Multichannel audio playing control method based on DSP operation

By adjusting the bit rate and sampling parameters of the audio stream data in the audio distributor, combining the equalization pipeline structure and unified clock processing, the multi-audio playback outage caused by the difference in speaker performance is solved, and the user experience and audio transmission efficiency are improved.

CN119012079AInactive Publication Date: 2024-11-22SHANGDE LIANSHENG TECH (ZHUHAI) CO LTD
View PDF 4 Cites 0 Cited by

Patent Information

Application Number
CN202411499438.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-10-25
Publication Date
2024-11-22
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

Due to the possible differences in the performance and hardware configuration of the speaker, the problem of multi-audio playback is not synchronized, which reduces the user's actual experience.

Method used

By setting a cache area in the audio distributor, the response time and response rate of the speaker are obtained, and the bit rate and sampling parameters of the audio stream data are adjusted according to the audio content type, ensuring that the decoding efficiency of each speaker is consistent, and a balanced pipeline structure and unified use The clock processes data to ensure smooth audio playback.

Benefits of technology

It effectively solves the problem of multi-audio playback that is not synchronized, improves user experience, ensures audio fluency and playback effect, and optimizes the transmission efficiency of audio content.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119012079A_ABST
    Figure CN119012079A_ABST
Patent Text Reader

Abstract

The invention discloses a multi-channel audio playing control method based on DSP operation, and belongs to the technical field of audio playing, and the method specifically comprises the steps: obtaining parameters of original audio stream data transmitted to an audio distributor by a cloud data server; setting a cache region, feeding back instructions to all the connected sound boxes when original audio stream data is received, obtaining response durations of all the sound boxes, and calculating response rates of all the sound boxes; and when original audio stream data with the duration of T is pre-stored in the cache region, if the response rate V of the loudspeaker box is smaller than the bit rate, compressing the bit rate of the original audio stream data, analyzing the audio data in the cache region, and adjusting the sampling rate and the sampling bit number of the audio stream data according to an analysis result. According to the invention, the transmission audio bit rate is dynamically adjusted according to the response time of the sound boxes, and different optimization strategies are respectively formulated according to the audio types, so that the synchronous smooth playing of the multiple sound boxes is realized, and the user experience is improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of audio playback, and in particular to a multi-channel audio playback control method based on DSP calculation. Background Art

[0002] When users use multi-channel audio playback speakers to play audio files such as symphonies or movies, they can feel as if they are in the scene, which greatly improves the user experience. Multi-channel audio playback speakers can play 5.1-channel audio files or stereo audio files, etc. However, since multi-channel audio playback speakers are expensive and have high requirements for placement, and the home environment of ordinary users does not meet such high requirements for placement, many users eventually give up multi-channel audio playback speakers.

[0003] In a Chinese patent with the patent announcement number of publication number CN107277691A, a cloud-based multi-channel audio playback method, system and audio gateway device are disclosed. The scheme mainly includes dividing a multi-channel audio bit stream into a corresponding number of mono audio bit streams and metadata information; sending the mono audio bit streams and corresponding metadata information belonging to the same subject identifier to the audio gateway device through the network; the audio gateway device synchronously generates corresponding multiple mono audio signals according to the received multiple mono audio bit streams and corresponding metadata information belonging to the same subject identifier; sending the multiple mono audio signals belonging to the same subject identifier to the corresponding wireless speakers through different wireless interfaces, providing high-quality music or video listening effects at low cost, but in actual situations, due to the differences in the performance and hardware configuration of the speakers, the time generated by each speaker when decoding the received audio channels is also different, so the multi-audio playback is not synchronized, thereby reducing the actual user experience. Summary of the invention

[0004] The purpose of the present invention is to provide a multi-channel audio playback control method based on DSP operation to solve the following technical problems: However, in actual situations, due to differences in speaker performance and hardware configuration, the time taken by each speaker to decode the received audio channels is also different, resulting in asynchronous multi-audio playback, which reduces the actual user experience.

