Audio processor remote network commissioning method and system

Abstract

The application discloses a kind of audio processor remote network debugging method and system.Method includes: the instantaneous spectral energy distribution of original program signal is acquired, and energy gap band that can be used for injection detection signal is identified;Through psychoacoustic model, the injection level of detection signal under the constraint of audible probability is determined;Based on injection level, mixed signal spectrum is obtained by injecting detection signal;Original program signal spectrum is acquired, and the difference processing is carried out to mixed signal spectrum and original program signal spectrum, and the signal-to-noise ratio of each band is enhanced processing;Through the impedance curve of enhanced signal-to-noise ratio reconstruction, the abnormal position of line impedance is confirmed, and the remote network debugging of audio is completed.The application provides a kind of audio processor remote network debugging method and system, to solve the technical problem that spectrum complementary strategy is beneficial in energy domain but harmful in perception domain in prior art, cannot realize high-precision impedance online detection under the premise of ensuring that detection signal is inaudible.

Description

Technical Field

[0001] This invention relates to the field of audio processing, and more particularly to a method and system for remote network debugging of an audio processor. Background Technology

[0002] With the widespread application of remote network debugging technology in the field of audio processing, it plays an increasingly important role in scenarios such as live performances, broadcasts, and conference systems. One of the core technologies for achieving high-quality remote debugging is online detection of speaker line impedance. Its goal is to accurately identify abnormal conditions in speaker lines under the complex acoustic environment of continuous program signal broadcasting, without affecting the listener's subjective listening experience.

[0003] Currently, the mainstream remote debugging methods for audio processors mainly employ a spectrum complementarity injection strategy. This method first analyzes the instantaneous spectral energy distribution of the program signal to identify low-energy frequency bands as injection gaps for the detection signal. Then, a detection signal of a specific frequency band is injected into this gap to obtain the impedance response of the speaker circuitry. The core assumption of this method is that selecting a frequency band complementary to the program signal spectrum for the injection of the detection signal allows for the utilization of the energy gap in that band to achieve a higher impedance measurement signal-to-noise ratio, thus theoretically balancing the contradiction between detection accuracy and signal concealment. However, this method does not fully consider the psychoacoustic masking effect of the human auditory system. Due to the frequency domain asymmetry of the masking effect, whether a frequency band can effectively mask the detection signal does not depend on the absolute energy level of that band, but rather on whether it is masked by adjacent high-energy frequency bands. On the contrary, the energy gap frequency band selected by the spectrum complementarity strategy is precisely the region with the lowest program signal energy and therefore the weakest masking ability. Injecting a detection signal into this band, even at an extremely low injection level, actually has the highest probability of being exposed as audible. To prevent the detection signal from being detected by the audience, the system has to further reduce the injection level in that frequency band. This causes the impedance anomaly information in that band to be drowned out by the measurement noise floor, making it impossible to accurately identify the fault condition of the speaker circuit. Therefore, the biggest drawback of the existing remote debugging method is that its spectrum complementarity strategy is beneficial in the energy domain but harmful in the perception domain. It cannot achieve high-precision online impedance detection while ensuring that the detection signal is inaudible. In other words, there is an inherent contradiction between the concealment of the detection signal and the detection accuracy. Summary of the Invention

[0004] This invention provides a method and system for remote network debugging of an audio processor, which solves the technical problem that the spectrum complementarity strategy in the prior art is beneficial in the energy domain but harmful in the perception domain, and cannot achieve high-precision online impedance detection while ensuring that the detection signal is inaudible.

[0005] The present invention discloses the following technical solutions: In a first aspect, embodiments of the present invention provide a method for remote network debugging of an audio processor, the method comprising: The instantaneous spectral energy distribution of the original program signal is obtained, and the energy gap frequency bands that can be used to inject the detection signal are identified. The masking energy deficiency in the energy gap frequency band is analyzed by psychoacoustic modeling to determine the injection level of the detection signal under the audible probability constraint; Based on the injection level, the detection signal is injected into the energy gap frequency band to obtain the mixed signal spectrum; The original program signal spectrum is obtained, the mixed signal spectrum and the original program signal spectrum are differentially processed to calculate the signal-to-noise ratio of each frequency band, and the signal-to-noise ratio of each frequency band is enhanced to obtain the enhanced signal-to-noise ratio of each frequency band. By reconstructing the impedance curve using the enhanced signal-to-noise ratio, the abnormal location of the line impedance is identified, and remote network debugging of the audio is completed.

