Bone conduction multi-terminal differentiated broadcast distribution method

Abstract

The application discloses a bone conduction multi-terminal differentiation broadcast distribution method and relates to the technical field of immersive sound field control, which comprises the following steps: combining multiple sound element signals according to the immersive sound field configuration of each bone conduction terminal to form a target immersive sound field reference signal and then issuing and correlatively storing the target immersive sound field reference signal, and establishing a sound field interference corresponding relationship according to the acoustic influence direction between each bone conduction terminal; collecting mixed acoustic signals of each bone conduction terminal, extracting a source end reference signal segment and a target end reference signal segment according to a collection time identifier and a playing time identifier; filtering self playing feedback signals according to the target end reference signal segment and a self playing feedback path response to form a sound field interference residual signal; determining a leakage acoustic component and generating a sound field cancellation signal according to the source end reference signal segment and the sound field interference residual signal; synthesizing each sound field cancellation signal acting on the same target terminal, mixing the sound field cancellation signal with the target immersive sound field reference signal to form a sound field driving signal, and then issuing the sound field driving signal to the target terminal for playing.

Description

Technical Field

[0001] This invention relates to the field of immersive sound field control technology, specifically a bone conduction multi-terminal differentiated broadcast control and distribution method. Background Technology

[0002] Bone conduction audio technology transmits mechanical vibrations corresponding to sound content through the skull and related tissues to the auditory system via bone conduction sound-emitting components, allowing users to receive target audio information while maintaining awareness of surrounding sounds. With the application of bone conduction terminals in close-range shared spaces such as home rest areas, vehicle cabins, exhibition tours, and multi-person collaborative training, different users receive different sound content through corresponding bone conduction terminals. Differentiated sound presentations are created based on the sound element composition, playback gain, and playback timing corresponding to different users, thus creating an application demand for constructing differentiated immersive sound fields for multiple bone conduction terminals. Existing multi-terminal audio playback control methods typically use a central control device to independently distribute and route different audio content, sending different audio streams to corresponding terminals and achieving synchronized playback between multiple terminals through a unified clock, playback buffer, and time stamp. Some playback control methods can also arrange, adjust gain, and control playback timing of multiple sound elements according to the playback needs of different terminals to form playback signals corresponding to different terminals.

[0003] In a multi-terminal differentiated immersive sound field, the sound element composition, playback gain, and playback timing of each bone conduction terminal differ. Therefore, it is necessary to maintain the relative independence of the sound content corresponding to different bone conduction terminals in a close-range shared space. When multiple bone conduction terminals are playing simultaneously at close range, the mechanical vibrations generated by the bone conduction sound-generating components and terminal structures, in addition to acting on the corresponding user along the bone conduction path, can also form additional acoustic components through air propagation, structural contact propagation, or human proximity coupling propagation. This causes some of the sound generated by the source terminal to act on other target terminals along a directional acoustic leakage path. At the same time, the microphone set on the target terminal will also collect the feedback acoustic components generated by the target terminal's own playback. Thus, the acoustic signal collected by the target terminal simultaneously includes its own playback feedback, leakage acoustic components generated by different source terminals, and environmental noise.

[0004] Due to differences in spatial relationships, acoustic influence directions, acoustic leakage path responses, and propagation delays between different source terminals and target terminals, and because the same target terminal can be simultaneously affected by the acoustics of multiple source terminals, it is difficult to maintain the sound independence between the immersive sound fields of each target terminal simply by distributing audio content, configuring playback gain, and controlling playback timing for each bone conduction terminal separately. Existing multi-terminal audio playback control methods still have shortcomings in distinguishing between the playback feedback of the target terminal itself and the leaked acoustic components of different source terminals, separately processing acoustic leakage paths in different directions, coordinating the timing of multiple sound field cancellation signals, and coordinating the control between the sound field cancellation signals and the original playback signals of the target terminal. This results in some playback sounds corresponding to the source terminal entering the immersive sound field corresponding to other target terminals, thereby affecting the independent presentation of differentiated sound content of different terminals. Therefore, while maintaining the differentiated sound element composition, playback gain, and playback timing of each bone conduction terminal, and considering the directional acoustic leakage relationship between different bone conduction terminals, it is necessary to distinguish between the playback feedback of the target terminal itself and the leakage acoustic components of the source terminal, and to implement timing coordination and sound field cancellation for multiple acoustic leakage components acting on the same target terminal, so as to maintain the sound independence and playback coordination of the differentiated immersive sound field corresponding to each bone conduction terminal under the condition of simultaneous playback at close range. This constitutes a technical problem that needs to be solved in the field of bone conduction multi-terminal immersive sound field control. Summary of the Invention

[0005] Based on the shortcomings of the prior art described above, the purpose of this invention is to provide a bone conduction multi-terminal differentiated broadcast control and distribution method to solve the above-mentioned technical problems.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a bone conduction multi-terminal differentiated broadcast control and distribution method, applied to a broadcast control system including a central controller and multiple bone conduction terminals communicatively connected to the central controller, the method comprising: S1: Based on the immersive sound field configuration of each bone conduction terminal, multiple sound element signals are combined differentially to form the target immersive sound field reference signal for each bone conduction terminal; S2: Send the target immersive sound field reference signal to the corresponding bone conduction terminal respectively, store the target immersive sound field reference signal and the playback time identifier, establish the sound field interference correspondence relationship according to the acoustic influence direction between each bone conduction terminal, and determine the source terminal and target terminal in each sound field interference correspondence relationship; S3: Collect mixed acoustic signals through the microphones of each bone conduction terminal, and extract the source end reference signal segment and target end reference signal segment corresponding to the time of the mixed acoustic signal from the associated stored target immersive sound field reference signal based on the acquisition time identifier of the mixed acoustic signal and the playback time identifier of the target immersive sound field reference signal. S4: Determine the self-play feedback signal based on the target end reference signal segment and the target terminal's self-play feedback path response, filter out the self-play feedback signal from the target terminal's mixed acoustic signal, and form a sound field interference residual signal; S5: Determine the leakage acoustic component of the source terminal acting on the target terminal based on the source terminal reference signal segment and the sound field interference residual signal, and generate the corresponding sound field cancellation signal based on the leakage acoustic component; S6: Combine the sound field cancellation signals acting on the same target terminal, mix the combined sound field cancellation signal with the target immersive sound field reference signal of the target terminal to form a sound field driving signal, and send the sound field driving signal to the target terminal for playback.

[0007] The present invention is further configured such that the immersive sound field configuration includes sound element affiliation, sound element playback gain, and sound element playback timing; and differentiated combination of multiple sound element signals, including: Based on the attribution relationship of sound elements, determine the sound element signal corresponding to each bone conduction terminal; Based on the playback sequence of sound elements, extract the corresponding sound element signal segments from each sound element signal; Adjust the gain of each sound element signal segment according to the playback gain of the sound element; The gain-adjusted signal segments of each sound element corresponding to the same bone conduction terminal are mixed to generate the target immersive sound field reference signal for the corresponding bone conduction terminal.

[0008] The present invention is further configured to associate and store the target immersive sound field reference signal with a playback time identifier, including: Configure a ring-shaped reference signal buffer corresponding to each bone conduction terminal in the central controller; The immersive sound field reference signals of each target are processed in frames to obtain continuous reference signal frames. The audio sampling data, terminal identifier and playback time identifier of each reference signal frame are associated and encapsulated into a reference signal data unit. Based on the terminal identifier, each reference signal data unit is stored in the corresponding circular reference signal buffer, and the storage timing of each reference signal data unit in the circular reference signal buffer is determined according to the playback time identifier.

[0009] The present invention is further configured to establish a sound field interference correspondence based on the acoustic influence direction between each bone conduction terminal, including: An adjacency relation matrix is ​​constructed to characterize the correspondence of sound field interference between bone conduction terminals. The adjacency relation matrix is ​​a Boolean square matrix of order equal to the number of bone conduction terminals. The row index of the adjacency matrix corresponds one-to-one with the terminal identifier of the source terminal, and the column index corresponds one-to-one with the terminal identifier of the target terminal. Each matrix element is used to record the existence status of the acoustic leakage path to be canceled between the source terminal indicated by the corresponding row index and the target terminal indicated by the corresponding column index.

[0010] The present invention is further configured such that S3 includes: The mixed acoustic signal at the location of the microphone is continuously sampled by the microphone of each bone conduction terminal, and the sampled mixed acoustic signal is processed into a frame to form a mixed signal frame. The audio sampling data, acquisition time identifier, and terminal identifier of each mixed signal frame are associated and encapsulated into a mixed signal data unit, and the mixed signal data unit is uploaded to the central controller. The acquisition time identifier is generated based on the terminal clock synchronized with the reference clock of the central controller. Based on the signal transmission delay and acoustic propagation delay, delay compensation is performed on the acquisition time marker of each mixed signal frame to determine the reference signal extraction time range corresponding to each mixed signal frame; The source terminal and target terminal are determined based on the correspondence of sound field interference, and reference signal data units are read from the corresponding circular reference signal buffer according to the terminal identifiers of the source terminal and target terminal. Based on the playback time identifier of each reference signal data unit, select the reference signal data unit that corresponds to the playback time range and the reference signal extraction time range, and extract audio sampling data from the selected reference signal data unit to form the source end reference signal segment and the target end reference signal segment respectively.

