A method and system for optimizing sound field parameters

Abstract

This invention relates to the technical field of sound field optimization, and provides a method and system for optimizing sound field parameters. The method includes: activating a vibration sensing device; driving a loudspeaker to play a test signal to excite the loudspeaker enclosure to vibrate, and driving the vibration sensing device to collect vibration data of the loudspeaker enclosure; receiving the vibration data and performing frequency domain analysis to identify the structural resonant frequency and structural resonant intensity; storing the structural resonant frequency and structural resonant intensity as structural resonant information; determining a preliminary low-frequency compensation scheme based on the room's acoustic characteristics; comparing the preliminary low-frequency compensation scheme with the structural resonant information to determine whether the preliminary low-frequency compensation scheme poses a risk of exciting structural resonance in the loudspeaker enclosure, obtaining an excitation risk assessment result; and adjusting and updating the preliminary low-frequency compensation scheme based on the excitation risk assessment result to obtain a sound field parameter optimization scheme. This invention improves the sound field stability and listening experience of stereo systems.

Description

Technical Field

[0001] This invention relates to the technical field of sound field optimization, and specifically to a method and system for optimizing sound field parameters. Background Technology

[0002] In a home listening environment, high-fidelity stereo systems are designed to provide an accurate and immersive sound field experience. However, with prolonged use, the elastic modulus of internal components such as the tweeter diaphragm may increase slightly, leading to a decrease in high-frequency response; capacitors in the crossover may age due to accumulated electrical stress, causing capacitance drift, which in turn changes the crossover point and filter slope, resulting in phase distortion and frequency response dips. These physical aging and performance degradations cause a shift in the acoustic output characteristics of the speaker itself.

[0003] More complicated by the fact that incorrect low-frequency over-compensation may inadvertently excite the speaker cabinet's own weak structural resonance points, causing audible vibrations and coloration that further impair the clarity of the mid-low frequencies. In this case, no single parameter adjustment can completely solve this complex acoustic problem caused by the superposition of multiple factors such as physical aging, logical conflicts, and mechanical resonance.

[0004] To address the aforementioned issues, existing technologies urgently need improvement. Summary of the Invention

[0005] This application discloses a sound field parameter optimization method and system, which aims to solve complex acoustic problems such as the deviation of the acoustic output characteristics of loudspeakers due to long-term use, the difficulty of traditional sound field optimization methods in accurately distinguishing room acoustic characteristics from front-end system abnormalities, and the possibility that incorrect low-frequency compensation may excite resonance in the loudspeaker cabinet structure.

[0006] The technical solution of this application is as follows:

[0007] Firstly, this application discloses a method for optimizing sound field parameters, specifically including:

[0008] Activate the vibration sensing device preset on the speaker enclosure;

[0009] The speaker is driven to play a test signal to excite the speaker enclosure to vibrate, and the vibration sensing device is driven to collect the vibration data of the speaker enclosure.

[0010] The system receives vibration data and performs frequency domain analysis to identify the structural resonant frequency and intensity of the loudspeaker enclosure.

[0011] Store the structural resonant frequency and structural resonant intensity as structural resonant information;

[0012] Based on the acoustic characteristics of the room where the speaker enclosure is located, a preliminary low-frequency compensation scheme is determined;

[0013] The preliminary low-frequency compensation scheme is compared with the structural resonance information to determine whether the preliminary low-frequency compensation scheme has the risk of exciting the structural resonance of the loudspeaker cabinet, and the excitation risk judgment result is obtained. Based on the excitation risk judgment result, the preliminary low-frequency compensation scheme is adjusted and updated to obtain the sound field parameter optimization scheme.

[0014] This technical solution comprehensively considers the structural resonance characteristics of the speaker enclosure and the acoustic characteristics of the room, avoiding accidental excitation of enclosure resonance by the low-frequency compensation scheme. This effectively solves the sound field distortion problem caused by changes in the speaker's own characteristics and improper compensation in traditional sound field optimization, and improves the accuracy and stability of sound field optimization.

[0015] Furthermore, based on the above, the preliminary low-frequency compensation scheme is compared with the structural resonance information to determine whether the preliminary low-frequency compensation scheme poses a risk of exciting the structural resonance of the loudspeaker enclosure, thus obtaining the excitation risk assessment result. The steps for adjusting and updating the preliminary low-frequency compensation scheme based on the excitation risk assessment result to obtain the sound field parameter optimization scheme include:

[0016] It continuously and synchronously acquires vibration data of the speaker enclosure, real-time electrical signals of the speaker driver unit, and temperature and humidity data of the environment where the speaker is located;

[0017] Blind source separation is performed on real-time electrical signals and vibration data to decompose the inherent resonant vibration components of the loudspeaker enclosure.

[0018] The resonant frequency and quality factor are extracted from the inherent resonant vibration components. Combined with temperature and humidity data, the resonant frequency drift and resonant intensity drift of the loudspeaker enclosure are monitored, and the resonant characteristic records of the loudspeaker enclosure are dynamically updated.

[0019] The preliminary low-frequency compensation scheme was compared with the resonance characteristic records, and a multi-dimensional risk assessment was conducted on each adjustment frequency point in the preliminary low-frequency compensation scheme.

[0020] Based on the results of the multidimensional risk assessment, a compensation strategy is selected to adjust and update the initial low-frequency compensation scheme, resulting in an optimized sound field parameter scheme.

[0021] Through this technical solution, this application can dynamically monitor the resonance characteristic drift of the loudspeaker cabinet and, combined with multi-dimensional risk assessment, achieve fine-tuning of the low-frequency compensation scheme, thereby more effectively avoiding excitation of cabinet resonance and further improving the adaptability and robustness of sound field optimization.

[0022] In some preferred embodiments, the steps of receiving vibration data and performing frequency domain analysis on the vibration data to identify the structural resonant frequency and structural resonant intensity of the loudspeaker enclosure include:

[0023] The vibration data is subjected to spectral analysis at a preset resolution to obtain the spectrum at the preset resolution.

[0024] By performing multi-peak detection on the spectrum at a preset resolution, several potential resonant peaks are obtained;

[0025] Calculate the quality factor and bandwidth corresponding to each potential resonant peak, and determine whether each potential resonant peak belongs to an independent structural resonance based on the quality factor, bandwidth, frequency interval between each potential resonant peak and adjacent peaks, and relative intensity, thus obtaining the peak independence judgment result.

[0026] When the peak independence judgment result indicates that there are overlapping spectra, curve fitting is performed on the overlapping spectra to decompose them into independent resonance components, and the center frequency and intensity of the independent resonance components are obtained.

[0027] Harmonic resonance is obtained by analyzing the harmonic relationships of the center frequencies of independent resonant components.

[0028] By recording the harmonic resonance and the preset fundamental frequency resonance, the structural resonant frequency and structural resonant intensity of the loudspeaker enclosure are obtained.

[0029] Through this technical solution, this application can accurately identify the independent structural resonant components of the loudspeaker enclosure through precise spectrum analysis and peak detection, and can effectively decompose them even in the presence of overlapping spectra, thereby providing more accurate resonant information for subsequent sound field parameter optimization.

[0030] As an optional approach, the preliminary low-frequency compensation scheme is compared with the structural resonance information to determine whether the preliminary low-frequency compensation scheme poses a risk of exciting the structural resonance of the loudspeaker enclosure, thus obtaining the excitation risk assessment result. Based on the excitation risk assessment result, the preliminary low-frequency compensation scheme is adjusted and updated to obtain the sound field parameter optimization scheme. The steps include:

[0031] The transient power detector preset at the power amplifier output of the speaker driver unit is activated;

[0032] The transient power detector is driven to monitor the instantaneous output power of the speaker driver unit;

[0033] When the instantaneous output power is detected to exceed the preset instantaneous power threshold within the preset range corresponding to the adjustment frequency point of the initial low-frequency compensation scheme, the risk assessment result is determined to be the risk of transient excitation of the speaker cabinet's structural resonance.

[0034] Based on the determination of the excitation risk, a transient suppression filter is introduced at the corresponding adjustment frequency point;

[0035] The transient suppression filter is activated when the instantaneous output power exceeds a preset instantaneous power threshold, and the signal within the preset range corresponding to the adjusted frequency point is instantaneously attenuated.

[0036] The transient suppression filter is de-attenuated when the instantaneous output power is lower than the preset instantaneous power threshold.

[0037] Through this technical solution, this application can respond in real time and suppress transient high-power signals that may excite resonance in the cabinet structure by using transient power detection and transient suppression filters, thereby effectively protecting the speaker cabinet and avoiding transient sound coloration during dynamic playback.

[0038] In another embodiment, the steps of receiving vibration data and performing frequency domain analysis on the vibration data to identify the structural resonant frequency and structural resonant intensity of the loudspeaker enclosure include:

[0039] The electrical signals of the speaker driver unit are acquired while the speaker plays a test signal to excite the speaker cabinet to vibrate.

[0040] Time-frequency domain analysis was performed on vibration data and electrical signals to calculate the coherence of vibration data and electrical signals at different frequencies;

[0041] A coherence threshold is set. When the coherence between vibration data and electrical signal at different frequencies is lower than the coherence threshold, the vibration of the speaker enclosure is determined to be external interference vibration.