[0006] Optionally, acquiring the instantaneous spectral energy distribution of the original program signal and identifying the energy gap frequency band that can be used to inject the detection signal includes: Based on the instantaneous spectral energy distribution, the continuous frequency point regions in each frequency band where the spectral energy is lower than a preset threshold are identified; Extract the start frequency, cutoff frequency, lowest energy within the energy gap, and average energy around the energy gap of the continuous frequency region. Calculate the absolute value of the lowest energy within the energy gap minus the average energy around the energy gap to obtain the gap amplitude, and determine the energy gap position and energy gap amplitude for each frequency band. Based on the energy gap location and the energy gap amplitude, energy gap frequency bands that can be used to inject detection signals are selected.

[0007] Optionally, the step of analyzing the masking energy deficiency in the energy gap frequency band using a psychoacoustic model to determine the injection level of the detection signal under the constraint of audibility probability includes: The amount of masking energy missing in the energy gap frequency band is obtained by calculating the auditory masking threshold. Based on the amount of masking energy missing, the audibility probability of different test levels is evaluated using a psychoacoustic model; The audible probability is compared with a preset threshold, and the test level is iteratively tested from low to high. The upper limit of the level corresponding to when the audible probability does not exceed the preset threshold is determined as the injection level.

[0008] Optionally, the step of injecting the detection signal into the energy gap frequency band based on the injection level to obtain the mixed signal spectrum includes: Obtain impedance response amplitude data; Based on the impedance response amplitude data and the injection level, a detection signal is injected into the energy gap frequency band to obtain a preliminary mixed signal; The initial mixed signal is subjected to spectral conversion to obtain the mixed signal spectrum.

[0009] The process of acquiring the original program signal spectrum, performing differential processing on the mixed signal spectrum and the original program signal spectrum to calculate the signal-to-noise ratio (SNR) of each frequency band, and enhancing the SNR of each frequency band to obtain the enhanced SNR of each frequency band includes: The spectrum of the mixed signal and the spectrum of the original program signal are differentially processed, and the signal-to-noise ratio of each frequency band is calculated based on the differential results; The frequency bands with a signal-to-noise ratio below the noise floor threshold are marked as potential masking areas; Based on each potential obscured region, and in conjunction with the corresponding energy gap amplitude and the audible probability constraint, the actual injection amount of the potential obscured region is increased without exceeding the injection level, thereby obtaining the enhanced signal-to-noise ratio of each frequency band.

[0010] Optionally, based on each potential obscured region, and in conjunction with the corresponding energy gap amplitude and the audible probability constraint, the actual injection amount of the potential obscured region is increased without exceeding the injection level to obtain the enhanced signal-to-noise ratio for each frequency band, including: The initial injection boundary is obtained by weighted summing of the energy gap amplitude corresponding to the potential shielding region and the energy difference between the potential shielding region and the adjacent frequency band; Based on the initial injection boundary, the feasibility of injection is determined by the audible probability constraint, and the adjusted boundary threshold is obtained. Within the adjusted boundary threshold, combined with the spectral masking threshold, the injection level of the potential masking region is dynamically improved through the inter-band energy balance method, wherein the spectral masking threshold is extracted from the audio spectrum by the psychoacoustic model; Based on the enhanced injection level, the logarithmic ratio of the signal-to-noise ratio after injection is calculated for each frequency band to obtain the enhanced signal-to-noise ratio for each frequency band.

[0011] The process of reconstructing the impedance curve using the enhanced signal-to-noise ratio to identify abnormal locations in the line impedance and completing remote network debugging of the audio includes: Based on the enhanced signal-to-noise ratio, the noise compensation coefficient is obtained by calculating the ratio of noise level to signal strength in each frequency band, and the impedance curve is reconstructed. Based on the impedance curve, the frequency bands whose amplitude in the impedance curve exceeds a preset abnormal threshold are marked as abnormal frequency bands; Based on the abnormal frequency band, the amplitude difference between the abnormal frequency band and its adjacent frequency bands is calculated to obtain the adjacent frequency band comparison value; The location where the comparison value of the adjacent frequency bands exceeds the preset comparison threshold is identified as an abnormal location of the line impedance, thus completing the remote network debugging of the audio.

[0012] Secondly, embodiments of the present invention provide a remote network debugging system for an audio processor, comprising: The acquisition module is used to acquire the instantaneous spectral energy distribution of the original program signal and identify the energy gap frequency bands that can be used to inject the detection signal. The determination module is used to analyze the masking energy deficiency in the energy gap frequency band through a psychoacoustic model and determine the injection level of the detection signal under the audible probability constraint. A mixing module is used to inject the detection signal into the energy gap frequency band based on the injection level to obtain a mixed signal spectrum; An enhancement module is used to acquire the original program signal spectrum, perform differential processing on the mixed signal spectrum and the original program signal spectrum, calculate the signal-to-noise ratio of each frequency band, and enhance the signal-to-noise ratio of each frequency band to obtain the enhanced signal-to-noise ratio of each frequency band. The debugging module is used to reconstruct the impedance curve through the enhanced signal-to-noise ratio, identify the abnormal location of the line impedance, and complete the remote network debugging of the audio.

