High-quality audio data processing method and system

Abstract

The invention relates to the field of audio data processing, in particular to a high-tone-quality audio data processing method and system. The method comprises the following steps: identifying an audio signal to be processed; performing multi-time-window segmentation processing and adaptive noise filtering on the audio signal to be processed to obtain a noise filtering optimized audio signal; performing multi-scale spectral decomposition on the noise filtering optimized audio signal, capturing transient audio details, and identifying transient audio detail information of each scale; performing audio frequency range calculation according to the transient audio detail information of each scale, and performing dynamic full-band range adjustment to obtain transient information full-band optimized audio; and carrying out environmental acoustic effect mining on the transient information full-band optimized audio, and carrying out current audio scene inference to obtain a real-time audio propagation scene. According to the invention, the space sense, the precision and the fidelity of the audio are improved.

Description

Technical Field

[0001] The present invention relates to the field of audio data processing, and in particular to a high-quality audio data processing method and system. Background Art

[0002] With the continuous development of audio technology, high-quality audio plays an increasingly important role in various audio devices. From home audio to professional recording studios, from music players to virtual reality systems, the demand for high-quality audio continues to increase. In order to meet the diverse needs of modern society for audio experience, how to achieve clear, delicate, realistic and deep audio output has become a technical challenge. In the long-term audio processing process, audio signals are often affected by various factors, such as noise interference, frequency distortion, dynamic range compression, environmental acoustic effects, etc. These factors will cause the quality of audio signals to deteriorate and affect the user's listening experience.

[0003] Traditional audio data processing methods mainly rely on static signal processing technologies, such as spectrum analysis and equalizer adjustment. Although they can improve the sound quality to a certain extent, they are often powerless when faced with complex audio signals, dynamically changing audio characteristics and changing environmental conditions. More importantly, traditional methods cannot dynamically optimize according to real-time audio environment changes, resulting in poor flexibility and adaptability of audio output, and cannot fully meet users' pursuit of high-quality audio.

[0004] With the rapid development of artificial intelligence and deep learning technologies, more and more intelligent audio processing methods have been proposed. These methods can dynamically analyze and process audio signals and improve sound quality by simulating the auditory characteristics of the human ear and optimizing intelligent algorithms. However, existing intelligent audio processing technologies still have performance problems in some application scenarios, such as loss of transient details, excessive processing of low and high frequencies, and insufficient sense of space.

[0005] Therefore, how to break through the limitations of existing audio processing methods through innovative technical means and adopt more intelligent and real-time technologies to improve the quality of audio signals has become a challenge that needs to be urgently solved in the current audio processing field. Summary of the invention

[0006] In order to solve the above technical problems, the present invention proposes a high-quality audio data processing method and system to solve at least one of the above technical problems.

[0007] To achieve the above object, the present invention provides a high-quality audio data processing method, comprising the following steps: Step S1: Identify the audio signal to be processed; perform multi-time window segmentation processing and adaptive noise filtering on the audio signal to be processed to obtain a noise-filtered optimized audio signal; Step S2: performing multi-scale spectrum decomposition on the noise filtering optimized audio signal, and capturing transient audio details, and identifying transient audio detail information of each scale; Step S3: Calculate the audio frequency range according to the transient audio detail information of each scale, and perform dynamic full-band range adjustment to obtain transient information full-band optimized audio; Step S4: mining the environmental acoustic effect of the full-band optimized audio of transient information, and inferring the current audio scene to obtain a real-time audio propagation scene; Step S5: performing stereo field sound effect enhancement on the full-band optimized audio of transient information according to the real-time audio propagation scenario, thereby generating a stereo field enhanced audio signal; Step S6: performing propagation delay optimization on the stereo field enhanced audio signal and performing global distortion prediction optimization to generate global adaptive distortion optimized audio.

[0008] The present invention ensures that we process the correct audio signal through the recognition of audio signals, whether it is background music, voice signal or other types of audio. By classifying audio signals, different processing strategies can be adopted for different types of signals. The multi-time window segmentation technology is used to divide the audio signal into multiple small segments so that each segment can be processed independently, which can effectively avoid the calculation redundancy caused by too long time domain signals, and can more accurately identify and filter noise in a local range. Through adaptive noise filtering, the parameters of the filter are automatically adjusted for different noise types and intensities to achieve more effective noise elimination. Compared with traditional static noise suppression methods, this adaptive filtering can better cope with complex and dynamically changing noise environments and retain the original information of the signal to the maximum extent. Through multi-scale spectrum analysis, the audio signal is decomposed into different frequency bands and time scales. In this way, not only can the features in different frequency ranges be extracted, but also the transient audio details can be effectively captured, especially the delicate changes in the high-frequency part. The spectrum decomposition of each scale can help the system accurately distinguish the audio information of different frequencies and time periods, and ensure the accuracy of processing. Transient audio refers to the sudden change part of the audio signal, such as drum sounds, playing of musical instruments, etc. By accurately capturing the transient part, it is possible to ensure that these key details in the audio are preserved, making the audio more vivid and expressive. Traditional audio processing methods often lack in transient processing. The addition of this step significantly enhances the dynamics and details of the audio. After the transient audio is captured, the frequency range of each transient audio is calculated to help determine which frequency areas need to be strengthened or weakened. Through accurate analysis of the audio frequency, the spectral distribution of the audio can be identified and optimized to ensure that the frequency of each part of the audio is properly adjusted. Based on the frequency range of the transient audio, dynamic full-band adjustment is performed. This means that the low-frequency, mid-frequency, and high-frequency parts of the audio signal will be automatically adjusted according to the key information of each frequency band to optimize the sound quality performance of each frequency band. Through this refined frequency adjustment, the transparency and layering of the audio can be improved, ensuring that every detail of the transient audio can be optimized throughout the entire frequency band. By exploring the environmental acoustic effects in the audio (such as reverberation, reflection, etc.), we can understand the propagation environment of the audio signal and extract its impact on the sound quality. The extraction of environmental acoustic effects not only helps to understand the reflection path and propagation characteristics of sound, but also provides important information for subsequent spatial sound enhancement. Real-time scene inference based on environmental acoustic effects helps to understand the spatial propagation scenario of the current audio. Depending on the scene, the sound effects of audio can vary greatly. For example, the sound in a large space is relatively hollow and reverberant, while in a small space it is more concentrated and clear. By inferring the audio scene, more intelligent adjustments can be made for subsequent processing. According to the real-time audio propagation scenario, a spatial sound effect algorithm is used to enhance the stereo field of the audio.This process not only provides a richer stereo effect, but also optimizes the spatial sense of the audio, making the audio more realistic and vivid in hearing. Especially under headphones or surround sound systems, the enhancement of the stereo field sound effect can greatly enhance the auditory immersion. The spatial sense of the audio is greatly enhanced by the enhancement of the stereo field. By simulating the spatial distribution of the audio, the sense of positioning of the sound source and the width of the sound field are enhanced, further enriching the performance level of the audio. During the audio propagation process, delays can cause different parts of the audio signal to be misaligned in time, affecting the accuracy of hearing. Through propagation delay optimization, the audio can be time-aligned to ensure that all audio signals remain synchronized during the propagation process and eliminate audio distortion caused by delay. During the long-term propagation of audio signals, distortion may occur due to equipment or environmental reasons. Through the distortion prediction model, the possible distortion bands can be monitored and identified in real time, and adaptive gain optimization can be performed. Distortion compensation can ensure that the final audio has no obvious artifacts and unnatural sound effects, and improve the accuracy and fidelity of the sound quality.

[0009] In this specification, a high-quality audio data processing system is provided, which is used to execute the high-quality audio data processing method as described above, including: The noise filtering module is used to identify the audio signal to be processed; perform multi-time window segmentation processing and adaptive noise filtering on the audio signal to be processed to obtain a noise filtering optimized audio signal; The transient detail capture module is used to perform multi-scale spectrum decomposition on the noise filtering optimized audio signal, and to capture transient audio details, and identify transient audio detail information at each scale; A full-band range adjustment module is used to calculate the audio frequency range according to the transient audio detail information of each scale, and to perform dynamic full-band range adjustment to obtain transient information full-band optimized audio; The acoustic effect mining module is used to mine the environmental acoustic effects of the full-band optimized audio of transient information and infer the current audio scene to obtain the real-time audio propagation scene; A stereo field enhancement module is used to enhance the stereo field sound effect of the full-band optimized audio of transient information according to the real-time audio propagation scenario, thereby generating a stereo field enhanced audio signal; The distortion optimization module is used to optimize the propagation delay of the stereo field enhanced audio signal and perform global distortion prediction optimization to generate global adaptive distortion optimized audio.