[0011] The present invention is further configured such that the self-play feedback path response is obtained through offline calibration, wherein the offline calibration includes: According to the preset calibration order, the bone conduction terminal to be calibrated is determined from each bone conduction terminal in sequence, the bone conduction terminal to be calibrated is switched to offline calibration state, the bone conduction terminal to be calibrated is controlled to play the preset broadband test signal, and the acoustic response signal is synchronously collected through the microphone set on the bone conduction terminal to be calibrated. The preset broadband test signal is used as the system identification input, and the acoustic response signal is used as the system identification output. The self-play feedback path filter coefficients of the bone conduction terminal to be calibrated are obtained through linear time-invariant system identification. The terminal identifier of the bone conduction terminal to be calibrated is associated with its own playback feedback path filter coefficient to form a feedback path calibration record. The feedback path calibration record is stored, and the own playback feedback path filter coefficient is used to characterize the own playback feedback path response of the bone conduction terminal to be calibrated.

[0012] The present invention is further configured to form a sound field interference residual signal, including: The target terminal reference signal segment is convolved with the target terminal's own playback feedback path response to generate the target terminal's own playback feedback signal. Based on the time correspondence between the target end reference signal segment and the mixed acoustic signal, the sampling time reference of its own playback feedback signal is corrected so that each sampling position of its own playback feedback signal corresponds one-to-one with each sampling position of the mixed acoustic signal. For each sampling location, the corresponding self-play feedback signal sampling value is subtracted from the mixed acoustic signal sampling value of the target terminal to form the sound field interference residual signal.

[0013] The present invention is further configured to determine the leakage acoustic component of the source terminal acting on the target terminal based on the source-end reference signal segment and the acoustic field interference residual signal, including: Multi-band filtering is performed on the acoustic interference residual signal to obtain multiple frequency band residual signals. The short-time average power of each frequency band residual signal is calculated, and each short-time average power is compared with the dynamic decision threshold of the corresponding frequency band. When the short-time average power of the residual signal in at least one frequency band is greater than the dynamic decision threshold of the corresponding frequency band, the source-end reference signal segment is used as the reference input signal of the adaptive filter, and the acoustic field interference residual signal is used as the desired response signal of the adaptive filter. The path prediction signal is generated based on the current filter coefficients of the adaptive filter and the source-end reference signal segment. The path identification error signal between the path prediction signal and the desired response signal is determined. Based on the source-end reference signal segment and the path identification error signal, the filter coefficients of the adaptive filter are updated using the least mean square algorithm. The updated filter coefficients are then determined as the acoustic leakage path response from the source terminal to the target terminal. From the target immersive sound field reference signal corresponding to and associated with the source terminal, extract the reference signal segment whose playback time identifier is later than the current processing time of the central controller to form the source terminal reference signal segment to be played; perform convolution operation on the source terminal reference signal segment to be played and the acoustic leakage path response to generate the predicted leakage signal of the source terminal acting on the target terminal, and determine the predicted leakage signal as the leakage acoustic component.

[0014] The present invention is further configured to generate a corresponding sound field cancellation signal based on the leakage acoustic component, including: The target terminal's own playback feedback path response is determined as the offset playback path response, and a secondary path model is established based on the offset playback path response. The reference signal segment from the source to be played is input into the cancellation filter to generate a candidate sound field cancellation signal. The candidate sound field cancellation signal is then filtered using a secondary path model to obtain the predicted cancellation output signal corresponding to the microphone position of the candidate sound field cancellation signal at the target terminal. The predicted cancellation output signal is superimposed with the predicted leakage signal to form a cancellation error signal; the secondary path model is used to filter the reference signal segment at the source end to be played to form a secondary path reference signal; based on the secondary path reference signal and the cancellation error signal, the filter coefficients of the cancellation filter are updated using the filter-x least mean square algorithm. The updated cancellation filter is used to filter the reference signal segment at the source end of the playback source to generate a sound field cancellation signal.

[0015] The present invention is further configured such that S6 includes: Based on the playback time identifier of the reference signal segment of the source to be played used to generate each sound field cancellation signal and the output sampling position corresponding to each sample value in each sound field cancellation signal, the target playback time corresponding to each sample value in each sound field cancellation signal under the reference clock of the central controller is determined. Based on the target playback time corresponding to each sample value, time alignment is performed on each sound field cancellation signal acting on the same target terminal; The sampled values ​​of each time-aligned sound field cancellation signal at the same target playback time are summed to generate the total sound field cancellation signal of the target terminal, and the target playback time range corresponding to the total sound field cancellation signal is determined. Based on the target playback time range and the playback time identifier of the target immersive sound field reference signal of the target terminal, extract the target terminal reference signal segment corresponding to the playback time range from the target immersive sound field reference signal; The total sound field cancellation signal is mixed with the target end reference signal segment on a sample-by-sample basis to generate the sound field driving signal of the target terminal; Based on the target terminal's identifier, the sound field drive signal is sent to the target terminal for playback.

[0016] This invention provides a differentiated broadcast control and distribution method for bone conduction multi-terminals. The method comprises: S1: Differentiatedly combining multiple sound element signals according to the immersive sound field configuration of each bone conduction terminal to form a target immersive sound field reference signal for each bone conduction terminal; S2: Distributing each target immersive sound field reference signal to its corresponding bone conduction terminal, associating and storing the target immersive sound field reference signal with a playback time identifier, establishing a sound field interference correspondence based on the acoustic influence direction between each bone conduction terminal, and determining the source terminal and target terminal in each sound field interference correspondence; S3: Acquiring mixed acoustic signals through the microphones of each bone conduction terminal, and extracting from the associated stored target immersive sound field reference signal based on the acquisition time identifier of the mixed acoustic signal and the playback time identifier of the target immersive sound field reference signal. S4: Determine the self-play feedback signal based on the target terminal's self-play feedback path response and the target terminal's reference signal segment corresponding to the time of the mixed acoustic signal; filter out the self-play feedback signal from the target terminal's mixed acoustic signal to form a sound field interference residual signal; S5: Determine the leakage acoustic component of the source terminal acting on the target terminal based on the source terminal's reference signal segment and the sound field interference residual signal; generate a corresponding sound field cancellation signal based on the leakage acoustic component; S6: Synthesize the sound field cancellation signals acting on the same target terminal; mix the synthesized sound field cancellation signal with the target immersion sound field reference signal of the target terminal to form a sound field driving signal; send the sound field driving signal to the target terminal for playback. The beneficial effects include: By determining the target terminal's own playback feedback path response as the cancellation playback path response, and establishing a secondary path model based on the cancellation playback path response, the predicted cancellation output signal and the secondary path reference signal are obtained using the secondary path model. Based on the secondary path reference signal and the cancellation error signal, the filter coefficients of the cancellation filter are updated using the filter-x least mean square algorithm, so that the generated sound field cancellation signal is adapted to the cancellation playback path of the target terminal, thereby reducing the acoustic interference of the source terminal's leakage acoustic component on the target terminal and improving the independence of the sound content played by different bone conduction terminals.

[0017] Based on the target playback time corresponding to each sample value in each sound field cancellation signal, multiple sound field cancellation signals acting on the same target terminal are time-aligned, and the sample values ​​corresponding to the same target playback time are summed to generate the total sound field cancellation signal of the target terminal. The total sound field cancellation signal is then mixed with the target end reference signal segment corresponding to the playback time range on a sample-by-sample basis to form the sound field driving signal of the target terminal. This achieves the coordinated cancellation of leaked acoustic components from multiple source terminals while maintaining the differentiated immersive sound field playback of each bone conduction terminal, thereby improving the sound independence and playback coordination in the multi-terminal broadcast control and distribution process of bone conduction.

[0018] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, the following are specific embodiments of this application. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. In the drawings: Figure 1 The flowchart illustrates a bone conduction multi-terminal differentiated broadcast control and distribution method as an exemplary embodiment of the present invention. Detailed Implementation

[0020] The embodiments of the present invention will be described below with reference to the accompanying drawings and preferred embodiments. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be understood that the preferred embodiments are only for illustrating the present invention and not for limiting the scope of protection of the present invention.

[0021] It should be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of the present invention. Therefore, the drawings only show the components related to the present invention and are not drawn according to the actual number, shape and size of the components in the actual implementation. In the actual implementation, the form, quantity and proportion of each component can be arbitrarily changed, and the layout of the components may also be more complex.

[0022] In the following description, numerous details are explored to provide a more thorough explanation of embodiments of the invention. However, it will be apparent to those skilled in the art that embodiments of the invention may be practiced without these specific details. In other embodiments, well-known structures and devices are shown in block diagram form rather than in detail to avoid obscuring embodiments of the invention.

[0023] A bone conduction multi-terminal differentiated broadcast control and distribution method is applied to a broadcast control system including a central controller and multiple bone conduction terminals communicatively connected to the central controller, such as... Figure 1 As shown, the method includes: S1: Based on the immersive sound field configuration of each bone conduction terminal, multiple sound element signals are combined differentially to form the target immersive sound field reference signal for each bone conduction terminal; S2: Send the target immersive sound field reference signal to the corresponding bone conduction terminal respectively, store the target immersive sound field reference signal and the playback time identifier, establish the sound field interference correspondence relationship according to the acoustic influence direction between each bone conduction terminal, and determine the source terminal and target terminal in each sound field interference correspondence relationship; S3: Collect mixed acoustic signals through the microphones of each bone conduction terminal, and extract the source end reference signal segment and target end reference signal segment corresponding to the time of the mixed acoustic signal from the associated stored target immersive sound field reference signal based on the acquisition time identifier of the mixed acoustic signal and the playback time identifier of the target immersive sound field reference signal. S4: Determine the self-play feedback signal based on the target end reference signal segment and the target terminal's self-play feedback path response, filter out the self-play feedback signal from the target terminal's mixed acoustic signal, and form a sound field interference residual signal; S5: Determine the leakage acoustic component of the source terminal acting on the target terminal based on the source terminal reference signal segment and the sound field interference residual signal, and generate the corresponding sound field cancellation signal based on the leakage acoustic component; S6: Combine the sound field cancellation signals acting on the same target terminal, mix the combined sound field cancellation signal with the target immersive sound field reference signal of the target terminal to form a sound field driving signal, and send the sound field driving signal to the target terminal for playback.