[0013] Thirdly, another embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the audio processor remote network debugging method as described above.

[0014] Another embodiment of the present invention also provides a computer program product, including a computer program or instructions, which, when executed by a device, implement the steps of the aforementioned remote network debugging method for an audio processor.

[0015] Compared with the prior art, the present invention has the following beneficial effects: This invention acquires the instantaneous spectral energy distribution of the program signal through a spectral energy acquisition unit, identifies energy gap frequency bands, and provides candidate regions for covert injection of the detection signal. Further, an auditory masking threshold calculation unit analyzes the masking deficiencies of the energy gaps, uses a psychoacoustic model to evaluate the audibility probability of the detection signal, and determines the maximum level under the constraint of audibility probability as the injection level, fundamentally ensuring the inaudibility of the detection signal and solving the problem of harmful effects in the perceptual domain of traditional methods. By differentially processing the mixed signal and the original signal to calculate the signal-to-noise ratio (SNR) of each frequency band, and enhancing the injection amount for potential masking regions with excessively low SNR, the loss of detection accuracy caused by audibility constraints is effectively compensated. Finally, the impedance curve is reconstructed based on the enhanced SNR to locate anomalies, achieving accurate online detection of speaker line impedance while ensuring signal covertness. Attached Figure Description

[0016] To more clearly illustrate the technical solution of this application, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0017] Figure 1 This is a flowchart illustrating a remote network debugging method for an audio processor provided in an embodiment of the present invention; Figure 2 This is a schematic diagram illustrating the specific implementation process of a remote network debugging method for an audio processor provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of the injection level determination process of a remote network debugging method for an audio processor provided in an embodiment of the present invention; Figure 4 This is a schematic diagram of the mixed signal spectrum generation process of a remote network debugging method for an audio processor provided in an embodiment of the present invention; Figure 5 This is a schematic diagram of a remote network debugging device for an audio processor provided in an embodiment of the present invention. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains; the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the invention; the terms “comprising” and “having”, and any variations thereof, in the specification, claims, and foregoing description of the invention, are intended to cover non-exclusive inclusion.

[0020] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of the invention. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0021] See Figure 1 To address the technical problem that existing spectral complementarity strategies are advantageous in the energy domain but detrimental in the perception domain, failing to achieve high-precision online impedance detection while ensuring the detection signal remains audible, an embodiment of this invention provides a remote network debugging method for an audio processor, comprising: S1: Obtain the instantaneous spectral energy distribution of the original program signal and identify the energy gap frequency band that can be used to inject the detection signal.

[0022] The audio processor's built-in spectrum energy acquisition unit performs time-frequency transformation on the input raw program signal. Specifically, the signal is digitized at a sampling rate of 44.1kHz, and a Fast Fourier Transform (FFT) is used to convert the time-domain signal into a frequency-domain representation, thereby obtaining the instantaneous spectrum energy distribution, i.e., the energy intensity at each frequency point. Based on this, energy gap frequency bands that can be used to inject detection signals are identified. The spectrum is divided into several frequency bands, such as the low-frequency band 20Hz–200Hz, and the arithmetic mean E of the energy values ​​of all frequency points in the current frame is calculated. avg And set the energy threshold T=0.5×E avg The process iterates through each frequency band, marking frequency regions with continuously lower energy values ​​than T as energy gaps. The starting and ending frequencies, as well as the difference between the lowest energy within the gap and the surrounding average energy, are recorded as gap amplitudes. To improve the stability of identification, it is preferable to analyze the spectrum of at least three consecutive frames, retaining only energy gaps that consistently appear in all three frames. Finally, frequency bands with larger gap amplitudes and stable positions are selected as energy gap frequency bands that can be used to inject detection signals for subsequent steps.

[0023] S2, by analyzing the masking energy deficiency in the energy gap frequency band using a psychoacoustic model, the injection level of the detection signal under the constraint of audible probability is determined.