[0010] The present invention effectively removes background noise through adaptive noise filtering to ensure the clarity of the original audio signal. This provides a cleaner input signal for subsequent audio processing and helps to improve the overall quality of the audio. Through multi-time window processing, the changes of audio signals in different time periods can be analyzed more carefully, and customized processing can be performed for different signal features, avoiding the limitations of a single time window processing method. Through multi-scale spectrum decomposition, the audio details in different frequency intervals can be fully analyzed to ensure high-fidelity capture and reproduction of transient audio, and avoid dynamic details in the audio from being ignored or distorted. The capture of transient audio details allows every subtle change in the audio to be retained, especially for audio that requires a quick response (such as musical instrument performance, voice, etc.), which can provide a more realistic and natural sound quality. By dynamically adjusting the frequency band of the audio signal, the audio details of each frequency band are optimized, avoiding excessive suppression of high frequencies or distortion of low frequencies, and providing a more balanced and natural sound quality. Transient information optimization can prevent the transient signal of the audio from being smoothed excessively, ensure that the rapidly changing details in the audio (such as drum beats, piano hitting sounds, etc.) are accurately retained, and improve the clarity and realism of the audio. By mining the acoustic effect, the environmental background of audio propagation can be accurately restored. This is particularly important for recording and audio playback, and can provide data support for the spatial sense and positioning sense of audio. Scene inference enables audio signals to be simulated according to the characteristics of the environment, thereby enhancing the immersiveness of audio and the realism of hearing, which is particularly suitable for application scenarios such as virtual reality and game audio. Through stereo field sound enhancement, the spatial sense of audio signals is significantly improved. The sound is not only emitted from two headphones or speakers, but also simulates a rich three-dimensional sound space, which greatly enhances the auditory experience. The enhanced audio is not only directional in position, but also has a sense of depth and layering, especially in applications such as virtual reality and cinema audio, which can make the audience feel more natural and realistic sound effects. Through delay optimization, signal synchronization in real-time audio applications is ensured, audio asynchrony caused by propagation delay is avoided, and the audio interaction experience is improved. Through global distortion prediction, each frequency segment in the audio is optimized to avoid sound quality degradation caused by environmental interference or propagation problems, and ensure the high-fidelity effect of audio output. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] Figure 1 A schematic flow chart of the steps of a high-quality audio data processing method of the present invention; Figure 2 Detailed implementation flow chart of step S1; Figure 3 Detailed implementation flow chart of step S2; Figure 4 Detailed implementation flow chart of step S3. DETAILED DESCRIPTION

[0012] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0013] The present application example provides a high-quality audio data processing method and system. The execution subject of the high-quality audio data processing method and system includes but is not limited to: mechanical equipment, data processing platform, cloud server node, network upload device, etc. equipped with the system can be regarded as the general computing node of the present application, and the data processing platform includes but is not limited to: at least one of an audio image management system, an information management system, and a cloud data management system.

[0014] See also Figures 1 to 4 The present invention provides a high-quality audio data processing method, the high-quality audio data processing method comprising the following steps: Step S1: Identify the audio signal to be processed; perform multi-time window segmentation processing and adaptive noise filtering on the audio signal to be processed to obtain a noise-filtered optimized audio signal; Step S2: performing multi-scale spectrum decomposition on the noise filtering optimized audio signal, and capturing transient audio details, and identifying transient audio detail information of each scale; Step S3: Calculate the audio frequency range according to the transient audio detail information of each scale, and perform dynamic full-band range adjustment to obtain transient information full-band optimized audio; Step S4: mining the environmental acoustic effect of the full-band optimized audio of transient information, and inferring the current audio scene to obtain a real-time audio propagation scene; Step S5: performing stereo field sound effect enhancement on the full-band optimized audio of transient information according to the real-time audio propagation scenario, thereby generating a stereo field enhanced audio signal; Step S6: performing propagation delay optimization on the stereo field enhanced audio signal and performing global distortion prediction optimization to generate global adaptive distortion optimized audio.

[0015] The present invention ensures that we process the correct audio signal through the recognition of audio signals, whether it is background music, voice signal or other types of audio. By classifying audio signals, different processing strategies can be adopted for different types of signals. The multi-time window segmentation technology is used to divide the audio signal into multiple small segments so that each segment can be processed independently, which can effectively avoid the calculation redundancy caused by too long time domain signals, and can more accurately identify and filter noise in a local range. Through adaptive noise filtering, the parameters of the filter are automatically adjusted for different noise types and intensities to achieve more effective noise elimination. Compared with traditional static noise suppression methods, this adaptive filtering can better cope with complex and dynamically changing noise environments and retain the original information of the signal to the maximum extent. Through multi-scale spectrum analysis, the audio signal is decomposed into different frequency bands and time scales. In this way, not only can the features in different frequency ranges be extracted, but also the transient audio details can be effectively captured, especially the delicate changes in the high-frequency part. The spectrum decomposition of each scale can help the system accurately distinguish the audio information of different frequencies and time periods, and ensure the accuracy of processing. Transient audio refers to the sudden change part of the audio signal, such as drum sounds, playing of musical instruments, etc. By accurately capturing the transient part, it is possible to ensure that these key details in the audio are preserved, making the audio more vivid and expressive. Traditional audio processing methods often lack in transient processing. The addition of this step significantly enhances the dynamics and details of the audio. After the transient audio is captured, the frequency range of each transient audio is calculated to help determine which frequency areas need to be strengthened or weakened. Through accurate analysis of the audio frequency, the spectral distribution of the audio can be identified and optimized to ensure that the frequency of each part of the audio is properly adjusted. Based on the frequency range of the transient audio, dynamic full-band adjustment is performed. This means that the low-frequency, mid-frequency, and high-frequency parts of the audio signal will be automatically adjusted according to the key information of each frequency band to optimize the sound quality performance of each frequency band. Through this refined frequency adjustment, the transparency and layering of the audio can be improved, ensuring that every detail of the transient audio can be optimized throughout the entire frequency band. By exploring the environmental acoustic effects in the audio (such as reverberation, reflection, etc.), we can understand the propagation environment of the audio signal and extract its impact on the sound quality. The extraction of environmental acoustic effects not only helps to understand the reflection path and propagation characteristics of sound, but also provides important information for subsequent spatial sound enhancement. Real-time scene inference based on environmental acoustic effects helps to understand the spatial propagation scenario of the current audio. Depending on the scene, the sound effects of audio can vary greatly. For example, the sound in a large space is relatively hollow and reverberant, while in a small space it is more concentrated and clear. By inferring the audio scene, more intelligent adjustments can be made for subsequent processing. According to the real-time audio propagation scenario, a spatial sound effect algorithm is used to enhance the stereo field of the audio.This process not only provides a richer stereo effect, but also optimizes the spatial sense of the audio, making the audio more realistic and vivid in hearing. Especially under headphones or surround sound systems, the enhancement of the stereo field sound effect can greatly enhance the auditory immersion. The spatial sense of the audio is greatly enhanced by the enhancement of the stereo field. By simulating the spatial distribution of the audio, the sense of positioning of the sound source and the width of the sound field are enhanced, further enriching the performance level of the audio. During the audio propagation process, delays can cause different parts of the audio signal to be misaligned in time, affecting the accuracy of hearing. Through propagation delay optimization, the audio can be time-aligned to ensure that all audio signals remain synchronized during the propagation process and eliminate audio distortion caused by delay. During the long-term propagation of audio signals, distortion may occur due to equipment or environmental reasons. Through the distortion prediction model, the possible distortion bands can be monitored and identified in real time, and adaptive gain optimization can be performed. Distortion compensation can ensure that the final audio has no obvious artifacts and unnatural sound effects, and improve the accuracy and fidelity of the sound quality.

[0016] In the embodiment of the present invention, refer to Figure 1 , is a flowchart of a method for processing high-quality audio data according to the present invention. In this example, the method for processing high-quality audio data includes: Step S1: Identify the audio signal to be processed; perform multi-time window segmentation processing and adaptive noise filtering on the audio signal to be processed to obtain a noise-filtered optimized audio signal; In this embodiment, a high-quality recording device is used to collect the audio signal to be processed. A suitable sampling rate (such as 44.1 kHz or 48 kHz) is selected to ensure the clarity and details of the audio signal. Ensure that the recording environment is as quiet as possible to reduce the impact of background noise on subsequent processing. The collected audio signal is converted into a format suitable for processing, and lossless audio formats such as WAV or FLAC are usually selected. This step ensures that the audio quality will not be reduced due to compression during the processing process. Use audio analysis software or libraries (such as Librosa or Praat) to identify the audio signal to be processed. The identified content includes the duration, frequency range and audio events (such as speech, music, noise, etc.) of the audio. This step can be achieved by feature extraction of the audio signal, such as calculating features such as Mel frequency cepstral coefficients (MFCC) and zero crossing rate (ZCR). The recognition results are recorded in a database, and a visual chart is generated to show the characteristics of the audio signal. This will provide important information support for subsequent processing steps to ensure that subsequent processing can be carried out in a targeted manner. And overlap rate (such as 50% or 75%). This processing can divide a long audio signal into multiple short-time signals for subsequent analysis and processing. The choice of overlapping windows can be determined experimentally to balance the resolution in the time domain and frequency domain. Apply a short-time Fourier transform to each segmented audio segment and calculate its spectrum. STFT can provide information about the audio signal in time and frequency, making the subsequent noise filtering process more accurate. Select a suitable adaptive noise filtering algorithm, such as an adaptive filter (e.g., LMS algorithm or NLMS algorithm). The adaptive filter can dynamically adjust its parameters according to the characteristics of the input signal to eliminate noise in real time. Set the initial value of the noise estimate, which can be achieved by analyzing the frequency characteristics of the segmented signal. Apply an adaptive noise filter to each short-time audio segment to analyze and remove background noise in real time. Evaluate the filtering effect by comparing the signal-to-noise ratio (SNR) of the signal before and after filtering. A threshold value (e.g., 10 dB) can be set to determine whether the filtering is successful. Resynthesize the filtered short-time audio segment to generate a continuous noise-filtered optimized audio signal. Save the final audio signal in a lossless format (e.g., WAV) to ensure audio quality. At the same time, record the parameters and effects of the filtering for subsequent analysis and optimization.

