Sound effect optimization method, system and equipment based on adaptive algorithm

Abstract

The invention discloses a sound effect optimization method, system and equipment based on an adaptive algorithm, and relates to the technical field of stage sound equipment, and the method comprises the steps: obtaining stage three-dimensional sound field topological parameters, and constructing a reference sound field model through combining a preset sound equipment parameter table; obtaining audio source characteristic information, and generating a first optimized sound effect model according to the reference sound field model and the audio source characteristic information; generating a second optimized sound effect model through an acoustic compensation algorithm according to the reference sound field model and the first optimized sound effect model; based on the first sound effect optimization model and the second sound effect optimization model, triggering an adaptive regulation and control instruction, the adaptive regulation and control instruction comprising a sound field calibration instruction and a sound effect optimization instruction; the invention provides the stage sound equipment capable of dynamically optimizing the sound effect according to the site change.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of stage sound, in particular to a sound effect optimization method, system and device based on an adaptive algorithm. BACKGROUND

[0002] At present, in the traditional stage sound system, due to the lack of flexibility and adaptability, the system often cannot adjust the sound effect in time according to the real-time situation of the change of the venue environment and the distribution of the audience. This leads to many problems, such as uneven sound field distribution, interference of reverberation effect, and distortion of sound, which negatively affects the listening experience of the audience, making it difficult for the audience to enjoy the best auditory feast, and thus needs to be improved. SUMMARY

[0003] In order to provide a stage sound that can dynamically optimize sound effect according to the change of the venue, the present application provides a sound effect optimization method, system and device based on an adaptive algorithm.

[0004] In the first aspect, the application aims to achieve the following technical solutions: A sound effect optimization method based on an adaptive algorithm, comprising: Obtaining a three-dimensional sound field topology parameter of a stage, combining a preset sound equipment parameter table to construct a reference sound field model; Obtaining audio source characteristic information, generating a first optimized sound effect model according to the reference sound field model and the audio source characteristic information; Generating a second optimized sound effect model through an acoustic compensation algorithm according to the reference sound field model and the first optimized sound effect model; triggering an adaptive control instruction based on the first optimized sound effect model and the second optimized sound effect model, the adaptive control instruction including a sound field calibration instruction and a sound effect optimization instruction.

[0005] By adopting the above technical solutions, a stage sound that can dynamically optimize sound effect according to the change of the venue is provided; the reference sound field model realizes the digital representation of the stage acoustic environment, and the first optimized sound effect model is generated in combination with the audio source characteristic information, which can dynamically adapt to the physical properties of the sound source (such as spectral characteristics and dynamic range), solving the sound field distortion problem caused by the separation of sound source and environmental parameters in traditional methods; the second optimized sound effect model is generated by introducing an acoustic compensation algorithm, and time-varying compensation parameters (such as reverberation time correction and dynamic compression factor) are superimposed on the basis of the reference model, forming a double-layer optimization structure of "theoretical reference + real-time compensation", which significantly improves the uniformity of the sound field and the accuracy of the frequency response. Finally, through the adaptive control instruction, a closed-loop feedback is realized, solving the pain point that the static parameters of the traditional system cannot cope with the change of the complex acoustic environment.

[0006] In a preferred example of the present application, the acquiring audio source characteristic information, generating a first optimized sound effect model according to the reference sound field model and the audio source characteristic information specifically comprises: Acquiring audio source characteristic information, the audio source characteristic information including timbre characteristics, dynamic range parameters and spectral distribution parameters; obtaining initial sound effect quality evaluation parameters according to the audio source characteristic information and acoustic environment parameters, the initial sound effect quality evaluation parameters including frequency response balance parameters and sound field coverage uniformity parameters; Generating a first optimized sound effect model according to the initial sound effect quality evaluation parameters and the reference sound field model.

[0007] By adopting the above technical solution, the timbre characteristics, dynamic range and spectral distribution parameters of the audio source can be accurately identified, and the physical characteristics of the sound source (such as the harmonic components of musical instruments and the emotional frequency bands of human voices) can be accurately identified. Combined with the acoustic environment parameters (such as the sound absorption coefficient and the sound source positioning coordinates), the initial sound effect quality evaluation parameters (frequency response balance and sound field coverage uniformity) can quantitatively evaluate the adaptation degree of the sound source and the environment; the reference sound field model serves as a theoretical reference, and by comparing the deviation values of the evaluation parameters (such as ΔL≤3dB and CV≤0.15), the first optimized sound effect model can dynamically compensate for the mismatch between the sound source and the environment. For example, in a scene where the low-frequency attenuation is excessive, the model will automatically enhance the low-frequency gain, so that the sound field energy distribution tends to be reasonable, and the problem of poor environmental adaptability caused by the dependence on fixed parameters in traditional methods is solved.

[0008] In a preferred example of the present application, the generating a second optimized sound effect model according to the reference sound field model and the first optimized sound effect model specifically comprises: According to the initial sound effect quality evaluation parameters, obtaining acoustic characteristic compensation parameters through an acoustic compensation algorithm, and generating a second sound field model according to the reference sound field model and the acoustic characteristic compensation parameters; Generating a second optimized sound effect model according to the second sound field model and the first optimized sound effect model.

[0009] By adopting the above technical solution, the acoustic characteristic compensation parameters (such as the dynamic compression factor α and the reverberation time correction amount τ) can predict the dynamic evolution of the sound field over time (such as the reverberation accumulation effect) by combining the compensation parameters with the reference sound field model to generate the second sound field model. After the operation and processing of the first optimized sound effect model, the second optimized sound effect model has both static parameter optimization and time-varying compensation capabilities; for example, when it is detected that the early reflected sound is too strong, the model will automatically extend the reverberation time correction amount to offset the reflected sound interference.