[0024] After obtaining the energy gap frequency band, the auditory masking threshold calculation unit analyzes the masking energy loss in this band. Specifically, the ideal masking threshold is estimated based on the energy levels of adjacent frequency bands and compared with the actual masking energy within the energy gap frequency band to calculate the amount of masking energy loss. Subsequently, a psychoacoustic model is introduced to assess the probability that the detected signal is exposed as audible. This model divides the spectrum into critical bands, simulates the frequency selectivity of the basilar membrane in the human ear, and, combined with the frequency domain masking effect, calculates the degree to which the detected signal exceeds the masking threshold at a given injection level. Through statistical methods, such as assuming that auditory perception variability follows a Gaussian distribution, the excess is mapped to an audible probability. Finally, under the condition that the audible probability does not exceed a preset threshold (typically 0.05, 5% audible risk), an iterative testing method is adopted from low to high, gradually increasing the injection level and monitoring the corresponding audible probability until the maximum injection level that just meets the probability constraint is found, which is determined as the final injection level for the energy gap frequency band. This level is used in subsequent steps to generate the detection signal, ensuring that the injection intensity is maximized while remaining audibly imperceptible.

[0025] S3, based on the injection level, the detection signal is injected into the energy gap frequency band to obtain the mixed signal spectrum.

[0026] After determining the injection level, impedance response amplitude data is first obtained from the speaker lines connected to the audio processor. This data is calculated by outputting a test signal to the line and collecting voltage and current responses, reflecting the impedance magnitude distribution at different frequencies, and is used to assist in identifying the energy gap frequency band. Combined with the determined injection level (in dBFS), a detection signal is generated within the identified energy gap frequency band. The detection signal can be a single-tone or multi-tone sine wave, with its amplitude strictly controlled within the upper limit of the injection level to avoid exceeding the audible probability constraint. This detection signal is superimposed on the corresponding frequency band of the original program signal to form a preliminary mixed signal. Subsequently, a Fast Fourier Transform (FFT) is performed on the preliminary mixed signal to convert it into a frequency domain representation, obtaining the mixed signal spectrum. This spectrum contains the superposition of the original program signal spectrum and the detection signal spectrum, where the energy within the energy gap frequency band is filled by the detection signal, providing input for subsequent differential processing and signal-to-noise ratio calculation. The generation of the mixed signal spectrum ensures the quantizability and analyzability of the detection injection process.

[0027] S4, acquire the original program signal spectrum, perform differential processing on the mixed signal spectrum and the original program signal spectrum, calculate the signal-to-noise ratio of each frequency band, and enhance the signal-to-noise ratio of each frequency band to obtain the enhanced signal-to-noise ratio of each frequency band.

[0028] First, the spectrum of the original program signal is acquired. This spectrum can be calculated from the original program signal using the same Fast Fourier Transform (FFT) parameters as the mixed signal, ensuring frequency domain alignment. Then, the spectrum of the mixed signal and the spectrum of the original program signal are differentially processed. The difference value of the differential spectrum is calculated, which approximately reflects the energy increment introduced by the injected detection signal in each frequency band, and is used as a noise power estimate. Simultaneously, the power of the original program signal is substituted into the signal-to-noise ratio (SNR) formula, and the SNR is calculated for each frequency band to obtain the initial SNR distribution for each band. Based on this, the SNR is enhanced. Frequency bands with an SNR lower than a preset noise floor threshold (typically 10 dB) are marked as potential masking areas. For these areas, the energy gap amplitude and audibility probability constraint are fused: using the obtained energy gap amplitude and the energy difference between this area and adjacent frequency bands, the initial injection boundary is calculated using a weighted formula. Combined with the audibility probability constraint, the actual injection amount is iteratively adjusted, appropriately increasing the detection signal injection intensity in the potential masking areas without exceeding the original injection level. Finally, the signal-to-noise ratio (SNR) of these regions after injection is recalculated. By comparing the power changes before and after injection, the enhanced SNR of each frequency band is obtained, for example, improving regions that were originally below 10dB to above 15dB. The enhanced SNR is used for reliable reconstruction of subsequent impedance profiles, avoiding the omission of abnormal frequency bands in the speaker circuitry due to noise overload.

[0029] S5. By reconstructing the impedance curve using the enhanced signal-to-noise ratio, the abnormal location of the line impedance is confirmed, and the remote network debugging of the audio is completed.

