Computer-implemented method for audio control, and electronic device and storage medium

Abstract

The present disclosure relates to the technical field of audio control. Provided are a computer-implemented method for audio control, and an electronic device and a storage medium. The computer-implemented method for audio control comprises: on the basis of a specified measurement signal, performing acoustic effect testing on a target space, in order to obtain environmental response information, wherein the environmental response information comprises impulse response information and post-separation impulse response information; analyzing the environmental response information, in order to obtain acoustic characteristic information of the target space; on the basis of the acoustic characteristic information, performing compensation processing on audio to be processed, in order to obtain regulated audio; regulating gains of multiple channels corresponding to the regulated audio, in order to obtain re-regulated audio; and performing multi-channel collaborative optimization processing on the re-regulated audio, in order to obtain audio having a target effect. The method achieves precise compensation for spatial acoustic defects, resolves inter-channel crosstalk, improves the acoustic image localization and spatial perception, significantly improves the perceptual audio quality, and optimizes an output.

Description

Audio control computer-implemented method, electronic device and storage medium

[0001] The present application claims priority to the Chinese patent application No. 2024113272968, filed on September 23, 2024, and entitled "Audio control method, device, electronic device and storage medium", the whole content of which is incorporated herein by reference. TECHNICAL FIELD

[0002] The present disclosure relates to the technical field of audio control, and in particular to an audio control computer-implemented method, an electronic device and a storage medium. BACKGROUND

[0003] An audio system (Audio System), also known as a sound system (Sound System) or an audio system, is a collection of devices and technologies for generating, processing, transmitting and playing sound. Traditional audio systems often use preset audio parameters, which cannot effectively adapt to the characteristics of different playback environments, resulting in a significant gap between the actual performance of the audio quality and the ideal state, and thus affecting the overall listening experience of the audio. SUMMARY

[0004] The present disclosure provides an audio control computer-implemented method, an electronic device and a storage medium, which compensates, gain controls and optimizes multiple sound channels based on the analysis of the acoustic characteristic information obtained by the sound effect test of the specified measurement signal, accurately compensates for the acoustic defects of the target space, effectively solves the crosstalk problem between sound channels, significantly improves the positioning and spatial sense of the sound image, and greatly improves the overall listening experience of the audio, and continuously optimizes the audio output.

[0005] The first aspect of the present disclosure provides an audio control computer-implemented method, comprising:

[0006] Performing a sound effect test on a target space based on a specified measurement signal to obtain environmental response information, wherein the environmental response information includes impulse response information and information obtained by separating the impulse response information;

[0007] Analyzing the environmental response information to obtain acoustic characteristic information of the target space;

[0008] Compensating for the to-be-processed audio based on the acoustic characteristic information to obtain regulated audio;

[0009] Controlling the gain of multiple sound channels corresponding to the regulated audio to obtain audio after re-regulation;

[0010] The multi-channel collaborative optimization processing is performed on the audio after the re-regulation, and the audio with the target effect is obtained.

[0011] The second aspect of the present disclosure provides an electronic device, comprising a processor and a memory, the memory stores machine executable instructions capable of being executed by the processor, and the processor executes the machine executable instructions to implement the following steps:

[0012] The target space is subjected to acoustic effect testing based on a specified measurement signal, and environmental response information is obtained, wherein the environmental response information comprises impulse response information and information after separation of the impulse response;

[0013] The environmental response information is analyzed, and acoustic characteristic information of the target space is obtained;

[0014] The to-be-processed audio is compensated based on the acoustic characteristic information, and regulated audio is obtained;

[0015] The gain of the multi-channel corresponding to the regulated audio is regulated, and the audio after re-regulation is obtained;

[0016] The multi-channel collaborative optimization processing is performed on the audio after the re-regulation, and the audio with the target effect is obtained.

[0017] The third aspect of the present disclosure provides a computer readable storage medium, the computer readable storage medium stores a computer program, when the computer program runs on a computer, the computer program makes the computer execute the following steps:

[0018] The target space is subjected to acoustic effect testing based on a specified measurement signal, and environmental response information is obtained, wherein the environmental response information comprises impulse response information and information after separation of the impulse response;

[0019] The environmental response information is analyzed, and acoustic characteristic information of the target space is obtained;

[0020] The to-be-processed audio is compensated based on the acoustic characteristic information, and regulated audio is obtained;

[0021] The gain of the multi-channel corresponding to the regulated audio is regulated, and the audio after re-regulation is obtained;

[0022] The multi-channel collaborative optimization processing is performed on the audio after the re-regulation, and the audio with the target effect is obtained.

[0023] In the technical solution provided by the present disclosure, the target space is tested based on a specified measurement signal to obtain environmental response information, wherein the environmental response information includes impulse response information and information obtained after separation of the impulse response information; the environmental response information is analyzed to obtain acoustic characteristic information of the target space; the acoustic characteristic information is used to perform compensation processing on the audio to be processed to obtain regulated audio; the gain of the multi-channel corresponding to the regulated audio is regulated to obtain audio after secondary regulation; and the audio after secondary regulation is subjected to multi-channel collaborative optimization processing to obtain target-effect audio. In the embodiment of the present disclosure, compensation processing, gain regulation processing and multi-channel collaborative optimization processing are performed based on the acoustic characteristic information obtained by analyzing the environmental response information obtained by sound effect testing based on a specified measurement signal, thereby realizing accurate compensation for acoustic defects of the target space, effectively solving the problem of inter-channel crosstalk, significantly improving the positioning and spatial sense of the sound image, and greatly improving the overall listening experience of the audio and continuously optimizing the audio output. BRIEF DESCRIPTION OF DRAWINGS

[0024] FIG. 1 is a schematic diagram of one embodiment of an electronic device in the present disclosure;

[0025] FIG. 2 is a schematic diagram of one embodiment of a computer-implemented method for audio control in the present disclosure;

[0026] FIG. 3 is a schematic diagram of another embodiment of a computer-implemented method for audio control in the present disclosure. DETAILED DESCRIPTION

[0027] The present embodiment provides a computer-implemented method for audio control, an electronic device and a storage medium, which greatly improves the overall listening experience of the audio and continuously optimizes the audio output.

[0028] The terms "first", "second", "third", "fourth" and the like in the specification and claims of the present disclosure and the above-described drawings (if any) are used to distinguish similar objects, and do not necessarily have to be used to describe a specific order or sequence. It should be understood that the data thus used can be interchanged under appropriate circumstances, so that the embodiments described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "include" or "have" and any variations thereof are intended to cover non-exclusive inclusion, for example, a process, method, system, product or device including a series of steps or units does not have to be limited to only those steps or units clearly listed, but can include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0029] The present embodiment provides a computer-implemented method for audio control, an electronic device and a storage medium. It is mainly applied in an audio system.

[0030] It can be understood that the execution subject of the present disclosure can be a terminal or a server or an audio system, and the specific embodiments are not limited herein. The present embodiment takes the server as an example for description.

[0031] The present embodiment also provides an electronic device including a processor and a memory, the memory storing machine executable instructions capable of being executed by the processor, and the processor executes the machine executable instructions to implement the above-mentioned computer-implemented method for audio control. The electronic device can be a server or a terminal device.

[0032] Referring to FIG. 1, the electronic device includes a processor 100 and a memory 101, the memory 101 storing machine executable instructions capable of being executed by the processor 100, and the processor 100 executes the machine executable instructions to implement the above-mentioned method for audio control.

[0033] Further, the electronic device shown in FIG. 1 further includes a bus 102 and a communication interface 103, and the processor 100, the communication interface 103 and the memory 101 are connected through the bus 102.

[0034] The memory 101 can include a high-speed random access memory (RAM) and can also include a non-volatile memory such as at least one disk memory. The communication connection between the system network element and at least one other network element is realized through at least one communication interface 103 (which can be wired or wireless), and the Internet, a wide area network, a local area network, a metropolitan area network, etc. can be used. The bus 102 can be an ISA bus, a PCI bus or an EISA bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, only one bidirectional arrow is shown in FIG. 1, but it does not mean that there is only one bus or only one type of bus.

[0035] The processor 100 can be an integrated circuit chip with processing capability of signals. In the implementation process, each step of the above method can be completed by integrated logic circuit of hardware in the processor 100 or instruction in the form of software. The processor 100 described above can be a general processor, including a central processing unit (CPU), a network processor (NP), etc.; can also be a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic device, a discrete gate or transistor logic device, a discrete hardware component. Each method, step and logic block diagram disclosed in the embodiment can be implemented or executed. The general processor can be a microprocessor or the processor can also be any conventional processor. The steps of the method disclosed in combination with the embodiment can be directly embodied as a hardware code processor for execution, or a combination of hardware and software modules in the code processor for execution. The software module can be located in a random access memory, a flash memory, a read-only memory, a programmable read-only memory or an electrically erasable programmable memory, a register or other mature storage medium in the art. The storage medium is located in the memory 101, and the processor 100 reads the information in the memory 101, and combines the hardware to complete the following steps:

[0036] The sound effect test is performed on the target space based on the specified measurement signal to obtain environmental response information, wherein the environmental response information includes impulse response information and information obtained after the impulse response is separated;

[0037] The acoustic characteristic information of the target space is obtained by analyzing the environmental response information;

[0038] The audio to be processed is compensated based on the acoustic characteristic information to obtain regulated audio;

[0039] The gain of the multi-channel corresponding to the regulated audio is regulated to obtain audio after re-regulation;

[0040] The multi-channel collaborative optimization processing is performed on the audio after re-regulation to obtain audio with target effect.

[0041] In an embodiment, the sound effect test is performed on the target space based on the specified measurement signal to obtain environmental response information, including:

[0042] The response information of a plurality of measurement points in the target space to a specified measurement signal is collected by presetting an array of collection devices, wherein the plurality of measurement points constitute a measurement grid.

[0043] The response information of the plurality of measurement points is respectively subjected to information extraction to obtain impulse response information respectively corresponding to the plurality of measurement points.

[0044] The impulse response information respectively corresponding to the plurality of measurement points is respectively subjected to separation to obtain direct sound, early reflection and reverberation tail respectively corresponding to the plurality of measurement points.

[0045] The impulse response information, the direct sound, the early reflection and the reverberation tail respectively corresponding to the plurality of measurement points are determined as the environmental response information of the target space.