[0010] In a preferred example of the present application, the triggering an adaptive control instruction based on the first optimized sound effect model and the second optimized sound effect model specifically comprises: trigger a sound field calibration instruction based on the first optimized sound effect model and the second optimized sound effect model; trigger a sound effect optimization instruction based on the first optimized sound effect model and the second optimized sound effect model and a preset sound pressure level limit threshold, and obtain first optimized sound effect results and second optimized sound effect results respectively; obtain sound effect optimization results according to the first optimized sound effect results and the second optimized sound effect results.

[0011] By adopting the above technical solutions, the sound system can adjust its own settings in real time to achieve the best performance without manual intervention. This not only simplifies the operation process, but also ensures that the sound system can provide stable and high-quality sound effect output even in complex and variable environments; in addition, considering the preset sound pressure level limit threshold, it can also effectively avoid hearing damage or equipment overload caused by excessive volume.

[0012] In a preferred example of the present application: the adaptive control instruction triggered based on the first optimized sound effect model and the second optimized sound effect model further comprises: based on the frequency response characteristic curve of the first optimized sound effect model, dynamically adjust the following parameters through the preset sound pressure level limit threshold: beamforming parameters of the directional loudspeaker array, including a horizontal pointing angle θ ∈ [30°, 150°] and a vertical pointing angle φ ∈ [10°, 90°]; adjustable damping coefficient of the sound-absorbing material, with an adjustment range of ρ ∈ [0.2, 0.8]; delay compensation amount of the stage return listening system; when the sound field uniformity CV is detected to be greater than 0.15, start the multi-band equalizer for local compensation, and the compensation frequency bands are divided into 20Hz-80Hz, 80Hz-800Hz, and 800Hz-20kHz.

[0013] By adopting the above technical solutions, advanced functions such as dynamically adjusting the beamforming parameters of the directional loudspeaker array, the adjustable damping coefficient of the sound-absorbing material, and the delay compensation amount of the stage return listening system are introduced, especially when the sound field uniformity exceeds a certain threshold, the multi-band equalizer is started for local compensation; by precisely controlling these parameters, the uniformity and coverage of the sound field can be significantly improved, which is beneficial to ensure that every listener at each position can obtain a consistent auditory experience.

[0014] In a preferred example of the present application: the beamforming parameters of the directional loudspeaker array include: horizontal pointing angle vertical pointing angle wherein (x s , ys , z s ) is the sound source coordinate; (x a , y a , z a ) is the array center coordinate; Adjustable damping coefficient of sound-absorbing material wherein ρ0 is the initial damping coefficient, α is the attenuation factor, and ΔL(τ) is the change amount of frequency response uniformity over time; Delay compensation amount of stage return listening system wherein d sb is the distance from the sound source to the main loudspeaker; d sa is the distance from the sound source to the return listening system; v s is the sound speed; τ proc is the signal processing delay.

[0015] By adopting the above technical solutions, the geometric calculation model of the beamforming parameter realizes the spatial directional control of sound energy, and through dynamically adjusting the horizontal / vertical pointing angle, the main lobe of the sound wave can accurately cover the target area (such as a specific block of the audience seat), and the side lobe suppression ratio is increased by more than 15 dB. The exponential decay model of the damping coefficient of the sound-absorbing material gives the system the ability to adaptively adjust, and when the frequency response uniformity suddenly changes (such as low-frequency resonance), the damping coefficient can be automatically increased to suppress the standing wave effect; the multivariate fusion formula of the delay compensation amount of the return listening system ensures that the time synchronization accuracy of the direct sound and the return listening sound reaches the millisecond level, and eliminates the sound image positioning deviation; the Q value grading strategy (low-frequency quality factor Q≥2.5, high-frequency quality factor Q≤1.0) of the multi-band equalizer combined with the standard deviation of the frequency band energy realizes fine compensation.

[0016] In a preferred example of the present application: when the sound field unevenness CV is detected to be greater than 0.15, the multi-band equalizer is started to perform local compensation, which specifically includes: When the quality factor Q value is greater than or equal to 2.5, the compensation frequency band is divided into 20Hz-80Hz; When the quality factor Q value is in [1.2-1.8], the compensation frequency band is divided into 80Hz-800Hz; When the quality factor Q value is less than or equal to 1.0, the compensation frequency band is divided into 800Hz-20KHz; The compensation amount ΔG(f) of each frequency band is calculated by the following formula: P target (f) is the target sound pressure level; P measured(f) is a frequency weighting factor (low and mid frequencies γ = 0.8, high frequencies γ = 1.2) and σ(f) is the standard deviation of the energy distribution within the band.

[0017] By adopting the technical scheme, when it is detected that the sound field unevenness exceeds a specific threshold, how to select a suitable compensation frequency band according to a quality factor Q value and calculate a compensation amount of each frequency band by using a formula. The local compensation mechanism based on the multi-frequency equalizer can solve the sound imbalance problem of different frequency bands in a targeted manner.

[0018] In the second aspect, the application aims to achieve the following technical scheme: An acoustic effect optimization system based on an adaptive algorithm, the system comprising: A sound field modeling module configured to obtain three-dimensional sound field topological parameters of a stage and construct a reference sound field model in combination with a preset acoustic device parameter table; A first sound effect modeling module configured to obtain audio source characteristic information and generate a first optimized sound effect model according to the reference sound field model and the audio source characteristic information; An acoustic compensation module configured to generate a second optimized sound effect model by an acoustic compensation algorithm according to the reference sound field model and the first optimized sound effect model; An adaptive regulation module configured to trigger an adaptive regulation instruction based on the first optimized sound effect model and the second optimized sound effect model, the adaptive regulation instruction comprising a sound field calibration instruction and a sound effect optimization instruction, to adaptively adjust an output effect of an acoustic system.

[0019] By adopting the technical scheme, each module bears a specific function, from obtaining and analyzing sound field parameters to generating an optimized model, and finally triggering an adaptive regulation instruction to achieve optimal sound effect output; the modular design not only improves the scalability and maintainability of the system, but also enhances the practicality and flexibility of the system; through real-time monitoring and automatic adjustment, the system can quickly respond to environmental changes. In the third aspect, the application aims to achieve the following technical scheme: A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the above-mentioned acoustic effect optimization method based on an adaptive algorithm when executing the computer program.