[0042] Frequency domain analysis was performed on the vibration data of the loudspeaker enclosure, which was determined to be non-external interference vibration, to identify the structural resonant frequency and structural resonant intensity of the loudspeaker enclosure.

[0043] Through this technical solution, this application can effectively distinguish between the structural resonance of the speaker enclosure itself and external interference vibration by analyzing the coherence of vibration data and electrical signals, thereby improving the accuracy of structural resonance identification and avoiding misjudging external noise as enclosure resonance.

[0044] Furthermore, the steps of performing time-frequency domain analysis on the vibration data and electrical signals to calculate the coherence of the vibration data and electrical signals at different frequencies include:

[0045] The time delay between the electrical signal and the vibration data was determined by cross-correlation analysis.

[0046] The vibration data is time-aligned based on the time delay.

[0047] Fourier transform is performed on the aligned vibration data and electrical signal to obtain the frequency domain representation of the vibration data and the frequency domain representation of the electrical signal.

[0048] Phase compensation is performed on the frequency domain representation of the vibration data by calculating the phase difference between the frequency domain representation of the electrical signal and the frequency domain representation of the vibration data.

[0049] The coherence of the frequency domain representation of the vibration data after phase compensation and the frequency domain representation of the electrical signal at different frequencies is calculated.

[0050] Through this technical solution, this application can improve the accuracy of coherence calculation between vibration data and electrical signals by precise time alignment and phase compensation, thereby more reliably identifying the box vibration caused by the drive signal.

[0051] In some preferred embodiments, a coherence threshold is set. When the coherence between the vibration data and the electrical signal at different frequencies is lower than the coherence threshold, the step of determining that the vibration of the loudspeaker enclosure is external interference vibration includes:

[0052] When the vibration energy of the loudspeaker enclosure is significant, time-frequency domain analysis is performed on the vibration data and the electrical signal to obtain the time-frequency distribution.

[0053] In the time-frequency distribution, frequencies and time periods in which the vibration data and the electrical signal have high coherence are identified;

[0054] For frequencies and time periods where the vibration data and the electrical signal have high coherence, the transient characteristics of the vibration signal of the loudspeaker enclosure are analyzed; the transient characteristics include the steepness and duration of the vibration rise edge.

[0055] When the vibration signal of the speaker enclosure exhibits transient characteristics unique to non-drive signals within a frequency and time period in which the vibration data and the electrical signal are highly coherent, the vibration of the speaker enclosure is determined to be external interference vibration.

[0056] When the vibration signal of the speaker enclosure exhibits transient characteristics consistent with the driving signal within a frequency and time period in which the vibration data and the electrical signal are highly coherent, the vibration of the speaker enclosure is determined to be caused by the driving signal.

[0057] Through this technical solution, this application can further refine the judgment of vibrations with high coherence between the vibration data and the electrical signal within a frequency and time period by combining transient characteristic analysis, thereby more accurately distinguishing vibrations caused by the driving signal from external interference vibrations with similar frequencies but different transient characteristics.

[0058] Based on the above, the steps for performing spectral analysis on the vibration data according to a preset resolution to obtain the preset resolution spectrum include:

[0059] Before performing spectral analysis on the vibration data according to the preset resolution, the vibration data is subjected to real-time noise type identification to distinguish between transient impact noise and non-stationary background noise.

[0060] When identified as transient impact noise, adaptive threshold limiting is used to suppress vibration data that exceeds the instantaneous energy threshold.

[0061] When non-stationary background noise is identified, a noise suppression method based on time-varying filters is adopted. The filter parameters are dynamically adjusted according to the real-time spectral characteristics of the noise to filter the vibration data.

[0062] The vibration data after noise suppression is subjected to spectral analysis at a preset resolution to obtain the preset resolution spectrum.

[0063] Through this technical solution, this application can effectively remove transient impact noise and non-stationary background noise from vibration data by performing noise type identification and adaptive noise suppression before spectrum analysis, thereby improving the accuracy of spectrum analysis and providing a cleaner data foundation for subsequent resonance identification.

[0064] In some implementations, the step of performing multi-peak detection on a preset resolution spectrum to obtain several potential resonant peaks includes:

[0065] Peak shape analysis based on local features is performed on the spectrum at a preset resolution to identify potential resonant regions with broad tops or irregular shapes.

[0066] The potential resonant region is adaptively segmented, and each non-ideal peak shape is decomposed into several narrower sub-regions;

[0067] Perform local maxima search independently for each sub-region to identify local peak points within each sub-region;

[0068] Based on the frequency and intensity of the local peak points, as well as the frequency interval and relative intensity between the current local peak point and the adjacent local peak points, determine whether each local peak point belongs to an independent structural resonance;

[0069] When two or more independent structural resonances are identified, the independent structural resonances are recorded as potential resonance peaks.

[0070] Through this technical solution, this application can effectively identify and decompose non-ideal shaped resonance peaks by performing fine shape analysis and adaptive segmentation on the spectral peaks, thereby more accurately detecting multiple potential independent resonance peaks and improving the accuracy of resonance identification.

[0071] Secondly, this application also discloses a sound field parameter optimization system for performing sound field parameter optimization, specifically including:

[0072] The vibration sensing activation module is used to activate the vibration sensing device preset on the speaker enclosure;

[0073] The vibration data acquisition module is used to drive the loudspeaker to play a test signal to excite the loudspeaker enclosure to vibrate, and to drive the vibration sensing device to collect the vibration data of the loudspeaker enclosure.

[0074] The vibration data identification module is used to receive vibration data and perform frequency domain analysis on the vibration data to identify the structural resonant frequency and structural resonant intensity of the speaker enclosure.

[0075] The resonance information storage module is used to store the structural resonance frequency and structural resonance intensity as structural resonance information.

[0076] The compensation scheme determination module is used to determine a preliminary low-frequency compensation scheme based on the room acoustic characteristics of the room where the speaker cabinet is located;

[0077] The parameter optimization execution module is used to compare the preliminary low-frequency compensation scheme with the structural resonance information, determine whether the preliminary low-frequency compensation scheme has the risk of exciting the structural resonance of the loudspeaker cabinet, and obtain the excitation risk judgment result; based on the excitation risk judgment result, the preliminary low-frequency compensation scheme is adjusted and updated to obtain the sound field parameter optimization scheme.

[0078] Through this technical solution, this application provides a system capable of implementing the above-mentioned sound field parameter optimization method. Through modular design, it ensures the coordinated operation of each functional unit, thereby achieving effective identification of loudspeaker cabinet structure resonance and intelligent optimization of low-frequency compensation scheme, solving the sound field distortion problem caused by changes in loudspeaker characteristics and improper compensation in traditional sound field optimization.

[0079] Beneficial Effects: The sound field parameter optimization method disclosed in this application collects vibration data of the speaker cabinet by activating a vibration sensing device pre-installed on the speaker cabinet. Frequency domain analysis is then performed on the vibration data to identify the structural resonant frequency and intensity of the speaker cabinet, storing this information as structural resonance information. Simultaneously, a preliminary low-frequency compensation scheme is determined based on the room acoustic characteristics of the room where the speaker cabinet is located. Subsequently, the preliminary low-frequency compensation scheme is compared with the structural resonance information to determine whether it poses a risk of exciting structural resonance in the speaker cabinet. Based on the determination result, the preliminary low-frequency compensation scheme is adjusted and updated to obtain the final sound field parameter optimization scheme.

[0080] This method effectively addresses the problems in existing technologies, such as the shift in acoustic output characteristics caused by speaker aging and performance degradation, and the difficulty of accurately distinguishing between room acoustic characteristics and front-end system anomalies in traditional sound field optimization methods. By directly sensing and analyzing the vibration characteristics of the speaker enclosure, this application can accurately identify the inherent structural resonant points of the enclosure, avoiding misjudgments caused by a lack of understanding of the speaker's physical state in traditional methods. Furthermore, by comparing the preliminary low-frequency compensation scheme with the enclosure structural resonance information, the risk of the compensation scheme potentially triggering enclosure resonance can be predicted and avoided, thereby preventing problems such as low-frequency booming, hollow sound field, and impaired mid-low frequency clarity.