[0017] Step S2: performing multi-scale spectrum decomposition on the noise filtering optimized audio signal, and capturing transient audio details, and identifying transient audio detail information of each scale; In this embodiment, wavelet transform is used as a method for multi-scale spectrum decomposition. Wavelet transform can have good resolution in both time and frequency domains, and is suitable for capturing transient features of audio signals. Select a suitable wavelet basis (such as Daubechies wavelet) and decomposition level (such as 4 or 5 levels) to ensure that the details of the audio signal are fully captured. The audio signal after noise filtering optimization is normalized to ensure that the signal amplitude is within a reasonable range (such as -1 to 1). This process helps to improve the stability and effectiveness of subsequent spectrum decomposition. Wavelet transform is applied to the processed audio signal. By decomposing layer by layer, the audio signal is decomposed into low-frequency components (approximate coefficients) and high-frequency components (detail coefficients) of different scales. Setting a suitable number of decomposition layers, such as 4 layers, can ensure that the multi-level features of the audio signal are extracted. In each layer of decomposition, the corresponding frequency range and time information are recorded. The low-frequency components of each layer usually contain the main features of the signal, while the high-frequency components contain transient and detail information. By analyzing these components, the changing characteristics of the audio signal at different scales can be understood. Transient audio detail information is extracted from the wavelet transform results of each layer. For each high-frequency detail coefficient, a threshold (such as 0.1) is set to filter out tiny fluctuations and focus on obvious transient changes. The energy detection method can be used to calculate the energy of the detail coefficient to identify the transient part. For the captured transient details, its characteristics are further analyzed, including the duration, amplitude change and frequency characteristics of the transient. The start time, end time and frequency distribution of each transient in the spectrum can be calculated to fully describe the transient audio details. The identified transient audio detail information is recorded in the database, and a visual chart is generated to show the distribution of transient details in time and frequency. This will provide important data support for subsequent audio analysis and processing.

[0018] Step S3: Calculate the audio frequency range according to the transient audio detail information of each scale, and perform dynamic full-band range adjustment to obtain transient information full-band optimized audio; In this embodiment, the transient features of each scale are summarized from the transient audio detail information of each scale extracted in step S2. This includes the frequency range, duration and amplitude change of the transient. By calculating the high-frequency detail coefficient, the frequency range corresponding to each transient signal can be determined. For each transient signal, the main frequency component is calculated using the spectrum analysis method. A suitable frequency resolution (for example, 0.1 kHz) is set to ensure that the frequency range of each transient can be accurately identified. The upper and lower limits of the frequency range are determined by calculating the energy distribution of its spectrum. According to the transient audio detail information obtained by the analysis, the parameters for dynamic full-band range adjustment are set. These parameters include gain adjustment, frequency band enhancement and attenuation ratio. The dynamic gain control (DGC) method can be used to adjust the gain value in real time according to the characteristics of the transient signal. The audio signal is divided into multiple frequency bands (such as low frequency: 20-250 Hz, medium frequency: 250-2000 Hz, high frequency: 2000-20000 Hz) to ensure that the processing of each frequency band can be carried out in a targeted manner. For example, for the frequency range of transient information, a corresponding gain adjustment strategy is formulated to improve the audio signal quality of these frequency bands. In audio signal processing, the set gain adjustment strategy is applied in real time. By dynamically adjusting the gain of each frequency band, the performance of transient audio signals in each frequency band is optimized. For example, for rapidly changing transient signals, you can consider increasing the gain of the high frequency band to enhance clarity. Use a bandpass filter to process the transient signal to ensure that only the audio information within a specific frequency range is retained. Set an appropriate cutoff frequency to ensure that the main components of the transient signal are not weakened. For low and high frequency bands, appropriate attenuation can be applied to avoid unnecessary noise interference. Resynthesize the dynamically adjusted signals of each frequency band to generate a full-band optimized audio signal of transient information. During the synthesis process, ensure that the audio signals of different frequency bands are balanced to avoid a certain frequency band being too prominent or masked. Evaluate the quality of the generated full-band optimized audio signal of transient information. The quality of the audio signal can be objectively evaluated by calculating indicators such as signal-to-noise ratio (SNR) and spectrum uniformity. Set reasonable evaluation indicators (such as at least 5 dB increase in SNR) to judge the optimization effect. Generate the spectrum and waveform of the audio signal before and after optimization, and analyze the optimization effect by comparison. The visualization chart should clearly show the changes in audio characteristics of each frequency band, so as to intuitively understand the results of the optimization process.

[0019] Step S4: mining the environmental acoustic effect of the full-band optimized audio of transient information, and inferring the current audio scene to obtain a real-time audio propagation scene; In this embodiment, features are extracted from the full-band optimized audio of transient information to analyze the environmental acoustic effects. Common features include spectral features (such as Mel frequency cepstral coefficients (MFCC), short-time energy, frequency distribution, etc.) and time domain features (such as zero crossing rate and transient amplitude). This process can be implemented using signal processing tools (such as Librosa). Select a suitable environmental acoustic model, such as a geometric acoustic model or a statistical acoustic model. The geometric acoustic model uses the propagation path and reflection characteristics of sound waves to describe the propagation of sound in the environment, while the statistical acoustic model simulates the acoustic effect through statistical characteristics. Choosing a suitable model will affect the accuracy of subsequent analysis. The extracted audio signal features are analyzed to calculate acoustic characteristics such as reverberation time (RT60), clarity (C80) and speech transmission index (STI). The reverberation time can be estimated by analyzing the attenuation of the signal, and the clarity is calculated by the distribution of signal energy in different frequency bands. By analyzing the attenuation and reflection characteristics of the audio signal at different frequencies, the geometry and material of the environment are evaluated. You can use sound field simulation software (such as EASE or Odeon) to model the acoustic environment more accurately and simulate the reflection and diffraction characteristics of sound waves on the surfaces of different materials (such as concrete, wood, etc.). Based on the calculated acoustic characteristics, mark the key acoustic effect locations in the environment. These locations may be reflection surfaces, sound source locations, and receiving points. Use visualization tools to display the distribution of acoustic effects to help understand the propagation process of audio signals in the environment. Based on the extracted acoustic characteristics and environmental acoustic effects, build a real-time audio propagation scenario model. The model should include information such as the layout of sound sources and receiving points, environmental characteristics, and sound wave propagation paths. You can use 3D modeling software (such as SketchUp) to generate an environmental model to help intuitively display the audio propagation scenario. Through sound field simulation software, combine the real-time optimized audio signal with the established acoustic environment model to perform dynamic acoustic simulation. During the simulation process, update the locations of the sound source and receiving points in real time to observe the propagation effect of the audio signal under different conditions. This process can reveal the specific impact of the environment on the audio signal. Based on the simulation results, evaluate the performance of the real-time audio propagation scenario. If it is found that the audio quality degrades under certain conditions, the sound source position or environmental parameters (for example, adding sound-absorbing materials) can be adjusted to optimize the audio propagation effect. Ensure the dynamics and adaptability of the audio propagation scenario model to cope with different audio scenarios. Record the analyzed environmental acoustic characteristics, acoustic effects, and real-time audio propagation scenario models in the database for subsequent query and analysis. At the same time, save the audio signals and sound field data generated during the simulation process to provide a basis for subsequent research. Generate visual charts and 3D models to display environmental acoustic effects and audio propagation scenarios. The changes in acoustic characteristics are intuitively displayed through the waveform, spectrum, and sound field distribution diagram of the audio signal. This process can help understand the propagation characteristics of audio in complex environments and provide visualization support for subsequent audio processing and applications.