[0024] The present invention is further configured such that the immersive sound field configuration includes sound element affiliation, sound element playback gain, and sound element playback timing; and differentiated combination of multiple sound element signals, including: Based on the attribution relationship of sound elements, determine the sound element signal corresponding to each bone conduction terminal; Based on the playback sequence of sound elements, extract the corresponding sound element signal segments from each sound element signal; Adjust the gain of each sound element signal segment according to the playback gain of the sound element; The gain-adjusted signal segments of each sound element corresponding to the same bone conduction terminal are mixed to generate a target immersive sound field reference signal for the corresponding bone conduction terminal. Specifically, the central controller pre-configures a unique terminal identifier for each bone conduction terminal and establishes an immersive sound field configuration for each terminal. The immersive sound field configuration includes sound element attribution relationships, sound element playback gain, and sound element playback timing. The sound element attribution relationships include terminal identifiers and sound element identifiers, used to characterize the attribution relationship between each bone conduction terminal and each sound element signal. The sound element playback gain is used to characterize the amplitude adjustment parameters of each sound element signal in the corresponding bone conduction terminal. The playback timing sequence of each sound element signal is used to characterize the start and end times, start and end positions of each signal segment. The central controller acquires multiple sound element signals, decodes each signal to obtain audio sample data arranged in the sampling sequence, performs sampling rate conversion, bit depth conversion, and channel format conversion on each signal to ensure all signals have the same sampling rate, bit depth, and channel format, and associates and stores the sound element identifier of each signal with the format-converted audio sample data. The central controller reads the terminal identifier of each bone conduction terminal and retrieves the corresponding immersion device based on the terminal identifier. The sound field configuration involves obtaining sound element identifiers associated with terminal identifiers from the sound element affiliation relationship, and reading the corresponding audio sampling data based on the sound element identifiers to determine the sound element signals corresponding to each bone conduction terminal. When the same sound element signal belongs to multiple bone conduction terminals, it is processed according to the sound element playback gain and playback timing of each bone conduction terminal. The central controller uses its reference clock as a unified time base, reading the playback start time, playback end time, signal segment start position, and signal segment end position from the sound element playback timing. Based on the sampling rate of the sound element signal, the signal segment is... The start position and signal segment end position are converted into the start sampling position and end sampling position, respectively. The sampling values ​​between the start sampling position and the end sampling position are continuously read from the audio sampling data of the corresponding sound element signal to form the corresponding sound element signal segment. The central controller reads the sound element playback gain from the immersive sound field configuration of the corresponding bone conduction terminal according to the terminal identifier and the sound element identifier. The sound element playback gain is converted into a linear amplitude adjustment coefficient. The amplitude of each sampling value in the sound element signal segment is adjusted using the linear amplitude adjustment coefficient. The amplitude-adjusted sampling values ​​are arranged according to the original sampling time sequence to form the gain-adjusted sound element signal segment.The central controller, based on the start and end times of each sound element signal segment, configures the gain-adjusted sound element signal segments corresponding to the same bone conduction terminal onto a unified playback time axis. This ensures that sampled values ​​at the same playback time are located at the same sampling position. The sampled values ​​at the same sampling position are then superimposed to obtain the mixed sampled values ​​for each position. These mixed sampled values ​​are then arranged sequentially according to the playback time sequence to generate the target immersive sound field reference signal for the corresponding bone conduction terminal. Due to differences in the sound element affiliation, sound element playback gain, and sound element playback time sequence among different bone conduction terminals, the generated target immersive sound field reference signals have different sound element compositions, playback gains, and playback times.

[0025] The present invention is further configured to associate and store the target immersive sound field reference signal with a playback time identifier, including: Configure a ring-shaped reference signal buffer corresponding to each bone conduction terminal in the central controller; The immersive sound field reference signals of each target are processed in frames to obtain continuous reference signal frames. The audio sampling data, terminal identifier and playback time identifier of each reference signal frame are associated and encapsulated into a reference signal data unit. Based on the terminal identifier, each reference signal data unit is stored in its corresponding circular reference signal buffer, and the storage timing of each reference signal data unit in the circular reference signal buffer is determined according to the playback time identifier. Specifically, the central controller pre-configures a unique terminal identifier for each bone conduction terminal and configures a corresponding circular reference signal buffer in the central controller, establishing a correspondence between the terminal identifier and the circular reference signal buffer, which is used to determine the corresponding circular reference signal buffer based on the terminal identifier. The central controller determines the number of reference signal data units that can be stored in each circular reference signal buffer based on the preset reference signal storage duration, the sampling rate of the target immersive sound field reference signal, and the preset frame length, and configures a write position marker and an earliest valid position marker for each circular reference signal buffer. The write position marker is used to indicate the storage location of the reference signal data unit to be written, and the earliest valid position marker is used to indicate the storage location of the earliest valid reference signal data unit of the playback time identifier. The central controller performs frame processing on each target immersive sound field reference signal according to the sampling timing of each target immersive sound field reference signal, starting from the initial sampling position of each target immersive sound field reference signal and according to the preset frame length. The system obtains continuous reference signal frames and ensures that the sampling positions of adjacent reference signal frames are continuous to avoid audio sampling data overlap or omission between adjacent reference signal frames. The central controller uses its reference clock as a unified time base. Based on the planned playback start time corresponding to each target immersive sound field reference signal and the sampling position of each reference signal frame in the corresponding target immersive sound field reference signal, it determines the playback time identifier corresponding to each reference signal frame. Based on the number of sampling points contained before the starting sampling position of each reference signal frame and the sampling rate of the corresponding target immersive sound field reference signal, it determines the time offset of each reference signal frame relative to the planned playback start time. Based on the planned playback start time and the time offset, it determines the playback time identifier of each reference signal frame. The central controller extracts the audio sampling data of each reference signal frame, reads the terminal identifier of the bone conduction terminal corresponding to each target immersive sound field reference signal, and associates and encapsulates the audio sampling data, terminal identifier, and playback time identifier of each reference signal frame into a reference signal data unit. The terminal identifier is used to determine the bone conduction terminal corresponding to the reference signal data unit, and the playback time identifier is used to characterize the planned playback time corresponding to the starting sampling position of the reference signal data unit.The central controller stores each reference signal data unit into its corresponding circular reference signal buffer based on the terminal identifier. It then determines the storage sequence of each reference signal data unit in the circular reference signal buffer according to the playback time identifier. Following this sequence, it writes each reference signal data unit sequentially into its corresponding circular reference signal buffer and updates the write position marker upon completion. When the circular reference signal buffer is full, it replaces the reference signal data unit with the earliest playback time identifier with a new one, and simultaneously updates the earliest valid position marker. This ensures that the circular reference signal buffer cyclically stores reference signal data units according to the playback time sequence represented by the playback time identifier.