[0046] In an embodiment, the environmental response information is analyzed to obtain acoustic characteristic information of the target space, including:

[0047] The impulse response information in the environmental response information is subjected to spectral analysis to obtain a spectral analysis result, wherein the spectral analysis result includes a frequency response curve and a power spectral density.

[0048] The environmental mode parameters of the target space are identified based on the power spectral density, wherein the environmental mode parameters include modal density and modal overlap.

[0049] The reverberation time respectively corresponding to the plurality of measurement points is calculated based on the reverberation tail respectively corresponding to the plurality of measurement points.

[0050] The early reflection and the direct sound respectively corresponding to the plurality of measurement points are analyzed to obtain early reflection analysis results respectively corresponding to the plurality of measurement points.

[0051] The spectral analysis result, the environmental mode parameters of the target space, and the reverberation time and the early reflection analysis results respectively corresponding to the plurality of measurement points are determined as the acoustic characteristic information of the target space.

[0052] In an embodiment, a to-be-processed audio is compensated based on the acoustic characteristic information to obtain a regulated audio, including:

[0053] An adaptive equalization model is created based on the frequency response curve in the acoustic characteristic information, wherein the adaptive equalization model is used for adaptive filtering of frequency band limitation.

[0054] An adaptive modal control model is created, wherein the adaptive modal control model is used for adaptive dynamic control of environmental modes.

[0055] The early reflection analysis results in the acoustic characteristic information are subjected to optimization processing to obtain optimized early reflection information.

[0056] Adjust the reverberation time in the acoustic characteristic information to obtain an adjusted reverberation time;

[0057] The to-be-processed audio is processed and regulated by the adaptive equalization model, the adaptive modal control model, the optimized early reflection information and the adjusted reverberation time to obtain regulated audio.

[0058] In an implementation, the adaptive equalization model is created based on a frequency response curve in the acoustic characteristic information, including:

[0059] The frequency response curve in the acoustic characteristic information is adjusted by a specified standard curve to obtain a target response curve;

[0060] The parameter equalization model is constructed based on the target response curve;

[0061] The parameters of the parameter equalization model are adjusted by a preset adaptive algorithm to obtain the adaptive equalization model.

[0062] In an implementation, the adaptive modal control model is created, including:

[0063] The mode type of the target space is identified;

[0064] The filter corresponding to the mode type is set;

[0065] The parameters of the filter corresponding to the mode type are adjusted to obtain the adaptive modal control model.

[0066] In an implementation, the early reflection analysis result in the acoustic characteristic information is optimized to obtain optimized early reflection information, including:

[0067] Based on the early reflection analysis result in the acoustic characteristic information, the delay time and the energy attenuation value of each early reflection path are calculated;

[0068] The early reflections are time-aligned based on the delay time, and the impulse response information is phase-corrected to obtain processed information;

[0069] The target sound field characteristic is obtained, wherein the target sound field characteristic is an ideal environmental acoustic characteristic;

[0070] Based on the energy attenuation value and the target sound field characteristic, the reflection energy in the early reflection analysis result is adjusted to obtain adjusted reflection energy;

[0071] The processed information and the adjusted reflection energy are determined as the optimized early reflection information.

[0072] In an implementation, the reverberation time in the acoustic characteristic information is adjusted to obtain an adjusted reverberation time, including:

[0073] determine an ideal reverberation time according to a scene type;

[0074] construct a multi-channel feedback delay network, wherein the multi-channel feedback delay network is used to simulate natural reverberation;

[0075] adjust a reverberation time in the acoustic characteristic information based on the ideal reverberation time and the multi-channel feedback delay network, to obtain an adjusted reverberation time.

[0076] In an implementation, the gain of the multi-channels corresponding to the regulated audio is regulated to obtain the audio after the second regulation, including:

[0077] obtain noise prediction information of the target space, wherein the noise prediction information is used to indicate noise levels in different time periods;

[0078] adjust the compression ratio and the compression threshold of the multi-channels corresponding to the regulated audio based on the noise prediction information, to obtain adjusted multi-channels;

[0079] perform gain control based on masking on the regulated audio based on the adjusted multi-channels, to obtain the audio after the second regulation.

[0080] In an implementation, the audio after the second regulation is subjected to multi-channel collaborative optimization processing to obtain the audio with the target effect, including:

[0081] analyze channel cross-influence characteristics between the adjusted multi-channels;

[0082] perform phase compensation and amplitude compensation on the audio after the second regulation based on the channel cross-influence characteristics, to obtain balanced sound effects;

[0083] perform processing on the balanced sound effects based on a preset adaptive cross-cancellation algorithm, to obtain the audio with the target effect.

[0084] In an implementation, before obtaining the environmental response information based on the specified measurement signal for sound effect testing of the target space, further including:

[0085] identify a current audio use scenario and automatically match a control strategy corresponding to the audio use scenario, wherein the control strategy includes content of the acoustic characteristic information, compensation processing content, and a compensation processing flow.

[0086] In an implementation, after the multi-channel collaborative optimization processing on the audio after the second regulation to obtain the audio with the target effect, further including:

[0087] obtain analysis data and perform mechanism updating and / or strategy optimization of the audio control method based on the analysis data, wherein the analysis data includes performance test data and user evaluation data.

[0088] For ease of understanding, the specific process of the present embodiment is described below. Referring to FIG. 2, one embodiment of the computer-implemented method for audio control in the present embodiment includes the following steps:

[0089] 201, performing acoustic effect testing on the target space based on a specified measurement signal to obtain environmental response information, wherein the environmental response information includes impulse response information and information obtained after separation of the impulse response information;

[0090] The specified measurement signal can be understood as a preset audio signal used to evaluate the performance of an audio system or the acoustic characteristics of an environment. The specified measurement signal is a measurement signal that has controllability, stability, and analyzability, but is not limited thereto. The impulse response information can be understood as information obtained after extraction of target response information. The target response information is response information fed back by the target space after acoustic effect testing based on the specified measurement signal and recorded by a preset recording device. The information obtained after separation of the impulse response information can be understood as information obtained after processing of the impulse response information by a preset window separation function (or a preset window separation algorithm). By way of example but not limitation, the target space can be a room in which audio playback is to be performed.

[0091] By way of example but not limitation, the specified measurement signal is determined; the specified measurement signal is emitted to the target space by a preset emission device; response information fed back by a specific position in the target space based on the specified measurement signal is collected by a preset recording device, and the response information is extracted to obtain impulse response information; the impulse response information is processed by separation to obtain separated information; and the impulse response information and the separated information are determined as the environmental response information.

[0092] 202, analyzing the environmental response information to obtain acoustic characteristic information of the target space;

[0093] By way of example but not limitation, at least one acoustic characteristic factor required for analysis is determined, wherein the at least one acoustic characteristic factor is a key factor for spatial acoustic performance; a preset search algorithm is used to search a preset algorithm library based on the at least one acoustic characteristic factor to obtain a target algorithm, wherein the search algorithm is used to search or find an algorithm corresponding to an optimal path in the preset algorithm library, for example, the search algorithm can be a Dijkstra algorithm, an A* algorithm, a depth-first search DFS, or a breadth-first search BFS, the preset algorithm library stores a plurality of algorithms (models) for analyzing acoustic characteristic factors, and the target algorithm includes algorithms corresponding to each acoustic characteristic factor; and the environmental response information is processed by the target algorithm to obtain the acoustic characteristic information of the target space. By matching the target algorithm based on the at least one acoustic characteristic factor, the efficiency and accuracy of the analysis of the environmental response information are improved, and the accuracy and reliability of the acoustic characteristic information of the target space are improved.

[0094] 203、compensate the to-be-processed audio based on the acoustic characteristic information to obtain regulated audio;

[0095] The to-be-processed audio is audio information corresponding to a target use scenario, for example, music appreciation, movie watching, and game playing.

[0096] By way of example but not limitation, a corresponding target compensation algorithm is matched based on the acoustic characteristic information, and the to-be-processed audio is compensated based on the target compensation algorithm to obtain the regulated audio, wherein the target compensation algorithm includes but is not limited to an adaptive equalization algorithm, an ambient mode suppression algorithm (also referred to as an intelligent ambient adaptive processing algorithm, which is used to analyze ambient information of a target space in real time and dynamically adjust audio parameters), an early reflection optimization algorithm, an automatic gain control algorithm, a wide dynamic range compression algorithm, an equalizer, a noise suppression algorithm, and an echo cancellation algorithm.

[0097] 204、regulate the gains of multiple channels corresponding to the regulated audio to obtain audio after re-regulation;

[0098] By way of example but not limitation, a gain adjustment strategy for each channel is determined, wherein specifically, the content of the to-be-processed audio is obtained, and the gain adjustment strategy for each channel corresponding to the regulated audio is determined based on the content of the to-be-processed audio and the acoustic characteristic information, wherein the content of the gain adjustment strategy includes but is not limited to loudness matching, dynamic range control, spatial balance, spectral balance, and scene optimization; the gains of multiple channels corresponding to the regulated audio are regulated based on the gain adjustment strategy for each channel to obtain audio after re-regulation.

[0099] 205、perform multi-channel collaborative optimization processing on the audio after re-regulation to obtain target-effect audio.

[0100] By way of example but not limitation, the multi-channel layout of the audio after re-regulation can be determined, and the audio after re-regulation is subjected to channel analysis, adaptive filtering, and multi-channel complementary equalization processing based on the multi-channel layout to obtain the target-effect audio.

[0101] By way of example but not limitation, after the audio after re-regulation is subjected to channel analysis, adaptive filtering, and multi-channel complementary equalization processing, the processed audio can be subjected to sound effect enhancement to obtain enhanced audio; the enhanced audio is tested, and the result of the test is evaluated, and in a case where the result of the evaluation meets a preset condition, the enhanced audio is determined as the target-effect audio. Through the test and evaluation, the reliability of the target-effect audio can be improved, thereby providing a high-quality audio experience.

[0102] In this embodiment, by performing compensation processing, gain control processing, and multi-channel collaborative optimization processing based on the acoustic characteristic information obtained from the analysis of the environmental response information of the sound effect test based on the specified measurement signal, the acoustic defects of the target space are accurately compensated, the crosstalk problem between channels is effectively solved, the localization and spatial sense of the sound image are significantly improved, and the overall listening experience of the audio is greatly improved, thus continuously optimizing the audio output.