[0020] In the fourth aspect, the application aims to achieve the following technical scheme: A computer-readable storage medium storing a computer program, wherein the computer program is executable by a processor to implement the steps of the above-mentioned acoustic effect optimization method based on an adaptive algorithm.

[0021] In summary, the present application includes at least one of the following beneficial technical effects: 1. Detailed definition of the beamforming parameters of the directional loudspeaker array, the adjustable damping coefficient of the sound-absorbing material, and the specific calculation method of the delay compensation amount of the stage return listening system. By accurately adjusting these parameters, the system can dynamically optimize the sound field distribution and sound performance. For example, adjusting the horizontal and vertical pointing angles of the directional loudspeaker can effectively control the direction and range of sound propagation, thereby reducing unnecessary reflections and interference; a stage sound system that can dynamically optimize sound effects according to changes in the venue; 2. When detecting that the sound field unevenness exceeds a certain threshold, how to select the appropriate compensation frequency band according to the quality factor Q value, and calculate the compensation amount of each frequency band using the formula; Based on the local compensation mechanism of the multi-band equalizer, it can specifically solve the sound imbalance problem of different frequency bands; The system not only effectively improves the sound clarity and smoothness of low, medium and high frequency bands, but also significantly improves the consistency and stability of the overall sound field. BRIEF DESCRIPTION OF DRAWINGS

[0022] Figure 1 is a flowchart of an audio effect optimization method based on an adaptive algorithm in an embodiment of the present application; Figure 2 is a device schematic diagram in an embodiment of the present application. DETAILED DESCRIPTION

[0023] The present application will be further described in detail below in conjunction with the accompanying drawings.

[0024] In an embodiment, as shown in Figure 1 The present application discloses an audio effect optimization method based on an adaptive algorithm, specifically including the following steps: S1: Obtain the three-dimensional sound field topology parameters of the stage, and construct a reference sound field model combined with a preset audio equipment parameter table.

[0025] In this embodiment, the three-dimensional sound field topology parameters include the sound wave reflection coefficient matrix, the sound source positioning coordinate system, and the environmental noise spectrum characteristics; wherein the sound wave reflection coefficient matrix describes the absorption, reflection and scattering characteristics of different material surfaces (such as walls, floors, ceilings) to sound energy when sound waves propagate in the stage space, mainly including the sound absorption coefficient and the geometric diffusion coefficient, wherein the proportion of material absorbing sound energy (dimensionless) is calculated by impulse response measurement (such as MLS signal): wherein α dB is the sound pressure level attenuation value. The geometric diffusion coefficient measures the diffusion degree of sound waves in space, which is related to the shape and roughness of the material surface, and the formula is: wherein A is the diffusion area; S is the surface area; λ is the wavelength.

[0026] The sound source positioning coordinate system establishes the coordinate mapping relationship between the stage space and the audience space, is used for accurately positioning the sound source and the sound equipment, adopts the SLAM technology, combines the laser radar and the IMU inertial sensor, and constructs the three-dimensional space coordinate system in real time; and the environmental noise spectrum characteristic is the energy distribution characteristic of the background noise in the stage area, including the frequency band range and the intensity, mainly including the air conditioner noise, the crowd noise and the equipment fan noise.

[0027] The sound equipment parameter table includes the loudspeaker parameter, the signal processing equipment parameter and the sound absorption material parameter (type and installation parameter); wherein the loudspeaker parameter includes the directivity pattern, the frequency division point and the rated power, wherein the directivity pattern includes the on-axis frequency response curve, and the frequency division point is the bass unit cutoff frequency and the treble unit starting frequency. The signal processing equipment parameter mainly includes the DSP parameter and the power amplifier parameter, wherein the DSP parameter is embodied by the ED curve, the delay amount and the amplitude limiting threshold, and the power amplifier parameter includes the damping coefficient.

[0028] Specifically, the distributed microphone array (such as MEMS digital microphone, sampling rate ≥96kHz) and the acoustic camera (containing 32-channel ring microphone array) are arranged in the stage area to collect full-band (such as 20Hz-20kHz) sound field data. The preset sound equipment parameter table includes the device database of the parameters of all sound equipment in the stage pre-constructed and the parameter matching rule. The device database includes the directivity pattern (on-axis frequency response curve, off-axis attenuation slope) of the loudspeaker, the rated power, the frequency division point (such as bass unit cutoff 80Hz), the DSP parameter (EQ curve, delay amount) of the processor, the damping coefficient (such as ξ=0.3-0.7) of the power amplifier and the type, thickness (such as 5cm-30cm) and installation angle (vertical, inclined) of the sound absorption material. The parameter matching rule refers to the preset equipment combination rule according to the function of the stage area (such as the main expansion area and the ear return area), for example, the linear array speaker is selected for the main expansion area, and the compact active monitoring speaker is configured for the ear return area.

[0029] Specifically, the acoustic simulation software (such as EASE Focus 3) based on the finite element method (FEM) is adopted, the stage CAD model (containing material properties and equipment layout) is imported, the boundary conditions and the equipment parameters are set based on the actual stage area, wherein the boundary conditions are such as the wall sound absorption coefficient, the ground material (concrete, wooden floor) and the ceiling height; the equipment parameters are such as the loudspeaker coordinates and the directivity file; then the model training is performed to obtain the reference sound field model.

[0030] S2: acquiring audio source characteristic information, generating a first optimized sound effect model according to the reference sound field model and the audio source characteristic information.

[0031] In this embodiment, the first optimized sound effect model solves the initial mismatch problem of sound source and environmental parameters, emphasizes the determinacy of frequency response equalization and sound field coverage, for example, before a theater performance, by measuring the frequency response of the loudspeaker and the sound field data of the audience, a static optimization model is generated to ensure that the full-band sound pressure level deviation is ≤3dB and the sound field uniformity meets the standard.