[0020] Step S5: performing stereo field sound effect enhancement on the full-band optimized audio of transient information according to the real-time audio propagation scenario, thereby generating a stereo field enhanced audio signal; In this embodiment, the real-time audio propagation scene is analyzed to understand the location of its sound source and the layout of the receiving points. By analyzing the distance, angle and reflection characteristics between the sound source and the listener, the parameters that can enhance the stereo field are determined. This step can be achieved through sound field simulation software to ensure a comprehensive understanding of the environment. According to the characteristics of the audio content (such as music, dialogue or ambient sound), the goal of stereo field enhancement is set. For example, for music content, the spatial sense and clarity of the stereo field are enhanced; while for dialogue, the clarity and localization of the voice are emphasized. Specific target parameters are set, such as a 10% increase in the spatial sense of sound and a 5 dB increase in voice clarity. The full-band optimized audio of transient information is preprocessed to ensure that the amplitude of the signal is within a reasonable range (such as -1 to 1) to avoid distortion introduced during subsequent processing. Dynamic range compression (DRC) technology can be used to balance the dynamic range of the audio signal to ensure that each part of the signal is balanced. Based on the information of the real-time audio propagation scene, virtual sound source localization technology is used to distribute the audio signal to different positions in the stereo field. A more natural sense of space can be achieved by setting the direction and distance of the virtual sound source (for example, the gain difference between the left and right channels). The HRTF (Head-Related Transfer Function) model can be used to simulate the human ear's perception of the direction of sound. Apply reverberation and delay effects to enhance the experience of the stereo field. By setting appropriate reverberation time (such as 1 second) and pre-delay (such as 20 ms), the sound is more layered in the space. Select the appropriate reverberation type (such as room reverberation or hall reverberation) and adjust it according to the characteristics of the audio content. During the processing, the quality and effect of the audio signal are monitored in real time. The enhancement effect is evaluated by calculating the signal-to-noise ratio (SNR) and spectral characteristics. If the audio signal of a certain frequency band is found to be too prominent or distorted, the gain and effect parameters can be adjusted dynamically to ensure the naturalness and clarity of the audio. The audio signals after virtual sound source localization and sound field effect enhancement are synthesized to generate the final stereo field enhanced audio signal. During the synthesis process, ensure that the audio signals of the left and right channels are balanced to avoid uneven distribution of audio signals. Perform final processing on the synthesized audio signal, such as normalization and compression, to ensure that the amplitude of the output signal is suitable for the playback device. You can set the target volume of the output signal (such as -3 dBFS) to ensure that there is no distortion during playback.

[0021] Step S6: performing propagation delay optimization on the stereo field enhanced audio signal and performing global distortion prediction optimization to generate global adaptive distortion optimized audio.

[0022] In this embodiment, propagation delay analysis is performed on the stereo field enhanced audio signal. In this process, the propagation delay of the signal at different frequencies is calculated to understand the time required for the sound wave to propagate in the environment. The cross-correlation function can be used to measure the delay between the signals and determine the time difference between the left and right channels of the audio signal. According to the analysis results, a suitable delay parameter is set. For the stereo field, it is usually desired to control the delay within 20 milliseconds to ensure the naturalness and spatial sense of the audio. A reasonable delay range (such as 0-20 ms) is set so that it can be optimized in subsequent processing. For each channel, the audio signal is delayed and compensated according to the set delay parameters. This process can be implemented using a digital delay line to ensure that the signal is synchronized in the left and right channels. By adjusting the time offset of each channel, the phase relationship of the audio signal is kept consistent to avoid phase interference. During the delay compensation process, the phase and amplitude changes of the signal are monitored in real time. The effect of delay compensation is evaluated by calculating the phase difference and signal energy. If it is found that the delay of a certain frequency band is too large, the delay parameter can be dynamically adjusted to ensure the clarity and spatial sense of the signal. Select a suitable global distortion prediction model. Common methods include regression models based on machine learning (such as support vector regression SVR) or deep learning models (such as LSTM networks). These models can analyze distortion features and predict distortion levels, providing a basis for subsequent optimization. Extract distortion features from stereo field enhanced audio signals, such as audio spectrum features, dynamic range, transient features, etc. Prepare a training dataset for the model, including audio samples with known distortion levels, to ensure the diversity and representativeness of feature vectors. Use the labeled training dataset to train the selected distortion prediction model. During the training process, use the cross-validation method to evaluate the accuracy of the model and optimize the model parameters. Ensure that the model has good generalization ability when predicting distortion levels. Input the stereo field enhanced audio signal into the trained distortion prediction model to generate audio distortion prediction data. According to the prediction results, set the gain and filter parameters of the distortion compensation to achieve global adaptive distortion optimization. Set a reasonable target distortion level (such as reducing the total harmonic distortion THD to less than 1%). Synthesize the audio signals after distortion optimization to generate a global adaptive distortion optimized audio signal. During the synthesis process, ensure that the various frequency bands of the signal are balanced to avoid any one frequency band being too prominent or distorted. You can use the Dynamic Equalizer to fine-tune the frequency bands.

[0023] In this embodiment, refer to Figure 2 , is a flowchart of detailed implementation steps of step S1. In this embodiment, the detailed implementation steps of step S1 include: Step S11: Identify the audio signal to be processed; perform multi-time window segmentation processing on the audio signal to be processed to obtain audio signals of multiple time windows; Step S12: performing audio noise recognition on the audio signals of multiple time windows segment by segment, marking the audio noise points of each time window; Step S13: Calculate the power spectrum density of the audio noise point in each time window, and extract the power spectrum density of each noise point; Step S14: classifying noise according to the power spectrum density of each noise point, thereby generating a type feature of each noise point; Step S15: Adaptively perform noise filtering based on the type characteristics of each noise point to obtain a noise-filtered optimized audio signal.

[0024] In this embodiment, suitable audio signal processing tools and libraries (such as Librosa, PyDub, etc.) are selected to identify the audio signal to be processed. Load the audio file, read its sampling rate, duration and channel information for subsequent processing. Divide the audio signal into multiple time windows. Set the duration of each window (for example, 100 milliseconds) and select an appropriate overlap rate (such as 50%) to ensure continuity and accuracy of analysis. The window function in the short-time Fourier transform (STFT) can be used to achieve smoothing. By traversing the sample data of the audio signal, the audio signal of each time window is extracted according to the set window size and overlap rate. Each window is stored to make it the basis for subsequent noise identification and analysis. The multiple time windows obtained by segmentation are recorded in a data structure (such as a list or array) for subsequent processing. Each time window will contain the corresponding audio signal data and its timestamp information for subsequent noise identification. Select a suitable audio noise recognition algorithm, and usually a machine learning-based classifier (such as a support vector machine SVM, a random forest, or a deep learning model such as a CNN) can be used to identify noise points. The audio signal of each time window is subjected to feature extraction, usually Mel-frequency cepstral coefficients (MFCC), audio energy, zero-crossing rate and other features. These features can effectively reflect the characteristics of the audio signal and help with subsequent noise identification. The extracted features are input into the selected noise identification model for inference. The model will output whether there is noise in each time window and the specific location of the noise (such as the start and end time). According to the output of the model, the audio noise points of each time window are marked and the information is recorded in the database. Each noise point will contain its position in the time window and its corresponding noise features for subsequent analysis. Select a suitable power spectral density calculation method, usually using the Welch method or Fourier transform (FFT) to calculate the power spectral density. The Welch method is usually more popular because of its better smoothing effect. The audio noise points of each time window are processed to extract the audio signal data in the time window. Make sure that the length of the data is sufficient for effective spectrum analysis and perform zero padding when necessary. Using the Welch method, the audio signal of each noise point is processed by window function and the power spectral density is calculated in segments. According to the set window width (such as 256 points) and overlap rate (such as 50%), the power spectrum of each window is calculated and the average is taken to obtain a smooth power spectrum density. The power spectrum density calculated for each noise point is recorded in the database, and a visualization chart is generated to show the spectrum characteristics of each noise point. Through visualization, the frequency component of each noise point can be intuitively analyzed to provide a basis for subsequent classification. Select a suitable noise classification model, usually using a machine learning-based classifier (such as KNN, decision tree, or deep learning model). Choose to classify based on the characteristics of the noise (such as power spectrum density).Normalize the power spectral density features of each noise point to ensure that the feature values ​​are in a similar range to improve the performance of the classifier. You can use Z-score normalization or Min-Max normalization. Input the standardized power spectral density features into the selected noise classification model for inference. The model will output the category of each noise point, such as "machine sound", "human voice", "environmental noise", etc. Record the classification results of each noise point in the database and perform statistical analysis. Analyze the distribution of different noise types for subsequent processing and optimization of noise filtering strategies. Select a suitable adaptive noise filter, usually using a minimum mean square error (LMS) filter or an adaptive filter array, and filter according to the real-time noise characteristics. Set different filtering parameters according to the type characteristics of each noise point. For example, for high-frequency noise (such as machine sound), you can choose a high-pass filter to reduce low-frequency components; for low-frequency noise (such as environmental noise), you can choose a low-pass filter. Apply the corresponding adaptive filter to the audio signal in each time window. According to the noise type characteristics, adjust the filter parameters in real time to optimize the filtering effect. Record the audio signal after noise filtering in the database, and generate a visual chart to show the audio waveform and spectrum before and after filtering to evaluate the filtering effect. Through comparison, the effectiveness of filtering can be analyzed intuitively, providing a basis for subsequent processing.