[0030] The process involves reconstructing the impedance curve of the speaker circuit based on the enhanced signal-to-noise ratio (SNR) data for each frequency band. Specifically, the speaker circuit's response signal is acquired, and Fourier transform is used to separate the components of each frequency band. Noise compensation is applied to frequency bands with an SNR below 15dB, and the signal power of these bands is boosted to bring the SNR closer to the 15dB target. Using the compensated frequency band data, combined with the frequency domain ratio of voltage to current, the impedance magnitude is calculated point-by-point, and a complete impedance curve is generated. Then, the amplitude of each frequency band in the impedance curve is evaluated to identify any anomalies. A preset anomaly threshold is 1.5 times the normal impedance mean (a typical value, adjustable between 1.2 and 2.0 times depending on the speaker model). The curve is iterated through, recording frequency bands with amplitudes exceeding this threshold, and comparing the amplitude difference between these bands and adjacent bands. If the difference is significant, such as exceeding 30% of the mean, it is marked as a potential anomaly location. To accurately locate the specific location of the circuit impedance anomaly, Time Domain Reflectometry (TDR) is introduced. A pulse signal is sent to the speaker circuitry, the reflection time is measured, and the anomaly distance is calculated based on the signal propagation speed in the cable (typically 60%-80% of the speed of light). This distance, accurate to the centimeter level, is used to determine the physical location of faults such as open circuits, short circuits, or component aging. Finally, the graphical representation of the impedance curve, the list of abnormal frequency bands, the coordinates of the abnormal location, and suggested handling measures are integrated into a complete online impedance detection result for the speaker circuitry, which is output via a remote debugging terminal. Output methods may include a visual interface, log files, or alarm information, allowing audio engineers to remotely confirm the circuit status and perform corresponding debugging operations, thereby completing remote network debugging of the audio processor.

[0031] Furthermore, such as Figure 2As shown, this application provides a specific implementation flow of a remote network debugging method for an audio processor. First, the instantaneous spectral energy distribution of the program signal is acquired through the audio processor's spectral energy acquisition unit, identifying the energy gap positions and amplitudes of each frequency band to obtain energy gap frequency bands that can be used to inject detection signals. Then, the masking energy deficiency in the energy gap frequency band is analyzed through an auditory masking threshold calculation unit, and a psychoacoustic model is used to assess the probability that the detection signal in this frequency band will be exposed as audible. The upper limit of the level corresponding to the audible probability not exceeding a preset threshold is determined as the injection level. Next, impedance response amplitude data is obtained from the speaker lines connected to the audio processor, and combined with the injection level, a detection signal is injected into the energy gap frequency band to obtain a mixed signal. The signal spectrum is then processed by differential processing between the mixed signal spectrum and the original program signal spectrum. The signal-to-noise ratio (SNR) of each frequency band is calculated using the Fourier transform algorithm. Frequency bands with SNR below the noise floor threshold are marked as potential masking areas. For potential masking areas, the energy gap amplitude and audibility probability constraints are fused, and the actual injection amount in the area is increased without exceeding the injection level to obtain the enhanced SNR of each frequency band. Finally, the impedance curve is reconstructed using the enhanced SNR of each frequency band, and the frequency bands with amplitudes exceeding the preset abnormal threshold in the curve are evaluated to confirm the abnormal location of the speaker line impedance and output complete online detection results of the speaker line impedance.

[0032] In one embodiment, the instantaneous spectral energy distribution of the original program signal is obtained, and the energy gap frequency band that can be used to inject the detection signal is identified, including steps S201 to S203, each step of which is as follows: S201, based on the instantaneous spectral energy distribution, identify continuous frequency point regions in each frequency band where the spectral energy is lower than a preset threshold.

[0033] After obtaining the instantaneous spectral energy distribution of the program signal, the arithmetic mean of the energy of all frequency points in the current frame is first calculated, and an energy threshold is set accordingly. The formula for calculating the energy threshold is as follows:

[0034] In the formula, The energy value is the arithmetic mean, with a coefficient of 0.5 that can be adjusted between 0.3 and 0.7 depending on the program type, typically set to 0.5. Then, the entire spectrum is divided into several continuous frequency bands, such as the low-frequency band (20-200Hz), the mid-frequency band (200-2000Hz), and the high-frequency band (2000-20000Hz). Within each band, the energy values ​​of each frequency point are scanned. Frequency sequences with continuously lower energy values ​​than a threshold T are identified and marked as a continuous frequency region.

[0035] S202, extract the starting frequency, cutoff frequency, lowest energy within the energy gap, and average energy around the energy gap of the continuous frequency point region, calculate the absolute value of the lowest energy within the energy gap minus the average energy around the energy gap to obtain the gap amplitude, and determine the energy gap position and energy gap amplitude of each frequency band.