[0081] In summary, this application achieves more precise and comprehensive optimization of sound field parameters by introducing the perception and analysis of speaker enclosure structural resonance and incorporating it into the optimization process of the low-frequency compensation scheme. This not only overcomes the complex acoustic problems caused by the superposition of multiple factors such as physical aging, logical conflicts, and mechanical resonance in existing technologies, but also significantly improves the sound field stability and immersiveness of high-fidelity stereo systems, providing users with a more accurate and immersive listening experience. Attached Figure Description

[0082] Figure 1 This is a flowchart of a sound field parameter optimization method in one embodiment of the present invention;

[0083] Figure 2 This is a flowchart of a sound field parameter optimization method according to another embodiment of the present invention;

[0084] Figure 3 This is a system block diagram of a sound field parameter optimization system according to another embodiment of the present invention;

[0085] Explanation of reference numerals in the attached figures:

[0086] 1. Sound field parameter optimization system; 11. Vibration sensing start-up module; 12. Vibration data acquisition module; 13. Vibration data identification module; 14. Resonance information storage module; 15. Compensation scheme determination module; 16. Parameter optimization execution module. Detailed Implementation

[0087] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this application, and not all embodiments. The components of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0088] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this application, terms such as "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0089] Traditional sound field parameter optimization methods, when used in home listening environments, are susceptible to deviations in speaker acoustic output characteristics due to factors such as speaker aging, performance degradation, and user-initiated adjustments. This makes it difficult for optimization algorithms to accurately distinguish between room acoustic problems and speaker malfunctions, potentially leading to the generation of incorrect digital filters, especially excessive and inappropriate compensation in the low-frequency range. Such inappropriate compensation not only disrupts the balance of high-frequency reflections but may also induce incorrect low-frequency energy distribution within the room, resulting in a hollow sound field, booming bass with erratic positioning, severely compromising the stability and immersive feel of the stereo image. Further complicating matters, excessive and incorrect low-frequency compensation may inadvertently excite weak, inherent structural resonance points within the speaker cabinet, causing audible vibrations and coloration, further contaminating the clarity of the mid-to-low frequencies.

[0090] To address this, this application proposes a sound field parameter optimization method, combining... Figure 1 As shown, it includes:

[0091] S1, activate the vibration sensing device preset on the speaker cabinet;

[0092] S2 drives the loudspeaker to play a test signal to excite the loudspeaker enclosure to vibrate, and drives the vibration sensing device to collect vibration data of the loudspeaker enclosure.

[0093] S3 receives vibration data and performs frequency domain analysis on the vibration data to identify the structural resonant frequency and structural resonant intensity of the loudspeaker enclosure.

[0094] S4, store the structural resonant frequency and structural resonant intensity as structural resonant information;

[0095] S5. Determine a preliminary low-frequency compensation scheme based on the room acoustic characteristics of the room where the speaker cabinet is located;

[0096] S6. Compare the preliminary low-frequency compensation scheme with the structural resonance information to determine whether the preliminary low-frequency compensation scheme has the risk of exciting the structural resonance of the speaker cabinet, and obtain the excitation risk judgment result; based on the excitation risk judgment result, adjust and update the preliminary low-frequency compensation scheme to obtain the sound field parameter optimization scheme.

[0097] To better understand the sound field parameter optimization method proposed in this application, some key terms involved will be explained first.

[0098] "Vibration sensing device" refers to a sensor that can detect and quantify the vibration of an object, such as an accelerometer, laser vibrometer, or piezoelectric sensor. Its function is to acquire physical vibration information of the speaker enclosure. "Vibration data" refers to the raw signal collected by the vibration sensing device that reflects the vibration state of the speaker enclosure, usually a time-domain signal. "Frequency domain analysis" refers to converting the time-domain signal into a frequency-domain representation using mathematical methods such as Fourier transform to reveal the energy distribution of the signal at different frequencies. "Structural resonant frequency" refers to the natural frequency at which the speaker enclosure exhibits its maximum vibration response; these frequencies are determined by the enclosure's structural characteristics. "Structural resonant intensity" refers to the amplitude or energy of the speaker enclosure's vibration at the structural resonant frequency. "Structural resonant information" refers to a set of information including structural resonant frequencies and structural resonant intensities, used to describe the inherent vibration characteristics of the speaker enclosure. "Room acoustic characteristics" refer to the influence of the room on sound propagation, including reverberation time, frequency response, and standing wave modes; these characteristics affect the perception of sound in the listening environment. "Preliminary low-frequency compensation scheme" refers to an initial equalization or filtering strategy developed based on the room's acoustic characteristics, aimed at improving the low-frequency response. "Risk assessment results" refers to the conclusion that assesses whether the initial low-frequency compensation scheme may cause harmful structural resonance in the speaker cabinet. "Sound field parameter optimization scheme" refers to the final low-frequency compensation strategy that, after adjustment and updating, can effectively improve sound field performance and avoid cabinet resonance.

[0099] Specifically, the method first activates a vibration sensing device pre-installed on the speaker enclosure. This vibration sensing device can be a miniature accelerometer installed on the surface or inside the speaker enclosure to monitor the enclosure's vibration in real time. Alternatively, a non-contact laser vibrometer can be used to acquire vibration data by scanning the enclosure surface with a laser beam.

[0100] Subsequently, the loudspeaker is driven to play a test signal to excite the loudspeaker cabinet to vibrate, while a vibration sensing device collects the vibration data of the loudspeaker cabinet. The test signal can be a swept-frequency signal, powder noise, or a pulse signal of a specific frequency, the purpose of which is to comprehensively excite the vibration response of the loudspeaker cabinet at different frequencies. The vibration sensing device converts these vibrations into electrical signals and records them.

[0101] Next, vibration data is received and analyzed in the frequency domain to identify the structural resonant frequency and intensity of the speaker enclosure. For example, the acquired time-domain vibration data can be converted into a frequency-domain spectrum using a Fast Fourier Transform (FFT), and then a peak detection algorithm can be used to identify significant peaks in the spectrum. The frequencies corresponding to these peaks are the structural resonant frequencies, while the height or area of ​​the peaks reflects the structural resonant intensity.

[0102] The identified structural resonant frequencies and intensities are stored as structural resonant information. This information can be stored in the internal memory of the loudspeaker system or uploaded to a cloud server for management. For example, it can be stored in the form of a data table, which includes parameters such as frequency values, intensity values, and corresponding quality factors.

[0103] Simultaneously, based on the room's acoustic characteristics where the speaker enclosure is located, a preliminary low-frequency compensation scheme is determined. Room acoustic characteristics can be obtained by collecting and analyzing impulse response data from professional acoustic measurement equipment (such as a measuring microphone). The preliminary low-frequency compensation scheme can be an equalizer curve designed to correct problems such as low-frequency standing waves or uneven absorption in the room.

[0104] The key is to compare the preliminary low-frequency compensation scheme with the structural resonance information to determine whether the preliminary low-frequency compensation scheme poses a risk of triggering the structural resonance of the speaker cabinet, thus obtaining a risk assessment result. For example, one can check whether there are gain boost points in the preliminary low-frequency compensation scheme that are close to or overlap with the structural resonance frequency. If the compensation scheme has a significant low-frequency gain near a certain structural resonance frequency, there may be a risk of triggering.

[0105] Finally, based on the risk assessment, the initial low-frequency compensation scheme is adjusted and updated to obtain an optimized sound field parameter scheme. If a risk of excitation is identified, a notch filter or a reduction in gain can be introduced at the corresponding frequency point in the initial low-frequency compensation scheme to avoid exciting cabinet resonance. For example, if the initial compensation scheme has a gain of +6dB at 50Hz, while the cabinet exhibits strong structural resonance at 52Hz, the gain at 50Hz can be reduced, or a narrowband attenuator can be introduced near that frequency to obtain the final optimized sound field parameter scheme.

[0106] Optional, combined Figure 2As shown, S6 compares the preliminary low-frequency compensation scheme with the structural resonance information to determine whether the preliminary low-frequency compensation scheme poses a risk of exciting the structural resonance of the speaker cabinet, thus obtaining the excitation risk assessment result. Based on the excitation risk assessment result, the steps of adjusting and updating the preliminary low-frequency compensation scheme to obtain the sound field parameter optimization scheme include:

[0107] S61 continuously and synchronously acquires vibration data of the speaker enclosure, real-time electrical signals of the speaker driver unit, and temperature and humidity data of the speaker's environment;

[0108] S62 performs blind source separation on real-time electrical signals and vibration data to decompose the inherent resonant vibration components of the loudspeaker enclosure.

[0109] S63 extracts the resonant frequency and quality factor from the inherent resonant vibration components, combines temperature and humidity data, monitors the resonant frequency drift and resonant intensity drift of the speaker enclosure, and dynamically updates the resonant characteristic records of the speaker enclosure.

[0110] S64 compares the preliminary low-frequency compensation scheme with the resonance characteristic record and performs a multi-dimensional risk assessment on each adjustment frequency point in the preliminary low-frequency compensation scheme.

[0111] S65. Based on the results of the multi-dimensional risk assessment, a compensation strategy is selected to adjust and update the initial low-frequency compensation scheme, resulting in an optimized sound field parameter scheme.

[0112] Specifically, the system continuously and synchronously acquires vibration data of the speaker enclosure, real-time electrical signals of the speaker driver unit, and temperature and humidity data of the speaker's environment, aiming to provide comprehensive real-time input for subsequent dynamic analysis. Vibration data directly reflects the mechanical vibration state of the speaker enclosure; real-time electrical signals can be used to distinguish between vibrations caused by the speaker driver unit's own excitation and external interference; and temperature and humidity data are used to capture the potential impact of environmental changes on the speaker enclosure's material properties and resonant frequencies.

[0113] Furthermore, blind source separation is performed on the real-time electrical signals and vibration data to effectively separate the actual inherent resonant vibration components of the speaker enclosure from the mixed signal. In practical applications, the vibration data of the speaker enclosure may contain multiple components, including vibrations caused by the signal played from the speaker driver unit, the inherent resonant vibrations of the enclosure itself, and environmental noise. Blind source separation technology can decompose these mixed signals into independent source signals without relying on prior information, based on principles such as statistical independence. This allows for the accurate identification of the inherent resonant vibration components of the speaker enclosure, avoiding interference from other signal components in the resonance characteristic analysis.