[0025] In this embodiment, refer to Figure 3 , is a flowchart of detailed implementation steps of step S2. In this embodiment, the detailed implementation steps of step S2 include: Step S21: performing multi-scale spectrum decomposition on the noise filtering optimized audio signal to obtain audio spectrum data of different scales; Step S22: Calculating the frequency component distribution of the audio spectrum data at different scales to generate the audio frequency component distribution characteristics of each scale; Step S23: dynamically adjusting the time-frequency resolution according to the distribution characteristics of the audio frequency components at each scale to generate the time-frequency resolution at each scale; Step S24: capturing transient audio details of audio spectrum data of different scales based on the time-frequency resolution of each scale, and identifying transient audio detail information of each scale.

[0026] In this embodiment, a suitable multi-scale spectrum decomposition method is selected, and commonly used methods include wavelet transform (WaveletTransform) and short-time Fourier transform (STFT). Wavelet transform is usually more suitable for processing non-stationary signals due to its localized characteristics in time and frequency. Determine the basis function (such as Daubechies wavelet, Morlet wavelet, etc.) used for wavelet transform. The selection of basis function will affect the decomposition effect, and is usually selected according to the signal characteristics. For example, the use of Daubechies wavelet is suitable for processing signals with high-frequency components. The optimized audio signal is subjected to wavelet transform, and the number of decomposition layers (such as 5 layers) is set to obtain audio spectrum data of different scales. Each layer will correspond to a different frequency range, and can effectively capture multiple frequency components of the signal. The audio spectrum data of each scale is recorded in a database, and a visual chart is generated to show the spectrum characteristics of different scales. Through visualization, the changes of the signal in different frequency ranges can be intuitively observed for subsequent analysis. Select a suitable frequency component distribution calculation method, and a spectrum statistical analysis method can usually be used, such as calculating the spectrum energy distribution or spectrum density of each scale. The audio spectrum data of each scale is analyzed to calculate the energy value of each frequency component. E(f)= , where E(f) is the energy value of frequency f, is the value of the spectrum at time t and frequency f. Based on the calculated spectrum energy, the frequency component distribution characteristics of each scale are generated. Normalization can be performed to make the total energy of the frequency component of each scale 1 for comparison. The frequency component distribution characteristics of each scale are recorded in the database, and a bar chart or heat map is generated using a visualization tool to show the frequency distribution of different scales. This helps to quantitatively analyze the performance of each frequency component at different scales. Select a suitable dynamic time-frequency resolution adjustment model, which is usually adaptively adjusted based on signal characteristics (such as transient characteristics, frequency components, etc.). An adaptive window function or a dynamic window width strategy can be used. Set the time-frequency resolution adjustment parameters according to the frequency component distribution characteristics of each scale. For example, for low-frequency signals, a wider window width is set to improve the time resolution; for high-frequency signals, a narrower window width is set to improve the frequency resolution. The audio spectrum data of each scale is dynamically adjusted according to the set resolution adjustment parameters. The parameters of short-time Fourier transform or wavelet transform can be adjusted to optimize the time-frequency resolution. The adjusted time-frequency resolution is recorded in the database to evaluate the time-frequency resolution effect at different scales. By comparing the resolutions of different scales, the performance of the signal in the time-frequency domain can be analyzed to provide a basis for subsequent processing. Select a suitable transient audio detail capture model. Usually, transient detection algorithms (such as energy threshold detection, transient analysis method, etc.) are used to identify transient features in audio signals. Transient audio detail capture is performed on the audio spectrum data of each scale. According to the adjusted time-frequency resolution, the signal is analyzed to identify the parts with significant energy changes in a short period of time, which usually contain transient features. Transient features are extracted from the audio signal of each scale, such as transient start time, duration, and amplitude. By analyzing these features, important details in the audio signal can be identified, such as percussion instrument sounds, instantaneous sound effects, etc. The extracted transient audio detail information is recorded in the database, and the distribution of transient features is displayed through visualization tools. This will help with subsequent analysis and processing and provide a basis for the enhancement of audio signals.

[0027] In this embodiment, refer to Figure 4 , is a flowchart of detailed implementation steps of step S3. In this embodiment, the detailed implementation steps of step S3 include: Step S31: performing transient detail audio fineness quantification analysis on the transient audio detail information of each scale to generate an audio detail fineness quantification value; Step S32: evaluating the clarity of each scale of transient audio information to obtain the clarity of each scale; Step S33: performing a quantitative analysis of the auditory effect according to the information clarity and audio detail fineness value of each scale, thereby generating a transient detail auditory effect of each scale; Step S34: Calculate the audio frequency range of the audio spectrum data of different scales, and extract the audio frequency range of each scale; Step S35: dynamically adjusting the audio frequency range of each scale in the full frequency range based on the transient detail auditory effect of each scale to obtain transient information full frequency band optimized audio.

[0028] In this embodiment, a suitable transient detail exquisiteness quantitative analysis method is selected, usually using a combination of psychoacoustic models and subjective evaluation. Some relevant parameters can be set for quantification with reference to the study of auditory perception. Relevant features are extracted from the transient audio detail information of each scale, such as the clarity, duration, richness of frequency components and attack time of the transient. These features can usually reflect the exquisiteness of audio details. According to the extracted features, these features are converted into exquisiteness quantitative values ​​using a weighted average method or a linear regression model. Set weight parameters (for example, transient clarity weight 0.5, frequency richness weight 0.3, attack time weight 0.2), and calculate the exquisiteness quantitative value of each scale. The calculated audio detail exquisiteness quantitative value is recorded in a database and statistically analyzed. By comparing the exquisiteness values ​​of different scales, the exquisiteness of audio at different scales can be evaluated, providing a basis for subsequent processing. Select a suitable audio clarity evaluation method, commonly used including signal-to-noise ratio (SNR), the ratio of short-time energy to short-time average energy, etc. These indicators can reflect the clarity of the audio signal. Analyze the transient audio signal at each scale and extract its short-time energy and noise components. The energy spectrum of each transient can be calculated using short-time Fourier transform, and the background noise level can be determined. Based on the extracted short-time energy and background noise level, calculate the clarity value of each scale. SNR=10log10 (signal energy / noise energy), record the clarity value of each scale in the database, and generate a visual chart to show the clarity distribution of different scales. This helps to intuitively evaluate the performance of audio at different scales. Select a suitable auditory effect analysis model, usually using a multivariate linear regression model, and use clarity and delicate perception values ​​as input to predict the auditory effect. Arrange the audio clarity value and delicate perception value of each scale to build a data set containing multidimensional features to ensure the integrity and accuracy of the data set. Input the sorted data into the selected analysis model for training and prediction. Through the model, calculate the transient detail auditory effect value of each scale. An output range (such as 0 to 100) can be set to facilitate the quantification of the auditory effect. Record the transient detail auditory effect value in the database and perform statistical analysis. By comparing the auditory effect values ​​at different scales, the overall performance of the audio signal can be evaluated to provide a basis for subsequent optimization. Select a suitable frequency range calculation method. Usually, the frequency range can be extracted by calculating the frequency boundaries (lowest frequency and highest frequency) of the spectrum data. Analyze the audio spectrum data of each scale and extract its frequency distribution. The maximum and minimum values ​​of the frequency components of each scale can be calculated to determine the frequency range. Based on the spectrum data, calculate the frequency range of each scale. For example, a threshold (such as a -3 dB point) can be set to determine the effective frequency range in order to exclude unimportant frequency components. Record the frequency range of each scale in the database and generate a visual chart to show the frequency ranges of different scales.Through visualization, the frequency characteristics of audio signals at each scale can be observed intuitively, providing a basis for subsequent processing. Select a suitable dynamic full-band adjustment model, usually using a parametric equalizer or a dynamic range compressor to adjust the gain of the audio signal in different frequency ranges. Set the gain parameters of the frequency band according to the transient detail auditory effect value of each scale. For example, for frequency bands with better effects, the gain can be increased; and for frequency bands with poor effects, the gain can be reduced. Perform dynamic full-band range adjustment on the audio spectrum data of each scale. Use digital signal processing (DSP) technology to adjust the gain of each frequency range according to the set parameters to optimize transient information. The adjusted audio signal is recorded in the database, and a visual chart is generated to show the audio waveform and spectrum characteristics before and after the adjustment. By comparison, the adjustment effect can be intuitively analyzed, providing a basis for subsequent optimization.

[0029] In this embodiment, step S4 includes the following steps: Step S41: identifying the ambient sound source of the full-band optimized audio of transient information and separating the features of the ambient sound source; Step S42: Analyze the size of the ambient space based on the ambient sound source characteristics to generate the size of the ambient space; Step S43: estimating the reflection characteristics of the ambient sound source based on the ambient sound source characteristics, thereby estimating the ambient reflection material; Step S44: performing sound reflection path evolution on the ambient sound source characteristics to generate an ambient sound reflection path; Step S45: mining the environmental acoustic effect according to the size of the environmental space, the environmental reflection material and the environmental sound reflection path, thereby generating environmental acoustic effect parameters; Step S46: Infer the current audio scene based on the environmental acoustic effect parameters to obtain a real-time audio propagation scene.