[0032] Specifically, step S2 includes: S21: Obtain audio source characteristic information, which includes timbre characteristics, dynamic range parameters, and spectral distribution parameters.

[0033] In this embodiment, timbre characteristics refer to the unique auditory attributes of an audio source, determined by harmonic components and spectral envelope; high-precision microphones (such as Neumann KM184) are used to collect audio source signals (such as a violin solo). The spectrum is calculated by short-time Fourier transform (STFT), and MFCC coefficients and harmonic component proportion (harmonic energy / total energy≥0.6 for human voice) are extracted. Dynamic range (DR) reflects the contrast between strong and weak audio signals, and compression parameters / expansion parameters control the signal amplitude change rate; to measure the dynamic range parameters: apply -20dBFS pink noise, simulate the actual performance signal through the CCIR noise generator, to measure the ratio of the maximum peak value (A_peak) to the effective value (A_rms), and calculate the dynamic range; spectral distribution parameters describe the energy proportion of audio signals at different frequencies, which can be recorded using a 1 / 24 octave analyzer to record the full-band spectrum of the signal (20Hz-20kHz), for example, the high-frequency overtone of the piano is concentrated in the 5kHz-8kHz interval, and the high-frequency response needs to be optimized.

[0034] S22: Obtain initial sound quality evaluation parameters based on audio source characteristic information and acoustic environment parameters, including frequency response equalization parameters and sound field coverage uniformity parameters.

[0035] In this embodiment, the frequency response equalization parameter is used to measure the deviation of the actual frequency response of the audio source from the target frequency response; the sound field coverage uniformity parameter is used to evaluate the energy distribution uniformity of the sound field in the audience area.

[0036] Specifically, the calculation of the frequency response equalization parameter includes: Load the AES-1986 standard target curve as the target frequency response curve, collect 20Hz-20kHz frequency response data as the measured frequency response curve through the acoustic measurement system (such as SoundCheck), and calculate the frequency response equalization ΔL(f) = |(L measured (f)-L target (f))|, where L measured (f) is the actual measured sound pressure level at frequency f; L target(f) is the target sound pressure level at frequency f.

[0037] The calculation of the sound field coverage uniformity parameter includes: 25 measurement points are arranged in the audience stand (spacing ≤ 2 m), the sound pressure level (SPL) of each point is recorded, and the standard deviation is calculated: Wherein, i is the identification of the measurement point; N is the total number of measurement points (such as 25 points arranged in the audience stand, then N = 25); L i is the actual sound pressure level of the i-th measurement point; is the average sound pressure level of all measurement points (unit: dB); the evaluation formula of the sound field coverage uniformity CV is If CV > 0.15, it is determined that the sound field is not uniform.

[0038] S23: generating a first optimized sound effect model according to the initial sound effect quality evaluation parameter and the reference sound field model.

[0039] In this embodiment, based on the reference sound field model (FEM simulation result), the acoustic transfer function of each loudspeaker to the measurement point is calculated: Wherein, f is the current analysis frequency point; G j (f) is the frequency response gain of the j-th loudspeaker; j is the identification of the loudspeaker, and M is the total number of loudspeakers; Q j (f) is the quality factor; f 0j is the center frequency; j ′ is the imaginary unit, indicating the phase information.

[0040] The optimization objectives of the frequency response equalization degree and the sound field coverage uniformity are quantified by a mathematical model, and the target optimization function is: Wherein, ω1 is the regularization weight coefficient; threshold is the maximum allowed CV value (such as 0.15); ΔL(f) is the absolute difference value between the actual sound pressure level and the target sound pressure level at frequency f, that is, the frequency response deviation; ΔL max is the maximum allowed sound pressure level deviation; the integral interval [f1, f2] is the frequency range covered by the optimization objective.

[0041] Then the gradient descent method combined with the genetic algorithm is used to balance the convergence speed and global optimality to generate the first optimized sound effect model: first set the initialization parameters, wherein the loudspeaker gain (G_i = 1), the delay amount (τ_i = 0), and the EQ parameters (such as 31 segment parameter equalization); then perform model iteration optimization: wherein the gradient descent adjusts the parameters in the negative gradient direction of the frequency response deviation ΔL(f); the genetic algorithm uses random perturbation parameter combination to select the optimal solution (such as the minimum fitness function J) to train the parameter set of the first optimized sound effect model.

[0042] Specifically, the training and verification method of the first optimized sound effect model plays back the optimized signal through an acoustic measurement system, and measures the frequency response curve and the sound field uniformity. If ΔL(f)≤3dB and CV≤0.15, the model passes the verification.

[0043] S3: generating a second optimized sound effect model through an acoustic compensation algorithm according to the reference sound field model and the first optimized sound effect model.

[0044] In this embodiment, the second optimized sound effect model is a dynamic compensation mechanism that adjusts acoustic parameters (such as compression factor and reverberation time) in real time to cope with environmental changes, thereby improving the anti-interference ability and robustness of the system. For example, if the detected sound field non-uniformity (CV=0.18) exceeds the threshold in a performance, the dynamic compensation is triggered.

[0045] Specifically, step S3 includes: S31: obtaining acoustic characteristic compensation parameters through an acoustic compensation algorithm according to the initial sound effect quality evaluation parameters, and generating a second sound field model according to the reference sound field model and the acoustic characteristic compensation parameters.

[0046] In this embodiment, the initial sound effect quality evaluation parameters include the frequency response balance and the sound field coverage uniformity; the second sound field model is a dynamic model with dynamic compensation capability, which is a dynamic model fused from the reference sound field model and the acoustic characteristic compensation parameters, and is used to describe the compensated sound field characteristics (such as the corrected frequency response curve and the reverberation time distribution); the acoustic compensation algorithm is a dynamic adjustment algorithm designed for the initial sound effect defects, including frequency response balance compensation based on FIR / IIR filter design, reverberation time correction based on RT60 adjustment, and directivity optimization by adjusting the speaker pointing angle; the acoustic characteristic compensation parameters are specific parameters output by the acoustic compensation algorithm for correcting acoustic defects, including filter coefficients (such as 5kHz gain+3dB and 250Hz gain-2dB), absorption curtain opening degree, and speaker pointing angle offset.