[0103] Referring to Figure 3, another embodiment of the computer-implemented method for audio control in this example includes:

[0104] 301. Conduct sound effect tests on the target space based on the specified measurement signal to obtain environmental response information, wherein the environmental response information includes impulse response information and information after the impulse response is separated;

[0105] In one implementation, when conducting sound effect testing on a target space based on a specified measurement signal to obtain environmental response information, a preset acquisition device array can be used to acquire response information of multiple measurement points in the target space to the specified measurement signal, wherein the multiple measurement points constitute a measurement grid; information extraction is performed on the response information of the multiple measurement points to obtain impulse response information corresponding to each measurement point; the impulse response information corresponding to each measurement point is separated to obtain direct sound, early reflection, and reverberant tail sound corresponding to each measurement point; the impulse response information, direct sound, early reflection, and reverberant tail sound corresponding to each measurement point are determined as the environmental response information of the target space.

[0106] Here, as an example and not a limitation, the specified measurement signal is a wideband swept frequency signal capable of covering the entire frequency band from 20Hz to 20kHz. Before acquiring the response information of multiple measurement points in the target space to the specified measurement signal through a preset acquisition device array, the specified measurement signal can be generated by a preset exponential swept frequency signal algorithm. The exponential swept frequency signal algorithm can be as follows: x(t)=sin[2πf1T(e ∧(t / L) [-1 / ln(f2 / f1)], where f1 is the starting frequency, f2 is the ending frequency, T is the sweep time, and L = T / ln(f2 / f1). The specified measurement signal is generated using an exponential sweep frequency signal algorithm, which improves the signal-to-noise ratio and time resolution of the specified measurement signal.

[0107] In an implementation, before collecting the response information of the multiple measuring points in the target space to the specified measuring signal by the preset array of collecting devices, the environmental characteristics of the target space can be acquired, including but not limited to geometric structure and acoustic characteristics; according to the environmental characteristics of the target space, the multiple measuring points are determined to be distributed at different positions of the listening area in the target space, wherein the multiple measuring points form a three-dimensional measuring grid in the X, Y and Z axis directions, and the three-dimensional measuring grid can cover multiple spatial positions in the target space so as to collect comprehensive acoustic data. Specifically, the environmental characteristics of the target space can be acquired, including but not limited to geometric structure (for example, spatial shape and size), environmental type, environmental purpose and acoustic characteristics; the measuring point layout information corresponding to the environmental characteristics can be matched from a preset database, wherein the measuring point layout information includes but is not limited to key positions, number and combined shape of the measuring points; and the multiple measuring points are determined to be distributed at different positions of the listening area in the target space based on the measuring point layout information. By determining the multiple measuring points of the target space based on the measuring point layout information, the reliability of the multiple measuring points is improved, which is conducive to the measurement of spaces in various environments, and further helps to improve the accuracy control of the target effect audio.

[0108] As an example but not limitation, high-precision measuring-grade microphones can be used in the preset array of collecting devices to ensure the accuracy of the measurement. When collecting the response information of the multiple measuring points in the target space to the specified measuring signal by the preset array of collecting devices, the response information corresponding to each measuring point in the target space can be collected by the preset array of collecting devices based on a preset measurement number, to obtain the collection information corresponding to the preset measurement number of each measuring point; and the arithmetic mean of the collection information corresponding to the preset measurement number of each measuring point is calculated to obtain the response information of the multiple measuring points in the target space to the specified measuring signal. The calculation of the arithmetic mean of the collection information corresponding to the preset measurement number of each measuring point improves the signal-to-noise ratio of the response information.

[0109] In an implementation, when the response information of the multiple measuring points is respectively subjected to information extraction to obtain the impulse response information respectively corresponding to the multiple measuring points, the response information of the multiple measuring points can be respectively subjected to signal extraction by a preset frequency domain analysis algorithm to obtain the impulse response information respectively corresponding to the multiple measuring points, wherein the frequency domain analysis algorithm can be as follows: h(t) = IFFT(H(ω)), H(ω) = Y(ω) / X(ω), H(ω) is the transfer function (frequency response function) corresponding to each measuring point, Y(ω) is the response information of each measuring point, X(ω) is the specified measuring signal, and h(t) is the impulse response information corresponding to each measuring point.

[0110] In an implementation, the impulse response information corresponding to each of the plurality of measurement points can be separated by a preset time window algorithm to obtain direct sound, early reflections and reverberation tail corresponding to each of the plurality of measurement points, wherein the time window algorithm can be a sliding window algorithm, a window function (for example, a Hann window) and an overlapping window algorithm to reduce spectral leakage and improve the accuracy of frequency domain analysis.

[0111] By obtaining the impulse response information, direct sound, early reflections and reverberation tail of the specified measurement signal based on the plurality of measurement points in the form of a measurement grid in the target space, the measurement process is simplified, the accuracy of the environmental response information analysis is improved, and the propagation effect and listening quality of the audio are improved.

[0112] In an implementation, before the sound effect test of the target space based on the specified measurement signal is performed to obtain the environmental response information, the current audio use scenario can be identified, and a control strategy corresponding to the audio use scenario can be automatically matched, wherein the control strategy includes the content of the acoustic characteristic information, the compensation processing content and the compensation processing flow.

[0113] In an implementation, the current audio use scenario can be identified by a preset machine learning algorithm, wherein the machine learning algorithm can be a support vector machine (SVM) classification algorithm, and the specific implementation is as follows:

[0114] f(x) = sign(∑(α i *y i *K(x i ,x) + b), α i is a Lagrange multiplier, y i is a class label, K is a kernel function, x i is an input audio to be processed, and b is a bias term.

[0115] By way of example and not limitation, the audio use scenario can be music appreciation, movie watching, and gaming. By adjusting the parameters and algorithms of the audio processing system according to the audio use scenario, the audio experience in each scenario can be optimized to meet the needs and preferences of different users.

[0116] In an implementation, the control strategy can be understood as a specific implementation process of performing a sound effect test of a target space based on a specified measurement signal to obtain environmental response information, wherein the environmental response information includes impulse response information and separated information of the impulse response information; analyzing the environmental response information to obtain acoustic characteristic information of the target space; performing compensation processing on the audio to be processed based on the acoustic characteristic information to obtain regulated audio; regulating the gain of the multi-channel of the regulated audio to obtain audio after secondary regulation; and performing multi-channel collaborative optimization processing on the audio after secondary regulation to obtain target effect audio.

[0117] By way of example and not limitation, the automatic matching of the control strategy corresponding to the audio usage scenario can be as follows: (1) when the audio usage scenario is music (characteristics: requires high-fidelity sound quality, strong low-frequency response, and clear mid-high frequencies, good spatial sense and stereo effect of audio), the corresponding control strategy includes a room acoustic compensation strategy (adjusting the frequency response to reduce standing waves and mode interference in the room, making the low frequency more balanced), a reverberation time control strategy (setting a moderate reverberation time to preserve the natural continuity of music without causing excessive reverberation tails), and a dynamic range processing strategy (using slight dynamic range compression when listening to music to preserve the details and dynamic changes of the music); (2) when the audio usage scenario is movie (characteristics: requires clear dialogue, rich sound effects, and appropriate low frequencies to enhance immersion, and is more sensitive to spatial positioning and effects of sound, such as surround sound and foreground sound effects), the corresponding control strategy includes a room acoustic compensation strategy (optimizing the frequency response and phase to ensure clear dialogue and good positioning of sound effects), a reverberation time control strategy (setting a longer reverberation time to simulate the surround sound effect of a movie while controlling low-frequency reverberation), and a dynamic range processing strategy (using moderate dynamic range compression to balance the dynamic range of dialogue and sound effects to avoid overly jarring sounds).

[0118] For example, the audio system develops a touch screen interface through which users can visually view and adjust the sound field through a 3D room model. By analyzing the audio content and user operations, the audio system can distinguish between music, movie, and game modes with 95% accuracy and automatically apply the corresponding optimization settings. The audio system observes that users often increase low-frequency output when watching action movies, so it automatically increases the low-frequency response by 2dB when it recognizes an action scene.

[0119] Through intelligent scene recognition and matching of control strategies, intelligent dynamic optimization is achieved.

[0120] In an implementation, the computer-implemented method of audio control further comprises: displaying, through a preset three-dimensional (3D) spatial acoustic visualization tool, a sound field distribution of the audio to be processed and / or a sound field distribution of the audio of the target effect, and a comparison result, wherein the comparison result is a result of comparison between the real-time frequency response curve and the optimization effect. Through the three-dimensional intuitive visualization of the three-dimensional spatial acoustic visualization tool, high performance is ensured while also ensuring ease of use and scalability, greatly improving the user experience.

[0121] In an implementation, the computer-implemented method of audio control further comprises: in response to the touch operation of the three-dimensional spatial acoustic visualization tool, obtaining adjustment data, and performing sound field adjustment on the to-be-processed audio and / or the target effect based on the adjustment data. Through intuitive control of the three-dimensional spatial acoustic visualization tool, high performance is ensured while ensuring ease of use and scalability, greatly improving user experience.

[0122] In an implementation, the computer-implemented method of audio control further comprises: in response to the optimization instruction, obtaining adjustment operation data based on the user interface; and performing parameter optimization based on the adjustment operation data through the preset reinforcement learning algorithm. As an example but not limitation, the reinforcement learning algorithm can be specifically as follows:

[0123] Q(s,a)←Q(s,a)+α*[r+γ*max a Q(s',a')-Q(s,a)], s is a state, a is an action, a is a learning rate, r is a reward, and g is a discount factor. By performing parameter optimization based on the adjustment operation data through the preset reinforcement learning algorithm, the system parameters can be continuously optimized to meet user preferences, realizing user preference learning and dynamic optimization of individualization.

[0124] 302, analyze the environmental response information to obtain acoustic characteristic information of the target space;

[0125] In an implementation, when the environmental response information is analyzed to obtain the acoustic characteristic information of the target space, the impulse response information in the environmental response information can be subjected to spectral analysis to obtain a spectral analysis result, wherein the spectral analysis result includes a frequency response curve and a power spectral density; the environmental mode parameters of the target space can be identified based on the power spectral density, wherein the environmental mode parameters include modal density and modal overlap; the reverberation time of the plurality of measurement points can be calculated based on the early reflections and the direct sound of the plurality of measurement points, respectively; the early reflection analysis results of the plurality of measurement points can be obtained by analyzing the early reflections of the plurality of measurement points, respectively; and the spectral analysis result, the environmental mode parameters of the target space, and the reverberation time and the early reflection analysis results of the plurality of measurement points are determined as the acoustic characteristic information of the target space.