[0030] In this embodiment, a suitable sound source recognition algorithm is selected, and a model based on deep learning (such as a convolutional neural network CNN or a recurrent neural network RNN) is usually used. These models can effectively process the time domain and frequency domain features of the audio signal to perform sound source recognition. Features such as Mel frequency cepstral coefficients (MFCC), short-time Fourier transform (STFT) features, and spectrograms are extracted from the full-band optimized audio signal. By extracting these features, the model's ability to recognize different sound sources can be enhanced. The sound source recognition model is trained using a prepared audio data set (including labeled environmental sound sources). During the training process, cross-validation technology is used to optimize the model parameters. After the training is completed, the optimized audio signal is input into the model for reasoning, the environmental sound source is identified, and its features are extracted. The identified environmental sound source features are recorded in a database, and a visual chart is generated to display the spectral features of different sound sources. This process facilitates subsequent sound source analysis and processing to ensure the effective separation of environmental sound sources. Select a suitable space size analysis method, and an estimation method based on acoustic measurement, such as a room acoustic model or a reflection model, can usually be used. From the environmental sound source features extracted in step S41, its frequency distribution and time delay characteristics are analyzed. By analyzing the reflected sound, the distance between the sound source and the reflecting surface can be inferred. Using the known speed of sound (such as 343 m / s) and the time delay between the sound source and the reflecting surface, the size of the environmental space is calculated. According to the reflection characteristics of different sound sources, the overall size of the environmental space is calculated comprehensively. The calculated size of the environmental space is recorded in the database, and a visual chart is generated to show the spatial structure. This will help in the subsequent analysis of acoustic effects and ensure the accuracy of the environmental model. Select a suitable method for estimating the reflection characteristics of the environmental sound source. Usually, the acoustic characteristic analysis method (such as impulse response measurement) is used to obtain the reflection characteristics. Analyze the reflection components in the characteristics of the environmental sound source, such as the intensity, delay and frequency response of the reverberation. These characteristics can reflect the properties of the environmental surface, such as hardness, smoothness, etc. According to the reflection characteristics of the sound source, combined with the existing acoustic database (such as the acoustic characteristics table of different materials), the reflective material of the environment is inferred. For example, if the reflection intensity is high and clear, it may be a hard material (such as concrete, glass); if the reflection is blurred, it may be a soft material (such as carpet, wall). The inferred environmental reflection material is recorded in the database to generate statistical data on the environmental reflection characteristics. This will help in the subsequent acoustic effect analysis and ensure the accuracy of the reflective material. Select a suitable sound reflection path model, usually using a method based on geometric acoustics or wave acoustics to simulate the propagation and reflection path of sound in the environment. The sound source and reflective material features extracted from step S41 and step S43 provide basic data for the simulation of the sound reflection path. Ensure the accuracy of the data to improve the reliability of the path estimation. Use the acoustic model to simulate the sound reflection path and calculate the path of the sound wave from the sound source to the receiving point.According to the size of the environmental space and the reflective material, the specific path of the sound wave reflection is derived, and the angle and distance of each reflection are recorded. The generated environmental sound reflection path is recorded in the database, and a visual chart is generated to show the path of sound wave propagation. This will help the subsequent analysis of acoustic effects and provide a basis for model optimization. Select a suitable acoustic effect mining model. Usually, the reverberation time (RT60) calculation model, sound energy distribution model, etc. can be used to evaluate the characteristics of the acoustic environment. Comprehensive information such as the size of the environmental space, reflective materials and reflection paths is used to construct a data set containing multidimensional features. These features will be used to calculate acoustic effect parameters. According to the selected acoustic effect model, calculate the environmental acoustic effect parameters, such as reverberation time, clarity, sound pressure level, etc. The reverberation time can be calculated using the following formula: RT60=A / (0.161×V), where V is the room volume and A is the sum of the acoustic absorption coefficients. The calculated acoustic effect parameters are recorded in the database and statistically analyzed. By comparing the effect parameters under different environmental conditions, the propagation characteristics of the audio signal are evaluated, providing a basis for subsequent audio scene inference. Select a suitable audio scene inference model, usually using a Bayesian inference or machine learning method combined with acoustic effect parameters for inference. Integrate the environmental acoustic effect parameters obtained in step S45 into the inference model. Ensure the integrity and accuracy of the data to improve the reliability of scene inference. Provide input data to the inference model to infer the real-time audio propagation scene. The model will infer the characteristics of the current audio scene, such as indoors, outdoors, open spaces or closed spaces, based on the input acoustic features and environmental information. Record the inferred audio propagation scene in the database, and generate a visual chart to display the inference results. This will help with subsequent audio processing and optimization to ensure that the propagation effect of the audio signal meets the requirements.

[0031] In this embodiment, the specific steps of step S5 are: Step S51: performing audio space propagation evolution on environmental acoustic effect parameters to generate a spatial propagation acoustic effect evolution law; Step S52: performing spatial propagation simulation processing on the real-time audio propagation scene according to the evolution law of spatial propagation acoustic effect, and collecting audio spatial propagation simulation data; Step S53: performing iterative learning of deep stereo field features on the audio space propagation simulation data to generate deep stereo field features; Step S54: performing stereo field sound effect enhancement on the transient information full-band optimized audio according to the deep stereo field characteristics, thereby generating a stereo field enhanced audio signal.

[0032] In this embodiment, a suitable spatial propagation model is selected, and a method based on geometric acoustics or wave acoustic simulation is usually used. These models can accurately simulate the characteristics of sound propagation in different environments. The acoustic effect parameters (such as reverberation time, clarity, sound pressure level, etc.) extracted from step S45 are integrated into the spatial propagation model. These parameters will be used as input to affect the propagation characteristics of the sound. The selected model is used to simulate the evolution of the spatial propagation acoustic effect. According to the environmental characteristics (such as space size, reflective material, sound source position, etc.), the propagation path and change law of the sound in the space are calculated. Through multiple simulations, the changes of acoustic effects over time and space are observed, and the evolution laws of key parameters are recorded. The generated spatial propagation acoustic effect evolution law is recorded in the database, and a visual chart is generated to show the changes in acoustic effects under different environmental conditions. This will provide an important reference for subsequent audio processing. Select a suitable audio space propagation simulation processing method, usually using acoustic simulation software (such as EASE, Odeon, etc.) for more accurate simulation. Collect the current real-time audio signal, and set the simulation parameters according to the acoustic effect evolution law in step S51. These parameters include the location of the sound source, the location of the receiving point, the environmental reflection characteristics, etc. The real-time audio signal is input into the acoustic simulation software to perform spatial propagation simulation processing. The software will generate audio propagation simulation data based on the set environmental parameters and acoustic effects, and record the behavior of sound waves in space. The collected audio spatial propagation simulation data is recorded in the database, and a visual chart is generated to show the sound wave propagation process. This helps to analyze and optimize the audio effect in the future and ensure the accuracy of the simulation results. Select a suitable deep learning model, usually using a convolutional neural network (CNN) or a recurrent neural network (RNN) to process the feature learning of the audio signal. The audio spatial propagation simulation data extracted from step S52 is organized into a data set suitable for the input format of the deep learning model. Ensure that the data set contains rich samples to enhance the generalization ability of the model. Train the selected deep learning model to extract the stereo field features in the audio spatial propagation data through iterative learning. Use data enhancement techniques (such as time expansion, frequency perturbation, etc.) to improve the robustness of the model. Record the generated deep stereo field features in the database and evaluate the learning effect of the model. The learned features can be displayed by visualization means to analyze their potential effects in audio enhancement. Select appropriate sound enhancement technology, which usually includes sound processing methods based on deep learning (such as GAN generative adversarial network) or traditional spatial sound processing technology (such as HRTF). According to the deep stereo field features extracted in step S53, set the parameters for sound enhancement. This includes gain adjustment, frequency modulation, phase difference adjustment, etc. to improve the performance of the stereo field. Combine the full-band optimized audio signal of transient information with the set enhancement parameters to perform sound enhancement processing. By adjusting the spatial properties of the audio signal, an audio signal with a richer sense of stereo field is generated.The generated stereo field enhanced audio signal is recorded in the database, and a comparative visualization chart is generated to show the audio waveform and spectrum characteristics before and after the enhancement. Through comparison, the enhancement effect can be analyzed intuitively, providing a basis for subsequent optimization.

[0033] In this embodiment, the specific steps of step S6 are: Step S61: performing real-time propagation parameter monitoring on the audio signal to be processed to generate audio real-time propagation monitoring parameters; Step S62: Calculate the propagation path delay of the audio real-time propagation monitoring parameters and extract the propagation path delay parameters; Step S63: performing audio delay compensation calculation on the propagation path delay parameter, thereby obtaining an audio delay compensation parameter; Step S64: optimizing the propagation delay of the stereo field enhancement audio signal according to the audio delay compensation parameter, thereby generating a delay optimized audio signal; Step S65: performing global distortion prediction optimization on the delay-optimized audio signal to generate a global adaptive distortion-optimized audio.

