying the ratio of the sound pressure gradient dispersion to the dynamic range. An EQ value close to 1 indicates that the sound field distribution is highly balanced.

[0049] S3. During the dynamic testing phase:

[0050] Based on the speaker position data calibrated in real time by the laser rangefinder, the initial sound field model is dynamically corrected to generate a corrected sound field model; vibration data collected by the triaxial accelerometer is recorded simultaneously.

[0051] In S3, dynamic correction includes: reconstructing the sound wave propagation path function based on the real-time drift of the speaker position data, and marking abnormal mechanical jitter in the vibration data that exceeds the threshold as sound field distortion compensation factors to be input into the corrected sound field model; the vibration data is synchronously bound to the surface of each speaker unit housing through a triaxial accelerometer, and the recorded vibration spectrum is used to identify abnormal areas of sound field interference caused by equipment resonance;

[0052] When marking the sound field distortion compensation factor in S3, calculate the vibration-frequency offset coupling coefficient γ:

[0053] Among them, A k Let ΔF be the acceleration amplitude at the k-th abnormal vibration frequency point. kIt is the offset of the abnormal frequency band corresponding to the vibration frequency point, and M is the total number of abnormal frequency points in the vibration spectrum. This formula reveals the degree of linear correlation between mechanical vibration and sound frequency offset. When |γ|>0.8, it is determined that there is sound field distortion caused by hardware resonance, and the weight of the compensation factor needs to be strengthened.

[0054] S4. Perform frequency band energy analysis on the audio signal during the dynamic testing phase using a spectrum analyzer, extract the energy distribution characteristics of each frequency band, and mark the abnormal frequency band offset;

[0055] Meanwhile, by using the head rotation trajectory data of the listeners captured by the high-speed camera and combining it with the corrected sound field model, a sound image positioning trajectory map is generated, and the deviation angle between the actual sound image position and the theoretical position is calculated.

[0056] In S4, frequency band energy analysis includes: dividing the frequency response curves captured by the spectral analyzer into critical frequency band groups, detecting frequency response flatness by calculating the energy variance between groups, and marking abnormal frequency band offsets in the corresponding frequency bands when the variance exceeds the limit; the marking of abnormal frequency band offsets is associated with the timestamps of the dynamic testing phase and is spatially-frequency coupled with the abnormal sound pressure regions in the corrected sound field model; the generation of the sound image positioning trajectory map includes: calculating the spatial pointing vector of the binaural auditory axis based on head motion data captured by a high-speed camera, mapping the virtual sound source position in the corrected sound field model, and generating a topology map of the sound image movement path; the calculation of the deviation angle adopts a spatial spherical coordinate system, which is achieved by comparing the difference between the azimuth and pitch angles of the virtual sound source position and the pre-coded theoretical coordinates of the spatial sound effect test signal;

[0057] A high-speed camera simultaneously captures the head motion data of at least three listeners at a frame rate of 60fps or higher. The coupling analysis of their head rotation trajectory data and the modified sound field model includes the calculation of the sound image position following delay time.

[0058] When calculating the deviation angle in S4, the dynamic acoustic positioning error angle θ is defined as follows:

[0059] in, The orientation vector of the theoretical coordinates for precoding the spatial sound effect test signal. The actual acoustic image orientation vector is calculated based on the head rotation trajectory. This formula directly quantifies the virtual sound source positioning accuracy by inversely solving the error angle through the cosine value of the spatial angle between the two orientation vectors. θ < 5° is considered as excellent spatial perception performance.

[0060] S5. Based on the three-dimensional sound pressure distribution heat map, the corrected sound field model, the abnormal frequency band offset, vibration data and sound image positioning trajectory map, determine the sound quality score for each test stage;

[0061] Calculate the rate of change of scores between the dynamic testing phase and the final verification phase, and use this rate of change as the core indicator for evaluating the spatial sound performance of the audio system.

[0062] In S5, the sound quality rating includes a base sound field rating and a dynamic response rating:

[0063] The basic sound field score is generated based on the uniformity index of the three-dimensional sound pressure distribution heatmap;

[0064] The dynamic response score is calculated by weighting the amplitude of the abnormal frequency band offset, the effective peak-to-average ratio of the vibration data, and the standard deviation of the deviation angle.