[0126] In an implementation manner, when the impulse response information in the environment response information is subjected to frequency spectrum analysis, the frequency spectrum analysis result is obtained, and the frequency response curve of each measurement point can be calculated based on the impulse response information of each measurement point in the environment response information, wherein the frequency response curve is used to present the response of the audio system at different frequencies; the frequency response curve is subjected to smoothing processing to obtain the smoothed frequency response curve; the power spectral density is calculated based on the smoothed frequency response curve (the amplitude square of the smoothed frequency response curve is subjected to normalization processing to obtain the power spectral density); and the frequency response curve and the power spectral density are determined as the frequency spectrum analysis result.

[0127] For example, the frequency response curve can be subjected to smoothing processing based on a one-third octave smoothing processing algorithm to obtain the smoothed frequency response curve, so as to reduce the influence of slight fluctuations; and the smoothed frequency response curve is subjected to operation based on a preset power spectral density estimation formula to obtain the power spectral density, wherein the power spectral density is used to indicate the distribution of the signal power of the environment response information at different frequencies, and the power spectral density estimation formula is as follows: PSD(ω) = |FFT(h(t))| 2 / N, PSD(ω) is the power spectral density, h(t) is the smoothed frequency response curve, and N is the FFT length.

[0128] In an implementation manner, when the impulse response information in the environment response information is subjected to frequency spectrum analysis, the frequency spectrum analysis result is obtained, and the frequency response curve of each measurement point can be calculated based on the impulse response information of each measurement point in the environment response information, wherein the frequency response curve is used to present the response of the audio system at different frequencies; the frequency response curve is subjected to smoothing processing to obtain the smoothed frequency response curve; the power spectral density is calculated based on the smoothed frequency response curve (the amplitude square of the smoothed frequency response curve is subjected to normalization processing to obtain the power spectral density); and the frequency response curve and the power spectral density are determined as the frequency spectrum analysis result. 3 2 ​, D(f) is a modal density, N is the number of modes in a specific frequency range, Delta f is the width of the frequency range, V is the volume of the target space, c is the sound speed, and f is the frequency; analyzing the frequency response information of the target space to determine the bandwidth of each mode of the target space; based on the bandwidth of each mode of the target space, drawing the frequency range of different modes of the target space on a spectrum corresponding to the impulse response information to obtain a target spectrum; based on the target spectrum, calculating the overlapping area of different modes of the target space; through the overlapping area of different modes, calculating the area of the overlapping of different modes in the target spectrum and the total spectral area of all modes; based on the area of the overlapping of different modes in the target spectrum and the total spectral area of all modes, calculating the modal overlap degree, specifically, the modal overlap degree can be calculated by dividing the area of the overlapping of different modes in the target spectrum by the total spectral area of all modes to obtain the modal overlap degree; and determining the modal density and the modal overlap degree as the environmental mode parameters of the target space. Through the calculation of the environmental mode parameters of the modal density and the modal overlap degree, the environmental mode of the target space can be preliminarily understood, the degree of interaction between modes and the complexity of the dynamic response of the audio system can be reflected, and the acoustic characteristics of the target space can be more accurately analyzed.

[0129] In an implementation manner, when the reverberation time corresponding to each of the plurality of measurement points is calculated based on the reverberation tail corresponding to each of the plurality of measurement points, the reverberation time corresponding to each of the plurality of measurement points can be calculated by using a preset Schroeder integral algorithm, wherein the Schroeder integral algorithm is as follows:

[0130] EDC(t) = ∫[t→∞]h 2 (τ)dτ, RT60 = time when EDC(t) drops by 60dB, EDC(t) is an energy decay curve, t is a lower limit time of integration, h 2 (τ) is impulse response information, dτ is a small increment of the integral variable τ, and RT60 is the reverberation time corresponding to each measurement point. By calculating the reverberation time in different frequency bands, the decay characteristics of different frequencies can be effectively analyzed, and the reliability of the acoustic characteristic information is improved.

[0131] In one implementation, when the early reflections and direct sound corresponding to each of the plurality of measurement points are analyzed to obtain early reflection analysis results corresponding to each of the plurality of measurement points, the early reflections and direct sound corresponding to each of the measurement points can be detected by a preset threshold detection algorithm to obtain reflection peak values and direct sound peak values corresponding to each of the measurement points, where the reflection peak value is the peak value of the early reflection, and the direct sound peak value is the peak value of the direct sound; the time delay of the early reflection relative to the direct sound is calculated based on the distance difference between the reflection peak value and the direct sound peak value corresponding to each of the measurement points; the total sound energy in the entire measurement period (including the direct sound and all reflected sound) is calculated; the direct sound energy is calculated based on the direct sound peak value; the reflection energy sum is calculated, where specifically, for each reflection peak value, a window around the center time of each peak can be determined, the square sum of the signal in each window can be calculated to obtain the energy of a single reflection peak, and the energies of all reflection peaks can be added to obtain the reflection energy sum; the first energy ratio is obtained by dividing the direct sound energy by the reflection energy sum, and the second energy ratio is obtained by dividing the reflection energy sum by the total sound energy, and the first energy ratio and the second energy ratio are taken as the energy ratio of the early reflection to the direct sound; the time delay of the early reflection relative to the direct sound and the energy ratio of the early reflection to the direct sound corresponding to each of the plurality of measurement points are determined as the early reflection analysis results corresponding to each of the plurality of measurement points. By calculating the time delay of the early reflection relative to the direct sound and the energy ratio of the early reflection to the direct sound corresponding to each of the plurality of measurement points, the efficiency and accuracy of the early reflection analysis result analysis are improved, the spatial sense of the audio can be effectively improved, and the immersion of the listener can be enhanced.

[0132] The acoustic characteristic information including the frequency spectrum analysis result, the environmental mode parameter of the target space, the reverberation time corresponding to each of the plurality of measurement points, and the early reflection analysis result is obtained by analyzing the environmental response information, the accuracy of the acoustic characteristic analysis of the target space is improved, the negative impact of the acoustic characteristics of the target space on the sound quality can be effectively overcome, and the unique acoustic environment of different spaces can be effectively adapted to.

[0133] 303、creating an adaptive equalization model based on the frequency response curve in the acoustic characteristic information, where the adaptive equalization model is used for adaptive filtering of frequency band limitation;

[0134] In one implementation, when the adaptive equalization model is created based on the frequency response curve in the acoustic characteristic information, the frequency response curve in the acoustic characteristic information can be adjusted by a specified standard curve to obtain a target response curve; a parameter equalization model is constructed based on the target response curve; and the parameters of the parameter equalization model are adjusted by a preset adaptive algorithm to obtain the adaptive equalization model.

[0135] The preset standard curve can be understood as a standard curve or a reference curve in the industry that helps to design and calibrate an audio system to achieve a more natural or specific listening effect, that is, it can also be understood as an idealized frequency response curve with flatness, bandwidth coverage, phase consistency, and compliance with industry standards, such as Harman Target Curves, Diffuse Field Equalization (DFE) Curve, Equal Loudness Contours, and AES Standard Curve.

[0136] By way of example and not limitation, before the frequency response curve in the acoustic characteristic information is adjusted by the specified standard curve to obtain the target response curve, the use scenario and adjustment target of the audio to be processed can be obtained, and the corresponding standard curve can be determined based on the use scenario and adjustment target of the audio to be processed to obtain the specified standard curve. When the frequency response curve in the acoustic characteristic information is adjusted by the specified standard curve to obtain the target response curve, the specified standard curve and the frequency response curve in the acoustic characteristic information can be compared and analyzed to obtain a difference part, and the frequency response curve in the acoustic characteristic information can be adjusted based on the difference part by using a pre-set calibration tool to obtain the target response curve. By adjusting the frequency response curve in the acoustic characteristic information, the diffuse sound field response of the listening position can be considered.

[0137] In an implementation, when the parameter equalization model is constructed based on the target response curve, the parameter equalization model can be constructed by cascading the target response curve and an IIR filter bank in a second-order filter structure, where the parameter equalization model can implement the following algorithm: H(s) = (s 2 +A*ω0 / Q*s+ω0 2 ) / (s 2 +ω0 / Q*s+ω0 2 ), where s is a complex frequency domain variable, s 2 reflects the second derivative influence of the system on the time domain signal, A is a gain, ω0 is a center frequency in the target response curve, and Q is a quality factor. By using the IIR filter bank, accurate frequency response correction is achieved, and by using the second-order filter cascade structure, the smoothness of the phase response is ensured.

[0138] Wherein, as an example but not limited, the adaptive algorithm can be a least mean squares (LMS) algorithm, and the adaptive algorithm is specifically as follows: W(k+1) = W(k) + μ*X(k)*E(k), W(k) is a filter coefficient, μ is a step size, X(k) is a frequency response corresponding to a target response curve, and E(k) is an error signal. By adjusting the parameters of the parameter equalization model through the least mean squares (LMS) algorithm, the adaptive filtering of the frequency band limitation is realized, and the overcorrection is avoided.

[0139] By adjusting the frequency response curve in the acoustic characteristic information by specifying a standard curve, and adjusting the parameters of the parameter equalization model through a preset adaptive algorithm, the reliability and accuracy of the adaptive equalization model are improved, which is helpful to output high-quality audio effects and meet the use requirements in various scenes (various complex environments).

[0140] 304, creating an adaptive modal control model, wherein the adaptive modal control model is used for adaptive dynamic control of an environment mode;

[0141] In an implementation manner, when the adaptive modal control model is created, the mode type of the target space can be identified, a filter corresponding to the mode type is set, and the parameters of the filter corresponding to the mode type are adjusted to obtain the adaptive modal control model.

[0142] Wherein, in an implementation manner, when the mode type of the target space is identified, a preset mode frequency algorithm can be used for calculation to obtain a calculation result, and the mode frequency algorithm is specifically as follows:

[0143] f nx,ny,nz = (c / 2)*sqrt(nx / Lx) 2 + (ny / Ly) 2 + (nz / Lz) 2 , c is, nx, ny, nz is a mode order, Lx, Ly, Lz is a room size; the mode type of the target space is determined based on the calculation result, and the mode type includes an axial mode, a tangential mode and an oblique mode. By identifying the mode type of the target space, detailed information is provided for subsequent suppression and optimization. As an example but not limited, when there is only one non-zero mode order (f nx,ny,nz , f nx,0,0 , f 0,ny,0 , f 0,0,nz ) in the calculation result f nx,ny,nz , the mode type of the target space is an axial mode; when there are two non-zero mode orders (f nx,ny,0 , f nx,0,nz , f 0,ny,nzWhen the number of non-zero mode orders is three, the mode type of the target space is oblique mode. nx,ny,nz When the number of non-zero mode orders is three, the mode type of the target space is oblique mode.