[0047] In this embodiment, the type of acoustic compensation algorithm is determined first: If the frequency response balance deviation (<500Hz) is large (such as ΔL=4.2dB), a low-frequency improvement filter is selected (wherein an IIR notch filter suppresses resonance and an FIR equalizer improves weak frequencies); If the sound field coverage uniformity CV>0.15 (sound field non-uniformity), a directivity optimization algorithm is selected (such as adjusting the speaker beam to point to the center area of the audience seat); For example, for the problem of ΔL=4.2dB in the 250Hz frequency band, an IIR biquadratic filter (center frequency 250Hz, Q=1.5, gain+3dB) is selected; for the problem of CV=0.18, a beamforming algorithm is selected (narrowing the main lobe width from 120° to 90° and pointing to the center of the audience seat).

[0048] Then, the acoustic characteristic compensation parameters are obtained through the following acoustic compensation algorithm, taking the FIR filter as an example: Frequency response L after FIR filter compensation comp (f) = L messured (f)+H(f), where H(f) is the filter frequency response, satisfying: max f |L comp (f)-L target (f)|≤ΔL max , where L comp (f) is the measured frequency response, L target (f) is the target frequency response; using minimum phase FIR design, input the target frequency response L target (f) and the measured frequency response L comp (f) Output filter coefficient h(n) (n=0, 1, ..., N-1).

[0049] Directivity optimization parameter calculation, taking beamforming as an example: Assuming that the goal is to adjust the delay and gain of 16 line array speakers so that the main lobe of the synthesized beam points to the center of the audience (angle θ = 30°), then based on the array element position (x i ,y i , z i ) and target direction θ, calculate the delay τ of each array element i =(x i × sinθ) / c (c is the speed of sound 343m / s), and set the gain G i =1 (uniform excitation).

[0050] Then substitute the acoustic characteristic compensation parameters into the reference sound field model to update the acoustic characteristics: The frequency response correction is to load the filter coefficient h(n) in the simulation software and recalculate the frequency response L of each measurement point. comp (f) = L baseline (f)×H(f); Directivity correction is to set the delay τ of each speaker in the simulation model i and gain G i , recalculate the sound field coverage, and then generate the parameter set of the second sound field model (L comp (f), τ i , G i ) and save it as a callable file.

[0051] S32: Generate a second optimized sound effect model according to the second sound field model and the first optimized sound effect model.

[0052] In the embodiment, the first optimized sound effect model includes fixed parameters (such as the initial debugging base gain G0, initial delay τ0, and initial EQ parameter EQ0) of the initial debugging; the second sound field model is a sound field model including dynamic compensation parameters, which describes the compensated acoustic environment characteristics; and the second optimized sound effect model is a final model fusing the static parameters and the dynamic compensation parameters, and the parameter form is: P final = P static + P danamic (t), wherein P static is the fixed parameter of the first optimized model; P danamic (t) is the time-varying compensation parameter of the second sound field model; and the parameter fusion strategy of the second sound effect optimization model can be preset.

[0053] The parameter fusion strategy of the second sound effect optimization model includes fixed parameter reservation and dynamic parameter superposition, wherein the fixed parameter reservation is to reserve the parameters (such as the gain G0=1.0 of the main expansion channel) in the first optimized model that do not need to be dynamically adjusted, and the dynamic parameter superposition is to add the compensation parameters of the second sound field model to the static parameters, such as the filter coefficients h(n) and the delay adjustment amount Δτ(t). For example, the EQ parameter of the first optimized model is EQ0 (31 paragraphs of parametric equalization, 80 Hz gain-2 dB).

[0054] The dynamic EQ parameter of the second sound field model is EQ dynamic (n) (FIR filter, 250 Hz gain+3 dB); and the final EQ parameter is EQ final (f) = EQ0(f) × EQ dynamic (f) (frequency domain multiplication, corresponding to time domain convolution). A mapping table of the static parameters and the dynamic parameters is established through experiments or simulations to clearly show the parameter combination in different scenes: Scene 1: regular performance (without severe environmental changes): The dynamic parameters take default values (such as the filter coefficients h_default(n) and the sound absorption curtain opening degree 60%).

[0055] Scene 2: actor walking (sound source position change ±2 m): Trigger dynamic compensation: adjust the delay of each loudspeaker τ_dynamic(i) = τ0(i) + Δτ (Δτ=0.1 ms, compensate for the sound path difference 0.343 m).

[0056] S4: trigger adaptive control instructions based on the first optimized sound effect model and the second optimized sound effect model, and the adaptive control instructions include sound field calibration instructions and sound effect optimization instructions.

[0057] In the embodiment, step S4 includes: S41: trigger sound field calibration instructions based on the first optimized sound effect model and the second optimized sound effect model.

[0058] In this embodiment, the sound field calibration instruction refers to a trigger signal that triggers the sound system to perform calibration operations, including calibration mode (such as "fast calibration" and "fine calibration"), calibration parameter range (such as gain adjustment step ΔG = 0.5 dB), and calibration termination condition (such as calibration error ≤ 0.5 dB).

[0059] Specifically, the microphone array (such as 16 omnidirectional microphones with a sampling rate of 48 kHz) deployed in the audience area collects sound field response signals x(t) in real time, and compares them with the predicted sound field of M1 (the ideal response calculated based on M1 parameters) to calculate the deviation Based on the preset calibration trigger threshold δ (such as the absolute value of frequency response deviation ≥ 1 dB, or the sound field uniformity CV ≥ 0.15), when |e(f)| > δ or CV > δ, the sound field calibration instruction is triggered.

[0060] The sound field calibration instruction includes calibration mode, calibration parameter range (limiting the upper limit of gain adjustment (such as G_max = G0 + 3dB to avoid overload) and delay adjustment range (such as τ max = ±1ms to avoid excessive sound path difference)), and termination condition (such as stopping when the deviation |e(f)| ≤ δ / 2 or CV ≤ δ / 2 after calibration).