[0114] This process involves extracting resonant frequency and quality factor from inherent resonant vibration components, and combining this with temperature and humidity data to monitor the resonant frequency drift and resonant intensity drift of the speaker enclosure, dynamically updating the enclosure's resonant characteristic records. Resonant frequency and quality factor are key parameters characterizing structural resonant properties. Resonant frequency drift may be caused by factors such as material aging and changes in environmental temperature and humidity, while resonant intensity drift may reflect issues such as enclosure structural fatigue or loose connections. By combining real-time temperature and humidity monitoring, a correlation model between resonant characteristics and environmental factors can be established, allowing for more accurate prediction and tracking of dynamic changes in resonant characteristics, and timely updates to the speaker enclosure's resonant characteristic records, ensuring the accuracy of subsequent risk assessments.

[0115] In practical applications, the preliminary low-frequency compensation scheme is compared with the resonance characteristic records, and a multi-dimensional risk assessment is performed on each adjusted frequency point in the preliminary low-frequency compensation scheme. This multi-dimensional risk assessment considers not only the proximity of the compensated frequency point to the resonant frequency, but also comprehensively considers multiple dimensions such as resonance intensity, quality factor, resonant frequency drift trend, and the gain of the compensation scheme. For example, even if a compensated frequency point does not completely coincide with the resonant frequency, if it is within the bandwidth of the resonance peak, has a high compensation gain, and exhibits a large resonance intensity or a tendency to drift towards that frequency, its excitation risk will be assessed as high.

[0116] Therefore, based on the results of the multi-dimensional risk assessment, a compensation strategy is selected to adjust and update the initial low-frequency compensation scheme, resulting in an optimized sound field parameter scheme. The choice of compensation strategy can be diverse. For example, for high-risk adjustment frequencies, measures such as reducing the compensation gain at that frequency, introducing a narrowband notch filter for attenuation, or adjusting the phase to avoid resonance superposition can be taken. Through this dynamic adjustment based on real-time monitoring and multi-dimensional assessment, it can be ensured that the low-frequency compensation scheme effectively avoids the risk of exciting resonance in the speaker enclosure structure while optimizing the sound field.

[0117] Optionally, the steps of receiving vibration data and performing frequency domain analysis on the vibration data to identify the structural resonant frequency and structural resonant intensity of the loudspeaker enclosure include:

[0118] The vibration data is subjected to spectral analysis at a preset resolution to obtain the spectrum at the preset resolution.

[0119] By performing multi-peak detection on the spectrum at a preset resolution, several potential resonant peaks are obtained;

[0120] Calculate the quality factor and bandwidth corresponding to each potential resonant peak, and determine whether each potential resonant peak belongs to an independent structural resonance based on the quality factor, bandwidth, frequency interval between each potential resonant peak and adjacent peaks, and relative intensity, thus obtaining the peak independence judgment result.

[0121] When the peak independence judgment result indicates that there are overlapping spectra, curve fitting is performed on the overlapping spectra to decompose them into independent resonance components, and the center frequency and intensity of the independent resonance components are obtained.

[0122] Harmonic resonance is obtained by analyzing the harmonic relationships of the center frequencies of independent resonant components.

[0123] By recording the harmonic resonance and the preset fundamental frequency resonance, the structural resonant frequency and structural resonant intensity of the loudspeaker enclosure are obtained.

[0124] Specifically, spectral analysis is performed on the vibration data according to a preset resolution, aiming to convert the time-domain vibration signal into a frequency-domain representation to facilitate the identification of its frequency components. The preset resolution determines the level of detail in the spectral analysis; a higher resolution can more accurately identify closely adjacent resonance peaks.

[0125] The purpose of multi-peak detection on the preset resolution spectrum is to identify all possible resonant points from the complex spectrum. These potential resonant peaks are preliminary indicators of the resonant characteristics exhibited by the speaker enclosure at specific frequencies.

[0126] In practical applications, the quality factor and bandwidth corresponding to each potential resonant peak are calculated. Based on the quality factor, bandwidth, frequency interval between each potential resonant peak and adjacent peaks, and relative intensity, it is determined whether each potential resonant peak belongs to an independent structural resonance, thus obtaining the peak independence judgment result. The quality factor (Q value) reflects the sharpness of the resonance, while the bandwidth represents the resonant frequency range. By comprehensively considering these parameters, it is possible to effectively distinguish between true structural resonances and noise or non-independent resonances. For example, peaks with high Q values, narrow bandwidths, and sufficient frequency intervals from other peaks are more likely to represent independent structural resonances.

[0127] When the peak independence assessment indicates the presence of overlapping spectra, curve fitting is performed on the overlapping spectra to decompose them into independent resonant components, yielding the center frequency and intensity of each independent resonant component. This addresses the difficulty in accurately identifying each independent resonance when multiple resonant frequencies are very close, causing their spectral peaks to overlap. Through curve fitting (e.g., using the Lorentz function or Gaussian function), overlapping composite peaks can be decomposed into multiple independent single peaks, thereby accurately obtaining the center frequency and intensity of each independent resonance.

[0128] Furthermore, harmonic relationship analysis is performed on the center frequencies of the independent resonant components to obtain harmonic resonances. The structural resonance of a loudspeaker enclosure often exhibits a fundamental frequency and its integer multiples of harmonics. By identifying these harmonic relationships, it can be further confirmed whether the detected resonant components truly originate from the enclosure structure, and it helps to distinguish between primary and secondary resonances.

[0129] Therefore, the harmonic resonances and the preset fundamental frequency resonance are recorded to obtain the structural resonant frequency and structural resonant intensity of the loudspeaker enclosure. This step systematically records all confirmed and decomposed independent structural resonances (including the fundamental frequency and harmonics) and their corresponding intensities, forming complete structural resonance information, which provides an accurate basis for subsequent low-frequency compensation scheme adjustments.

[0130] Optionally, the steps of comparing the preliminary low-frequency compensation scheme with the structural resonance information to determine whether the preliminary low-frequency compensation scheme poses a risk of exciting the structural resonance of the loudspeaker enclosure, and obtaining the excitation risk assessment result; and adjusting and updating the preliminary low-frequency compensation scheme based on the excitation risk assessment result to obtain the sound field parameter optimization scheme include:

[0131] The transient power detector preset at the power amplifier output of the speaker driver unit is activated;

[0132] The transient power detector is driven to monitor the instantaneous output power of the speaker driver unit;

[0133] When the instantaneous output power is detected to exceed the preset instantaneous power threshold within the preset range corresponding to the adjustment frequency point of the initial low-frequency compensation scheme, the risk assessment result is determined to be the risk of transient excitation of the speaker cabinet's structural resonance.

[0134] Based on the determination of the excitation risk, a transient suppression filter is introduced at the corresponding adjustment frequency point;

[0135] The transient suppression filter is activated when the instantaneous output power exceeds a preset instantaneous power threshold, and the signal within the preset range corresponding to the adjusted frequency point is instantaneously attenuated.

[0136] The transient suppression filter is de-attenuated when the instantaneous output power is lower than the preset instantaneous power threshold.

[0137] Specifically, a transient power detector can be understood as a device capable of monitoring the electrical signal power at the output of a speaker driver amplifier in real time. It is preset at the amplifier output to directly acquire the instantaneous electrical power information driving the speaker. This detector can be implemented using a high-sampling-rate analog-to-digital converter (ADC) combined with a digital signal processor (DSP) to quickly and accurately capture instantaneous power changes. Its purpose is to identify signals that may reach extremely high amplitudes within a short period of time; these signals, although short in duration, possess enormous energy sufficient to excite structural resonance. Instantaneous output power refers to the actual electrical power output of the speaker driver unit at a given moment, reflecting the instantaneous energy of the driving signal. This power value is dynamically changing, especially when playing audio content with a wide dynamic range, where its peak value may be much higher than the average power. The preset instantaneous power threshold is a safety upper limit pre-set based on the structural resonance characteristics of the speaker enclosure, material tolerance, and desired sound quality performance. When the instantaneous output power exceeds this threshold, it is considered that there is a risk of transiently exciting structural resonance. This threshold can be determined through experimental testing, finite element analysis, or empirical data.

[0138] A transient suppression filter is a dynamically responding signal processing module that instantaneously attenuates signals within a specific frequency range when a transient excitation risk is detected. This filter can be implemented using digital filters, such as DSP-based adaptive filters or multi-band dynamic equalizers. Its key characteristic is that its activation and attenuation are instantaneous, intervening only when a risk occurs to minimize the impact on the normal audio signal. The preset range corresponding to the adjusted frequency points refers to the frequency points in the initial low-frequency compensation scheme that may overlap with or approach the resonant frequencies of the speaker cabinet structure, as well as a certain bandwidth around these frequency points. Within this range, even brief high-power signals can induce resonance. Instantaneous attenuation refers to the rapid, temporary reduction in signal amplitude when the transient suppression filter is activated. This attenuation is instantaneous and aims to quickly reduce the energy that may cause resonance, thereby avoiding or mitigating structural resonance. De-attenuation means that when the instantaneous output power falls below a preset instantaneous power threshold, the transient suppression filter stops attenuating the signal, allowing the audio signal to resume normal transmission. This ensures that the audio signal is not subjected to unnecessary processing when there is no transient excitation risk, thus maintaining the integrity of the sound quality.