[0026] The present invention is further configured to establish a sound field interference correspondence based on the acoustic influence direction between each bone conduction terminal, including: An adjacency relation matrix is ​​constructed to characterize the correspondence of sound field interference between bone conduction terminals. The adjacency relation matrix is ​​a Boolean square matrix of order equal to the number of bone conduction terminals. The adjacency matrix has row indices that correspond one-to-one with the terminal identifiers of source terminals and column indices that correspond one-to-one with the terminal identifiers of target terminals. Each matrix element records the existence status of the acoustic leakage path to be canceled between the source terminal indicated by the corresponding row index and the target terminal indicated by the corresponding column index. Specifically, the central controller acquires the terminal identifiers of each bone conduction terminal participating in differentiated broadcast control, arranges the terminal identifiers according to a preset index order, establishes a one-to-one correspondence between terminal identifiers and row indices, and a one-to-one correspondence between terminal identifiers and column indices, and constructs an adjacency matrix based on the number of bone conduction terminals to characterize the acoustic field interference correspondence between each bone conduction terminal. The adjacency matrix has an order equal to the bone conduction terminal number. A Boolean matrix representing the number of acoustic transmission terminals is used, and each element in the adjacency relation matrix is ​​initialized to a Boolean value indicating that there is no acoustic leakage path to be canceled. The row indices of the adjacency relation matrix correspond one-to-one with the terminal identifiers of the source terminals, and the column indices correspond one-to-one with the terminal identifiers of the target terminals. Each matrix element records the existence status of the acoustic leakage path to be canceled between the source terminal indicated by the corresponding row index and the target terminal indicated by the corresponding column index. The central controller sequentially identifies each bone conduction terminal as a source terminal to be detected and identifies each bone conduction terminal other than the source terminals to be detected as a target terminal to be detected. It controls the source terminals to play a preset acoustic detection signal and controls the target terminals to stop playing the signal. The acoustic response signal at the microphone location is synchronously acquired by microphones installed at each target terminal under test. The central controller uses its reference clock as a unified time base. Based on the playback time marker of the preset acoustic detection signal and the acquisition time marker of the acoustic response signal, it extracts the response signal segment corresponding to the preset acoustic detection signal from the acoustic response signal. The preset acoustic detection signal and the response signal segment are time-aligned to determine the correlation between them. Based on the acoustic response components in the response signal segment that have waveform correlation with the preset acoustic detection signal, the acoustic leakage intensity of the source terminal acting on the target terminal under test is determined. The central controller stops at each bone conduction terminal. In the playback state, ambient acoustic signals are collected by microphones set on each target terminal to be tested. The ambient noise level corresponding to each target terminal to be tested is determined based on the ambient acoustic signals. The acoustic leakage path judgment threshold is determined based on the ambient noise level. If the correlation between the preset acoustic detection signal and the response signal segment meets the preset correlation conditions and the acoustic leakage intensity is greater than the acoustic leakage path judgment threshold, it is determined that there is an acoustic leakage path to be canceled between the source terminal to be tested and the target terminal to be tested. If the correlation does not meet the preset correlation conditions or the acoustic leakage intensity is not greater than the acoustic leakage path judgment threshold, it is determined that there is no acoustic leakage path to be canceled between the source terminal to be tested and the target terminal to be tested.The central controller determines the corresponding row index based on the terminal identifier of the source terminal to be detected, and the corresponding column index based on the terminal identifier of the target terminal to be detected. If there is an acoustic leakage path to be canceled between the source terminal and the target terminal, the matrix element determined by the corresponding row and column indices is set to a Boolean value indicating the existence of such an acoustic leakage path. If there is no acoustic leakage path to be canceled between the source terminal and the target terminal, the matrix element remains a Boolean value indicating the absence of such an acoustic leakage path. The central controller detects acoustic leakage paths from the source terminal to the target terminal and from the target terminal to the source terminal, respectively, and sets the adjacency relationship matrix according to the detection results for different acoustic influence directions. For the corresponding matrix elements, where both row and column indices correspond to the same bone conduction terminal, the matrix element is set to a Boolean value indicating the absence of an acoustic leakage path to be canceled. The acoustic feedback generated by the bone conduction terminal itself is processed through the corresponding bone conduction terminal's own playback feedback path response. After the central controller completes the acoustic leakage path detection between each bone conduction terminal, it associates and stores the adjacency matrix with the terminal identifiers of the source terminals corresponding to each row index and the target terminals corresponding to each column index. The bone conduction terminal indicated by the row index corresponding to the matrix element indicating an acoustic leakage path to be canceled is identified as the source terminal, and the bone conduction terminal indicated by the corresponding column index is identified as the target terminal. This establishes a sound field interference correspondence based on the acoustic influence direction between each bone conduction terminal.

[0027] The present invention is further configured such that S3 includes: The mixed acoustic signal at the location of the microphone is continuously sampled by the microphone of each bone conduction terminal, and the sampled mixed acoustic signal is processed into a frame to form a mixed signal frame. The audio sampling data, acquisition time identifier, and terminal identifier of each mixed signal frame are associated and encapsulated into a mixed signal data unit, and the mixed signal data unit is uploaded to the central controller. The acquisition time identifier is generated based on the terminal clock synchronized with the reference clock of the central controller. Based on the signal transmission delay and acoustic propagation delay, delay compensation is performed on the acquisition time marker of each mixed signal frame to determine the reference signal extraction time range corresponding to each mixed signal frame; The source terminal and target terminal are determined based on the correspondence of sound field interference, and reference signal data units are read from the corresponding circular reference signal buffer according to the terminal identifiers of the source terminal and target terminal. Based on the playback time identifier of each reference signal data unit, reference signal data units corresponding to the playback time range and the reference signal extraction time range are selected. Audio sampling data is extracted from the selected reference signal data units to form source-end reference signal segments and target-end reference signal segments, respectively. Specifically, each bone conduction terminal is equipped with a microphone, signal acquisition circuit, and terminal clock. The mixed acoustic signal at the microphone location is continuously sampled through the microphone of each bone conduction terminal, and the signal acquisition circuit converts the mixed acoustic signal into audio sampling data arranged according to the sampling time sequence. The mixed acoustic signal includes the feedback acoustic component generated by the target terminal's own playback and the source terminal's playback. The system detects the leakage acoustic components and ambient noise at the microphone location of the target terminal. Each bone conduction terminal performs frame processing on the sampled mixed acoustic signal according to a preset frame length, forming a mixed signal frame. The sampling positions of adjacent mixed signal frames are made continuous to avoid audio sampling data overlap or omission between adjacent frames. The central controller sends clock synchronization information to each bone conduction terminal according to a preset synchronization period. Each bone conduction terminal corrects its clock based on the clock synchronization information, ensuring that its clock is synchronized with the central controller's reference clock. When forming a mixed signal frame, each bone conduction terminal uses the reference clock of the central controller as a reference. A synchronized terminal clock generates an acquisition time identifier for each mixed signal frame, which represents the acquisition time corresponding to the start sampling position of the mixed signal frame. Each bone conduction terminal associates and encapsulates the audio sampling data, acquisition time identifier, and terminal identifier of each mixed signal frame into a mixed signal data unit, and uploads each mixed signal data unit to the central controller according to the sampling sequence of the mixed signal frames. The central controller determines the bone conduction terminal acquiring the mixed acoustic signal based on the terminal identifier in each mixed signal data unit, determines the acquisition start time of the mixed signal frame based on the acquisition time identifier, and determines the mixing time based on the number of sampling points contained in the mixed signal frame and the sampling rate of the mixed acoustic signal. The acquisition termination time of the combined signal frame; the central controller determines the signal transmission delay from the central controller to the corresponding bone conduction terminal based on the transmission time of the target immersive sound field reference signal and the playback time of the target immersive sound field reference signal output by the corresponding bone conduction terminal, and stores the signal transmission delay in association with the terminal identifier of the corresponding bone conduction terminal, controls the source terminal to play the preset acoustic detection signal, collects the acoustic response signal through the microphone of the target terminal, determines the acoustic propagation delay from the source terminal to the target terminal based on the playback time identifier of the preset acoustic detection signal and the acquisition time identifier of the acoustic response signal, and stores the acoustic propagation delay in association with the terminal identifier of the source terminal and the terminal identifier of the target terminal;The central controller determines the target terminal based on the terminal identifier in the mixed signal data unit, and determines the source terminal acting on the target terminal based on the sound field interference correspondence. Based on the signal transmission delay corresponding to the source terminal and the acoustic propagation delay from the source terminal to the target terminal, it performs delay compensation on the acquisition time identifier of each mixed signal frame to determine the reference signal extraction time range corresponding to the source terminal. Similarly, based on the signal transmission delay corresponding to the target terminal and the acoustic propagation delay corresponding to the target terminal's own playback feedback path, it performs delay compensation on the acquisition time identifier of each mixed signal frame to determine the reference signal extraction time range corresponding to the target terminal. The central controller determines the source and target terminals based on the sound field interference correspondence, and determines the corresponding circular reference signal buffers for the source and target terminals based on their terminal identifiers. It reads reference signal data units from the circular reference signal buffer corresponding to the source terminal and from the circular reference signal buffer corresponding to the target terminal. The central controller determines the playback start time of each reference signal data unit based on its playback time identifier and, based on the information contained in each reference signal data unit... The playback termination time of each reference signal data unit is determined by the number of sampling points and the sampling rate of the target immersive sound field reference signal. The playback time range of each reference signal data unit is determined by the playback start time and playback termination time. The playback time range of each reference signal data unit is compared with the corresponding reference signal extraction time range, and the reference signal data unit corresponding to the playback time range and the reference signal extraction time range is selected. The central controller determines the start extraction position and end extraction position of the audio sampling data in the selected reference signal data unit according to the start time and end time of the reference signal extraction time range. Audio sampling data located within the corresponding reference signal extraction time range is extracted from the reference signal data unit corresponding to the source terminal and arranged continuously according to the playback timing represented by the playback time identifier to form the source end reference signal segment. Audio sampling data located within the corresponding reference signal extraction time range is extracted from the reference signal data unit corresponding to the target terminal and arranged continuously according to the playback timing represented by the playback time identifier to form the target end reference signal segment. When multiple source terminals act on the same target terminal, corresponding source end reference signal segments are formed for each source terminal.