[0061] S42: Trigger sound effect optimization instruction based on the first and second optimized sound effect models and the preset sound pressure level limit threshold, to obtain the first and second optimized sound effect results, respectively.

[0062] In this embodiment, the sound pressure level limit threshold is to avoid excessive sound pressure level leading to distortion or hearing damage, and the maximum allowed sound pressure level L max (such as L max = 90dB in the center area of the audience area and L max = 85dB in the edge area); the sound effect optimization instruction is an instruction that triggers the sound system to perform optimization operations, including optimization goals (such as "improve high frequency clarity" and "reduce low frequency boom"), constraint conditions (such as L max ), and optimization algorithm parameters (such as filter order N = 5); the first optimized sound effect refers to the sound field response optimized only using M1 parameters (static gain, EQ, delay), representing the "basic optimization effect". The second optimized sound effect refers to the sound field response optimized by combining M1 parameters and M2 dynamic compensation parameters (such as adaptive filter, real-time delay adjustment), representing the "dynamic adaptive comprehensive effect".

[0063] Specifically, according to the calibrated sound field state and the preset L max, generate sound effect optimization instructions, including: optimization target is to design frequency weighting for deviation e(f) (such as low frequency band weight w1=0.3, medium and high frequency band weight w2=0.7, improve speech intelligibility). The constraint condition is that the sound pressure level L_Z i ≤L max (Z i )(Z i is the seat number; the sound effect optimization algorithm selects to optimize M1 parameters using the least mean square (LMS) algorithm, and to optimize M2 dynamic parameters using the recursive least square (RLS) algorithm.

[0064] Specifically, the first optimized sound effect result (R1) calculation is performed: only the static parameters (G0, τ0, EQ0) of M1 are called, the optimized sound field response R1(f) is calculated through acoustic simulation software (such as EASE) or real-time DSP, and it is verified whether L_Z i (R1)≤L max .

[0065] The second optimized sound effect result (R2) calculation is performed: the static parameters of M1 and the dynamic parameters (h_dynamic(n), Δτ(t)) of M2 are called, the optimized sound field response R2(f) is calculated through real-time signal processing, and it is verified whether L_Z i (R2)≤L max .

[0066] S43: Obtain the sound effect optimization result according to the first optimized sound effect result and the second optimized sound effect result.

[0067] Specifically, the sound effect optimization result is the result of synthesizing R1 (static optimization) and R2 (dynamic optimization), selecting or fusing the optimal parameters to generate the final sound effect optimization result that adapts to the current scene; it contains static parameter correction values (ΔG_final, Δτ_final) and dynamic parameter activation strategies (such as "enable M2 filter when ambient noise > 60dB").

[0068] In this embodiment, the strategy of fusing R1 and R2 is set in advance, and the set strategy includes: Performance priority strategy: compare the target achievement degrees (such as frequency response deviation, CV, sound pressure level exceeding rate) of R1 and R2, and select the better one as R final .

[0069] Robustness priority strategy: if the dynamic compensation of R2 can cover more potential interference (such as audience movement, equipment noise), R2 is preferred, and R1 is recorded as a backup.

[0070] Combined with the parameter validity verification step for verification: through the acoustic measurement system (such as SoundCheck), the test signal (such as sweep sine wave) is played back in real time, and it is verified whether Rfinal Actual effects: Measure the frequency response L final (f) of each seat, ensure that |L final (f)-L target (f) |≤0.5dB.

[0071] Measure the sound field uniformity CV final , ensure that CV final ≤0.12. The final R final is encoded as an executable parameter set.

[0072] In an embodiment, triggering the adaptive regulation instruction based on the first optimized sound effect model and the second optimized sound effect model further comprises: S401: Based on the frequency response characteristic curve of the first optimized sound effect model, dynamically adjust the following parameters through the preset sound pressure level limit threshold: Beamforming parameters of directional loudspeaker array, including horizontal pointing angle θ∈[30°, 150°], vertical pointing angle φ∈[10°, 90°]; Adjustable damping coefficient of sound-absorbing material, adjustment range ρ∈[0.2, 0.8]; Delay compensation amount of stage return listening system; When the sound field uniformity CV>0.15 is detected, start the multi-band equalizer for local compensation, and the compensation frequency band is divided into 20Hz-80Hz, 80Hz-800Hz, 800Hz-20kHz.

[0073] In this embodiment, the frequency response characteristic curve of the first optimized sound effect model comes from M1, the pre-stored "ideal frequency response-actual frequency response" comparison curve of M1, which contains the target sound pressure level L target (f) of each frequency point (such as 20Hz-20kHz) and the current measured sound pressure level L current (f). The deviation e(f)=L current (f)-L target (f). The sound pressure level limit threshold refers to the preset maximum allowed sound pressure level L max (f) of each area (such as the front row L max (5kHz)=90dB, the back row L maxThe beamforming parameter of the directional loudspeaker array refers to a parameter for controlling the sound wave propagation direction, including a horizontal pointing angle, which is the coverage angle of the sound beam in the horizontal plane, and a vertical pointing angle, which is the coverage angle of the sound beam in the vertical plane. The adjustable damping coefficient (a) of the sound-absorbing material refers to a parameter of the sound-absorbing capacity of the sound-absorbing material (such as polyester fiber board, adjustable acoustic curtain), and the greater a is, the higher the sound energy absorption rate of the material. The delay compensation amount (Δτ) of the stage return listening system refers to the signal delay difference between the return listening speaker and the main speaker, which is used to compensate for the position deviation of the performer and the return listening speaker.