[0139] In some preferred embodiments, a specific example is illustrated below. Suppose a speaker system is playing music containing strong bass drum beats, and the initial low-frequency compensation scheme has been preliminarily optimized based on the room acoustics and the structural resonance information of the speaker cabinet. However, when a certain bass drum beat occurs, the instantaneous output power of the speaker driver unit spikes instantaneously at 80Hz (a frequency close to a structural resonance frequency of the speaker cabinet), exceeding a preset instantaneous power threshold. At this moment, a transient power detector preset at the amplifier output immediately detects this instantaneous power exceedance and determines that there is a risk of transiently exciting structural resonance in the speaker cabinet. The system then activates a transient suppression filter within 80Hz and its preset frequency range, instantaneously attenuating the signal within this frequency range, for example, by 3dB, to reduce the instantaneous energy. After the drum beat impact ends, the instantaneous output power quickly drops below the threshold, and the transient suppression filter then de-attenuates, restoring the signal to normal. In this way, even when playing music with a wide dynamic range, the speaker cabinet can avoid structural resonance caused by instantaneous high-power impacts, thereby maintaining the clarity of the bass and the purity of the overall sound quality, while protecting the speaker cabinet from excessive vibration.

[0140] Optionally, the steps of receiving vibration data and performing frequency domain analysis on the vibration data to identify the structural resonant frequency and structural resonant intensity of the loudspeaker enclosure include:

[0141] The electrical signals of the speaker driver unit are acquired while the speaker plays a test signal to excite the speaker cabinet to vibrate.

[0142] Time-frequency domain analysis was performed on vibration data and electrical signals to calculate the coherence of vibration data and electrical signals at different frequencies;

[0143] A coherence threshold is set. When the coherence between vibration data and electrical signal at different frequencies is lower than the coherence threshold, the vibration of the speaker enclosure is determined to be external interference vibration.

[0144] Frequency domain analysis was performed on the vibration data of the loudspeaker enclosure, which was determined to be non-external interference vibration, to identify the structural resonant frequency and structural resonant intensity of the loudspeaker enclosure.

[0145] During the process of playing a test signal to excite the loudspeaker enclosure to vibrate, the electrical signal of the loudspeaker driver unit is simultaneously acquired. The purpose is to obtain excitation source information directly related to the loudspeaker enclosure vibration. This electrical signal typically refers to the voltage applied to the voice coil of the loudspeaker driver unit or the current signal flowing through the voice coil, which directly reflects the instantaneous operating state and energy input of the loudspeaker driver unit.

[0146] Furthermore, time-frequency domain analysis is performed on the collected vibration data and electrical signals, and the coherence of the two at different frequencies is calculated. Time-frequency domain analysis can employ methods such as Short-Time Fourier Transform (STFT), Wavelet Transform, or Hilbert-Huang Transform to reveal the distribution characteristics of the signals in time and frequency. Coherence is typically calculated as the ratio of the cross-power spectrum to the self-power spectrum, with a value between 0 and 1, used to quantify the degree of linear correlation between two signals at a specific frequency. A coherence of 1 indicates that the two signals are perfectly correlated at that frequency, while a coherence of 0 indicates that they are completely uncorrelated. Here, calculating the coherence of the vibration data and electrical signals aims to assess the correlation between the vibration of the loudspeaker enclosure and the electrical excitation of the loudspeaker driver unit.

[0147] Based on this, a coherence threshold is set. When the coherence between vibration data and electrical signals at different frequencies is lower than the coherence threshold, the vibration of the speaker enclosure can be determined to be external interference vibration. This coherence threshold can be set based on experience, experimental data, or preset system requirements; for example, it can be set to 0.7 or 0.8. Coherence below this threshold indicates that there is a lack of sufficient linear correlation between the vibration at that frequency and the electrical excitation of the speaker driver unit, and therefore it is likely not caused by the speaker itself, but by external factors.

[0148] Ultimately, frequency domain analysis was performed only on the vibration data of the speaker enclosure that was determined to be non-external interference vibrations in order to identify the structural resonant frequency and intensity of the speaker enclosure. This means that vibration components identified as external interference will be excluded from subsequent analysis, thereby ensuring that the identified structural resonances are inherent characteristics of the speaker enclosure itself.

[0149] In some preferred embodiments, a specific example is given below. Assume a user is optimizing sound field parameters in a home theater environment. While the speaker plays a test signal to excite the speaker cabinet to vibrate, a vibration sensing device collects vibration data from the speaker cabinet. Simultaneously, the system also synchronously collects real-time electrical signals from the speaker driver unit. During the test, external environmental factors, such as a neighbor suddenly opening or closing a door, or a vehicle passing by, may cause additional vibrations in the speaker cabinet.

[0150] At this point, the system performs time-frequency domain analysis on the collected vibration data and electrical signals. For example, by calculating the cross-power spectrum and auto-power spectrum at different frequency points, the coherence between the two is obtained. Suppose that at a certain frequency point F1, the calculated coherence between the vibration data and the electrical signal is 0.95, indicating that the vibration at frequency F1 is highly correlated with the electrical excitation of the speaker driver unit, and therefore is determined to be structural vibration of the speaker enclosure itself. At another frequency point F2, the calculated coherence between the vibration data and the electrical signal is 0.2, far below the preset coherence threshold (e.g., 0.7). This indicates that the vibration at frequency F2 has a very low correlation with the electrical excitation of the speaker driver unit, and therefore is determined to be external interference vibration.

[0151] Based on the above judgment, the system will only perform subsequent frequency domain analysis on the vibration data at frequency point F1 to identify the structural resonant frequency and structural resonant intensity of the speaker enclosure at that frequency. The vibration data at frequency point F2, being identified as external interference, will be excluded from structural resonance identification. In this way, even in complex environments with external interference, the proposed solution can accurately identify the true structural resonant characteristics of the speaker enclosure, avoiding misjudgments caused by external interference, thus providing reliable basic data for subsequent sound field parameter optimization.

[0152] Optionally, the step of performing time-frequency domain analysis on the vibration data and electrical signal to calculate the coherence of the vibration data and electrical signal at different frequencies may include the following:

[0153] The time delay between the electrical signal and the vibration data was determined by cross-correlation analysis.

[0154] The vibration data is time-aligned based on the time delay.

[0155] Fourier transform is performed on the aligned vibration data and electrical signal to obtain the frequency domain representation of the vibration data and the frequency domain representation of the electrical signal.

[0156] Phase compensation is performed on the frequency domain representation of the vibration data by calculating the phase difference between the frequency domain representation of the electrical signal and the frequency domain representation of the vibration data.

[0157] The coherence of the frequency domain representation of the vibration data after phase compensation and the frequency domain representation of the electrical signal at different frequencies is calculated.

[0158] Specifically, cross-correlation analysis is a mathematical method used to measure the similarity between two signals at different time offsets. By performing cross-correlation analysis on electrical signals and vibration data, the time delay between them, i.e., the time delay, can be accurately determined. This time delay may be caused by factors such as the signal transmission path and sensor response time.

[0159] The process of time-aligning the vibration data based on the determined time delay aims to eliminate the time deviation between the electrical signal and the vibration data, ensuring that the two signals are synchronized on the time axis, thus providing an accurate basis for subsequent frequency domain analysis. Time alignment can be achieved by time shifting or interpolating one of the signals.

[0160] In practical applications, Fourier transform is performed on the aligned vibration data and electrical signals to convert the time-domain signal into a frequency-domain representation. The Fourier transform reveals the distribution of the signal across different frequency components, yielding frequency-domain representations of both the vibration data and the electrical signal.

[0161] Furthermore, by calculating the phase difference between the frequency domain representation of the electrical signal and the frequency domain representation of the vibration data, phase compensation can be performed on the frequency domain representation of the vibration data. The purpose of phase compensation is to eliminate phase distortion introduced during signal transmission or processing, ensuring that the phase relationship between the two signals in the frequency domain is accurate.

[0162] Therefore, after completing time alignment and phase compensation, the coherence between the frequency domain representation of the phase-compensated vibration data and the frequency domain representation of the electrical signal at different frequencies can be calculated. Coherence is an index that measures the degree of linear correlation between two signals in the frequency domain, with a value between 0 and 1, where 1 indicates perfect correlation and 0 indicates no correlation.

[0163] Optionally, a coherence threshold is set. When the coherence between the vibration data and the electrical signal at different frequencies is lower than the coherence threshold, the steps to determine that the vibration of the speaker enclosure is external interference vibration include:

[0164] When the vibration energy of the loudspeaker enclosure is significant, time-frequency domain analysis is performed on the vibration data and the electrical signal to obtain the time-frequency distribution.

[0165] In the time-frequency distribution, frequencies and time periods in which the vibration data and the electrical signal have high coherence are identified;

[0166] For frequencies and time periods where the vibration data and the electrical signal have high coherence, the transient characteristics of the vibration signal of the loudspeaker enclosure are analyzed; the transient characteristics include the steepness and duration of the vibration rise edge.