[0144] In an implementation manner, when the filter corresponding to the mode type is set, narrow-band notch filters can be set for the axial mode, the tangential mode and the oblique mode in the mode type, wherein the narrow-band notch filter can realize H(z) = (1-2r*cosω0*z -1 +r 2 *z -2 ) / (1-2*cosω0*z -1 +z -2 ), r is a control bandwidth, ω0 is a center frequency, z is a complex variable in Z transform, representing a frequency domain characteristic of a discrete-time system; and a quality factor value and a depth of each narrow-band notch filter are adjusted to achieve an optimal suppression effect.

[0145] In an implementation manner, when the parameters of the filter corresponding to the mode type are adjusted to obtain the adaptive modal control model, the parameters of the filter corresponding to the mode type can be adjusted by using a recursive least squares (RLS) algorithm to obtain the adaptive modal control model, wherein the recursive least squares (RLS) algorithm is specifically as follows:

[0146] K(n) = [λ -1 *P(n-1)*x(n)] / (1+λ -1 *x∧T(n)*P(n-1)*x(n)), e(n) = d(n)-w∧T(n)*x(n), w(n) = w(n-1)+K(n)*e(n), P(n) = λ -1 *P(n-1)-λ -1*K(n)*x∧T(n)*P(n-1), where K(n) is the gain vector, controlling the magnitude of weight updates; λ is the forgetting factor, controlling the influence of historical data on the current iteration; P(n) is the error covariance matrix, measuring the filter's error and uncertainty; x(n) is the input signal, the signal entering the filter; T(n) represents the transpose of vector x at time point n; e(n) is the error signal, the difference between the desired signal and the filter output; d(n) is the desired signal, the ideal value the system wants the filter to output; and w(n) is the filter weight, determining how the filter processes the input signal. Using the Recursive Least Squares (RLS) algorithm, it can handle low-frequency response fluctuations under environmental changes, especially addressing low-frequency resonance problems caused by changes in spatial state (such as opening doors or moving furniture). It enables more refined and faster real-time dynamic adjustments to low-frequency spatial patterns, real-time monitoring of low-frequency response changes, and adaptation to changes in the target space's state.

[0147] By setting the corresponding filter according to the mode type of the target space and performing adaptive modal control on the filter, it is possible to cope with low-frequency response fluctuations under environmental changes, make more precise and faster real-time dynamic adjustments to the spatial mode in the low-frequency band, improve the flexibility of audio control, and maintain the flatness of the low-frequency response.

[0148] 305. Optimize the early reflection analysis results in the acoustic characteristic information to obtain optimized early reflection information;

[0149] In one implementation, when optimizing the early reflection analysis results in the acoustic characteristic information to obtain optimized early reflection information, the following steps can be taken: Based on the early reflection analysis results in the acoustic characteristic information, calculate the delay time and energy attenuation value of each early reflection path; perform time alignment processing on the early reflections based on the delay time, and perform phase correction on the impulse response information to obtain processed information; acquire the target sound field characteristics, wherein the target sound field characteristics are ideal environmental acoustic characteristics; adjust the reflection energy in the early reflection analysis results based on the energy attenuation value and the target sound field characteristics to obtain adjusted reflection energy; and determine the processed information and adjusted reflection energy as the optimized early reflection information.

[0150] In an implementation, in the process of calculating the delay time and energy attenuation value of each early reflection path based on the early reflection analysis result in the acoustic characteristic information, the early reflection path model can be constructed based on the geometric information of the target space and the response information of the multiple measurement points in the target space to the specified measurement signal; the delay time and energy attenuation value of each early reflection path can be calculated through the early reflection path model, specifically, ray tracing is performed through the early reflection path model, and the total path length of the ray of each early reflection path is calculated, and the total path length of the ray of each early reflection path is divided by the sound speed to obtain the delay time of each early reflection path, and the continuous product of the sound absorption coefficients of all reflecting surfaces along the early reflection path is calculated, and the energy attenuation value of each early reflection path is calculated based on the continuous product of the sound absorption coefficients of all reflecting surfaces along the early reflection path.

[0151] In the process of ray tracing through the early reflection path model, the mirror position of the sound source for each reflecting surface can be calculated based on the mirror source algorithm through the early reflection path model, and the ray path is traced, specifically, the structure of the measurement target space (including the length, width, height of the target space and the position of the reflecting surface) is measured, the position of the sound source receiver is set, the mirror sound source position of each reflecting surface is calculated, the path from the mirror sound source to the receiver is traced, the path length, time delay and energy attenuation are calculated, the multiple reflections are repeatedly processed based on the preset reflection times, the impulse responses of all reflected sounds are synthesized, the final room impulse response is generated, and the input signal (impulse response information) is convolved to simulate the sound propagation effect in the target space.

[0152] In an implementation, in the process of performing time alignment processing on the early reflections based on the delay time and performing phase correction on the impulse response information to obtain the processed information, the FIR filter can be set, wherein the FIR filter is specifically as follows: h(n) = sinc(n-D) * w(n), n is a time point in the delay time, D is a fractional delay, representing the delay amount of the early reflection sound and the direct sound, and w(n) is a window function; the early reflections are processed through the FIR filter to perform time alignment processing. Through the FIR filter, the time alignment processing is performed on the early reflections, which realizes time alignment and makes the early reflection sound and the direct sound accurately synchronized in time. The phase information is obtained by extracting the impulse response information; the phase compensation is performed based on the phase information through the pre-set all-pass filter to adjust the phase of the corresponding frequency component. The phase correction is performed through the all-pass filter, which optimizes the sound image positioning.

[0153] In an implementation, in the process of calculating the delay time and energy attenuation value of each early reflection path based on the early reflection analysis result in the acoustic characteristic information, the early reflection path model can be constructed based on the geometric information of the target space and the response information of the multiple measurement points in the target space to the specified measurement signal; the delay time and energy attenuation value of each early reflection path can be calculated through the early reflection path model, specifically, ray tracing is performed through the early reflection path model, and the total path length of the ray of each early reflection path is calculated, and the total path length of the ray of each early reflection path is divided by the sound speed to obtain the delay time of each early reflection path, and the continuous product of the sound absorption coefficients of all reflecting surfaces along the early reflection path is calculated, and the energy attenuation value of each early reflection path is calculated based on the continuous product of the sound absorption coefficients of all reflecting surfaces along the early reflection path.

[0154] In one implementation, when adjusting the reflected energy in the early reflection analysis results based on the energy attenuation value and the target sound field characteristics to obtain the adjusted reflected energy, the following can be done: Based on the energy attenuation value, the target sound field characteristics, and the reflection energy optimization algorithm, adjust the energy of each early reflection in the early reflection analysis results to make it conform to the target energy, thereby obtaining the adjusted reflected energy. The specific reflection energy optimization algorithm is as follows: minimize∑(E i -T i ) 2 ,subject to∑E i =E total E i Let T be the energy of the i-th early reflection. i For the target energy, E total This represents the total reflected energy. By adjusting the energy distribution of early reflections based on energy attenuation values, target sound field characteristics, and a reflection energy optimization algorithm, controllable diffuse sound field and precise acoustic control within the target space are achieved, optimizing the clarity, spatiality, and localization of the audio sound.

[0155] By adjusting time alignment, phase correction, and reflected energy, a controllable diffuse sound field and precise acoustic control within the target space are achieved, optimizing the clarity, spatiality, and localization of audio sound, creating a natural diffuse sound field, and significantly improving the stability and spatiality of the sound image.

[0156] 306. Adjust the reverberation time in the acoustic characteristic information to obtain the adjusted reverberation time;

[0157] In one implementation, when adjusting the reverberation time in the acoustic characteristic information to obtain the adjusted reverberation time, the following steps can be taken: determine the ideal reverberation time according to the scene type; construct a multi-channel feedback delay network, wherein the multi-channel feedback delay network is used to simulate natural reverberation; and adjust the reverberation time in the acoustic characteristic information based on the ideal reverberation time and the multi-channel feedback delay network to obtain the adjusted reverberation time.

[0158] Here, as an example rather than a limitation, it is possible to obtain the scene type of the target space and the content type of the audio to be processed. The ideal reverberation time is determined based on the scene type of the target space, the content type of the audio to be processed, and a preset calculation function. The ideal reverberation time takes into account the optimal reverberation time ratio for different frequency bands. The specific calculation function is as follows:

[0159] RT o pt = k * (V / S) (1 / 3) k is a coefficient determined based on the purpose (scene type) of the target space, V is the volume of the target space, and S is the total surface area of ​​the target space.

[0160] Among them, as an example but not limited to, the constructed multi-channel feedback delay network can be specifically as follows:

[0161] y(n) = b T *z(n) + d*x(n), z(n) is the state vector of the feedback delay network, indicating the signal input after passing through several delay units, b, d, c are gain coefficients, x(n) is the input signal, that is, the data that needs to be mixed with the sound, A is the feedback matrix, and D is the delay matrix. By constructing a multi-channel feedback delay network, frequency-dependent reverb decay is achieved, and real space characteristics are simulated.

[0162] In an implementation, when adjusting the reverb time in the acoustic characteristic information based on the ideal reverb time and the multi-channel feedback delay network, the adjusted reverb time is obtained, the actual reverb characteristics of the target space can be analyzed in real time, wherein the actual reverb characteristics can be the actual reverb time, and the actual reverb characteristics of the target space can be estimated from the impulse response information by a preset reverb time estimation algorithm to obtain the actual reverb characteristics of the target space, and the reverb time estimation algorithm is specifically as follows: RT60 = argmax_RTp(h|RT), h is the impulse response information, and p(h|RT) is the probability of observing h under a given reverb time RT; the artificial reverb parameters of the multi-channel feedback delay network are adjusted based on the ideal reverb time and the actual reverb characteristics of the target space, and the adjusted reverb time is obtained. Through the adaptive condition of the reverb parameter, the natural reverb is compensated or enhanced.

[0163] By adjusting the reverb time in the acoustic characteristic information based on the ideal reverb time and the multi-channel feedback delay network, the control of the reverb time is realized, the natural reverb and the real space characteristics are simulated, and the fullness of the low-frequency response is maintained.