[0028] The present invention is further configured such that the self-play feedback path response is obtained through offline calibration, wherein the offline calibration includes: According to the preset calibration order, the bone conduction terminal to be calibrated is determined from each bone conduction terminal in sequence, the bone conduction terminal to be calibrated is switched to offline calibration state, the bone conduction terminal to be calibrated is controlled to play the preset broadband test signal, and the acoustic response signal is synchronously collected through the microphone set on the bone conduction terminal to be calibrated. The preset broadband test signal is used as the system identification input, and the acoustic response signal is used as the system identification output. The self-play feedback path filter coefficients of the bone conduction terminal to be calibrated are obtained through linear time-invariant system identification. The terminal identifier of the bone conduction terminal to be calibrated is associated with its own playback feedback path filter coefficients to form a feedback path calibration record. This record is stored, and the playback feedback path filter coefficients characterize the self-playback feedback path response of the bone conduction terminal to be calibrated. Specifically, the central controller pre-establishes an offline calibration task and records the terminal identifier and preset calibration order of each bone conduction terminal. When executing the offline calibration task, the bone conduction terminal to be calibrated is sequentially determined from the bone conduction terminals according to the preset calibration order. The bone conduction terminal to be calibrated is switched to offline calibration state, and all bone conduction terminals except the one to be calibrated are controlled to stop playback. The offline calibration state indicates that the bone conduction terminal to be calibrated stops performing normal audio playback tasks and plays a preset broadband test signal under the control of the central controller. Simultaneously, acoustic response signals are collected synchronously through a microphone located on the bone conduction terminal to be calibrated. The self-playback feedback path of the bone conduction terminal to be calibrated includes the playback processing circuit, the bone conduction sound-generating component, the acoustic propagation path between the bone conduction sound-generating component and the microphone, and the signal transmission path formed by the microphone acquisition circuit. The central controller pre-generates a preset broadband test signal covering the operating frequency band of the bone conduction terminal to be calibrated and determines the allowable playback amplitude range of the bone conduction terminal to be calibrated. The playback gain of the preset broadband test signal is set to keep the playback processing circuit, bone conduction sound generation component, and microphone acquisition circuit within their linear operating range. This controls the bone conduction terminal to be calibrated to play the preset broadband test signal and synchronously acquires the acoustic response signal through the microphone located on the terminal. The central controller uses its reference clock as a unified time base. Based on the playback time marker of the preset broadband test signal and the acquisition time marker of the acoustic response signal, the central controller performs time alignment on the preset broadband test signal and the acoustic response signal, and converts the preset broadband test signal into audio sample data with the same sampling rate and bit depth as the acoustic response signal. The system establishes a sampling position correspondence between the sampled values ​​in the preset broadband test signal and the acoustic response signal that are in the same feedback propagation process. The central controller takes the preset broadband test signal as the system identification input and the acoustic response signal as the system identification output. It establishes a sampling position correspondence between the system identification input and the system identification output according to the sampling timing. It determines the filter length according to the propagation delay range corresponding to its own playback feedback path and the duration of the acoustic response signal. It configures the filter coefficients to be identified according to the filter length and uses the filter coefficients to be identified to construct a finite impulse response filter for simulating the playback feedback path of the bone conduction terminal to be calibrated.When obtaining the self-playback feedback path filter coefficients of the bone conduction terminal to be calibrated through linear time-invariant system identification, the central controller extracts the input sampling sequence corresponding to the filter length sequentially according to the sampling timing of the system identification input. Each input sample value in the input sampling sequence is multiplied with the corresponding filter coefficient to be identified, and the results of each multiplication are accumulated to obtain the path prediction output value. The actual output sample value at the same sampling position as the path prediction output value is read from the system identification output, and the difference between the actual output sample value and the path prediction output value is determined as the path identification error. The central controller updates each filter coefficient to be identified based on the path identification error, the input sampling sequence, and the preset update step size to reduce the difference between the path prediction output value generated by the updated filter coefficients and the corresponding actual output sample value. This process of extracting the input sampling sequence is repeated. The process involves determining the path prediction output value, determining the path identification error, and updating the coefficients of the filter to be identified until the change in the path identification error does not exceed a preset convergence threshold. The coefficients of each filter to be identified at the convergence state are then arranged according to their temporal position to obtain the self-playback feedback path filter coefficients of the bone conduction terminal to be calibrated. The central controller associates the terminal identifier of the bone conduction terminal to be calibrated with its self-playback feedback path filter coefficients to form a feedback path calibration record. This record is stored, and the self-playback feedback path filter coefficients characterize the self-playback feedback path response of the bone conduction terminal to be calibrated. After completing the offline calibration of the bone conduction terminal to be calibrated, the terminal is taken out of the offline calibration state, and the system continues to identify other bone conduction terminals that have not yet completed offline calibration according to a preset calibration order, until the self-playback feedback path response of each bone conduction terminal is obtained through offline calibration.

[0029] The present invention is further configured to form a sound field interference residual signal, including: The target terminal reference signal segment is convolved with the target terminal's own playback feedback path response to generate the target terminal's own playback feedback signal. Based on the time correspondence between the target end reference signal segment and the mixed acoustic signal, the sampling time reference of its own playback feedback signal is corrected so that each sampling position of its own playback feedback signal corresponds one-to-one with each sampling position of the mixed acoustic signal. For each sampling location, the corresponding self-playback feedback signal sampling value is subtracted from the mixed acoustic signal sampling value of the target terminal to form a sound field interference residual signal. Specifically, the central controller determines the target terminal based on the terminal identifier in the mixed signal data unit, reads the audio sampling data and acquisition time identifier corresponding to the mixed acoustic signal of the target terminal from the mixed signal data unit, determines the acquisition time range corresponding to the mixed acoustic signal of the target terminal based on the acquisition time identifier, and reads the target end reference signal segment that has a time correspondence with the acquisition time range corresponding to the mixed acoustic signal of the target terminal. The central controller retrieves the feedback path calibration record based on the terminal identifier of the target terminal, and reads the target terminal reference signal segment from the feedback path calibration record. The self-playback feedback path filter coefficients corresponding to the terminal represent the self-playback feedback path response of the target terminal. These coefficients are arranged according to their temporal positions, and the filter length corresponding to the self-playback feedback path response of the target terminal is determined based on the number of coefficients. Before performing convolution operations, the central controller compares the sampling rate of the target terminal's reference signal segment with the sampling rate of the target terminal's mixed acoustic signal. If the sampling rates are inconsistent, the target terminal's reference signal segment undergoes sampling rate conversion according to the sampling rate of the target terminal's mixed acoustic signal, ensuring that the target terminal's reference signal segment matches the sampling rate of the target terminal's mixed acoustic signal. The mixed acoustic signals of the target terminal have the same sampling rate. When convolving the target terminal reference signal segment with the target terminal's own playback feedback path response, the central controller extracts the reference sampling sequence corresponding to the filter length according to the sampling timing of the target terminal reference signal segment. Each reference sample value in the reference sampling sequence is multiplied by the corresponding timing position of the own playback feedback path filter coefficient, and the multiplication results are accumulated to obtain the own playback feedback signal sample value. These sample values ​​are then arranged continuously according to the sampling timing of the target terminal reference signal segment to generate the target terminal's own playback feedback signal. For the reference samples required before the start position of the target terminal reference signal segment... The central controller reads the corresponding reference sample value from the target reference signal segment whose sampling position is continuous with the target reference signal segment and whose playback time is earlier. In the case that there is no target reference signal segment whose playback time is earlier, zero-value sampled data is used to supplement the reference sampling sequence required for convolution operation. The central controller determines the time correspondence between the target reference signal segment and the target terminal's mixed acoustic signal based on the playback time identifier of the target reference signal segment, the acquisition time identifier of the target terminal's mixed acoustic signal, and the propagation delay represented by the target terminal's own playback feedback path response. Based on the time correspondence, the central controller determines the sampling time offset of its own playback feedback signal relative to the target terminal's mixed acoustic signal.The central controller converts the sampling time offset into a sampling position offset based on the sampling rate, and corrects the sampling time reference of its own playback feedback signal based on the sampling position offset. For the remaining time offset less than the sampling interval, it corrects it through interpolation between adjacent self-playing feedback signal sample values. Based on the acquisition time range and number of sampling points of the target terminal's mixed acoustic signal, the central controller shifts, truncates, or adds zeros to the self-playing feedback signal, ensuring that it has the same sampling rate, number of sampling points, and sampling time range as the target terminal's mixed acoustic signal, and that each sampling position of the self-playing feedback signal corresponds one-to-one with each sampling position of the target terminal's mixed acoustic signal. When forming the sound field interference residual signal, the central controller subtracts the corresponding self-playing feedback signal sample value from the target terminal's mixed acoustic signal sample value for each sampling position, and arranges the subtracted sample values ​​continuously according to the sampling time sequence to form the sound field interference residual signal. The sound field interference residual signal includes the leakage acoustic components and environmental noise from each source terminal acting on the target terminal, retained after filtering out the self-playing feedback signal from the target terminal's mixed acoustic signal.