[0065] The calculation of the basic sound field score includes three sub-indicators: dynamic range of sound pressure level, regional uniformity of heat map, and low-frequency attenuation slope. Regional uniformity is evaluated by the sound pressure gradient values ​​of adjacent grids in the heat map. In the final verification stage: based on the frequency response curve data of pink noise, it is verified whether the abnormal frequency band offset marked in the dynamic test stage has been suppressed, and its correction effectiveness is converted into the frequency response stability score of the final verification stage.

[0066] The calculation of the rate of change in ratings includes:

[0067] Extract the final sound quality score at the end of the dynamic testing phase;

[0068] Obtain the frequency response stability score during the final verification phase;

[0069] By comparing the direction and magnitude of change between the final sound quality score and the frequency response stability score, attenuation / enhancement rate of change indicators are generated.

[0070] As should be noted, the spatial audio test signals played during the dynamic testing phase include:

[0071] Sweep-frequency chirped signals for sound field width testing; moving-point sound source pulse sequences for sound image localization testing; multi-channel white noise burst signal sets for reverberation characteristic testing.

[0072] The method adds an anomaly tracing step after the sound quality score calculation: when the score change rate exceeds the tolerance threshold, hardware fault location is performed based on the mechanical resonance characteristics in the vibration data, the frequency range corresponding to the abnormal frequency band offset, and the direction of the sudden increase in the deviation angle.

[0073] The initial sound field model and the corrected sound field model adopt a sound field reconstruction algorithm based on spherical harmonic functions. During the dynamic testing phase, the speaker position compensation coefficient is updated every 0.5 seconds to maintain the real-time performance of the model.

[0074] Calculate the dynamic response score S in S5 d At that time, a frequency-time domain hybrid weighted model is adopted:

[0075] Where, ΔF max σ represents the absolute value of the maximum abnormal frequency band offset during the dynamic testing phase. θ Let RMS be the standard deviation of the deviation angle θ. A α, β, κ, and Q are the effective root mean square values ​​of acceleration for vibration data, and α, β, κ, and Q are system constants calibrated using standard acoustic scenarios. The exponential term of this formula suppresses the influence of frequency offset, while the logarithmic term suppresses vibration and positioning jitter, achieving nonlinear cross-dimensional degradation suppression and making the score more consistent with subjective listening sensitivity to degradation.

[0076] This implementation also proposes a spatial sound effect testing system for an audio system, the system being used to implement the above-mentioned method, including:

[0077] The multimodal perception subsystem, configured in the test environment, includes:

[0078] The microphone array module is used to collect sound field distribution data during the initial calibration phase, dynamic testing phase, and final verification phase.

[0079] The laser ranging module is used to acquire speaker position data in real time. The laser ranging module triggers position recalibration at a preset period. The position drift output is input into the dynamic correction unit to reconstruct the sound wave propagation path function. The abnormal mechanical vibration data of the vibration sensing module is marked as sound field distortion compensation factor and stored in the corrected sound field model.

[0080] The vibration sensing module includes a triaxial accelerometer for recording speaker vibration data;

[0081] The spectral analysis module is used to capture the frequency response curve data of audio signals;

[0082] The motion capture module includes a high-speed camera for capturing head movement data of the audience.

[0083] The sound field modeling subsystem, connected to the multimodal perception subsystem, includes:

[0084] The heatmap generation unit performs sound pressure level mapping processing on the sound field distribution data in the initial calibration stage to generate a three-dimensional sound pressure distribution heatmap;

[0085] The dynamic correction unit corrects the initial sound field model based on the real-time position data of the laser ranging module and outputs the corrected sound field model.

[0086] The analysis and evaluation subsystem includes:

[0087] The spectrum feature analysis unit extracts the energy distribution characteristics of each frequency band based on the frequency response curve data of the spectrum analysis module and marks the offset of abnormal frequency bands; the spectrum feature analysis unit performs frequency band energy analysis: divides the frequency response curve into critical frequency band groups, detects the frequency response flatness through the energy variance between groups, marks the offset of abnormal frequency bands by associating the timestamp of the dynamic test stage, and performs spatial-frequency domain coupling with the sound pressure anomaly region in the corrected sound field model.

[0088] The sound image trajectory generation unit combines the head motion data from the motion capture module with the corrected sound field model to generate a sound image positioning trajectory map and calculates the deviation angle between the actual sound image position and the theoretical position. The sound image trajectory generation unit is equipped with a spatial spherical coordinate system solver, which calculates the deviation angle by comparing the difference between the virtual sound source azimuth / pitch angle and the pre-encoded theoretical coordinates of the spatial sound effect test signal, and displays the head rotation trajectory and the sound image movement path topology map in real time.