[0167] When the vibration signal of the speaker enclosure exhibits transient characteristics unique to non-drive signals within a frequency and time period in which the vibration data and the electrical signal are highly coherent, the vibration of the speaker enclosure is determined to be external interference vibration.

[0168] When the vibration signal of the speaker enclosure exhibits transient characteristics consistent with the driving signal within a frequency and time period in which the vibration data and the electrical signal are highly coherent, the vibration of the speaker enclosure is determined to be caused by the driving signal.

[0169] Specifically, "significant vibration energy in the loudspeaker enclosure" refers to a situation where the energy level of the vibration data collected by the vibration sensing device is significantly higher than the background noise level within a certain time period, indicating the occurrence of one or more vibration events. Time-frequency domain analysis at this point can more effectively capture the details of the vibration events. "Time-frequency distribution" can be obtained through methods such as Short-Time Fourier Transform (STFT), wavelet transform, or Wigner-Ville distribution, which can simultaneously display the energy distribution of the signal in time and frequency, thus revealing the dynamic characteristics of the signal. "Transient characteristics" describe the rapid changes in the vibration signal in the time domain. Among them, "steepness of the rise edge" refers to the speed at which the vibration signal rises rapidly from a low-energy state to a high-energy state, usually measured by calculating the slope or rate of change of the signal during the rise phase; "duration" refers to the effective time length of the vibration signal from start to finish. These characteristics are crucial for distinguishing vibrations from different sources. For example, transient impacts typically have a very steep rise edge and a short duration, while resonances caused by driving signals typically have a relatively gentle rise edge and a longer duration. "Transient characteristics unique to non-drive signals" refer to transient features that are significantly different from the vibration characteristics produced by the electrical signals of the speaker driver unit. For example, an extremely short and high-energy impact vibration may have a much steeper rise edge and a very short duration than the vibration produced by the drive signal. "Transient characteristics consistent with drive signals" refer to transient characteristics that are highly similar to the vibration characteristics produced by the electrical signals of the speaker driver unit in terms of rise edge steepness and duration.

[0170] In some preferred embodiments, it is assumed that the loudspeaker is playing a test signal and the vibration sensing device is collecting vibration data from the loudspeaker enclosure. At this time, if the loudspeaker enclosure is unexpectedly subjected to a brief but strong external impact, such as a slight knock from an object, in conventional methods relying solely on coherence thresholds, this impact may temporarily increase coherence because it could couple with the drive signal or have a measurement correlation for a very short time. This could then be incorrectly identified as vibration caused by the drive signal.

[0171] However, according to the solution of this application, when significant vibration energy of the speaker enclosure is detected and the vibration data and electrical signal exhibit high coherence in a certain time-frequency region, the system further analyzes the transient characteristics of the vibration signal within the frequency and time period where the vibration data and electrical signal exhibit high coherence. For the aforementioned external impact, the rising edge of its vibration signal will be extremely steep and its duration very short, which is significantly different from the transient characteristics of the continuous resonant vibration generated by the speaker driver unit driving the enclosure through electrical signals (which typically has a relatively flat rising edge and a longer duration). Therefore, even if the coherence is temporarily high, the system can accurately determine that the vibration is an external interference vibration based on its "transient characteristics unique to the non-drive signal," thereby excluding it from the structural resonance analysis.

[0172] Conversely, if the loudspeaker enclosure does indeed exhibit structural resonance at a certain frequency caused by the driving signal, its vibration signal will show high coherence in its time-frequency distribution, and its transient characteristics (such as the steepness and duration of the rise time) will be highly consistent with the characteristics of the driving signal. In this case, the system will determine that the vibration is caused by the driving signal and include it in the structural resonance identification range. In this way, this application can effectively avoid misjudgment and ensure the accuracy of structural resonance identification.

[0173] Optionally, the step of performing spectral analysis on the vibration data according to a preset resolution to obtain the preset resolution spectrum includes:

[0174] Before performing spectral analysis on the vibration data according to the preset resolution, the vibration data is subjected to real-time noise type identification to distinguish between transient impact noise and non-stationary background noise.

[0175] When identified as transient impact noise, adaptive threshold limiting is used to suppress vibration data that exceeds the instantaneous energy threshold.

[0176] When non-stationary background noise is identified, a noise suppression method based on time-varying filters is adopted. The filter parameters are dynamically adjusted according to the real-time spectral characteristics of the noise to filter the vibration data.

[0177] The vibration data after noise suppression is subjected to spectral analysis at a preset resolution to obtain the preset resolution spectrum.

[0178] Specifically, "real-time noise type identification" refers to the system's ability to automatically determine the type of noise based on the time-domain and frequency-domain characteristics of vibration data. For example, transient impact noise typically manifests as short-duration, high-energy pulse signals, characterized by a steep rise time and rapid decay in the time domain; while non-stationary background noise may exhibit a time-varying spectral distribution, such as fan noise or air conditioner noise in the environment. This identification can be achieved by analyzing statistical characteristics of the signal, such as kurtosis, skewness, short-duration energy, zero-crossing rate, and spectral change rate.

[0179] "Transient impact noise" can be understood as short-duration, high-amplitude vibration signals caused by sudden external events, such as objects hitting a speaker enclosure or sudden knocking sounds. When identified as transient impact noise, "adaptive threshold limiting processing" is used to suppress vibration data exceeding the "instantaneous energy threshold." Adaptive threshold limiting processing refers to dynamically adjusting the limiting threshold based on the statistical characteristics of the real-time signal (such as mean, variance, or peak value), rather than using a fixed threshold. When the instantaneous energy of the vibration data exceeds this dynamically adjusted threshold, the excess portion is reduced or zeroed out to eliminate or significantly reduce the impact noise. Its purpose is to effectively remove sudden, non-periodic high-energy noise components and avoid misleading the spectral analysis results.

[0180] In practical applications, "non-stationary background noise" specifically refers to background noise whose statistical characteristics (such as mean, variance, and spectral distribution) change over time, such as human voices, traffic noise, and equipment operating noise in the environment. When identified as non-stationary background noise, a "noise suppression method based on time-varying filters" is used. This method dynamically adjusts filter parameters according to the real-time spectral characteristics of the noise to filter the vibration data. A time-varying filter is a filter whose filtering characteristics (such as cutoff frequency, bandwidth, and gain) can be adjusted according to the real-time characteristics of the input signal. For example, the noise spectrum can be analyzed in real time using methods such as Short-Time Fourier Transform (STFT) or Wavelet Transform, and then an adaptive filter (such as a Wiener filter, Kalman filter, or spectral subtraction) can be designed or adjusted to suppress the noise while preserving as much effective signal as possible. The goal is to specifically remove persistent background noise with constantly changing spectral characteristics, thereby improving the signal-to-noise ratio of the effective vibration signal.

[0181] In some preferred embodiments, a specific example is given below. Suppose that during the playback of a test signal to excite the speaker enclosure to vibrate and to collect vibration data, two types of noise exist simultaneously: one is transient impact noise caused by someone accidentally touching the speaker enclosure in the room; the other is a low-frequency humming sound generated by the continuous operation of the room's air conditioner, whose spectral characteristics fluctuate slightly with changes in compressor load and belong to non-stationary background noise.

[0182] First, the raw vibration data collected by the vibration sensing device is sent to the real-time noise type identification module. This module analyzes the instantaneous energy, kurtosis, and spectral change rate of the signal to accurately distinguish short-duration high-energy impact signals as transient impact noise, while continuous low-frequency signals with slowly changing spectra are non-stationary background noise.

[0183] When transient impact noise is detected, the system immediately initiates adaptive threshold limiting processing. For example, if impact noise causes the instantaneous amplitude of the vibration signal to suddenly spike to levels far exceeding normal vibration levels at a certain point in time, this processing dynamically calculates an instantaneous energy threshold based on the statistical characteristics of the current signal. Once the signal exceeds this threshold, the excess portion is reduced, thereby effectively suppressing the interference of impact noise on subsequent spectrum analysis.

[0184] Simultaneously, for identified non-stationary background noise, the system activates a noise suppression method based on a time-varying filter. This method continuously monitors the real-time spectral characteristics of the air conditioner noise. For example, when the air conditioner compressor starts, its noise spectrum may increase at a certain frequency point. The time-varying filter dynamically adjusts its filtering parameters (such as center frequency, bandwidth, and attenuation) according to these real-time spectral changes to accurately filter out or attenuate background noise without causing excessive damage to the structural vibration signals of the speaker enclosure itself.

[0185] After the above two noise suppression processes, the obtained vibration data will be significantly cleaner. Subsequently, these noise-suppressed vibration data are subjected to spectral analysis at a preset resolution. The resulting preset resolution spectrum will more accurately reflect the true structural resonance characteristics of the speaker cabinet, thus providing reliable basic data for subsequent sound field parameter optimization.

[0186] Optionally, the step of performing multi-peak detection on the preset resolution spectrum to obtain several potential resonant peaks includes:

[0187] Peak shape analysis based on local features is performed on the spectrum at a preset resolution to identify potential resonant regions with broad tops or irregular shapes.