[0030] The present invention is further configured to determine the leakage acoustic component of the source terminal acting on the target terminal based on the source-end reference signal segment and the acoustic field interference residual signal, including: Multi-band filtering is performed on the acoustic interference residual signal to obtain multiple frequency band residual signals. The short-time average power of each frequency band residual signal is calculated, and each short-time average power is compared with the dynamic decision threshold of the corresponding frequency band. When the short-time average power of the residual signal in at least one frequency band is greater than the dynamic decision threshold of the corresponding frequency band, the source-end reference signal segment is used as the reference input signal of the adaptive filter, and the acoustic field interference residual signal is used as the desired response signal of the adaptive filter. The path prediction signal is generated based on the current filter coefficients of the adaptive filter and the source-end reference signal segment. The path identification error signal between the path prediction signal and the desired response signal is determined. Based on the source-end reference signal segment and the path identification error signal, the filter coefficients of the adaptive filter are updated using the least mean square algorithm. The updated filter coefficients are then determined as the acoustic leakage path response from the source terminal to the target terminal. From the target immersive sound field reference signal corresponding to and associated with the source terminal, extract the reference signal segment whose playback time identifier is later than the current processing time of the central controller to form the source terminal reference signal segment to be played. Convolve the source terminal reference signal segment to be played with the acoustic leakage path response to generate the predicted leakage signal of the source terminal acting on the target terminal, and determine the predicted leakage signal as the leakage acoustic component. Specifically, the central controller determines the target terminal based on the terminal identifier of the target terminal associated with the sound field interference residual signal, and determines the source terminal acting on the target terminal based on the sound field interference correspondence. For each source terminal and target terminal, read the acquired sound field interference residual signal. The time range has a time-corresponding source-end reference signal segment; the central controller configures multiple frequency band filters according to the working frequency band of the bone conduction terminal, inputs the sound field interference residual signal into each frequency band filter respectively, retains the signal components within the corresponding frequency band range and attenuates the signal components outside the corresponding frequency band range, obtaining multiple frequency band residual signals, performs frame processing on each frequency band residual signal according to the preset analysis frame length, performs square processing and accumulation processing on the sampled values ​​in each frequency band residual signal, and performs average processing on the accumulation result according to the number of sampling points to obtain the short-time average power of each frequency band residual signal; the central controller obtains the historical data of the source terminal that has not played the target immersive sound field reference signal. Historical short-time average power of residual signals in each frequency band is extracted from historical data. Based on the distribution and range of variation of the historical short-time average power, the noise power benchmark and decision margin for the corresponding frequency band are determined. The dynamic decision threshold for the corresponding frequency band is then determined based on the noise power benchmark and decision margin. The dynamic decision threshold for each frequency band is adjusted synchronously based on the updated historical short-time average power. The central controller compares the short-time average power of the residual signals in each frequency band with the dynamic decision threshold for the corresponding frequency band. If the short-time average power of the residual signals in each frequency band is not greater than the dynamic decision threshold for the corresponding frequency band, the current filter coefficients of the adaptive filter remain unchanged. This process continues until at least one frequency band is reached. When the short-time average power of the differential signal is greater than the dynamic decision threshold of the corresponding frequency band, the source-end reference signal segment is used as the reference input signal of the adaptive filter, and the acoustic interference residual signal is used as the desired response signal of the adaptive filter. The central controller performs time alignment of the reference input signal and the desired response signal according to the playback time marker of the source-end reference signal segment, the acquisition time marker of the acoustic interference residual signal, and the acoustic propagation delay from the source terminal to the target terminal. It determines the filter length of the adaptive filter according to the acoustic propagation delay range from the source terminal to the target terminal and the preset acoustic leakage path response duration, and reads the current filter coefficients of the adaptive filter.The central controller extracts the reference input sampling sequence corresponding to the filter length sequentially according to the sampling timing of the source-end reference signal segment. Each sample value in the reference input sampling sequence is multiplied with the current filter coefficient at the corresponding timing position, and the results are accumulated to generate a path prediction signal. The difference between the path prediction signal and the sample values ​​at the same sampling position in the desired response signal is determined, and these differences are arranged continuously according to the sampling timing to form a path identification error signal. Based on the source-end reference signal segment and the path identification error signal, the central controller updates the filter coefficients of the adaptive filter using the least mean square algorithm. Based on the path identification error signal sample values ​​at each sampling position, the sample values ​​in the reference input sampling sequence, and the preset update step size, the adjustment amount of each filter coefficient is determined, and the corresponding filter coefficients are updated using these adjustment amounts until the change in the path identification error signal does not exceed a preset convergence threshold. The updated filter coefficients at the convergence state are then determined as the acoustic leakage path response from the source terminal to the target terminal. The central controller determines the target immersive sound field used to associate and store the source terminal based on the terminal identifier of the source terminal. A circular reference signal buffer is established, and reference signal data units formed by the target immersive sound field reference signal corresponding to the source terminal are read. The playback time identifier of each reference signal data unit is compared with the current processing time of the central controller. Reference signal data units with playback time identifiers later than the current processing time of the central controller are selected. Continuous audio sampling data is extracted from the selected reference signal data units according to the playback sequence to form the source-end reference signal segment to be played. The central controller performs a convolution operation on the source-end reference signal segment to be played and the acoustic leakage path response. According to the sampling sequence of the source-end reference signal segment to be played, the reference sampling sequence corresponding to the filter length of the acoustic leakage path response is extracted sequentially. Each sample value in the reference sampling sequence to be played is multiplied with the filter coefficient at the corresponding time position in the acoustic leakage path response. The results of each multiplication are accumulated to obtain the predicted leakage signal sampling value. The predicted leakage signal sampling values ​​are arranged continuously according to the sampling sequence to generate the predicted leakage signal of the source terminal acting on the target terminal. The predicted leakage signal is determined as the leakage acoustic component of the source terminal acting on the target terminal.

[0031] The present invention is further configured to generate a corresponding sound field cancellation signal based on the leakage acoustic component, including: The target terminal's own playback feedback path response is determined as the offset playback path response, and a secondary path model is established based on the offset playback path response. The reference signal segment from the source to be played is input into the cancellation filter to generate a candidate sound field cancellation signal. The candidate sound field cancellation signal is then filtered using a secondary path model to obtain the predicted cancellation output signal corresponding to the microphone position of the candidate sound field cancellation signal at the target terminal. The predicted cancellation output signal is superimposed with the predicted leakage signal to form a cancellation error signal; the secondary path model is used to filter the reference signal segment at the source end to be played to form a secondary path reference signal; based on the secondary path reference signal and the cancellation error signal, the filter coefficients of the cancellation filter are updated using the filter-x least mean square algorithm. The updated cancellation filter filters the reference signal segment at the source end to be played, generating a sound field cancellation signal. Specifically, the central controller determines the source and target terminals corresponding to the leaked acoustic component based on the terminal identifiers of the source and target terminals associated with the leaked acoustic component, and reads the predicted leakage signal as the leaked acoustic component and the reference signal segment at the source end to be played used to generate the predicted leakage signal. The central controller retrieves the feedback path calibration record based on the terminal identifier of the target terminal, reads the self-playing feedback path filter coefficients of the target terminal, and uses the self-playing feedback path filter coefficients to characterize the self-playing feedback path response of the target terminal, and converts the self-playing feedback of the target terminal into a signal. The path response is determined as the cancellation playback path response, and a secondary path model is established based on the cancellation playback path response, ensuring that the model coefficients, filter length, and sampling rate of the secondary path model are consistent with the cancellation playback path response. The central controller configures cancellation filters for each source and target terminal, determines the filter length of the cancellation filter based on the duration and propagation delay of the acoustic leakage path response and the cancellation playback path response, and configures the current filter coefficients of the cancellation filter. If historical cancellation filter coefficients for the corresponding source and target terminals are stored, the historical cancellation filter coefficients are used as the current filter coefficients of the cancellation filter. The central controller inputs the reference signal segment from the source to be played into the cancellation path. The cancellation filter, according to the sampling timing of the reference signal segment at the source end to be played, sequentially extracts the reference sampling sequence corresponding to the filter length of the cancellation filter. Each sample value in the reference sampling sequence is multiplied by the current filter coefficient at the corresponding timing position in the cancellation filter, and the results of each multiplication are accumulated to obtain the candidate sound field cancellation signal sample values. These candidate sound field cancellation signal sample values ​​are then arranged continuously according to the sampling timing to generate the candidate sound field cancellation signal. The central controller uses the secondary path model to filter the candidate sound field cancellation signal. According to the sampling timing of the candidate sound field cancellation signal, it sequentially extracts the candidate cancellation sampling sequence corresponding to the filter length of the secondary path model, and then... Each sampled value in the desampling sequence is multiplied with the model coefficients at the corresponding time positions in the secondary path model, and the results of each multiplication are accumulated to obtain the predicted cancellation output signal corresponding to the microphone position of the candidate sound field cancellation signal at the target terminal. The central controller performs time alignment on the predicted cancellation output signal and the predicted leakage signal based on the playback time identifier of the reference signal segment at the source end to be played, the propagation delay corresponding to the acoustic leakage path response, and the propagation delay corresponding to the cancellation playback path response. The time-aligned predicted cancellation output signal and the sampled values ​​of the predicted leakage signal at the same sampling position are superimposed, and the superposition results are arranged continuously according to the sampling time sequence to form the cancellation error signal.The central controller uses the secondary path model to filter the reference signal segment at the source end to be played. According to the sampling timing of the reference signal segment at the source end to be played, it sequentially extracts the reference sampling sequence corresponding to the filter length of the secondary path model. Each sample value in the reference sampling sequence is multiplied by the model coefficients at the corresponding timing positions in the secondary path model, and the results of each multiplication are accumulated to form the secondary path reference signal. Based on the secondary path reference signal and the cancellation error signal, the central controller updates the filter coefficients of the cancellation filter using the filter-x least mean square algorithm. Based on the sample values ​​of the cancellation error signal corresponding to each sampling position, the sample values ​​in the secondary path reference signal, and the preset update step size, the central controller determines the coefficients of each step of the cancellation filter. The filter coefficients are adjusted accordingly, and the corresponding filter coefficients are updated in the direction of reducing the cancellation error signal. This process of generating candidate sound field cancellation signals, generating predicted cancellation output signals, determining cancellation error signals, and updating filter coefficients is repeated until the degree of change in the cancellation error signal does not exceed a preset convergence threshold. The central controller filters the reference signal segment to be played using the updated cancellation filter. According to the sampling timing of the reference signal segment to be played, it convolves each sampled value in the reference signal segment with the filter coefficients at the corresponding timing positions in the updated cancellation filter to obtain the sound field cancellation signal sampled values. These sampled values ​​are then arranged consecutively according to the sampling timing to generate the corresponding sound field cancellation signal.