[0005] The purpose of the present invention can be achieved through the following technical solutions: A multi-channel audio playback control method based on DSP operation, wherein an audio distributor is connected to a plurality of speakers, characterized in that the method comprises the following steps: S1, obtaining parameters of the original audio stream data transmitted by the cloud data server to the audio distributor; the parameters include bit rate, sampling rate, sampling bit number and number of channels; S2, a buffer area is set in the audio distributor. When the buffer area receives the original audio stream data, a feedback instruction with a data volume of Q is sent to all connected speakers and response signals of all connected speakers are received. The response time of all speakers is obtained and marked as t1, t2, ..., tn, where n is the number of speakers. The response rate of all speakers is calculated according to the calculation formula V=Q / t and marked as V1, V2, ..., Vn; S3, when the buffer area pre-stores the original audio stream data with a duration of T, if there is any speaker whose response rate Vi is less than the bit rate, compress the bit rate of the original audio stream data with a duration of T, set the compressed bit rate as the response rate Vi of the speaker, and send the compressed audio stream data to the speaker; S4, analyzing the original audio stream data in the buffer area, determining whether the original audio within the current duration T is human voice audio or vocal audio, and adjusting the sampling rate and sampling bit number of the original audio stream data according to the type of the original audio within the current duration T.

[0006] As a further solution of the present invention: it is characterized in that the audio distributor is connected to multiple speakers in a manner including but not limited to Bluetooth communication, wireless network communication, wired network communication and wireless transmission communication.

[0007] As a further solution of the present invention: it is characterized in that the audio distributor also includes a data adjustment module for connecting to the buffer area, the data adjustment module contains n audio matrix modules, the audio matrix modules all adopt a balanced pipeline structure, and the balanced pipeline structure is composed of a number of two-to-one selectors connected in series.

[0008] As a further solution of the present invention: it is characterized in that the n audio matrix modules all use a unified clock to select and process data.

[0009] As a further solution of the present invention, it is characterized in that the analysis process of the audio stream data is as follows: The original audio stream data is divided into frames according to a preset frame length to obtain a number of audio frames and marked as Y1, Y2, ..., Yn, where n is a positive integer. Each frame of the audio stream data after the linear prediction coding analysis is used to extract audio features, wherein the audio features include the short-time zero-crossing rate, the LPC prediction coefficient, and the skewness and kurtosis of the LPC prediction residual amplitude spectrum. A feature vector is formed according to the audio features, and the feature vector is brought into a preset recognition support vector model to identify whether each frame of the audio stream data contains human voice. If human voice is contained, the audio frame is marked as a special audio frame and marked as Q1, Q2, ..., Qz, where z is a positive integer. If z / n is greater than or equal to a preset threshold value, the original audio within the duration T is determined to be human voice audio. If z / n is less than the preset threshold value, the original audio within the duration T is determined to be vocal audio.

[0010] As a further solution of the present invention: it is characterized in that if the original audio type within the current duration T is human voice audio, the sampling bit number of the original audio data is preferentially reduced; if the original audio type within the current duration T is vocal audio, the sampling rate of the original audio data is preferentially reduced.

[0011] As a further solution of the present invention: it is characterized in that the specific calculation process of the reduction amount is: Ci'=C-(B-Vi) / S*W; Wi'=W-(B-Vi)*S*C; Among them, B is the bit rate of the original audio stream data, S is the number of channels of the original audio stream data, C is the sampling rate of the original audio stream data, W is the number of sampling bits of the original audio stream data, Vi is the response rate of any speaker, Ci' is the sampling rate of the adjusted audio stream data, and Wi' is the number of sampling bits of the adjusted audio stream data.

[0012] As a further solution of the present invention: it is characterized in that, in the S3, it also includes maintaining the bit rate of the original audio stream data if there is a response rate Vi of any speaker that is greater than or equal to the bit rate.