[0074] Specifically, the frequency band of the low-frequency deviation is 20-80 Hz, the deviation e(f) is -3 dB, and the frequency band of the high-frequency deviation is 800 Hz-20 kHz, and the deviation e(f) is -3 dB. In the embodiment, the beamforming parameter of the directional loudspeaker array includes: a horizontal pointing angle a vertical pointing angle wherein (x s , y s , z s ) is the sound source coordinate; (x a , y a , z a ) is the array center coordinate; The adjustable damping coefficient of the sound-absorbing material wherein ρ0 is the initial damping coefficient, a is the attenuation factor, and ΔL(τ) is the change amount of the frequency response equalization degree over time; since the sound-absorbing performance of the sound-absorbing material decreases in a high-temperature and high-humidity environment, ρ(t) is dynamically adjusted by real-time monitoring of ΔL(τ) to maintain the target reverberation time.

[0075] The delay compensation amount of the stage return listening system wherein d sb is the distance from the sound source to the main speaker; d sa is the distance from the sound source to the return listening system; v s is the sound speed; τ proc is the signal processing delay; applied to the scene where the performer wears the return listening earphone, the sound image deviation is eliminated by adjusting Δt to ensure that the return listening signal is synchronized with the main sound field.

[0076] Specifically, when the sound source deviates from the array center (such as the singer moving to the left side of the stage), the linear array beam direction is adjusted by real-time calculation of θ and φ to ensure that the sound energy is concentrated and covers the audience seats.

[0077] In the embodiment, when the sound field unevenness CV is detected to be greater than 0.15, a multi-band equalizer is started to perform local compensation, specifically including: When the quality factor Q value is greater than or equal to 2.5, the compensation frequency band is divided into 20Hz-80Hz; When the quality factor Q value is in [1.2-1.8], the compensation frequency band is divided into 80Hz-800Hz; When the quality factor Q value is less than or equal to 1.0, the compensation frequency band is divided into 800Hz-20KHz; The compensation amount AG(f) of each frequency band is calculated by the following formula: P target (f) is the target sound pressure level; P measured (f) is the measured sound pressure level, γ(f) is the frequency weighting coefficient (γ=0.8 for low frequency band and middle frequency band, γ=1.2 for high frequency band), and σ(f) is the standard deviation of energy distribution in the frequency band.

[0078] Specifically, the quality factor Q is a ratio of the frequency band width to the center frequency, Q=fc / BW (fc is the center frequency and BW is the bandwidth.

[0079] It should be understood that the sequence numbers of the steps in the above embodiments do not mean the order of execution, and the execution order of the processes should be determined according to their functions and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0080] In an embodiment, a sound effect optimization system based on an adaptive algorithm is provided, which corresponds to the sound effect optimization method based on an adaptive algorithm in the above embodiment.

[0081] A sound effect optimization system based on an adaptive algorithm includes a sound field modeling module, a first sound effect modeling module, an acoustic compensation module, and an adaptive control module. The detailed description of each functional module is as follows: The sound field modeling module is configured to obtain three-dimensional sound field topological parameters of a stage, and construct a reference sound field model in combination with a preset sound equipment parameter table; The first sound effect modeling module is configured to obtain audio source characteristic information, and generate a first optimized sound effect model according to the reference sound field model and the audio source characteristic information; The acoustic compensation module is configured to generate a second optimized sound effect model through an acoustic compensation algorithm according to the reference sound field model and the first optimized sound effect model; The adaptive control module is configured to trigger an adaptive control instruction based on the first optimized sound effect model and the second optimized sound effect model, the adaptive control instruction including a sound field calibration instruction and a sound effect optimization instruction, to adaptively adjust the output effect of the sound system.

[0082] The specific limitation of the sound effect optimization system based on the adaptive algorithm can refer to the limitation of the sound effect optimization method based on the adaptive algorithm, which will not be repeated here; each module in the sound effect optimization system based on the adaptive algorithm can be realized by software, hardware, and their combination; each module can be embedded in or independent of the processor in the computer device in hardware form, or stored in the memory in the computer device in software form, so that the processor can call and execute the operation corresponding to each module.

[0083] In one embodiment, a computer device, which can be a server, is provided, and an internal structure diagram of the computer device can be as shown in Figure 2 The computer device includes a processor, a memory, a network interface, and a database connected by a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store audio source characteristic information, a reference sound field model, sound field calibration instructions, and sound effect optimization instructions. The network interface of the computer device is used to communicate with external terminals through network connection. The computer program is executed by the processor to implement a sound effect optimization method based on an adaptive algorithm.

[0084] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements the following steps when executing the computer program: S1: Obtain stage three-dimensional sound field topology parameters, and construct a reference sound field model in combination with a preset sound equipment parameter table; S2: Obtain audio source characteristic information, and generate a first optimized sound effect model according to the reference sound field model and the audio source characteristic information; S3: Generate a second optimized sound effect model through an acoustic compensation algorithm according to the reference sound field model and the first optimized sound effect model; and S4: Trigger adaptive control instructions based on the first optimized sound effect model and the second optimized sound effect model, wherein the adaptive control instructions include sound field calibration instructions and sound effect optimization instructions.

[0085] In one embodiment, a computer readable storage medium is provided, and a computer program is stored on the computer readable storage medium, and the computer program is executed by a processor to implement the following steps: S1: Obtain stage three-dimensional sound field topology parameters, and construct a reference sound field model in combination with a preset sound equipment parameter table; S2: obtaining audio source characteristic information, generating a first optimized sound effect model according to the reference sound field model and the audio source characteristic information; S3: generating a second optimized sound effect model according to the reference sound field model and the first optimized sound effect model through an acoustic compensation algorithm; S4: triggering an adaptive control instruction based on the first optimized sound effect model and the second optimized sound effect model, the adaptive control instruction including a sound field calibration instruction and a sound effect optimization instruction.

[0086] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by a computer program instructing related hardware, and the computer program can be stored in a non-volatile computer readable storage medium. When the computer program is executed, it can include the processes of the above-mentioned embodiments. Any reference to memory, storage, database or other medium used in the embodiments provided by the present application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration but not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0087] Those skilled in the art can clearly understand that, for the convenience and brevity of description, only the above-mentioned division of functional units and modules is exemplified, and in actual application, the above-mentioned functions can be completed by different functional units and modules according to needs, that is, the internal structure of the device is divided into different functional units or modules to complete all or part of the functions described above.