[0032] The present invention is further configured such that S6 includes: Based on the playback time identifier of the reference signal segment of the source to be played used to generate each sound field cancellation signal and the output sampling position corresponding to each sample value in each sound field cancellation signal, the target playback time corresponding to each sample value in each sound field cancellation signal under the reference clock of the central controller is determined. Based on the target playback time corresponding to each sample value, time alignment is performed on each sound field cancellation signal acting on the same target terminal; The sampled values ​​of each time-aligned sound field cancellation signal at the same target playback time are summed to generate the total sound field cancellation signal of the target terminal, and the target playback time range corresponding to the total sound field cancellation signal is determined. Based on the target playback time range and the playback time identifier of the target immersive sound field reference signal of the target terminal, extract the target terminal reference signal segment corresponding to the playback time range from the target immersive sound field reference signal; The total sound field cancellation signal is mixed with the target end reference signal segment on a sample-by-sample basis to generate the sound field driving signal of the target terminal; Based on the target terminal's terminal identifier, the sound field drive signal is sent to the target terminal for playback. Specifically, the central controller groups the sound field cancellation signals according to the terminal identifier of the target terminal associated with each sound field cancellation signal, grouping the sound field cancellation signals acting on the same target terminal into the same sound field cancellation signal group. The central controller compares the sampling rate of each sound field cancellation signal in the same sound field cancellation signal group with the playback sampling rate of the target terminal. If the sampling rate of the sound field cancellation signal is inconsistent with the playback sampling rate of the target terminal, the sampling rate of the sound field cancellation signal is converted according to the playback sampling rate of the target terminal, so that the sound field cancellation signals acting on the same target terminal have the same sampling rate and sampling interval. The central controller reads... The playback time identifier of the reference signal segment used to generate each sound field cancellation signal is taken. The output sampling position corresponding to each sample value in each sound field cancellation signal is determined according to the sampling timing of each sound field cancellation signal. Based on the number of sampling points preceding each output sampling position and the sampling rate of the sound field cancellation signal, the time offset of each sample value relative to the playback time identifier of the reference signal segment is determined. The central controller uses its reference clock as a unified time base. Based on the playback time identifier of the reference signal segment and the time offset corresponding to each sample value, the target playback time of each sample value in each sound field cancellation signal under the central controller's reference clock is determined. The central controller then determines the target playback time corresponding to each sample value under the central controller's reference clock based on the playback time identifier of the reference signal segment and the time offset corresponding to each sample value. A unified target playback time axis is established for each sampled value in the sound field cancellation signals of the target terminal corresponding to the target playback time. Each sampled value is then assigned to its corresponding sampling position within the unified target playback time axis, ensuring that sampled values ​​with the same target playback time are located at the same sampling position. For time offsets smaller than the sampling interval, the central controller corrects this through interpolation between adjacent sampled values ​​and assigns zero-value sampled data to sampling positions in the unified target playback time axis that do not belong to the target playback time of the corresponding sound field cancellation signal, thus completing the time alignment of the sound field cancellation signals acting on the same target terminal. The central controller then applies the time-aligned sound field cancellation signals to the same target terminal according to the sampling position order in the unified target playback time axis. The sampled values ​​corresponding to the playback time are summed to obtain the total sound field cancellation signal sampled value corresponding to each target playback time. The total sound field cancellation signal sampled values ​​are then arranged continuously according to the target playback time sequence to generate the total sound field cancellation signal for the target terminal. The central controller determines the target playback time range corresponding to the total sound field cancellation signal based on the target playback time corresponding to the starting sampling position in the total sound field cancellation signal, the number of sampling points contained in the total sound field cancellation signal, and the playback sampling rate of the target terminal. The central controller determines the circular reference signal buffer corresponding to the target terminal based on the terminal identifier of the target terminal and reads the reference signal data unit formed by the target immersive sound field reference signal of the target terminal from the circular reference signal buffer corresponding to the target terminal.The central controller determines the playback time range corresponding to each reference signal data unit based on the playback time identifier of each reference signal data unit, and selects reference signal data units whose playback time range corresponds to the target playback time range corresponding to the total sound field cancellation signal. Based on the target playback time range corresponding to the total sound field cancellation signal and the playback time identifier of the target immersive sound field reference signal of the target terminal, the central controller determines the start and end extraction positions of the audio sampling data in the selected reference signal data units. It extracts audio sampling data located within the target playback time range from the target immersive sound field reference signal of the target terminal, and arranges them continuously according to the playback sequence represented by the playback time identifier, forming a target-end reference signal segment whose playback time range corresponds to the target playback time range corresponding to the total sound field cancellation signal. The central controller compares the sampling rate of the total sound field cancellation signal with the sampling rate of the target-end reference signal segment. If the sampling rates are inconsistent, it performs sampling rate conversion according to the playback sampling rate of the target terminal, and then adjusts the sampling rate according to the target playback time range corresponding to the total sound field cancellation signal. The system aligns the total sound field cancellation signal and the target reference signal segment with the playback time identifier corresponding to the time range and target playback time, ensuring a one-to-one correspondence between the sampling positions of the total sound field cancellation signal and the sampling positions of the target reference signal segment with the same target playback time. The central controller mixes the total sound field cancellation signal and the target reference signal segment point by point according to the order of the target playback time. It sequentially reads the sample values ​​of the total sound field cancellation signal and the target reference signal segment corresponding to the same sampling position, performs mixing processing on the sample values ​​of the total sound field cancellation signal and the target reference signal segment, and arranges the mixing results continuously according to the target playback time sequence to generate the sound field drive signal for the target terminal. The central controller sends the sound field drive signal to the target terminal according to the terminal identifier of the target terminal. After receiving the sound field drive signal, the target terminal, according to its terminal clock synchronized with the reference clock of the central controller, sequentially outputs each sample value in the sound field drive signal at the target playback time corresponding to each sample value, thus playing the sound field drive signal.

[0033] Specifically, a bone conduction multi-terminal differentiated playback control and distribution method is applied to a playback control system. The playback control system includes a central controller and multiple playback terminals communicatively connected to the central controller. These playback terminals include at least multiple bone conduction terminals, as well as Bluetooth playback terminals formed by Bluetooth headsets and Bluetooth speakers, and wired local headphones. The central controller provides multiple different audio data and distributes these different audio data to their respective playback terminals, enabling each playback terminal to independently play audio data different from other playback terminals. In this embodiment, the central controller is integrated into the bone conduction terminal, which serves as the central control node. This bone conduction terminal, acting as the central control node, possesses bone conduction playback capabilities and acts as a multi-functional audio gateway, establishing communication connections with Bluetooth headsets, Bluetooth speakers, wired local headphones, and bone conduction terminals distinct from the central control node. The bone conduction terminal, acting as the central control node, manages communication, processes audio data, and controls playback for each playback terminal through the central controller. The central controller receives audio input from external audio sources, including mobile phones, music players, and dedicated sleep aid apps. It identifies the received audio input as the audio data source to be distributed. The central controller distinguishes between different audio streams from the same external source or multiple audio streams from different external sources, and distributes the different audio streams to their respective playback terminals according to their audio distribution configurations. Each playback terminal then independently plays the corresponding audio data according to its configuration. When constructing the sleep aid sound field, the central controller distributes the audio data corresponding to the guided meditation voice to the user's bone conduction terminal and the audio data corresponding to the ambient white noise to a Bluetooth speaker. This allows the bone conduction terminal to play individual audio content tailored to the user, while the Bluetooth speaker plays ambient audio content for the shared space, thus forming a multi-layered sleep aid sound field that includes individual and public audio areas. The central controller's local hardware is equipped with a multi-channel audio processing unit. This unit is used to simultaneously receive, decode, and process multiple independent audio channels. Each audio channel carries a sound element signal, which includes rain sound signals, wind sound signals, stream sound signals, preset frequency harmonic signals, non-rhythmic music signals, and guiding voice signals. The central controller configures playback parameters for each audio channel. These playback parameters include at least sound element playback gain and sound element playback timing. The sound element playback gain is used to adjust the playback amplitude of the corresponding sound element signal, and the sound element playback timing is used to control the playback start time, playback end time, and playback progress of the corresponding sound element signal.The central controller determines the sound element signals to be allocated to each playback terminal based on the sound element affiliation relationships. It adjusts the amplitude of each sound element signal according to the sound element playback gain and configures its playback position according to the playback sequence. The central controller uses a multi-channel audio processing unit to mix sound element signals belonging to the same playback terminal in the digital domain, forming the output signal for the corresponding playback terminal. It also adjusts the amplitude and playback position of each sound element signal based on changes in playback parameters. By independently configuring the sound element playback gain and playback sequence of different sound element signals, the central controller mixes thunder, rain, and wind signals according to preset amplitude relationships and playback sequences, creating sleep-aid sound content with different sound element compositions, amplitude relationships, and playback sequences. This sleep-aid sound content is then distributed to the corresponding playback terminals to construct a dynamic sleep-aid sound field corresponding to the sleep-aid sound configurations of different users.