[0013] As a further solution of the present invention: S3 further includes obtaining the best frequency response interval [M, N] of any speaker, performing spectrum analysis on the original audio stream data in the buffer area, extracting the audio stream data in the best frequency response interval of the speaker and marking it as target audio stream data, and sending the target audio stream data to the speaker. Beneficial effects of the present invention: The present invention acquires and caches original audio stream data of a certain length by setting a buffer area, and sends a feedback instruction with a data volume of Q to all connected speakers during the caching process and receives response signals from all connected speakers, thereby obtaining the response duration of all speakers, and calculating the response rate of all speakers based on the response duration; it can be understood that the response time represents the hardware performance of the speakers, and the faster the response rate, the higher the decoding efficiency. When the buffer area pre-stores original audio stream data of a length of T, if there is any speaker whose response rate Vi is less than the bit rate, the bit rate of the original audio stream data of a length of T is compressed, and the compressed bit rate is set to the bit rate of the speaker. The response rate Vi is used, and the compressed audio stream data is sent to the speaker. It can be understood that when the audio distributor sends the original audio stream data, the decoding efficiency of the speaker is not enough to support the current original audio stream bit rate, resulting in a delay for the speaker relative to other speakers. Therefore, the bit rate of the audio data is reduced and analyzed based on the extracted audio stream data, and it is determined whether the original audio within the current duration T is human voice audio or vocal audio, and based on the determination result, the sampling rate and sampling bit number of the original audio data in the buffer area are adjusted, which not only ensures the smooth playback of the audio, but also optimizes the audio playback effect according to the characteristics of the audio content, thereby improving the user experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] The present invention will be further described below in conjunction with the accompanying drawings.

[0015] Figure 1 The present invention is a flowchart of a multi-channel audio playback control method based on DSP operation. DETAILED DESCRIPTION

[0016] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0017] See also Figure 1 As shown, the present invention is a multi-channel audio playback control method based on DSP operation, wherein an audio distributor is connected to a plurality of speakers, and is characterized in that the method comprises the following steps: S1, obtaining parameters of the original audio stream data transmitted by the cloud data server to the audio distributor; the parameters include bit rate, sampling rate, sampling bit number and number of channels; S2, a buffer area is set in the audio distributor. When the buffer area receives the original audio stream data, a feedback instruction with a data volume of Q is sent to all connected speakers and response signals of all connected speakers are received. The response time of all speakers is obtained and marked as t1, t2, ..., tn, where n is the number of speakers. The response rate of all speakers is calculated according to the calculation formula V=Q / t and marked as V1, V2, ..., Vn; S3, when the buffer area pre-stores the original audio stream data with a duration of T, if there is any speaker whose response rate Vi is less than the bit rate, compress the bit rate of the original audio stream data with a duration of T, set the compressed bit rate as the response rate Vi of the speaker, and send the compressed audio stream data to the speaker; S4, analyzing the original audio stream data in the buffer area, determining whether the original audio within the current duration T is human voice audio or vocal audio, and adjusting the sampling rate and sampling bit number of the original audio stream data according to the type of the original audio within the current duration T.

[0018] The present invention acquires and caches audio stream data of a certain length in advance by setting a buffer area, and sends a feedback instruction with a data volume of Q to all connected speakers during the caching process and receives response signals from all connected speakers, thereby obtaining the response duration of all speakers, and calculating the response rate of all speakers based on the response duration; it can be understood that the response time represents the hardware performance of the speakers, and the faster the response rate, the higher the decoding efficiency; when the buffer area pre-stores original audio stream data of a length of T, if there is any speaker with a response rate Vi that is less than the bit rate, the bit rate of the original audio stream data of a length of T is compressed, and the compressed bit rate is set as the bit rate of the speaker. The response rate Vi of the speaker is adjusted, and the compressed audio stream data is sent to the speaker. It can be understood that when the audio distributor sends the original audio stream data, the decoding efficiency of the speaker is not enough to support the current original video stream bit rate, resulting in a delay for the speaker relative to other speakers. Therefore, the bit rate of the audio data is reduced and analyzed based on the extracted audio stream data, and it is determined whether the original audio within the current duration T is human voice audio or vocal audio, and based on the determination result, the sampling rate and sampling bit number of the original audio data in the buffer area are adjusted, which not only ensures the smooth playback of the audio, but also optimizes the audio playback effect according to the characteristics of the audio content, thereby improving the user experience.