[0089] The scoring decision unit generates sound quality scores for each stage based on a 3D sound pressure distribution heatmap, a corrected sound field model, abnormal frequency band offsets, vibration data, and acoustic image localization trajectory maps. It also evaluates spatial sound performance by assessing the rate of change in scores between the dynamic testing and final verification stages. The scoring decision unit includes a dual-channel calculation engine for both basic sound field scoring and dynamic response scoring.

[0090] The basic sound field score is based on the uniformity index of the heatmap, including the dynamic range of sound pressure level, regional uniformity, and low-frequency attenuation slope sub-indices.

[0091] The dynamic response score is output by weighting the amplitude of the abnormal frequency band offset, the peak-to-average ratio of vibration, and the standard deviation of the deviation angle.

[0092] The pink noise frequency response curve captured in the final verification phase is used to verify the effect of suppressing abnormal frequency band offsets and generate a frequency response stability score.

[0093] In addition, the system adds an anomaly tracing actuator: when the rate of change of the score exceeds the tolerance threshold, a hardware fault location report is generated based on the mechanical resonance characteristics of the vibration spectrum, the frequency range of the abnormal frequency band offset, and the direction of the sudden increase in the deviation angle.

[0094] In summary, the present invention has the following effects:

[0095] Real-time accurate modeling of dynamic sound field is achieved. Through the closed-loop linkage of laser ranging and sound field correction unit, the sound wave propagation path is automatically reconstructed when the speaker is displaced or vibrates, eliminating the sound pressure distribution distortion caused by temperature drift / mechanical resonance, and ensuring that the three-dimensional sound field model continuously matches the physical environment.

[0096] Establish a cross-dimensional fault tracing mechanism, couple and analyze vibration spectrum data, abnormal frequency band offset and acoustic image positioning deviation angle, accurately pinpoint the causal relationship between mechanical resonance and acoustic distortion, and provide quantifiable diagnostic basis for loudspeaker hardware faults.

[0097] The study quantifies the impact of human-factor interaction on spatial sound effects, drives the generation of sound image trajectories based on head rotation data from motion capture, objectively characterizes the degree of degradation in positioning perception under the dynamic behavior of the audience, and fills the gap in the lack of human-factor data in traditional testing.

[0098] Generate a comprehensive sound quality evaluation system that integrates multiple indicators such as sound pressure level, frequency response stability, vibration suppression, and positioning accuracy. Output a comprehensive score that reflects both the static characteristics and dynamic response of the sound field, avoiding the biased evaluation caused by single-point testing.

[0099] Optimize the calibration efficiency of digital playback devices by verifying abnormal frequency band offset and injecting sound field distortion compensation factors to improve the adaptability of the room calibration algorithm of the surround sound decoder and reduce the number of manual adjustments.

[0100] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for testing the spatial sound effects of an audio system, characterized in that, Includes the following steps: S1. Synchronously collect multimodal sensing data of the test environment in a preset stage; in step S1, the preset stage includes an initial calibration stage, a dynamic testing stage, and a final verification stage. The initial calibration phase plays a benchmark test signal, the dynamic test phase plays a spatial sound effect test signal, and the final verification phase plays pink noise. The multimodal sensing data includes: sound field distribution data collected by the microphone array, speaker position data acquired by the laser rangefinder, vibration data recorded by the triaxial accelerometer, frequency response curve data captured by the spectral analyzer, and listener head movement data captured by the high-speed camera. S2. Perform sound pressure level mapping processing on the sound field distribution data in the initial calibration stage to generate a three-dimensional sound pressure distribution heat map, and construct the initial sound field model based on this. S3. Based on the speaker position data calibrated in real time by the laser rangefinder, the initial sound field model is dynamically corrected to generate a corrected sound field model, and the vibration data collected by the triaxial accelerometer is recorded simultaneously. In step S3, the dynamic correction includes: reconstructing the sound wave propagation path function according to the real-time drift of the speaker position data, and marking abnormal mechanical vibrations above the threshold in the vibration data as sound field distortion compensation factors and inputting them into the corrected sound field model. The vibration data is synchronously bound to the surface of each speaker unit housing via a triaxial accelerometer, and the recorded vibration spectrum is used to identify abnormal areas of sound field interference caused by device resonance. S4. The audio signal in the dynamic test phase is analyzed by a spectrum analyzer to extract the energy distribution characteristics of each frequency band and mark the offset of abnormal frequency bands. At the same time, the head rotation trajectory data of the listener captured by the high-speed camera is combined with the modified sound field model to generate a sound image positioning trajectory map and calculate the deviation angle between the actual sound image position and the theoretical position. S5. Based on the three-dimensional sound pressure distribution heat map, the corrected sound field model, the abnormal frequency band offset, the vibration data, and the sound image positioning trajectory map, determine the sound effect quality score for each test stage.