[0164] 307、Through the adaptive equalization model, the adaptive modal control model, the optimized early reflection information, and the adjusted reverb time, the to-be-processed audio is regulated and processed to obtain regulated audio;

[0165] By creating the adaptive equalization model, the adaptive modal control model, the early reflection optimization, and the reverb time adaptive control, the reliability and accuracy of the target acoustic characteristic information are improved, which is beneficial to effectively overcome the negative impact of space acoustics on sound quality, so that close-to-ideal sound playback effects can be obtained in different environments.

[0166] 308、The gain of the multi-channel corresponding to the regulated audio is regulated to obtain the audio after being regulated again;

[0167] In one implementation, after adjusting the gain of the multi-channel corresponding to the modulated audio to obtain the re-modulated audio, the following can be further performed: acquiring noise prediction information of the target space, wherein the noise prediction information is used to indicate the noise level at different time periods; adjusting the compression ratio and compression threshold of the multi-channel corresponding to the modulated audio based on the noise prediction information to obtain the adjusted multi-channel; and performing masking-based gain control on the modulated audio based on the adjusted multi-channel to obtain the re-modulated audio.

[0168] In one implementation, when obtaining noise prediction information for the target space, the background noise of the target space can be identified using a preset long-time average spectrum algorithm, wherein the long-time average spectrum algorithm is as follows: LTAS(f)=(1 / M)*∑|X m (f)| 2 M is the total number of frames, X m (f) is the short-time Fourier transform of the m-th frame; a noise model is established based on the background noise of the target space; the noise level of different time periods is predicted by the noise model to obtain the noise prediction information of the target space.

[0169] In one implementation, the compression ratio is used to indicate the degree of compression of the input signal after it exceeds a threshold. It defines the input-output relationship of the signal when the signal exceeds the set threshold; the higher the compression ratio, the stronger the signal compression. The compression threshold is used to indicate the trigger point for dynamic range compression, defining when the signal begins to be compressed. When adjusting the compression ratio and compression threshold of the multi-channel audio corresponding to the modulated audio based on noise prediction information to obtain the adjusted multi-channel audio, the following can be done: The compression ratio and compression threshold of the multi-channel audio corresponding to the modulated audio can be adjusted using a preset multi-band dynamic range compression algorithm and noise prediction information to obtain the adjusted multi-channel audio. Specifically, the multi-band dynamic range compression algorithm is as follows: y(n) = x(n) * G(n), G(n) = min(1, (T / |x(n)|) (1-1 / R) Let y(n) be the output signal after dynamic range compression, G(n) be the gain (or attenuation) coefficient of the dynamic range compressor used to control the signal amplitude, T be the compression threshold, |x(n)| be the signal amplitude, and R be the compression ratio. When the input signal amplitude |x(n)| exceeds the compression threshold T, the compressor starts working, reducing the increment of the input signal. When the input signal amplitude |x(n)| is less than or equal to the compression threshold T, it remains in its original state and no compression is performed. By using a preset multi-band dynamic range compression algorithm and noise prediction information, the compression ratio and compression threshold of the multi-channel audio corresponding to the regulated audio are adjusted. This controls the signal amplitude, keeping it within a reasonable range. It controls signals with large intensity variations, reducing their dynamic range and preventing signals from being too strong or too weak, ensuring auditory comfort, which is especially important in environments with high spatial noise.

[0170] In an implementation manner, when the adjusted multi-channel pair is used to control the post-control audio based on the masking-based gain control to obtain the re-controlled audio, the post-control audio can be processed by a preset masking-based gain control algorithm to obtain the re-controlled audio, wherein the masking-based gain control algorithm is specifically as follows: G_opt = argmax_G(SNR_perceptual(G)-a*|G-G_prev|), SNR_perceptual(G) is a perceptual signal-to-noise ratio, a is a smoothing factor, G_prev is a gain of a previous frame, and G is a gain of a current frame. By using the preset masking-based gain control algorithm to process the post-control audio based on the adjusted multi-channel pair, automatic adjustment of the volume is realized, a proper ratio of the signal and the noise is maintained, the hearing masking effect is considered, and the gain adjustment strategy is optimized.

[0171] By adjusting the compression ratio and the compression threshold of the multi-channel pair corresponding to the post-control audio based on the noise prediction information, and controlling the post-control audio based on the adjusted multi-channel pair based on the masking-based gain control, multi-band compression is realized, the spectral balance of the signal is protected, automatic adjustment of the volume is realized, a proper ratio of the signal and the noise is maintained, the hearing masking effect is considered, the gain adjustment strategy is optimized, and it is ensured that the dynamic range of the audio is fully exhibited in a quiet environment, while avoiding too stimulating sound.

[0172] 309、The re-controlled audio is processed by multi-channel collaborative optimization to obtain audio with a target effect.

[0173] In an implementation manner, when the re-controlled audio is processed by multi-channel collaborative optimization to obtain audio with a target effect, the channel cross-influence characteristics between the adjusted multi-channel pairs can be analyzed; the re-controlled audio is phase-compensated and amplitude-compensated based on the channel cross-influence characteristics to obtain balanced sound effects; and the balanced sound effects are processed based on a preset adaptive cross-cancellation algorithm to obtain audio with a target effect.

[0174] In an implementation manner, when the channel cross-influence characteristics between the adjusted multi-channel pairs are analyzed, the acoustic coupling characteristics between the adjusted channels can be measured and analyzed; a multi-channel interaction model is established based on the acoustic coupling characteristics; and the channel cross-influence characteristics between the adjusted multi-channel pairs are analyzed based on a multi-channel transfer function matrix by using the multi-channel interaction model, wherein the multi-channel transfer function matrix is specifically as follows:

[0175] H = [H 11 H 12 ...H 1N ;H 21 H22 ...H 2N ;...H N1 H N2 ...H NN ],H ij is the transfer function from jth channel to ith position.

[0176] In an implementation, when the phase compensation and amplitude compensation are performed on the re-regulated audio based on the inter-channel cross effect characteristics, the equalized sound effect is obtained, and the multi-channel joint equalizer can be created based on the inter-channel cross effect characteristics, wherein the multi-channel joint equalizer can implement: W o pt=(H H *H+λI) -1 *H H , H is a multi-channel transfer function matrix used in the inter-channel cross effect characteristics, λ is a regularization parameter, I is a unit matrix, is a weight calculation formula of the multi-channel Wiener filter, is a matrix whose diagonal elements are 1 and other elements are 0, and the size is the same as H H , which is used for regularization; the phase compensation and amplitude compensation are performed on the re-regulated audio by the multi-channel joint equalizer, and the equalized sound effect is obtained. Through the multi-channel joint equalizer, the phase and amplitude compensation between channels are realized.

[0177] As an example but not limitation, the adaptive cross cancellation algorithm can be:

[0178] w i (n+1)=w i (n)+μ*e(n)*x i (n) / (||x(n)|| 2 +δ),w i (n) is the coefficient of the ith filter, μ is the step size, e(n) is the error signal, x i (n) is the input signal of each channel, and δ is a small positive number.

[0179] By performing the phase compensation and amplitude compensation on the re-regulated audio based on the inter-channel cross effect characteristics, and processing the equalized sound effect based on the preset adaptive cross cancellation algorithm, the phase and amplitude compensation between channels are realized, the adaptive interference cancellation between channels is realized, the negative interference between channels is minimized, the comb filtering effect caused by crosstalk is reduced, and the intelligibility of the dialogue and the separation degree of the front sound field in the audio are improved.

[0180] In an implementation, after the multi-channel collaborative optimization of the re-regulated audio is performed to obtain the audio with the target effect, the analysis data is obtained, and the mechanism update and / or strategy optimization of the computer-implemented method of audio control are performed based on the analysis data, wherein the analysis data includes performance test data and user evaluation data.

[0181] In an implementation, the performance test data includes, but is not limited to, an acoustic parameter comprehensive score, and the performance test data can be obtained by calculating the acoustic parameter comprehensive score through an acoustic parameter comprehensive score algorithm, wherein the acoustic parameter comprehensive score algorithm is as follows:

[0182] Score = w1*F_flatness + w2*T_60_deviation + w3*C_50 + w4*D_50 + w5*LF, w1, w2, w3, w4, and w5 are weight coefficients, F_flatness is the frequency response flatness, T_60_deviation is the reverberation time deviation, C_50 is the intelligibility, D_50 is the definition, and LF is the lateral energy fraction.

[0183] In an implementation, the user evaluation data includes data after double-blind testing by professional audio engineers and ordinary users, and data for evaluating the overall sound quality performance of the system by using the MUSHRA method.

[0184] In an implementation, the cloud big data can be self-optimized based on the analysis data, or the algorithm performance (mechanism update and / or strategy optimization of the computer-implemented method of audio control) can be improved through remote firmware update based on the analysis data, wherein the cloud big data self-optimized based on the analysis data can be achieved through a parameter optimization algorithm based on swarm intelligence, and the parameter optimization algorithm based on swarm intelligence is as follows: i (t+1) = p i (t) + v i (t+1), v i (t+1) = w*v i (t) + c1*r1*(p best -p i (t)) + c2*r2*(g best -p i (t)), p i (t+1) is a parameter vector, v i (t+1) is a velocity vector, w, c1, and c2 are control parameters, r1 and r2 are random numbers, p best is an individual optimal solution, p i (t) is an individual optimization strategy, g best is a global optimal solution.

[0185] For example, in 50 different home environments, the system improved the frequency response flatness to within ±3 dB in 90% of the scenarios, and the reverberation time control error was less than 10%. In MUSHRA tests, the system achieved an average score of 85, significantly higher than the 65 in the unoptimized state. Through cloud data analysis, the system identified the trend of low-frequency response being generally weak in rooms with a volume of 56-60 m 3 The optimization strategy for the corresponding room type was updated accordingly, further improving the system performance.

[0186] Through rigorous performance testing and user evaluation, it demonstrates excellent performance and stability. Through cloud big data analysis and remote update mechanism, it can continuously optimize performance and make customized adjustments for special needs of different types of space, greatly improving the adaptability and long-term value of the audio system.