[0188] The potential resonant region is adaptively segmented, and each non-ideal peak shape is decomposed into several narrower sub-regions;

[0189] Perform local maxima search independently for each sub-region to identify local peak points within each sub-region;

[0190] Based on the frequency and intensity of the local peak points, as well as the frequency interval and relative intensity between the current local peak point and the adjacent local peak points, determine whether each local peak point belongs to an independent structural resonance;

[0191] When two or more independent structural resonances are identified, the independent structural resonances are recorded as potential resonance peaks.

[0192] Specifically, peak shape analysis based on local features refers to determining whether a peak exhibits an ideal, sharp shape by examining features such as local gradient, curvature, and symmetry of the spectral curve. When a potential resonant region with a broad top or irregular shape is identified, it indicates that the region may contain multiple overlapping or closely adjacent resonant components. Further, adaptive segmentation of the potential resonant region involves intelligently dividing these complex peak regions with non-ideal shapes into multiple narrower sub-regions based on features such as local minima, inflection points, or curvature changes. Each sub-region is considered to contain one or a few independent resonant components. Based on this, a local maximum search is performed independently on each sub-region to accurately identify local peak points within each sub-region. These local peak points are the highest-energy frequency points within the sub-region, initially representing potential resonant centers. Subsequently, based on the frequency and intensity of the local peak points, as well as the frequency interval and relative intensity between the current local peak point and its adjacent local peak points, it can be determined whether each local peak point belongs to an independent structural resonance. For example, if the frequency interval between two local peaks is too small and their relative intensity difference is not significant, it may indicate that they are different manifestations of the same broad resonant peak, or that they are tightly coupled resonances. Conversely, if the frequency interval is large enough and the intensity is independent, they can be identified as independent structural resonances. When two or more independent structural resonances are identified, these identified independent structural resonances will be recorded as potential resonant peaks for use in subsequent steps such as quality factor and bandwidth calculations.

[0193] This application also discloses a sound field parameter optimization system for performing sound field parameter optimization, combined with... Figure 3 As shown, the sound field parameter optimization system 1 includes:

[0194] Vibration sensing activation module 11 is used to activate the vibration sensing device preset on the speaker enclosure;

[0195] The vibration data acquisition module 12 is used to drive the loudspeaker to play a test signal to excite the loudspeaker enclosure to vibrate, and to drive the vibration sensing device to collect the vibration data of the loudspeaker enclosure.

[0196] The vibration data identification module 13 is used to receive vibration data and perform frequency domain analysis on the vibration data to identify the structural resonant frequency and structural resonant intensity of the speaker enclosure.

[0197] The resonance information storage module 14 is used to store the structural resonance frequency and structural resonance intensity as structural resonance information.

[0198] The compensation scheme determination module 15 is used to determine a preliminary low-frequency compensation scheme based on the room acoustic characteristics of the room where the speaker cabinet is located.

[0199] The parameter optimization execution module 16 is used to compare the preliminary low-frequency compensation scheme with the structural resonance information, determine whether the preliminary low-frequency compensation scheme has the risk of exciting the structural resonance of the loudspeaker cabinet, and obtain the excitation risk judgment result; based on the excitation risk judgment result, the preliminary low-frequency compensation scheme is adjusted and updated to obtain the sound field parameter optimization scheme.

[0200] To better understand the sound field parameter optimization system proposed in this application, some key modules involved are described in detail below.

[0201] A vibration sensing activation module is used to activate a vibration sensing device pre-installed on the speaker enclosure. This module can be a control circuit or a software instruction set, responsible for sending a activation signal to the vibration sensing device to bring it into operation. Specific implementation methods for activating the vibration sensing device can be found in the descriptions in the above method embodiments, and will not be repeated here.

[0202] The vibration data acquisition module is used to drive the loudspeaker to play a test signal to excite the loudspeaker enclosure to vibrate, and to drive the vibration sensing device to collect the vibration data of the loudspeaker enclosure. This module may include a signal generator to generate the test signal and drive the loudspeaker to play it through a power amplifier. Simultaneously, this module is also responsible for receiving the vibration signal from the vibration sensing device and performing analog-to-digital conversion and preliminary data preprocessing. For specific implementation methods of driving the loudspeaker to play the test signal and collecting vibration data, please refer to the relevant descriptions in the above method embodiments, which will not be repeated here.

[0203] The vibration data identification module receives vibration data and performs frequency domain analysis to identify the structural resonant frequency and intensity of the speaker enclosure. This module can integrate Fast Fourier Transform (FFT) algorithms, peak detection algorithms, and resonance characteristic analysis algorithms. Specific implementation methods for receiving vibration data and performing frequency domain analysis to identify the structural resonant frequency and intensity can be found in the descriptions in the above method embodiments and will not be repeated here.

[0204] A resonance information storage module is used to store the structural resonant frequency and structural resonant intensity as structural resonance information. This module can be a non-volatile memory or a database interface, used to store the identified structural resonance information in the form of structured data. Specific implementation methods for storing structural resonance information can be found in the relevant descriptions in the above method embodiments, and will not be repeated here.

[0205] The compensation scheme determination module is used to determine a preliminary low-frequency compensation scheme based on the room acoustic characteristics of the room where the speaker enclosure is located. This module can receive room impulse response data collected from external measuring equipment and use acoustic analysis algorithms to identify low-frequency problems in the room. Based on these analysis results, the module generates a preliminary equalizer curve or filter parameter set. For specific implementation details regarding the determination of the preliminary low-frequency compensation scheme based on room acoustic characteristics, please refer to the relevant descriptions in the above method embodiments, which will not be repeated here.

[0206] The parameter optimization execution module compares the preliminary low-frequency compensation scheme with the structural resonance information to determine whether the preliminary low-frequency compensation scheme poses a risk of exciting the structural resonance of the speaker enclosure, thus obtaining an excitation risk assessment result. Based on the excitation risk assessment result, the preliminary low-frequency compensation scheme is adjusted and updated to obtain the sound field parameter optimization scheme. This module is the core decision-making unit of the system. It receives the preliminary low-frequency compensation scheme and structural resonance information, uses a risk assessment algorithm to determine potential excitation risks, and adjusts the compensation scheme accordingly. Specific implementation methods for comparing, judging, adjusting, and updating the preliminary low-frequency compensation scheme to obtain the sound field parameter optimization scheme can be found in the relevant descriptions in the above method embodiments, and will not be repeated here.

[0207] The sound field parameter optimization system proposed in this application aims to solve the problem of low-frequency compensation errors caused by neglecting the structural resonance of the speaker enclosure in traditional sound field optimization. Traditional methods often only focus on the acoustic characteristics of the room and determine the low-frequency compensation scheme by measuring the room impulse response. However, this method is prone to the problem of "paying attention to one thing but losing sight of another" in practical applications. That is, while correcting the room acoustic problems, it inadvertently introduces the mechanical vibration noise of the speaker itself.

[0208] The core innovation of this application lies in its systematic incorporation of the structural resonance characteristics of the speaker enclosure into the sound field optimization considerations. Vibration data of the enclosure is acquired through a vibration sensing activation module and a vibration data acquisition module. The vibration data identification module accurately identifies the structural resonance frequency and intensity of the enclosure, which is then recorded by the resonance information storage module. Based on this, the parameter optimization execution module can intelligently compare the preliminary low-frequency compensation scheme with this structural resonance information, thereby predicting and avoiding the risk of exciting harmful resonances in the enclosure before the compensation scheme is implemented. This prediction and adjustment mechanism enables the system of this application to provide a more robust and optimized sound field solution.

[0209] Compared to the closest existing technology, the system of this application has the advantage of comprehensiveness and precision. Existing systems typically lack the ability to perceive and evaluate the physical characteristics of the speaker cabinet itself in real time, resulting in optimization results that may be theoretically effective but significantly compromised in actual listening experience due to cabinet resonance. The system of this application effectively overcomes this deficiency of existing technologies by integrating a dedicated module to address cabinet resonance issues. Through the perception, analysis, and risk assessment of vibration data, the system of this application can provide a purer, more stable, and uncolored sound field experience, significantly avoiding problems such as bass booming, hollow sound field, and loss of clarity in the mid-low frequencies. This comprehensive optimization strategy makes the system of this application a significant technological advancement and innovation in solving complex acoustic problems.

[0210] The above are merely embodiments of this application and are not intended to limit the scope of protection of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A method for optimizing sound field parameters, characterized in that, include: Activate the vibration sensing device preset on the speaker enclosure; The speaker is driven to play a test signal to excite the speaker enclosure to vibrate, and the vibration sensing device is driven to collect the vibration data of the speaker enclosure. The vibration data is received, and frequency domain analysis is performed on the vibration data to identify the structural resonant frequency and structural resonant intensity of the speaker enclosure. The structural resonant frequency and structural resonant intensity are stored as structural resonant information; Based on the acoustic characteristics of the room where the speaker enclosure is located, a preliminary low-frequency compensation scheme is determined; The preliminary low-frequency compensation scheme is compared with the structural resonance information to determine whether the preliminary low-frequency compensation scheme has the risk of exciting the structural resonance of the speaker cabinet, and the excitation risk judgment result is obtained. Based on the excitation risk assessment results, the preliminary low-frequency compensation scheme is adjusted and updated to obtain an optimized sound field parameter scheme.