[0019] It is understandable that in actual situations, due to differences in the performance and hardware configuration of the speakers, the time generated by each speaker when decoding the received audio channels is also different, resulting in different actual playback rates of each speaker when the same audio stream data is distributed at the same time. This is determined based on the performance and hardware configuration of each speaker. Therefore, there will be a phenomenon of asynchronous multi-audio playback, which reduces the actual user experience. In the present invention, first, a feedback instruction with a data volume of Q is sent to all connected speakers during the caching process, and response signals from all connected speakers are received to obtain the response time of all speakers, and the response rate of all speakers is calculated based on the response time. It is understandable that the response time represents the hardware performance of the speaker. The faster the response rate, the more likely the speaker can complete real-time audio stream playback. When the cache area pre-stores the original audio stream data with a duration of T, audio transmission is performed. By pre-caching the audio stream, the delay and jitter in the network transmission process can be effectively alleviated. When the network bandwidth fluctuates, the cache can provide a continuous audio stream to avoid audio interruption or freeze due to network instability. Audio transmission needs to be targeted according to the performance of each speaker, while ensuring that the transmission time nodes corresponding to each speaker are consistent. Therefore, the audio distributor also includes a data adjustment module for connecting to the cache area. The data adjustment module includes n audio matrix modules. The audio matrix modules all adopt a balanced pipeline structure. The balanced pipeline structure is composed of several two-choice selectors connected in series, and the n audio matrix modules all use a unified clock to select and process data to ensure that the output time of the transmission audio stream corresponding to each speaker is the same, avoiding the phenomenon of asynchronous playback of multiple audios due to different output times.

[0020] In a preferred embodiment of the present invention, the audio distributor is connected to the multiple speakers via, but not limited to, Bluetooth communication, wireless network communication, wired network communication, and wireless transmission communication.

[0021] In another preferred embodiment of the present invention, the audio distributor also includes a data adjustment module for connecting to the buffer area, the data adjustment module includes n audio matrix modules, the audio matrix modules all adopt a balanced pipeline structure, and the balanced pipeline structure is composed of a number of two-to-one selectors connected in series.

[0022] It is understandable that an equalization pipeline structure is adopted, which is composed of two-to-one selectors connected in series, so that each channel audio signal enters the output through the same number of selectors, ensuring that the n-channel audio delay is the same and avoiding the occurrence of asynchronous transmission of audio signals in different channels.

[0023] In another preferred embodiment of the present invention, it is characterized in that the n audio matrix modules all use a unified clock to select and process data.

[0024] It can be understood that the same clock is used to select and process data at the same time so that each of the n audio channels can be output after √n system clock cycles, ensuring that the delay of the n-channel audio is the same.

[0025] In another preferred embodiment of the present invention, the analysis process of the audio stream data is as follows: The original audio stream data is divided into frames according to a preset frame length to obtain a number of audio frames and marked as Y1, Y2, ..., Yn, where n is a positive integer. Each frame of the audio stream data after the linear prediction coding analysis is used to extract audio features, wherein the audio features include the short-time zero-crossing rate, the LPC prediction coefficient, and the skewness and kurtosis of the LPC prediction residual amplitude spectrum. A feature vector is formed according to the audio features, and the feature vector is brought into a preset recognition support vector model to identify whether each frame of the audio stream data contains human voice. If human voice is contained, the audio frame is marked as a special audio frame and marked as Q1, Q2, ..., Qz, where z is a positive integer. If z / n is greater than or equal to a preset threshold value, the original audio within the duration T is determined to be human voice audio. If z / n is less than the preset threshold value, the original audio within the duration T is determined to be vocal audio.