[0034] In this embodiment, the audio propagation parameters to be monitored are determined, including sound pressure level (SPL), spectrum characteristics, delay, phase difference, and frequency response. These parameters will help to comprehensively evaluate the propagation characteristics of the audio signal. Configure the audio signal processing module and use short-time Fourier transform (STFT) or fast Fourier transform (FFT) to analyze the real-time audio signal. The length and overlap rate of each data window need to be selected appropriately (for example, the window length is 1024 points and the overlap rate is 50%) to ensure the balance between frequency resolution and time resolution. The monitored real-time propagation parameters are recorded in the database, and real-time visualization charts are generated to facilitate monitoring and analysis of changes in audio signals. This process will provide important data support for subsequent steps to ensure that the propagation characteristics of the audio signal are effectively evaluated. Select a suitable delay calculation model, commonly used including phase difference method and delay estimation method. These methods can effectively analyze the propagation delay of audio signals on different paths. From the audio real-time propagation monitoring parameters obtained in step S61, extract the corresponding timestamp and sound pressure level data. Ensure the integrity of the data for accurate delay calculation. Use the selected delay calculation method to analyze the delay in the real-time monitoring parameters. For example, if the phase difference method is used, the phase difference of the same sound source signal received by different microphone arrays can be compared to calculate the delay of the propagation path. Delay = Δϕ / 2πf, where Δϕ is the phase difference and f is the signal frequency. The calculated propagation path delay parameters are recorded in the database, and a visual chart is generated to show the delay characteristics of different sound sources. This will provide basic data for subsequent delay compensation calculations to ensure the accuracy of delay compensation. Select a suitable delay compensation model, which can usually be compensated based on linear interpolation or adaptive filters. These models can dynamically adjust the playback time of the audio signal according to the delay parameters. According to the propagation path delay parameters extracted in step S62, set the parameters of audio delay compensation. For example, a benchmark delay (such as 10ms) can be set and adjusted according to the actual measured delay. Apply the selected compensation model to delay the real-time audio signal. Ensure the synchronization of the signal by moving the playback time of the audio signal forward or backward. For example, for a signal with a measured delay of 15ms, its playback time is postponed by 15ms. The calculated audio delay compensation parameters are recorded in the database, and a visual chart is generated to show the comparison of the audio waveforms before and after the delay compensation. This process will ensure that the playback quality of the audio signal is optimized and provide support for subsequent steps. Select a suitable propagation delay optimization model, usually using a delay filter based on digital signal processing (DSP) technology. This method can effectively handle the propagation delay of the audio signal. According to the audio delay compensation parameters obtained in step S63, set the parameters for propagation delay optimization. This includes delay time, gain adjustment, etc. to ensure the synchronization of the audio signal. Input the stereo field enhanced audio signal into the optimization model and perform propagation delay optimization processing.By adjusting the time parameters of the audio signal, the best propagation effect of the signal can be achieved. The audio signal after delay optimization is recorded in the database, and a visual chart is generated to show the audio waveform and spectral characteristics before and after optimization. This will help analyze the effect of delay optimization and provide a basis for subsequent processing. Select a suitable global distortion prediction optimization model, usually using a regression model based on machine learning or a deep learning model (such as LSTM or CNN) for distortion prediction. Extract feature data such as spectral features, time domain features and distortion metrics from the delay-optimized audio signal obtained in step S64. Ensure the diversity and richness of the data to improve the prediction accuracy of the model. Train the selected distortion prediction model and input the delay-optimized audio signal into the model for prediction. The model will output the best distortion compensation parameters based on the input features to optimize the audio quality. The audio signal after global adaptive distortion optimization is recorded in the database, and a visual chart is generated to show the audio waveform and spectral characteristics before and after distortion optimization. By comparison, the effect of distortion optimization can be intuitively analyzed, providing a basis for subsequent application and analysis.

[0035] In this embodiment, the specific steps of step S65 are: Perform full-band audio frequency mutation detection on the delay-optimized audio signal and mark the frequency mutation nodes; Perform mutation point distribution analysis on frequency mutation nodes to generate audio mutation point distribution features; Predict audio distortion based on the distribution characteristics of audio mutation points and generate audio distortion prediction data; The delay-optimized audio signal is globally adaptively distorted according to the audio distortion prediction data, thereby generating globally adaptive distortion-optimized audio.

[0036] In this embodiment, a suitable frequency mutation detection algorithm is selected, and commonly used methods include spectrum analysis based on short-time Fourier transform (STFT) and mutation detection algorithms based on energy changes. These methods can effectively identify points where the frequency changes significantly in a short time. The audio signal after delay optimization is preprocessed, including normalization and denoising, to improve the accuracy of mutation detection. A low-pass filter is used to remove high-frequency noise to ensure the smoothness of the signal. The spectrum of the processed audio signal is calculated using STFT, and a suitable window size (such as 1024 points) and overlap rate (such as 50%) are set to ensure a balance between frequency resolution and time resolution. The spectrum of each time frame will be saved and used for subsequent analysis. By analyzing the rate of change of the spectrum (such as mutation of spectrum energy or frequency components), the frequency mutation node is identified. A threshold (such as a change rate exceeding a certain percentage) is set to mark the mutation point and record its specific position (timestamp) in the audio signal. The marked frequency mutation node is recorded in the database, and a visual chart is generated to show the spectrum of the audio signal and the position of the mutation point. This process will provide basic data for subsequent analysis and ensure the reliability of mutation detection. From the frequency mutation node data extracted in step S66, calculate the distribution characteristics of the mutation points. Common characteristics include the number of mutation points, the time interval between mutation points, the amplitude change of mutation, etc. Perform statistical analysis on the mutation points to generate a histogram or distribution diagram to show the distribution of mutation points in time. The concentration and frequency distribution of mutation points can be analyzed to identify the regularity of mutation points. Set the analysis parameters, such as the minimum amplitude change of mutation points (such as 5 dB) and the time window (such as 100 ms) to filter out small mutations and ensure that important frequency changes are focused. Record the generated audio mutation point distribution characteristics in the database and generate a visual chart to show the statistical distribution of mutation points. This will help understand the changing characteristics of audio signals and provide data support for subsequent distortion prediction. Select a suitable audio distortion prediction model, usually using a machine learning-based regression model (such as support vector regression SVR) or a deep learning model (such as LSTM network) to predict audio distortion values. From the mutation point distribution characteristics extracted in step S67, construct a feature vector, including the number of mutation points, mutation time interval, amplitude change, etc. Ensure the diversity and representativeness of the feature vector. Use the labeled training data set (including audio samples with known distortion levels) to train the selected distortion prediction model. After training, input the audio mutation point distribution features into the model to generate audio distortion prediction data. Record the generated audio distortion prediction data in the database and perform statistical analysis. By comparing the distortion prediction values ​​of different samples, evaluate the prediction accuracy of the model and provide a basis for subsequent audio optimization. Select a suitable global adaptive distortion optimization model, and use nonlinear distortion compensation technology or adaptive filters to dynamically adjust the audio signal according to the distortion prediction data.According to the audio distortion prediction data obtained in step S68, the optimization parameters are set. This includes the gain of distortion compensation, the parameters of the filter, and the dynamic adjustment strategy to effectively reduce distortion. The audio signal after delay optimization is input into the adaptive distortion optimization model to perform global adaptive distortion optimization processing. By adjusting the frequency response and gain of the audio signal in real time, the degree of distortion is reduced and the overall quality of the audio is improved. The audio signal after global adaptive distortion optimization is recorded in the database, and a visual chart is generated to show the audio waveform and spectrum characteristics before and after optimization. Through comparative analysis, the effect of distortion optimization is evaluated to provide a basis for subsequent applications.

[0037] In this embodiment, a high-quality audio data processing system is provided, which is used to execute the high-quality audio data processing method as described above, including: The noise filtering module is used to identify the audio signal to be processed; perform multi-time window segmentation processing and adaptive noise filtering on the audio signal to be processed to obtain a noise filtering optimized audio signal; The transient detail capture module is used to perform multi-scale spectrum decomposition on the noise filtering optimized audio signal, and to capture transient audio details, and identify transient audio detail information at each scale; A full-band range adjustment module is used to calculate the audio frequency range according to the transient audio detail information of each scale, and to perform dynamic full-band range adjustment to obtain transient information full-band optimized audio; The acoustic effect mining module is used to mine the environmental acoustic effects of the full-band optimized audio of transient information and infer the current audio scene to obtain the real-time audio propagation scene; A stereo field enhancement module is used to enhance the stereo field sound effect of the full-band optimized audio of transient information according to the real-time audio propagation scenario, thereby generating a stereo field enhanced audio signal; The distortion optimization module is used to optimize the propagation delay of the stereo field enhanced audio signal and perform global distortion prediction optimization to generate global adaptive distortion optimized audio.