[0036] For each identified continuous region, the following characteristics are recorded: start frequency, cutoff frequency, and gap amplitude. The start frequency is the frequency corresponding to the first frequency point in the region that falls below a threshold; the cutoff frequency is the frequency corresponding to the last frequency point in the region that falls below the threshold; and the gap amplitude is the absolute value (in dB) of the difference between the lowest energy value in the region and the average energy value of the surrounding normal energy frequency bands, used to quantify the depth of the energy gap. This process ensures a quantitative description of the energy gap, providing fundamental data for dynamically adjusting the detection signal injection strategy.

[0037] S203, based on the energy gap position and the energy gap amplitude, select the energy gap frequency band that can be used to inject the detection signal.

[0038] After obtaining the energy gap location and amplitude of each frequency band, a joint screening process is performed using a preset amplitude threshold and a time stability criterion. The preset amplitude threshold can be a typical value of 60dB or set according to the dynamic range of the program signal. The time stability criterion screening requires that the energy gaps appear stably in at least three consecutive frames of the spectrum, with position and amplitude variations not exceeding the allowable tolerance. Only energy gap frequency bands that simultaneously meet the requirements of sufficiently large amplitude and good temporal continuity are determined as candidate frequency bands that can be used to inject detection signals, thereby ensuring that the selected frequency bands have sufficient energy depressions and masking margins, suitable for carrying subsequent detection signals without causing auditory perception.

[0039] This embodiment identifies continuous frequency regions sequentially, extracts gap feature parameters, and filters energy gap frequency bands based on amplitude and stability. This ensures that the selected frequency bands have the characteristics of deep energy depressions, large masking margins, and good time-domain continuity, thereby ensuring that the detection signal can be concealed into the program signal without causing auditory perception. At the same time, it provides a stable and reliable injection carrier for subsequent signal-to-noise ratio enhancement and impedance detection, effectively improving the feasibility and accuracy of online detection of speaker lines in remote debugging.

[0040] In one embodiment, the masking energy deficiency in the energy gap frequency band is analyzed using a psychoacoustic model to determine the injection level of the detection signal under the constraint of audible probability, including steps S301 to S303, each step of which is as follows: S301, by calculating the auditory masking threshold, the amount of masking energy missing in the energy gap frequency band is analyzed and obtained.

[0041] Specifically, the auditory masking threshold calculation unit estimates the ideal masking threshold of the current energy gap frequency band based on the energy levels of adjacent frequency bands and according to a psychoacoustic model. (Unit: dBSPL), and simultaneously extract the actual masking energy existing within this gap frequency band. (Unit: dBSPL), then calculate the difference between the two to obtain the masking energy missing amount. The formula for calculating the masking energy missing amount is as follows:

[0042] like This indicates that the frequency band has insufficient masking due to the low energy of the program signal, and the risk of the detected signal being exposed as audible is high. Therefore, the upper limit of the injection level needs to be constrained accordingly in subsequent steps.

[0043] S302, based on the amount of masking energy missing, the audibility probability of different test levels is evaluated using a psychoacoustic model.

[0044] In this process, after obtaining the amount of masking energy missing in the energy gap frequency band, the psychoacoustic model first divides the spectrum into several critical frequency bands to simulate the frequency selectivity of the basilar membrane in the human ear, and uses the actual masking threshold of the current frequency band as a benchmark. For a given test level, the model calculates the amount by which the injected energy of the detection signal exceeds the masking threshold in that frequency band, and maps the excess amount to the audible probability at that level based on the Gaussian distribution assumption of auditory perception variability. By iteratively changing the test level from low to high, the corresponding audible probabilities are calculated respectively, thereby obtaining the exposure risk distribution under different injection intensities, providing a quantitative basis for subsequently determining the maximum injection level that satisfies the audible probability constraint.

[0045] S303, compare the audible probability with a preset threshold, iteratively test the test level from low to high, and determine the upper limit of the level corresponding to when the audible probability does not exceed the preset threshold as the injection level.

[0046] The process involves comparing the audible probability obtained from the psychoacoustic model with a preset threshold. The test level is gradually increased starting from a lower injection level, with a typical preset threshold value of 0.05, meaning the audible risk is below 5%. After each increment, the audible probability at the current level is recalculated until it first reaches or slightly exceeds the preset threshold. Then, the level is reset to the previous level where the audible probability does not exceed the threshold, and this level is determined as the final injection level for that energy gap frequency band. This maximizes the injection intensity while ensuring the detected signal is not perceived by the listener.