[0026] By dividing the original audio stream data into frames according to the preset frame length, each frame can be analyzed independently, which improves the accuracy and reliability of audio feature extraction. This frame-level processing method can effectively capture the dynamic changes in the audio signal, making the difference between human voice and background sound more obvious. By extracting features such as short-time zero-crossing rate, LPC (linear predictive coding) prediction coefficient, and skewness and kurtosis of LPC prediction residual amplitude spectrum, a more comprehensive audio feature description can be obtained. These features not only help to identify human voices in audio, but also use support vector machines to classify audio frames, which can effectively cope with high-dimensional feature spaces and have strong classification capabilities for complex data. By extracting effective audio features, namely short-time zero-crossing rate, P-order LPC prediction coefficient, and skewness and kurtosis of LPC prediction residual amplitude spectrum to form a feature vector, and using machine learning to identify human voices in audio, high-precision and high-confidence recognition of human voices in audio can be achieved.

[0027] In another preferred embodiment of the present invention, if the original audio type within the current duration T is human voice audio, the sampling bit number of the original audio data is preferentially reduced; if the original audio type within the current duration T is vocal audio, the sampling rate of the original audio data is preferentially reduced.

[0028] In audio processing and transmission, human voice audio and vocal audio present their own unique characteristics. Therefore, it is necessary to make targeted considerations when adjusting the sampling rate and the number of sampling bits, in order to achieve effective data compression and transmission while ensuring audio quality. Human voice audio refers to audio signals containing speech, and its main purpose is to transmit information, so it has extremely high requirements for the clarity and intelligibility of speech. The frequency range of speech signals in human voice audio is usually between 300 Hz and 3400 Hz, which contains the key information required for clear pronunciation and intonation changes. Human voice audio usually needs to retain a wide dynamic range in order to show the strength changes and emotional expression of the sound. The vowels and consonants in the speech have a significant impact on its intelligibility. Maintaining the clarity of speech is particularly important. Prioritizing the reduction of the number of sampling bits can effectively maintain the clarity of speech. This is because the reduction of the number of sampling bits mainly affects the dynamic range and detail capture ability of the audio, while the retention of speech information in the audio is relatively less affected. In general, the dynamic range of 8 to 16 bits can meet most of the clarity requirements for speech transmission.

[0029] Vocal audio is usually used to create an atmosphere or environmental sound effect, and its composition often includes natural sounds, music, mid- and low-frequency components, etc. Vocal audio mainly conveys atmosphere and emotions rather than directly conveying information, so its clarity and discernibility requirements are obviously lower than those of human voice audio. At the same time, vocal audio generally contains more low-frequency components, and these low-frequency sound waves are relatively less sensitive to sampling rates. For example, many environmental sounds and bass instrument sounds can still be effectively accepted by the audience at low sampling rates. When reducing the bit rate, the strategy of reducing the sampling rate is preferred to reduce the amount of data transmission and have less impact on the background sound quality. In the application of vocal audio, the appropriate sampling rate allows not to lose too much low-frequency information and can maintain appropriate audio effects. Overall, for different types of audio content, the adjustment of sampling rate and sampling bit number should be guided by the user's auditory experience. For human voice audio, maintaining the clarity and intelligibility of voice information is the key, so the preference is to reduce the number of sampling bits; for vocal audio, the amount of data can be efficiently reduced by lowering the sampling rate, while having little impact on the overall sound quality. This not only improves the efficiency of audio transmission, but also enables smooth synchronous playback of multiple speakers under limited conditions, thereby improving the user experience.