[0038] The present invention effectively removes background noise through adaptive noise filtering to ensure the clarity of the original audio signal. This provides a cleaner input signal for subsequent audio processing and helps to improve the overall quality of the audio. Through multi-time window processing, the changes of audio signals in different time periods can be analyzed more carefully, and customized processing can be performed for different signal features, avoiding the limitations of a single time window processing method. Through multi-scale spectrum decomposition, the audio details in different frequency intervals can be fully analyzed to ensure high-fidelity capture and reproduction of transient audio, and avoid dynamic details in the audio from being ignored or distorted. The capture of transient audio details allows every subtle change in the audio to be retained, especially for audio that requires a quick response (such as musical instrument performance, voice, etc.), which can provide a more realistic and natural sound quality. By dynamically adjusting the frequency band of the audio signal, the audio details of each frequency band are optimized, avoiding excessive suppression of high frequencies or distortion of low frequencies, and providing a more balanced and natural sound quality. Transient information optimization can prevent the transient signal of the audio from being smoothed excessively, ensure that the rapidly changing details in the audio (such as drum beats, piano hitting sounds, etc.) are accurately retained, and improve the clarity and realism of the audio. By mining the acoustic effect, the environmental background of audio propagation can be accurately restored. This is particularly important for recording and audio playback, and can provide data support for the spatial sense and positioning sense of audio. Scene inference enables audio signals to be simulated according to the characteristics of the environment, thereby enhancing the immersiveness of audio and the realism of hearing, which is particularly suitable for application scenarios such as virtual reality and game audio. Through stereo field sound enhancement, the spatial sense of audio signals is significantly improved. The sound is not only emitted from two headphones or speakers, but also simulates a rich three-dimensional sound space, which greatly enhances the auditory experience. The enhanced audio is not only directional in position, but also has a sense of depth and layering, especially in applications such as virtual reality and cinema audio, which can make the audience feel more natural and realistic sound effects. Through delay optimization, signal synchronization in real-time audio applications is ensured, audio asynchrony caused by propagation delay is avoided, and the audio interaction experience is improved. Through global distortion prediction, each frequency segment in the audio is optimized to avoid sound quality degradation caused by environmental interference or propagation problems, and ensure the high-fidelity effect of audio output.

[0039] Therefore, the embodiments should be regarded as illustrative and non-restrictive from all points, and the scope of the present invention is limited by the appended claims rather than the above description, and it is therefore intended that all changes falling within the meaning and range of equivalent elements of the application documents are included in the present invention.

[0040] The above is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features invented herein.

Claims

1. A high-quality audio data processing method, characterized in that: The following steps are involved: Step S1: Identify the audio signal to be processed; perform multi-time window segmentation processing and adaptive noise filtering on the audio signal to be processed to obtain a noise-filtered optimized audio signal; Step S2: performing multi-scale spectrum decomposition on the noise filtering optimized audio signal, and capturing transient audio details, and identifying transient audio detail information of each scale; Step S3: Calculate the audio frequency range according to the transient audio detail information of each scale, and perform dynamic full-band range adjustment to obtain transient information full-band optimized audio; Step S4: mining the environmental acoustic effect of the full-band optimized audio of transient information, and inferring the current audio scene to obtain a real-time audio propagation scene; Step S5: performing stereo field sound effect enhancement on the full-band optimized audio of transient information according to the real-time audio propagation scenario, thereby generating a stereo field enhanced audio signal; Step S6: performing propagation delay optimization on the stereo field enhanced audio signal and performing global distortion prediction optimization to generate global adaptive distortion optimized audio.

2. The high-quality audio data processing method according to claim 1, characterized in that: The specific steps of step S1 are: Step S11: Identify the audio signal to be processed; perform multi-time window segmentation processing on the audio signal to be processed to obtain audio signals of multiple time windows; Step S12: performing audio noise recognition on the audio signals of multiple time windows segment by segment, marking the audio noise points of each time window; Step S13: Calculate the power spectrum density of the audio noise point in each time window, and extract the power spectrum density of each noise point; Step S14: classifying noise according to the power spectrum density of each noise point, thereby generating a type feature of each noise point; Step S15: Adaptively perform noise filtering based on the type characteristics of each noise point to obtain a noise-filtered optimized audio signal.

3. The high-quality audio data processing method according to claim 1, characterized in that: The specific steps of step S2 are: Step S21: performing multi-scale spectrum decomposition on the noise filtering optimized audio signal to obtain audio spectrum data of different scales; Step S22: Calculating the frequency component distribution of the audio spectrum data at different scales to generate the audio frequency component distribution characteristics of each scale; Step S23: dynamically adjusting the time-frequency resolution according to the distribution characteristics of the audio frequency components at each scale to generate the time-frequency resolution at each scale; Step S24: capturing transient audio details of audio spectrum data of different scales based on the time-frequency resolution of each scale, and identifying transient audio detail information of each scale.

4. The high-quality audio data processing method according to claim 1, characterized in that: The specific steps of step S3 are: Step S31: performing transient detail audio fineness quantification analysis on the transient audio detail information of each scale to generate an audio detail fineness quantification value; Step S32: evaluating the clarity of each scale of transient audio information to obtain the clarity of each scale; Step S33: performing a quantitative analysis of the auditory effect according to the information clarity and audio detail fineness value of each scale, thereby generating a transient detail auditory effect of each scale; Step S34: Calculate the audio frequency range of the audio spectrum data of different scales, and extract the audio frequency range of each scale; Step S35: dynamically adjusting the audio frequency range of each scale in the full frequency range based on the transient detail auditory effect of each scale to obtain transient information full frequency band optimized audio.

5. The high-quality audio data processing method according to claim 1, characterized in that: The specific steps of step S4 are: Step S41: identifying the ambient sound source of the full-band optimized audio of transient information and separating the features of the ambient sound source; Step S42: Analyze the size of the ambient space based on the ambient sound source characteristics to generate the size of the ambient space; Step S43: estimating the reflection characteristics of the ambient sound source based on the ambient sound source characteristics, thereby estimating the ambient reflection material; Step S44: performing sound reflection path evolution on the ambient sound source characteristics to generate an ambient sound reflection path; Step S45: mining the environmental acoustic effect according to the size of the environmental space, the environmental reflection material and the environmental sound reflection path, thereby generating environmental acoustic effect parameters; Step S46: Infer the current audio scene based on the environmental acoustic effect parameters to obtain a real-time audio propagation scene.

6. The high-quality audio data processing method according to claim 1, characterized in that: The specific steps of step S5 are: Step S51: performing audio space propagation evolution on environmental acoustic effect parameters to generate a spatial propagation acoustic effect evolution law; Step S52: performing spatial propagation simulation processing on the real-time audio propagation scene according to the evolution law of spatial propagation acoustic effect, and collecting audio spatial propagation simulation data; Step S53: performing iterative learning of deep stereo field features on the audio space propagation simulation data to generate deep stereo field features; Step S54: performing stereo field sound effect enhancement on the transient information full-band optimized audio according to the deep stereo field characteristics, thereby generating a stereo field enhanced audio signal.

7. The high-quality audio data processing method according to claim 1, characterized in that: The specific steps of step S6 are: Step S61: performing real-time propagation parameter monitoring on the audio signal to be processed to generate audio real-time propagation monitoring parameters; Step S62: Calculate the propagation path delay of the audio real-time propagation monitoring parameters and extract the propagation path delay parameters; Step S63: performing audio delay compensation calculation on the propagation path delay parameter, thereby obtaining an audio delay compensation parameter; Step S64: optimizing the propagation delay of the stereo field enhancement audio signal according to the audio delay compensation parameter, thereby generating a delay optimized audio signal; Step S65: performing global distortion prediction optimization on the delay-optimized audio signal to generate a global adaptive distortion-optimized audio.

8. The high-quality audio data processing method according to claim 1, characterized in that: The specific steps of step S65 are: Perform full-band audio frequency mutation detection on the delay-optimized audio signal and mark the frequency mutation nodes; Perform mutation point distribution analysis on frequency mutation nodes to generate audio mutation point distribution features; Predict audio distortion based on the distribution characteristics of audio mutation points and generate audio distortion prediction data; The delay-optimized audio signal is globally adaptively distorted according to the audio distortion prediction data, thereby generating globally adaptive distortion-optimized audio.

9. A high-quality audio data processing system, characterized in that: The method for processing high-quality audio data according to claim 1 comprises: The noise filtering module is used to identify the audio signal to be processed; perform multi-time window segmentation processing and adaptive noise filtering on the audio signal to be processed to obtain a noise filtering optimized audio signal; The transient detail capture module is used to perform multi-scale spectrum decomposition on the noise filtering optimized audio signal, and to capture transient audio details, and identify transient audio detail information at each scale; A full-band range adjustment module is used to calculate the audio frequency range according to the transient audio detail information of each scale, and to perform dynamic full-band range adjustment to obtain transient information full-band optimized audio; The acoustic effect mining module is used to mine the environmental acoustic effects of the full-band optimized audio of transient information and infer the current audio scene to obtain the real-time audio propagation scene; A stereo field enhancement module is used to enhance the stereo field sound effect of the full-band optimized audio of transient information according to the real-time audio propagation scenario, thereby generating a stereo field enhanced audio signal; The distortion optimization module is used to optimize the propagation delay of the stereo field enhanced audio signal and perform global distortion prediction optimization to generate global adaptive distortion optimized audio.

Images