[0187] This computer-implemented method of audio control uses advanced spatial acoustic measurement technology combined with artificial intelligence algorithms to accurately analyze room acoustic characteristics (including frequency response, reverberation time, early reflections, room modes, etc.). It introduces adaptive filtering and intelligent equalization technology to automatically adjust audio parameters based on analysis results, achieving precise compensation for spatial acoustic defects. The innovative room mode suppression algorithm effectively improves low-frequency response. In addition, it integrates real-time adaptive technology to dynamically respond to changes in room acoustic characteristics (such as opening and closing doors and windows, personnel movement, etc.), continuously optimizing audio output. Automatic adjustment brings significant sound quality improvement and user convenience. It effectively overcomes the negative impact of room acoustics on sound quality, allowing near-ideal sound reproduction in different environments. Intelligent and automated features allow users to easily obtain professional-level audio adjustment results without professional knowledge. By precisely controlling frequency response and phase characteristics, the system significantly improves sound clarity, localization, and spatiality. Dynamic adaptive functionality ensures optimal sound quality even when room usage changes. The system's room mode processing technology significantly improves low-frequency performance, providing more balanced and powerful low-frequency effects. Through innovative algorithm design and system architecture, it comprehensively addresses many challenges faced by traditional audio systems in real-world room environments. It not only accurately analyzes and compensates for the impact of room acoustics on sound quality but also implements intelligent dynamic optimization, providing strong technical support for users to achieve optimal audio experience in various environments. Adaptive learning ability and real-time optimization features enable it to adapt to different acoustic environments, audio content, and user preferences, representing a major leap in home audio system technology. Through rigorous performance testing and user evaluation, the system demonstrates excellent performance and stability, and is expected to be widely used in high-end audio markets, and to guide the development of future smart home audio technology.

[0188] In this embodiment, by performing compensation processing, gain control processing and multi-channel collaborative optimization processing based on the acoustic characteristic information obtained by analyzing the environment response information based on the specified measurement signal sound effect test, the precise compensation of the target space acoustic defects is realized, the inter-channel crosstalk problem is effectively solved, the sound image positioning and spatial sense are significantly improved, and the overall listening of the audio is greatly improved. Feel, can adapt to different listening environment and sound source type, continuously optimize the audio output, ensure that the best listening experience can be obtained in various environments.

[0189] The present disclosure also provides a computer readable storage medium, which can be a non-volatile computer readable storage medium or a volatile computer readable storage medium, and the computer readable storage medium stores a computer program, which, when running on a computer, causes the computer to perform the following steps:

[0190] Performing a sound effect test on the target space based on a specified measurement signal to obtain environment response information, wherein the environment response information includes impulse response information and information obtained by separating the impulse response information;

[0191] Analyzing the environment response information to obtain acoustic characteristic information of the target space;

[0192] Performing compensation processing on the audio to be processed based on the acoustic characteristic information to obtain regulated audio;

[0193] Controlling the gain of the multi-channel corresponding to the regulated audio to obtain audio after re-regulation;

[0194] Performing multi-channel collaborative optimization processing on the audio after re-regulation to obtain target effect audio.

[0195] In an embodiment, performing a sound effect test on the target space based on a specified measurement signal to obtain environment response information, comprising:

[0196] Collecting response information of multiple measurement points in the target space to the specified measurement signal by pre-setting an array of collection devices, wherein the multiple measurement points form a measurement grid;

[0197] Extracting the response information of the multiple measurement points respectively to obtain impulse response information corresponding to the multiple measurement points respectively;

[0198] Separating the impulse response information corresponding to the multiple measurement points respectively to obtain direct sound, early reflection and reverberation tail corresponding to the multiple measurement points respectively;

[0199] Determining the impulse response information, direct sound, early reflection and reverberation tail corresponding to the multiple measurement points respectively as the environment response information of the target space.

[0200] In an embodiment, the environmental response information is analyzed to obtain acoustic characteristic information of the target space, including:

[0201] The impulse response information in the environmental response information is subjected to spectral analysis to obtain spectral analysis results, wherein the spectral analysis results include a frequency response curve and a power spectral density;

[0202] The environmental mode parameters of the target space are identified based on the power spectral density, wherein the environmental mode parameters include modal density and modal overlap;

[0203] The reverberation times of the plurality of measurement points are calculated based on the reverb tails of the plurality of measurement points, respectively;

[0204] The early reflections and direct sounds of the plurality of measurement points are analyzed to obtain early reflection analysis results of the plurality of measurement points, respectively;

[0205] The spectral analysis results, the environmental mode parameters of the target space, and the reverberation times and early reflection analysis results of the plurality of measurement points, respectively, are determined as the acoustic characteristic information of the target space.

[0206] In an embodiment, the to-be-processed audio is compensated based on the acoustic characteristic information to obtain regulated audio, including:

[0207] An adaptive equalization model is created based on the frequency response curve in the acoustic characteristic information, wherein the adaptive equalization model is used for adaptive filtering of frequency band limitation;

[0208] An adaptive modal control model is created, wherein the adaptive modal control model is used for adaptive dynamic control of environmental modes;

[0209] The early reflection analysis results in the acoustic characteristic information are subjected to optimization processing to obtain optimized early reflection information;

[0210] The reverberation times in the acoustic characteristic information are adjusted to obtain adjusted reverberation times;

[0211] The to-be-processed audio is regulated by the adaptive equalization model, the adaptive modal control model, the optimized early reflection information, and the adjusted reverberation times to obtain the regulated audio.

[0212] In an embodiment, an adaptive equalization model is created based on the frequency response curve in the acoustic characteristic information, including:

[0213] The frequency response curve in the acoustic characteristic information is adjusted by a specified standard curve to obtain a target response curve;

[0214] Construct a parameter equalization model based on the target response curve;

[0215] Adjust the parameters of the parameter equalization model through a preset adaptive algorithm to obtain an adaptive equalization model.

[0216] In an implementation, creating an adaptive modal control model includes:

[0217] Identifying a mode type of the target space;

[0218] Setting a filter corresponding to the mode type;

[0219] Adjusting parameters of the filter corresponding to the mode type to obtain an adaptive modal control model.

[0220] In an implementation, the early reflection analysis result in the acoustic characteristic information is optimized to obtain optimized early reflection information, including:

[0221] Based on the early reflection analysis result in the acoustic characteristic information, the delay time and energy attenuation value of each early reflection path are calculated;

[0222] Based on the delay time, the early reflections are time-aligned, and the impulse response information is phase-corrected to obtain processed information;

[0223] Obtaining target sound field characteristics, wherein the target sound field characteristics are ideal environmental acoustic characteristics;

[0224] Based on the energy attenuation value and the target sound field characteristics, the reflection energy in the early reflection analysis result is adjusted to obtain adjusted reflection energy;

[0225] The processed information and the adjusted reflection energy are determined as the optimized early reflection information.

[0226] In an implementation, the reverberation time in the acoustic characteristic information is adjusted to obtain an adjusted reverberation time, including:

[0227] Determining an ideal reverberation time according to a scene type;

[0228] Constructing a multi-channel feedback delay network, wherein the multi-channel feedback delay network is used to simulate natural reverberation;

[0229] Based on the ideal reverberation time and the multi-channel feedback delay network, the reverberation time in the acoustic characteristic information is adjusted to obtain an adjusted reverberation time.

[0230] In an implementation, the gain of the multi-channel corresponding to the regulated audio is regulated to obtain the audio after the second regulation, including:

[0231] Obtaining noise prediction information of the target space, wherein the noise prediction information is used to indicate noise levels of different time periods;

[0232] Adjusting a compression ratio and a compression threshold of the multi-channel corresponding to the regulated audio based on the noise prediction information, to obtain an adjusted multi-channel;

[0233] Performing gain control based on masking on the regulated audio based on the adjusted multi-channel, to obtain audio after re-regulation.

[0234] In an embodiment, the audio after re-regulation is subjected to multi-channel collaborative optimization processing to obtain audio with a target effect, comprising:

[0235] Analyzing channel cross-influence characteristics between the adjusted multi-channels;

[0236] Performing phase compensation and amplitude compensation on the audio after re-regulation based on the channel cross-influence characteristics, to obtain balanced sound effects;

[0237] Processing the balanced sound effects based on a preset adaptive cross-cancellation algorithm to obtain audio with a target effect.

[0238] In an embodiment, before obtaining environmental response information by testing the sound effects of the target space based on a specified measurement signal, further comprising:

[0239] Identifying a current audio use scenario and automatically matching a control strategy corresponding to the audio use scenario, wherein the control strategy comprises acoustic characteristic information content, compensation processing content, and compensation processing flow.

[0240] In an embodiment, after the multi-channel collaborative optimization processing of the audio after re-regulation to obtain audio with a target effect, further comprising:

[0241] Obtaining analysis data and updating the mechanism and / or optimizing the strategy of the audio control method based on the analysis data, wherein the analysis data comprises performance test data and user evaluation data.

[0242] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the above-described system and unit can refer to the corresponding process in the foregoing method embodiments, which will not be described here.

[0243] The integrated unit, if implemented in the form of a software function unit and sold or used as an independent product, can be stored in a computer readable storage medium. Based on such understanding, the technical solutions of the present disclosure, essentially or in other words, the part that contributes to the prior art or the whole or part of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium, and includes several computer programs used to make a computer device (which can be a personal computer, a server, or a network device, etc.) execute all or part of the steps of the methods in the various embodiments of the present disclosure. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), magnetic disk or optical disk, and various media that can store program codes.

[0244] The above, the above embodiments are only to illustrate the technical solutions of the present disclosure, rather than limit them; although the present disclosure is described in detail with reference to the foregoing embodiments, those skilled in the art should understand that the technical solutions recorded in the foregoing embodiments can still be modified, or some technical features can be replaced by equivalents; and these modifications or replacements do not make the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present disclosure.