[0030] In another preferred embodiment of the present invention, the specific calculation process of the reduction amount is: Ci'=C-(B-Vi) / S*W; Wi'=W-(B-Vi)*S*C; Among them, B is the bit rate of the original audio stream data, S is the number of channels of the original audio stream data, C is the sampling rate of the original audio stream data, W is the number of sampling bits of the original audio stream data, Vi is the response rate of any speaker, Ci' is the sampling rate of the adjusted audio stream data, and Wi' is the number of sampling bits of the adjusted audio stream data.

[0031] In another preferred embodiment of the present invention, said S3 further includes maintaining the bit rate of the original audio stream data if there is any speaker whose response rate Vi is greater than or equal to the bit rate.

[0032] High-bitrate audio usually has a larger dynamic range and can better express subtle changes and dynamic characteristics of sound. When the response rate of the speaker is high enough, it can process higher bitrate audio signals, ensuring that the audio signal can respond instantly when playing without affecting the smoothness and accuracy of the playback. The bit rate directly affects the quality of the audio. Maintaining the original bit rate ensures that the details and clarity of the audio will not be compressed or reduced, thereby providing a better user experience. This can fully utilize the performance of the speaker, obtain better audio performance, and enhance the actual user experience.

[0033] In another preferred embodiment of the present invention, S3 also includes obtaining the best frequency response interval [M, N] of any speaker, performing spectrum analysis on the original audio stream data in the buffer area, extracting the audio stream data in the best frequency response interval of the speaker and marking it as target audio stream data, and sending the target audio stream data to the speaker.

[0034] It is worth noting that speakers are usually divided into full-range speakers, bass speakers and tweeters. Generally speaking, the frequency response range of full-range speakers should be between 20Hz and 20kHz, which covers all sound frequencies that the human ear can hear. Bass speakers usually focus on reproducing low-frequency sounds, and their optimal frequency response range is usually between 20Hz and 200Hz. Tweeters are usually responsible for reproducing high-frequency sounds, and their frequency response range is generally between 2kHz and 20kHz. It can be understood that the optimal frequency response intervals [M, N] corresponding to different speakers are different. In order to further improve the quality of audio playback, the transmitted audio signal is analyzed by fast Fourier transform (FFT) to decompose it into different frequency components to obtain multiple frequency bands, such as low frequency, intermediate frequency and high frequency. The audio signal is divided into different frequency bands, and these frequency bands are dynamically allocated according to the optimal working frequency band of each speaker. For example, the bass speaker specializes in processing low-frequency signals, while the midrange speaker processes intermediate frequency signals, thereby further improving the quality of audio playback and enhancing the actual user experience.

[0035] The above is a detailed description of an embodiment of the present invention, but the content is only a preferred embodiment of the present invention and cannot be considered to limit the scope of implementation of the present invention. All equivalent changes and improvements made within the scope of the present invention should still fall within the scope of the patent coverage of the present invention.

Claims

1. A multi-channel audio playback control method based on DSP operation, wherein an audio distributor is connected to a plurality of speakers, characterized in that: The following steps are involved: S1, obtaining parameters of the original audio stream data transmitted by the cloud data server to the audio distributor; the parameters include bit rate, sampling rate, sampling bit number and number of channels; S2, a buffer area is set in the audio distributor. When the buffer area receives the original audio stream data, a feedback instruction with a data volume of Q is sent to all connected speakers and response signals of all connected speakers are received. The response time of all speakers is obtained and marked as t1, t2, ..., tn, where n is the number of speakers. The response rate of all speakers is calculated according to the calculation formula V=Q / t and marked as V1, V2, ..., Vn; S3, when the buffer area pre-stores the original audio stream data with a duration of T, if there is any speaker whose response rate Vi is lower than the bit rate of the original audio stream data, the bit rate of the original audio stream data with a duration of T is compressed, the compressed bit rate is set as the response rate Vi of the speaker, and the compressed audio stream data is sent to the speaker; S4, analyzing the original audio stream data in the buffer area, determining whether the original audio within the current duration T is human voice audio or vocal audio, and adjusting the sampling rate and sampling bit number of the original audio stream data according to the type of the original audio within the current duration T.