[0047] This embodiment first calculates the masking energy loss in the energy gap frequency band, then evaluates the audibility probability at different test levels based on a psychoacoustic model, and finally determines the maximum injection level through an iterative method with the audibility probability not exceeding a preset threshold as a constraint. This ensures that the injection intensity of the detection signal can fully utilize the masking margin of the energy gap to improve the signal-to-noise ratio of impedance measurement, while also being strictly controlled below the human hearing perception threshold to avoid audible distortion. This effectively solves the contradiction between the energy domain and the perception domain in traditional spectrum complementarity strategies, and realizes concealed and high-precision online detection of speaker line impedance under live broadcast conditions.

[0048] Furthermore, such as Figure 3 As shown, this application provides a process for determining the injection level in a remote network debugging method for an audio processor. First, the injection level is iteratively tested from low to high, while simultaneously monitoring the audibility probability, i.e., the exposure probability. Then, it is determined whether the audibility probability exceeds a preset threshold (typically 0.05). If the audibility probability exceeds this threshold, the current test level is recorded as the upper limit of the candidate level. Through continuous iterative testing, the upper limit of the level corresponding to the audibility probability just below the preset threshold is determined. Finally, this upper limit of the level is determined as the injection level.

[0049] In one embodiment, the detection signal is injected into the energy gap frequency band based on the injection level to obtain the mixed signal spectrum, including steps S401 to S403, each step of which is as follows: S401, acquire impedance response amplitude data.

[0050] The audio processor outputs test signals, such as swept sine waves or broadband pulses, to the connected speaker circuitry. Simultaneously, voltage and current sensors acquire the response signals from the circuitry. A Fast Fourier Transform (FFT) is used to convert the time-domain response to a frequency-domain representation, and the impedance magnitude at each frequency point—the ratio of voltage amplitude to current amplitude—is calculated to obtain impedance response amplitude data. This data reflects the impedance distribution characteristics of the speaker circuitry at different frequencies, aiding in the identification of energy gap frequency bands and subsequent signal injection and impedance curve reconstruction.

[0051] S402, based on the impedance response amplitude data and the injection level, a detection signal is injected into the energy gap frequency band to obtain a preliminary mixed signal.

[0052] Based on the acquired impedance response amplitude data, the frequency range corresponding to the energy gap band is identified, and a detection signal of corresponding amplitude, such as a single-tone or multi-tone sine wave, is generated according to the determined injection level (in dBFS). This detection signal is superimposed on the same frequency band of the original program signal to form a preliminary mixed signal containing the original component and the injected component, ensuring that the injection intensity strictly does not exceed the upper limit of the injection level to avoid introducing audible distortion.

[0053] S403, the preliminary mixed signal is subjected to spectral conversion to obtain the mixed signal spectrum.

[0054] The initial mixed signal is framed using the same sampling rate and window function (e.g., Hanning window) as the original program signal. Then, a Fast Fourier Transform (FFT) is applied to transform each frame from the time domain to the frequency domain, obtaining the amplitude or power spectrum at each frequency point, thus generating the mixed signal spectrum. This spectrum simultaneously contains the spectral components of both the original program signal and the injected detection signal, providing a reliable frequency domain data foundation for subsequent differential processing and signal-to-noise ratio calculation.

[0055] This embodiment first acquires the impedance response amplitude data of the loudspeaker circuit to help identify the energy gap frequency band. Then, based on the determined injection level, a detection signal is injected into this frequency band to form a preliminary mixed signal. Finally, the spectrum of the mixed signal is obtained through spectrum conversion, so that the injection of the detection signal can match the actual impedance characteristics of the circuit. This ensures that the injection position and intensity conform to the energy gap distribution and are controlled by auditory masking constraints. This provides accurate frequency domain data containing the original signal and the injected components for subsequent differential processing and signal-to-noise ratio calculation, effectively improving the reliability of signal injection and the accuracy of impedance detection in remote debugging.

[0056] Furthermore, such as Figure 4 As shown, this application provides a mixed signal spectrum generation process for a remote network debugging method for an audio processor. First, impedance response amplitude data is obtained from the speaker lines connected to the audio processor, and the curve characteristics of this impedance response amplitude data are determined. Then, a detection signal is injected into the energy gap frequency band corresponding to the curve characteristics of the impedance response amplitude data, combined with the injection level, to obtain a preliminary mixed signal. Finally, the spectrum of the mixed signal is obtained by performing spectrum conversion on the preliminary mixed signal.

[0057] In one embodiment, the original program signal spectrum is obtained, the mixed signal spectrum and the original program signal spectrum are differentially processed to calculate the signal-to-noise ratio (SNR) of each frequency band, and the SNR of each frequency band is enhanced to obtain the enhanced SNR of each frequency band. This includes steps S501 to S503, each step of which is as follows: S501, perform differential processing on the spectrum of the mixed signal and the spectrum of the original program signal, and calculate the signal-to-noise ratio of each frequency band based on the differential results.