[0034] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A bone conduction multi-terminal differentiated broadcast control and distribution method, characterized in that, A broadcast control system including a central controller and multiple bone conduction terminals communicatively connected to the central controller, the method comprising: S1: Based on the immersive sound field configuration of each bone conduction terminal, multiple sound element signals are combined differentially to form the target immersive sound field reference signal for each bone conduction terminal; S2: Send the target immersive sound field reference signal to the corresponding bone conduction terminal respectively, store the target immersive sound field reference signal and the playback time identifier, establish the sound field interference correspondence relationship according to the acoustic influence direction between each bone conduction terminal, and determine the source terminal and target terminal in each sound field interference correspondence relationship; S3: Collect mixed acoustic signals through the microphones of each bone conduction terminal, and extract the source end reference signal segment and target end reference signal segment corresponding to the time of the mixed acoustic signal from the associated stored target immersive sound field reference signal based on the acquisition time identifier of the mixed acoustic signal and the playback time identifier of the target immersive sound field reference signal. S4: Determine the self-play feedback signal based on the target end reference signal segment and the target terminal's self-play feedback path response, filter out the self-play feedback signal from the target terminal's mixed acoustic signal, and form a sound field interference residual signal; S5: Determine the leakage acoustic component of the source terminal acting on the target terminal based on the source terminal reference signal segment and the sound field interference residual signal, and generate the corresponding sound field cancellation signal based on the leakage acoustic component; S6: Combine the sound field cancellation signals acting on the same target terminal, mix the combined sound field cancellation signal with the target immersive sound field reference signal of the target terminal to form a sound field driving signal, and send the sound field driving signal to the target terminal for playback.

2. The bone conduction multi-terminal differentiated broadcast control and distribution method according to claim 1, characterized in that, The immersive sound field configuration includes the sound element affiliation, sound element playback gain, and sound element playback timing. Differentiated combination of multiple sound element signals, including: Based on the attribution relationship of sound elements, determine the sound element signal corresponding to each bone conduction terminal; Based on the playback sequence of sound elements, extract the corresponding sound element signal segments from each sound element signal; Adjust the gain of each sound element signal segment according to the playback gain of the sound element; The gain-adjusted signal segments of each sound element corresponding to the same bone conduction terminal are mixed to generate the target immersive sound field reference signal for the corresponding bone conduction terminal.

3. The bone conduction multi-terminal differentiated broadcast control and distribution method according to claim 1, characterized in that, The target immersive sound field reference signal is associated and stored with the playback time identifier, including: Configure a ring-shaped reference signal buffer corresponding to each bone conduction terminal in the central controller; The immersive sound field reference signals of each target are processed in frames to obtain continuous reference signal frames. The audio sampling data, terminal identifier and playback time identifier of each reference signal frame are associated and encapsulated into a reference signal data unit. Based on the terminal identifier, each reference signal data unit is stored in the corresponding circular reference signal buffer, and the storage timing of each reference signal data unit in the circular reference signal buffer is determined according to the playback time identifier.

4. The bone conduction multi-terminal differentiated broadcast control and distribution method according to claim 1, characterized in that, The acoustic interference correspondence is established based on the acoustic influence direction between each bone conduction terminal, including: An adjacency relation matrix is ​​constructed to characterize the correspondence of sound field interference between bone conduction terminals. The adjacency relation matrix is ​​a Boolean square matrix of order equal to the number of bone conduction terminals. The row index of the adjacency matrix corresponds one-to-one with the terminal identifier of the source terminal, and the column index corresponds one-to-one with the terminal identifier of the target terminal. Each matrix element is used to record the existence status of the acoustic leakage path to be canceled between the source terminal indicated by the corresponding row index and the target terminal indicated by the corresponding column index.

5. The bone conduction multi-terminal differentiated broadcast control and distribution method according to claim 3, characterized in that, S3 includes: The mixed acoustic signal at the location of the microphone is continuously sampled by the microphone of each bone conduction terminal, and the sampled mixed acoustic signal is processed into a frame to form a mixed signal frame. The audio sampling data, acquisition time identifier, and terminal identifier of each mixed signal frame are associated and encapsulated into a mixed signal data unit, and the mixed signal data unit is uploaded to the central controller. The acquisition time identifier is generated based on the terminal clock synchronized with the reference clock of the central controller. Based on the signal transmission delay and acoustic propagation delay, delay compensation is performed on the acquisition time marker of each mixed signal frame to determine the reference signal extraction time range corresponding to each mixed signal frame; The source terminal and target terminal are determined based on the correspondence of sound field interference, and reference signal data units are read from the corresponding circular reference signal buffer according to the terminal identifiers of the source terminal and target terminal. Based on the playback time identifier of each reference signal data unit, select the reference signal data unit that corresponds to the playback time range and the reference signal extraction time range, and extract audio sampling data from the selected reference signal data unit to form the source end reference signal segment and the target end reference signal segment respectively.

6. The bone conduction multi-terminal differentiated broadcast control and distribution method according to claim 1, characterized in that, The self-playback feedback path response is obtained through offline calibration, which includes: According to the preset calibration order, the bone conduction terminal to be calibrated is determined from each bone conduction terminal in sequence, the bone conduction terminal to be calibrated is switched to offline calibration state, the bone conduction terminal to be calibrated is controlled to play the preset broadband test signal, and the acoustic response signal is synchronously collected through the microphone set on the bone conduction terminal to be calibrated. The preset broadband test signal is used as the system identification input, and the acoustic response signal is used as the system identification output. The self-play feedback path filter coefficients of the bone conduction terminal to be calibrated are obtained through linear time-invariant system identification. The terminal identifier of the bone conduction terminal to be calibrated is associated with its own playback feedback path filter coefficient to form a feedback path calibration record. The feedback path calibration record is stored, and the own playback feedback path filter coefficient is used to characterize the own playback feedback path response of the bone conduction terminal to be calibrated.

7. The bone conduction multi-terminal differentiated broadcast control and distribution method according to claim 1, characterized in that, The acoustic field interference residual signal is formed, including: The target terminal reference signal segment is convolved with the target terminal's own playback feedback path response to generate the target terminal's own playback feedback signal. Based on the time correspondence between the target end reference signal segment and the mixed acoustic signal, the sampling time reference of its own playback feedback signal is corrected so that each sampling position of its own playback feedback signal corresponds one-to-one with each sampling position of the mixed acoustic signal. For each sampling location, the corresponding self-play feedback signal sampling value is subtracted from the mixed acoustic signal sampling value of the target terminal to form the sound field interference residual signal.

8. The bone conduction multi-terminal differentiated broadcast control and distribution method according to claim 1, characterized in that, The leakage acoustic component of the source terminal acting on the target terminal is determined based on the source-end reference signal segment and the acoustic field interference residual signal, including: Multi-band filtering is performed on the acoustic interference residual signal to obtain multiple frequency band residual signals. The short-time average power of each frequency band residual signal is calculated, and each short-time average power is compared with the dynamic decision threshold of the corresponding frequency band. When the short-time average power of the residual signal in at least one frequency band is greater than the dynamic decision threshold of the corresponding frequency band, the source-end reference signal segment is used as the reference input signal of the adaptive filter, and the acoustic field interference residual signal is used as the desired response signal of the adaptive filter. The path prediction signal is generated based on the current filter coefficients of the adaptive filter and the source-end reference signal segment. The path identification error signal between the path prediction signal and the desired response signal is determined. Based on the source-end reference signal segment and the path identification error signal, the filter coefficients of the adaptive filter are updated using the least mean square algorithm. The updated filter coefficients are then determined as the acoustic leakage path response from the source terminal to the target terminal. From the target immersive sound field reference signal corresponding to and associated with the source terminal, extract the reference signal segment whose playback time identifier is later than the current processing time of the central controller to form the source terminal reference signal segment to be played; perform convolution operation on the source terminal reference signal segment to be played and the acoustic leakage path response to generate the predicted leakage signal of the source terminal acting on the target terminal, and determine the predicted leakage signal as the leakage acoustic component.

9. A bone conduction multi-terminal differentiated broadcast control and distribution method according to claim 8, characterized in that, Based on the leaked acoustic components, a corresponding sound field cancellation signal is generated, including: The target terminal's own playback feedback path response is determined as the offset playback path response, and a secondary path model is established based on the offset playback path response. The reference signal segment from the source to be played is input into the cancellation filter to generate a candidate sound field cancellation signal. The candidate sound field cancellation signal is then filtered using a secondary path model to obtain the predicted cancellation output signal corresponding to the microphone position of the candidate sound field cancellation signal at the target terminal. The predicted cancellation output signal is superimposed with the predicted leakage signal to form a cancellation error signal; the secondary path model is used to filter the reference signal segment at the source end to be played to form a secondary path reference signal; based on the secondary path reference signal and the cancellation error signal, the filter coefficients of the cancellation filter are updated using the filter-x least mean square algorithm. The updated cancellation filter is used to filter the reference signal segment at the source end of the playback source to generate a sound field cancellation signal.

10. A bone conduction multi-terminal differentiated broadcast control and distribution method according to claim 9, characterized in that, S6 includes: Based on the playback time identifier of the reference signal segment of the source to be played used to generate each sound field cancellation signal and the output sampling position corresponding to each sample value in each sound field cancellation signal, the target playback time corresponding to each sample value in each sound field cancellation signal under the reference clock of the central controller is determined. Based on the target playback time corresponding to each sample value, time alignment is performed on each sound field cancellation signal acting on the same target terminal; The sampled values ​​of each time-aligned sound field cancellation signal at the same target playback time are summed to generate the total sound field cancellation signal of the target terminal, and the target playback time range corresponding to the total sound field cancellation signal is determined. Based on the target playback time range and the playback time identifier of the target immersive sound field reference signal of the target terminal, extract the target terminal reference signal segment corresponding to the playback time range from the target immersive sound field reference signal; The total sound field cancellation signal is mixed with the target end reference signal segment on a sample-by-sample basis to generate the sound field driving signal of the target terminal; Based on the target terminal's identifier, the sound field drive signal is sent to the target terminal for playback.

Images