2. The method for controlling the playback of multi-channel audio based on DSP operation according to claim 1, characterized in that: The audio distributor is connected to the multiple speakers via, but not limited to, Bluetooth communication, wireless network communication, wired network communication, and wireless transmission communication.

3. The method for controlling the playback of multi-channel audio based on DSP operation according to claim 1, characterized in that: The audio distributor also includes a data adjustment module for connecting to the buffer area. The data adjustment module includes n audio matrix modules. The audio matrix modules all adopt a balanced pipeline structure, and the balanced pipeline structure is composed of a number of two-choice selectors connected in series.

4. The method for controlling the playback of multi-channel audio based on DSP operation according to claim 3, characterized in that: The n audio matrix modules all use a unified clock to select and process data.

5. The multi-channel audio playback control method based on DSP calculation according to claim 1 is characterized in that: The analysis process of audio stream data is as follows: The original audio stream data is divided into frames according to a preset frame length to obtain a number of audio frames and marked as Y1, Y2, ..., Yx, where x is a positive integer. Each frame of the audio stream data after the linear prediction coding analysis is used to extract audio features, wherein the audio features include the short-time zero-crossing rate, the LPC prediction coefficient, and the skewness and kurtosis of the LPC prediction residual amplitude spectrum. A feature vector is formed according to the audio features, and the feature vector is brought into a preset recognition support vector model to identify whether each frame of the audio stream data contains human voice. If human voice is contained, the audio frame is marked as a special audio frame and marked as Q1, Q2, ..., Qz, where z is a positive integer. If z / x is greater than or equal to a preset threshold value, the original audio within the duration T is determined to be human voice audio. If z / x is less than the preset threshold value, the original audio within the duration T is determined to be vocal audio.

6. The method for controlling the playback of multi-channel audio based on DSP operation according to claim 1, characterized in that: If the original audio type within the current duration T is human voice audio, the sampling bit number of the original audio data is reduced first; if the original audio type within the current duration T is vocal audio, the sampling rate of the original audio data is reduced first.

7. The multi-channel audio playback control method based on DSP calculation according to claim 6 is characterized in that: The specific calculation process of the reduction amount is: Ci'=C-(B-Vi) / S*W; Wi'=W-(B-Vi)*S*C; Among them, B is the bit rate of the original audio stream data, S is the number of channels of the original audio stream data, C is the sampling rate of the original audio stream data, W is the number of sampling bits of the original audio stream data, Vi is the response rate of any speaker, Ci' is the sampling rate of the adjusted audio stream data, and Wi' is the number of sampling bits of the adjusted audio stream data.

8. The multi-channel audio playback control method based on DSP calculation according to claim 1 is characterized in that: The S3 also includes maintaining the bit rate of the original audio stream data if the response rate Vi of any speaker is greater than or equal to the bit rate.

9. The method for controlling the playback of multi-channel audio based on DSP operation according to claim 1, characterized in that: The S3 also includes obtaining the best frequency response interval [M, N] of any speaker, performing spectrum analysis on the original audio stream data in the buffer area, extracting the audio stream data in the best frequency response interval of the speaker and marking it as target audio stream data, and sending the target audio stream data to the speaker.

Citation Information

Patent Citations

  • Cloud-based multi-channel audio playing method, system and audio gateway device

    CN107277691A

  • Display equipment and Bluetooth audio transmission method

    CN111385621A

  • Audio transmitting apparatus, audio receiving apparatus, and multi-channel audio transmission method

    CN117135534A

  • Singing scene-based implementation method and system

    CN117440209A