2. The spatial sound effect testing method for an audio system according to claim 1, characterized in that, The microphone array adopts a spherical distributed topology structure. The sound field distribution data it collects includes time-frequency domain sound pressure vector information. During the dynamic testing phase, the laser rangefinder is triggered to recalibrate the speaker position once every preset time interval.

3. The spatial sound effect testing method for an audio system according to claim 1, characterized in that, In step S2, the generation of the three-dimensional sound pressure distribution heat map includes: performing spatial interpolation processing on the sound field distribution data to complete the sound pressure values ​​of unmeasured points, and using an acoustic mapping algorithm to convert the sound pressure level into a heat map with spatial coordinate attributes.

4. The spatial sound effect testing method for an audio system according to claim 1, characterized in that, In S4, the frequency band energy analysis includes: dividing the frequency response curve captured by the spectral analyzer into critical frequency band groups, detecting the frequency response flatness by calculating the energy variance between groups, and marking the abnormal frequency band offset in the corresponding frequency band when the variance exceeds the limit; The markers for the abnormal frequency band offsets are associated with the timestamps of the dynamic testing phase and are spatially-frequency coupled with the abnormal sound pressure regions in the corrected sound field model.

5. The spatial sound effect testing method for an audio system according to claim 4, characterized in that, The generation of the acoustic image localization trajectory map includes: calculating the spatial pointing vector of the binaural auditory axis based on head motion data captured by a high-speed camera, mapping the virtual sound source position in the modified sound field model, and generating an acoustic image movement path topology map.

6. A spatial sound effect testing system for an audio system, characterized in that, The system is used to implement the method according to any one of claims 1-5, comprising: The multimodal sensing subsystem, configured in the test environment, includes: a microphone array module for acquiring sound field distribution data during the initial calibration, dynamic testing, and final verification phases; a laser ranging module for acquiring speaker position data in real time; a vibration sensing module, including a triaxial accelerometer, for recording speaker vibration data; a spectral analysis module for capturing frequency response curve data of audio signals; and a motion capture module, including a high-speed camera, for capturing listener head movement data. The sound field modeling subsystem, connected to the multimodal sensing subsystem, includes: a heat map generation unit, which performs sound pressure level mapping processing on the sound field distribution data in the initial calibration stage to generate a three-dimensional sound pressure distribution heat map; and a dynamic correction unit, which corrects the initial sound field model based on the real-time position data of the laser ranging module and outputs the corrected sound field model. The analysis and evaluation subsystem includes: a spectrum feature analysis unit, which extracts the energy distribution characteristics of each frequency band based on the frequency response curve data from the spectrum analysis module and marks the offset of abnormal frequency bands; a sound image trajectory generation unit, which combines head motion data from the motion capture module with the corrected sound field model to generate a sound image positioning trajectory map and calculates the deviation angle between the actual sound image position and the theoretical position; and a scoring decision unit, which generates sound quality scores for each stage based on the three-dimensional sound pressure distribution heat map, the corrected sound field model, the offset of abnormal frequency bands, vibration data, and the sound image positioning trajectory map, and evaluates the spatial sound performance by measuring the rate of change of scores during the dynamic testing stage and the final verification stage.

7. A spatial sound effect testing system for an audio system according to claim 6, characterized in that, The laser ranging module triggers position recalibration at a preset period, and the position drift output is input into the dynamic correction unit to reconstruct the sound wave propagation path function. The abnormal mechanical vibration data of the vibration sensing module is marked as sound field distortion compensation factor and stored in the corrected sound field model. The spectral feature analysis unit performs frequency band energy analysis: it divides the frequency response curve into critical frequency band groups, detects the frequency response flatness through the energy variance between groups, associates the timestamp of the dynamic test stage with the abnormal frequency band offset, and performs spatial-frequency domain coupling with the sound pressure anomaly region in the corrected sound field model.

Images