2. The sound field parameter optimization method according to claim 1, characterized in that, The preliminary low-frequency compensation scheme is compared with the structural resonance information to determine whether the preliminary low-frequency compensation scheme has the risk of exciting the structural resonance of the speaker cabinet, and the excitation risk judgment result is obtained. The steps for adjusting and updating the preliminary low-frequency compensation scheme based on the excitation risk assessment results to obtain the sound field parameter optimization scheme include: It continuously and synchronously acquires vibration data of the speaker enclosure, real-time electrical signals of the speaker driver unit, and temperature and humidity data of the environment where the speaker is located; Blind source separation is performed on the real-time electrical signal and vibration data to decompose the inherent resonant vibration components of the loudspeaker enclosure; From the inherent resonant vibration components, the resonant frequency and quality factor are extracted. Combined with the temperature and humidity data, the resonant frequency drift and resonant intensity drift of the speaker enclosure are monitored, and the resonant characteristic records of the speaker enclosure are dynamically updated. The preliminary low-frequency compensation scheme is compared with the resonance characteristic record, and a multi-dimensional risk assessment is performed on each adjustment frequency point in the preliminary low-frequency compensation scheme. Based on the results of the multidimensional risk assessment, a compensation strategy is selected to adjust and update the initial low-frequency compensation scheme, resulting in an optimized sound field parameter scheme.

3. The sound field parameter optimization method according to claim 1, characterized in that, The steps of receiving the vibration data, performing frequency domain analysis on the vibration data, and identifying the structural resonant frequency and structural resonant intensity of the speaker enclosure include: The vibration data is subjected to spectral analysis at a preset resolution to obtain a spectrum at the preset resolution. Multi-peak detection is performed on the preset resolution spectrum to obtain several potential resonance peaks; Calculate the quality factor and bandwidth corresponding to each potential resonant peak, and based on the quality factor, bandwidth, frequency interval between each potential resonant peak and adjacent peaks, and relative intensity, determine whether each potential resonant peak belongs to an independent structural resonance, and obtain the peak independence judgment result. When the peak independence judgment result indicates that there are overlapping spectra, curve fitting is performed on the overlapping spectra to decompose them into independent resonance components, and the center frequency and intensity of the independent resonance components are obtained. Harmonic resonance is obtained by analyzing the harmonic relationships of the center frequencies of independent resonant components. The harmonic resonance and the preset fundamental frequency resonance are recorded to obtain the structural resonant frequency and structural resonant intensity of the speaker enclosure.

4. The sound field parameter optimization method according to claim 1, characterized in that, The preliminary low-frequency compensation scheme is compared with the structural resonance information to determine whether the preliminary low-frequency compensation scheme has the risk of exciting the structural resonance of the speaker cabinet, and the excitation risk judgment result is obtained. The steps for adjusting and updating the preliminary low-frequency compensation scheme based on the excitation risk assessment results to obtain the sound field parameter optimization scheme include: The transient power detector preset at the power amplifier output of the speaker driver unit is activated; The transient power detector is driven to monitor the instantaneous output power of the speaker driver unit; When the instantaneous output power is detected to exceed a preset instantaneous power threshold within a preset range corresponding to the adjustment frequency point of the preliminary low-frequency compensation scheme, the risk assessment result is determined to be a risk of transient excitation of the speaker enclosure's structural resonance. Based on the determination of the excitation risk, a transient suppression filter is introduced at the corresponding adjustment frequency point; The transient suppression filter is activated when the instantaneous output power exceeds a preset instantaneous power threshold, and the signal within the preset range corresponding to the adjusted frequency point is instantaneously attenuated. The transient suppression filter is de-attenuated when the instantaneous output power is lower than a preset instantaneous power threshold.

5. The sound field parameter optimization method according to claim 1, characterized in that, The steps of receiving the vibration data, performing frequency domain analysis on the vibration data, and identifying the structural resonant frequency and structural resonant intensity of the speaker enclosure include: The electrical signals of the speaker driver unit are acquired while the speaker plays a test signal to excite the speaker cabinet to vibrate. Time-frequency domain analysis is performed on the vibration data and electrical signal to calculate the coherence of the vibration data and electrical signal at different frequencies. A coherence threshold is set. When the coherence between the vibration data and the electrical signal at different frequencies is lower than the coherence threshold, the vibration of the speaker enclosure is determined to be external interference vibration. Frequency domain analysis was performed on the vibration data of the loudspeaker enclosure, which was determined to be non-external interference vibration, to identify the structural resonant frequency and structural resonant intensity of the loudspeaker enclosure.

6. The sound field parameter optimization method according to claim 5, characterized in that, The step of performing time-frequency domain analysis on the vibration data and the electrical signal to calculate the coherence of the vibration data and the electrical signal at different frequencies includes: The time delay between the electrical signal and the vibration data is determined by cross-correlation analysis. The vibration data is time-aligned based on the time delay. Perform Fourier transform on the aligned vibration data and the electrical signal to obtain the frequency domain representation of the vibration data and the frequency domain representation of the electrical signal; Phase compensation is performed on the frequency domain representation of the vibration data by calculating the phase difference between the frequency domain representation of the electrical signal and the frequency domain representation of the vibration data; The coherence of the frequency domain representation of the vibration data after phase compensation and the frequency domain representation of the electrical signal at different frequencies is calculated.

7. The sound field parameter optimization method according to claim 5, characterized in that, The step of determining that the vibration of the speaker enclosure is an external interference vibration when the coherence between the vibration data and the electrical signal at different frequencies is lower than the set coherence threshold includes: When the vibration energy of the loudspeaker enclosure is significant, time-frequency domain analysis is performed on the vibration data and the electrical signal to obtain the time-frequency distribution. In the time-frequency distribution, frequencies and time periods in which the vibration data and the electrical signal have high coherence are identified; For frequencies and time periods where the vibration data and the electrical signal have high coherence, the transient characteristics of the vibration signal of the loudspeaker enclosure are analyzed; the transient characteristics include the steepness and duration of the vibration rise edge. When the vibration signal of the speaker enclosure exhibits transient characteristics unique to non-drive signals within a frequency and time period in which the vibration data and the electrical signal are highly coherent, the vibration of the speaker enclosure is determined to be external interference vibration. When the vibration signal of the speaker enclosure exhibits transient characteristics consistent with the driving signal within a frequency and time period in which the vibration data and the electrical signal are highly coherent, the vibration of the speaker enclosure is determined to be caused by the driving signal.

8. The sound field parameter optimization method according to claim 3, characterized in that, The step of performing spectral analysis on the vibration data according to a preset resolution to obtain a preset resolution spectrum includes: Before performing spectral analysis on the vibration data according to a preset resolution, the vibration data is subjected to real-time noise type identification to distinguish between transient impact noise and non-stationary background noise. When identified as transient impact noise, adaptive threshold limiting is used to suppress vibration data that exceeds the instantaneous energy threshold. When non-stationary background noise is identified, a noise suppression method based on time-varying filters is adopted. The filter parameters are dynamically adjusted according to the real-time spectral characteristics of the noise to filter the vibration data. The vibration data after noise suppression is subjected to spectral analysis at a preset resolution to obtain the preset resolution spectrum.

9. The sound field parameter optimization method according to claim 3, characterized in that, The step of performing multi-peak detection on the preset resolution spectrum to obtain several potential resonance peaks includes: Peak shape analysis based on local features is performed on the preset resolution spectrum to identify potential resonant regions with broad tops or irregular shapes; The potential resonant region is adaptively segmented, and each non-ideal shaped peak is decomposed into several narrower sub-regions; Perform local maxima search independently for each sub-region to identify local peak points within each sub-region; Based on the frequency and intensity of the local peak points, as well as the frequency interval and relative intensity between the current local peak point and the adjacent local peak points, determine whether each local peak point belongs to an independent structural resonance; When two or more independent structural resonances are identified, the independent structural resonances are recorded as potential resonance peaks.

10. A sound field parameter optimization system for performing sound field parameter optimization, characterized in that, include: The vibration sensing activation module is used to activate the vibration sensing device preset on the speaker enclosure; The vibration data acquisition module is used to drive the loudspeaker to play a test signal to excite the loudspeaker enclosure to vibrate, and to drive the vibration sensing device to collect the vibration data of the loudspeaker enclosure. The vibration data identification module is used to receive the vibration data and perform frequency domain analysis on the vibration data to identify the structural resonant frequency and structural resonant intensity of the speaker enclosure. A resonance information storage module is used to store the structural resonance frequency and structural resonance intensity as structural resonance information. The compensation scheme determination module is used to determine a preliminary low-frequency compensation scheme based on the room acoustic characteristics of the room where the speaker cabinet is located; The parameter optimization execution module is used to compare the preliminary low-frequency compensation scheme with the structural resonance information, determine whether the preliminary low-frequency compensation scheme has the risk of exciting the structural resonance of the speaker cabinet, and obtain the excitation risk judgment result; Based on the excitation risk assessment results, the preliminary low-frequency compensation scheme is adjusted and updated to obtain an optimized sound field parameter scheme.

Images