Claims

A computer-implemented method of audio control, wherein The method comprises the following steps: performing sound effect test on a target space based on a specified measurement signal to obtain environmental response information, wherein the environmental response information comprises impulse response information and information obtained after separation of the impulse response information; analyzing the environmental response information to obtain acoustic characteristic information of the target space; performing compensation processing on a to-be-processed audio based on the acoustic characteristic information to obtain regulated audio; regulating the gain of multiple sound channels corresponding to the regulated audio to obtain audio after secondary regulation; performing multi-sound channel cooperative optimization processing on the audio after secondary regulation to obtain audio with a target effect. The computer-implemented method of audio control of claim 1, wherein, The method of performing sound effect test on a target space based on a specified measurement signal to obtain environmental response information comprises the following steps: collecting response information of multiple measurement points in the target space to the specified measurement signal through a preset acquisition device array, wherein the multiple measurement points form a measurement grid; performing information extraction on the response information of the multiple measurement points respectively to obtain impulse response information corresponding to the multiple measurement points respectively; performing separation on the impulse response information corresponding to the multiple measurement points respectively to obtain direct sound, early reflection and reverberation tail sound corresponding to the multiple measurement points respectively; determining the impulse response information, the direct sound, the early reflection and the reverberation tail sound corresponding to the multiple measurement points respectively as the environmental response information of the target space. The computer-implemented method of audio control of claim 2, wherein, The method of analyzing the environmental response information to obtain acoustic characteristic information of the target space comprises the following steps: performing spectral analysis on the impulse response information in the environmental response information to obtain spectral analysis results, wherein the spectral analysis results comprise a frequency response curve and a power spectral density; identifying environmental mode parameters of the target space based on the power spectral density, wherein the environmental mode parameters comprise modal density and modal overlap; calculating reverberation time corresponding to the multiple measurement points respectively based on the reverberation tail sound corresponding to the multiple measurement points respectively; analyzing the early reflection and the direct sound corresponding to the multiple measurement points respectively to obtain early reflection analysis results corresponding to the multiple measurement points respectively; determining the spectral analysis results, the environmental mode parameters of the target space, and the reverberation time and the early reflection analysis results corresponding to the multiple measurement points respectively as the acoustic characteristic information of the target space. The computer-implemented method of audio control of claim 3, wherein, The method of performing compensation processing on a to-be-processed audio based on the acoustic characteristic information to obtain regulated audio comprises the following steps: creating an adaptive equalization model based on the frequency response curve in the acoustic characteristic information, wherein the adaptive equalization model is used for adaptive filtering of frequency band limitation; creating an adaptive modal control model, wherein the adaptive modal control model is used for adaptive dynamic control of environmental mode; performing optimization processing on the early reflection analysis results in the acoustic characteristic information to obtain optimized early reflection information; adjusting the reverberation time in the acoustic characteristic information to obtain adjusted reverberation time; performing regulation processing on the to-be-processed audio through the adaptive equalization model, the adaptive modal control model, the optimized early reflection information and the adjusted reverberation time to obtain regulated audio. The computer-implemented method of audio control of claim 4, wherein, The adaptive equalization model is created based on the frequency response curve in the acoustic characteristic information, and the adaptive equalization model includes: The frequency response curve in the acoustic characteristic information is adjusted by a specified standard curve to obtain a target response curve; A parameter equalization model is constructed based on the target response curve; The parameters of the parameter equalization model are adjusted by a preset adaptive algorithm to obtain an adaptive equalization model. The computer-implemented method of audio control of claim 4, wherein, The adaptive modal control model is created, and the adaptive modal control model includes: The mode type of the target space is identified; A filter corresponding to the mode type is set; The parameters of the filter corresponding to the mode type are adjusted to obtain an adaptive modal control model. The computer-implemented method of audio control of claim 4, wherein, The early reflection analysis result in the acoustic characteristic information is optimized to obtain optimized early reflection information, and the optimization includes: Based on the early reflection analysis result in the acoustic characteristic information, the delay time and energy attenuation value of each early reflection path are calculated; The early reflections are time-aligned based on the delay time, and the impulse response information is phase-corrected to obtain processed information; A target sound field characteristic is obtained, wherein the target sound field characteristic is an ideal environmental acoustic characteristic; Based on the energy attenuation value and the target sound field characteristic, the reflection energy in the early reflection analysis result is adjusted to obtain adjusted reflection energy; The processed information and the adjusted reflection energy are determined as the optimized early reflection information. The computer-implemented method of audio control of claim 4, wherein, The reverberation time in the acoustic characteristic information is adjusted to obtain an adjusted reverberation time, and the adjustment includes: An ideal reverberation time is determined according to the scene type; A multi-channel feedback delay network is constructed, wherein the multi-channel feedback delay network is used to simulate natural reverberation; Based on the ideal reverberation time and the multi-channel feedback delay network, the reverberation time in the acoustic characteristic information is adjusted to obtain an adjusted reverberation time. The computer-implemented method of audio control of claim 1, wherein, The gain of the multi-channel corresponding to the regulated audio is regulated to obtain the audio after the second regulation, and the regulation includes: Noise prediction information of the target space is obtained, wherein the noise prediction information is used to indicate the noise level of different time periods; Based on the noise prediction information, the compression ratio and compression threshold of the multi-channel corresponding to the regulated audio are adjusted to obtain adjusted multi-channels; Based on the adjusted multi-channels, the gain control based on masking is performed on the regulated audio to obtain the audio after the second regulation. The computer-implemented method of audio control of claim 9, wherein, The multi-channel collaborative optimization processing is performed on the audio after the second regulation to obtain the audio with target effect, and the processing includes: The channel cross-influence characteristics between the adjusted multi-channels are analyzed; Based on the channel cross-influence characteristics, the phase compensation and amplitude compensation are performed on the audio after the second regulation to obtain the equalized sound effect; The equalized sound effect is processed based on a preset adaptive cross-cancellation algorithm to obtain the audio with target effect. The computer-implemented method of audio control of any of claims 1-10, wherein, Before the environmental response information is obtained by testing the sound effect of the target space based on the specified measurement signal, the method further includes: Identify the current audio use scene, and automatically match the control strategy corresponding to the audio use scene, wherein the control strategy includes the content of acoustic characteristic information, compensation processing content, and compensation processing flow. The computer-implemented method of audio control of any of claims 1-10, wherein, After the multi-channel collaborative optimization processing of the re-regulated audio is performed, target effect audio is obtained. Obtain analysis data, and update the mechanism and / or optimize the strategy of the audio control method based on the analysis data, wherein the analysis data includes performance test data and user evaluation data. An electronic device, wherein, The electronic device includes a processor and a memory, the memory stores machine executable instructions executable by the processor, and the processor executes the machine executable instructions to implement the following steps: The environmental response information includes impulse response information and information obtained after separation of the impulse response information; Analyze the environmental response information to obtain acoustic characteristic information of the target space; Perform compensation processing on the to-be-processed audio based on the acoustic characteristic information to obtain regulated audio; Regulate the gain of the multi-channel corresponding to the regulated audio to obtain re-regulated audio; Perform multi-channel collaborative optimization processing on the re-regulated audio to obtain target effect audio. The electronic device of claim 13, wherein, The environmental response information is obtained by performing sound effect testing on the target space based on the specified measurement signal, including: Collect the response information of a plurality of measurement points in the target space to the specified measurement signal through a pre-set acquisition device array, wherein the plurality of measurement points form a measurement grid; Respectively extract the response information of the plurality of measurement points to obtain the impulse response information corresponding to the plurality of measurement points respectively; Respectively separate the impulse response information corresponding to the plurality of measurement points respectively to obtain direct sound, early reflection and reverberation tail corresponding to the plurality of measurement points respectively; Determine the impulse response information, direct sound, early reflection and reverberation tail corresponding to the plurality of measurement points respectively as the environmental response information of the target space. The electronic device of claim 14, wherein, The environmental response information is analyzed to obtain the acoustic characteristic information of the target space, including: Perform spectral analysis on the impulse response information in the environmental response information to obtain a spectral analysis result, wherein the spectral analysis result includes a frequency response curve and a power spectral density; Identify environmental mode parameters of the target space based on the power spectral density, wherein the environmental mode parameters include modal density and modal overlap; Calculate the reverberation time corresponding to the plurality of measurement points based on the reverberation tail corresponding to the plurality of measurement points respectively; Analyze the early reflection and direct sound corresponding to the plurality of measurement points respectively to obtain early reflection analysis results corresponding to the plurality of measurement points respectively; Determine the spectral analysis result, the environmental mode parameters of the target space, and the reverberation time and early reflection analysis results corresponding to the plurality of measurement points respectively as the acoustic characteristic information of the target space. The electronic device of claim 15, wherein, The to-be-processed audio is compensated based on the acoustic characteristic information to obtain regulated audio, including: create an adaptive equalization model based on the frequency response curve in the acoustic characteristic information, wherein the adaptive equalization model is used for adaptive filtering of band limitation; create an adaptive modal control model, wherein the adaptive modal control model is used for adaptive dynamic control of environmental modes; optimize early reflection analysis results in the acoustic characteristic information to obtain optimized early reflection information; adjust the reverberation time in the acoustic characteristic information to obtain an adjusted reverberation time; perform regulation and control processing on the to-be-processed audio through the adaptive equalization model, the adaptive modal control model, the optimized early reflection information, and the adjusted reverberation time to obtain regulated and controlled audio. The electronic device of claim 16, wherein, The adaptive equalization model based on the frequency response curve in the acoustic characteristic information comprises: adjust the frequency response curve in the acoustic characteristic information through a specified standard curve to obtain a target response curve; construct a parameter equalization model based on the target response curve; adjust the parameters of the parameter equalization model through a preset adaptive algorithm to obtain an adaptive equalization model. The electronic device of claim 16, wherein, The adaptive modal control model comprises: identify the mode type of the target space; set a filter corresponding to the mode type; adjust the parameters of the filter corresponding to the mode type to obtain an adaptive modal control model. The electronic device of claim 16, The optimization of the early reflection analysis results in the acoustic characteristic information comprises: based on the early reflection analysis results in the acoustic characteristic information, calculate the delay time and energy attenuation value of each early reflection path; based on the delay time, perform time alignment processing on the early reflection and phase correction on the impulse response information to obtain processed information; obtain target sound field characteristics, wherein the target sound field characteristics are ideal environmental acoustic characteristics; based on the energy attenuation value and the target sound field characteristics, adjust the reflection energy in the early reflection analysis results to obtain adjusted reflection energy; determine the processed information and the adjusted reflection energy as the optimized early reflection information. A computer-readable storage medium having stored thereon a computer program, wherein When the computer program is executed by the processor, the computer program causes the processor to implement the following steps: perform acoustic effect testing on the target space based on a specified measurement signal to obtain environmental response information, wherein the environmental response information includes impulse response information and information obtained after separation of the impulse response; analyze the environmental response information to obtain acoustic characteristic information of the target space; perform compensation processing on the to-be-processed audio based on the acoustic characteristic information to obtain regulated and controlled audio; regulate the gain of multiple sound channels corresponding to the regulated and controlled audio to obtain audio that has been regulated and controlled again; perform multi-channel collaborative optimization processing on the audio that has been regulated and controlled again to obtain audio with a target effect.

Images