[0088] The above-described embodiments are only used to illustrate the technical solutions of the present application, but not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that they can modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement for part of the features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application, and should be included in the protection scope of the present application.

Claims

1. A sound effect optimization method based on an adaptive algorithm, characterized in that: include: Obtain the three-dimensional sound field topology parameters of the stage and build a benchmark sound field model based on the preset audio equipment parameter table; Acquiring audio source characteristic information, and generating a first optimized sound effect model based on the reference sound field model and the audio source characteristic information; According to the reference sound field model and the first optimized sound effect model, a second optimized sound effect model is generated through an acoustic compensation algorithm; based on the first optimized sound effect model and the second optimized sound effect model, an adaptive control instruction is triggered, and the adaptive control instruction includes a sound field calibration instruction and a sound effect optimization instruction.

2. The sound effect optimization method based on the adaptive algorithm according to claim 1, characterized in that: The acquiring of audio source characteristic information and generating a first optimized sound effect model according to the reference sound field model and the audio source characteristic information specifically includes: Acquiring audio source characteristic information, the audio source characteristic information including timbre characteristics, dynamic range parameters, and spectrum distribution parameters; obtaining initial sound quality assessment parameters based on the audio source characteristic information and acoustic environment parameters, the initial sound quality assessment parameters including frequency response balance parameters and sound field coverage uniformity parameters; A first optimized sound effect model is generated according to the initial sound effect quality evaluation parameter and the reference sound field model.

3. The sound effect optimization method based on the adaptive algorithm according to claim 2, characterized in that: Generating a second optimized sound effect model by an acoustic compensation algorithm based on the reference sound field model and the first optimized sound effect model specifically includes: Acquiring acoustic characteristic compensation parameters through an acoustic compensation algorithm according to the initial sound quality evaluation parameters, and generating a second sound field model according to the reference sound field model and the acoustic characteristic compensation parameters; A second optimized sound effect model is generated according to the second sound field model and the first optimized sound effect model.

4. The sound effect optimization method based on the adaptive algorithm according to claim 1, characterized in that: The triggering of the adaptive control instruction based on the first optimized sound effect model and the second optimized sound effect model specifically includes: triggering a sound field calibration instruction based on the first optimized sound effect model and the second optimized sound effect model; Based on the first optimized sound effect model and the second optimized sound effect model and a preset sound pressure level limit threshold, triggering a sound effect optimization instruction to obtain a first optimized sound effect result and a second optimized sound effect result respectively; A sound effect optimization result is obtained according to the first sound effect optimization result and the second sound effect optimization result.

5. The sound effect optimization method based on the adaptive algorithm according to claim 1, characterized in that: The triggering of the adaptive control instruction based on the first optimized sound effect model and the second optimized sound effect model further includes: Based on the frequency response characteristic curve of the first optimized sound effect model, the following parameters are dynamically adjusted through the preset sound pressure level limit threshold: The beamforming parameters of the directional loudspeaker array include the horizontal directivity angle θ∈[30°, 150°], the vertical directivity angle The adjustable damping coefficient of the sound-absorbing material is in the range of ρ∈[0.2, 0.8]; Delay compensation amount of the stage monitoring system; When the sound field unevenness CV>0.15 is detected, the multi-band equalizer is started for local compensation, and the compensation bands are divided into 20Hz-80Hz, 80Hz-800Hz), and 800Hz-20kHz.

6. The sound effect optimization method based on the adaptive algorithm according to claim 5, characterized in that: The beamforming parameters of the directional loudspeaker array include: Horizontal pointing angle Vertical pointing angle Among them, (x s ,y s , z s ) is the coordinate of the sound source; (x a ,y a , z a ) is the array center coordinate; Adjustable damping coefficient of sound-absorbing materials Among them, ρ0 is the initial damping coefficient, α is the attenuation factor, and ΔL(τ) is the change of frequency response balance over time; Delay compensation of stage monitoring system Among them, d sb is the distance from the sound source to the main speaker; d sa is the distance from the sound source to the feedback system; v s is the speed of sound; τ proc Signal processing delay.

7. The sound effect optimization method based on adaptive algorithm according to claim 5 or 6, characterized in that: When the sound field unevenness CV>0.15 is detected, the multi-band equalizer is started to perform local compensation, specifically including: When the quality factor Q value is ≥2.5, the compensation frequency band is divided into 20Hz-80Hz; When the quality factor Q value is between [1.2-1.8], the compensation frequency band is divided into 80Hz-800Hz; When the quality factor Q value is ≤1.0, the compensation frequency band is divided into 800Hz-20KHz; The compensation amount ΔG(f) for each frequency band is calculated using the following formula: P target (f) is the target sound pressure level; P measured (f) is the measured sound pressure level, γ(f) is the frequency weighting coefficient (γ = 0.8 for the low and medium frequency bands, γ = 1.2 for the high frequency band), and σ(f) is the standard deviation of the energy distribution within the frequency band.

8. A sound effect optimization system based on an adaptive algorithm, characterized in that: The system comprises: The sound field modeling module is used to obtain the three-dimensional sound field topology parameters of the stage and build a reference sound field model based on the preset sound equipment parameter table; a first sound effect modeling module, configured to obtain audio source characteristic information and generate a first optimized sound effect model according to the reference sound field model and the audio source characteristic information; an acoustic compensation module, configured to generate a second optimized sound effect model by using an acoustic compensation algorithm according to the reference sound field model and the first optimized sound effect model; An adaptive control module is used to trigger adaptive control instructions based on the first optimized sound effect model and the second optimized sound effect model, and the adaptive control instructions include sound field calibration instructions and sound effect optimization instructions to adaptively adjust the output effect of the audio system.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the steps of the sound effect optimization method based on the adaptive algorithm as described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the sound effect optimization method based on the adaptive algorithm as claimed in any one of claims 1 to 7 are implemented.

Images