[0058] The differential spectrum is calculated by performing differential processing on the spectrum of the mixed signal and the spectrum of the original program signal. The calculation formula is as follows:

[0059] In the formula, For the mixed signal spectrum, The difference between the original program signal spectrum and the actual signal spectrum approximately reflects the energy increment introduced by the injected detection signal in each frequency band, and is used as a noise power estimate. Simultaneously, the power of the original program signal is also considered. As signal power Substituting the values ​​into the signal-to-noise ratio (SNR) formula, the SNR is calculated for each frequency band, thus obtaining the initial SNR distribution for each frequency band. The SNR formula is as follows:

[0060] In the formula, For signal power, This is for noise power estimation.

[0061] S502, mark the frequency bands with a signal-to-noise ratio lower than the noise floor threshold as potential masking areas.

[0062] The calculated signal-to-noise ratio (SNR) for each frequency band is compared one by one with a preset noise floor threshold. The typical value for the preset noise floor threshold is 10 dB, set according to the minimum requirements for signal purity in broadcast audio quality standards, and can be adjusted within the range of 5 dB to 15 dB depending on the program type. If the SNR of a certain frequency band is lower than this threshold, it indicates that the noise energy in that band is relatively close to or exceeds the signal energy, and the detection signal may be submerged by noise or already has good masking conditions. Therefore, this frequency band is marked as a potential masking area, and all frequency bands that meet the criteria are compiled into a marked frequency band set for subsequent injection enhancement processing.

[0063] S503, based on each of the potential masking regions, combined with the corresponding energy gap amplitude and the audible probability constraint, the actual injection amount of the potential masking region is increased without exceeding the injection level, thereby obtaining the enhanced signal-to-noise ratio of each frequency band.

[0064] Specifically, for each frequency band marked as a potential occlusion region, its corresponding energy gap amplitude and energy difference with adjacent frequency bands are extracted. The initial injection boundary is then calculated using a weighted fusion formula, as follows:

[0065] In the formula, For the energy gap amplitude, This represents the energy difference between adjacent frequency bands. and For the corresponding weighting coefficients, typical values ​​are... =0.6, =0.4, the sum of the two is 1, based on the empirical setting that the gap amplitude contributes more significantly to the injection capability, and can be calibrated according to the actual scenario. The unit is dB, representing the initial maximum injectable level estimate. Simultaneously, the initial boundary is verified and adjusted based on the audibility probability constraint (audibility probability not exceeding 5%) to ensure that the adjusted actual injection amount, without exceeding the original injection level upper limit, can appropriately increase the detection signal strength in the obscured area. The signal-to-noise ratio (SNR) after injection is recalculated to obtain the enhanced SNR for each frequency band, thereby improving the reliability of the previously noise-obscured frequency bands in subsequent impedance reconstruction.

[0066] This embodiment calculates the signal-to-noise ratio (SNR) of each frequency band by differentially processing the spectrum of the mixed signal and the spectrum of the original program signal. Frequency bands with SNR below the noise floor threshold are marked as potential masking areas. Then, by combining the energy gap amplitude and audibility probability constraints, the actual injection amount in these areas is increased without exceeding the original injection level, thereby obtaining an enhanced SNR. This process can effectively identify frequency bands whose SNR deteriorates due to insufficient masking and forced reduction in injection intensity. By targeting the masking areas, impedance anomaly information that might have been submerged by the measured noise floor can be recovered, thereby significantly improving the sensitivity and reliability of online detection of loudspeaker lines and solving the problem of decreased detection accuracy of the spectrum complementarity strategy under unfavorable conditions in the perception domain.

[0067] In one embodiment, based on each potential obscured region, and in conjunction with the corresponding energy gap amplitude and the audible probability constraint, the actual injection amount of the potential obscured region is increased without exceeding the injection level to obtain the enhanced signal-to-noise ratio for each frequency band. This includes steps S601 to S604, each step of which is as follows: S601, the initial injection boundary is obtained by weighted summing of the energy gap amplitude corresponding to the potential shielding region and the energy difference between the potential shielding region and the adjacent frequency band.

[0068] Specifically, for each potential shading region, its energy gap amplitude is extracted. (Characterizing the depth of the spectral dip in this region) and the energy difference between this region and adjacent frequency bands. (Reflecting the local spectral slope), then setting the weighting coefficients. and Using the weighted summation formula The initial injection boundary is calculated. (Unit: dB) This boundary, which serves as the basis for subsequent adjustments to the audible probability constraint, represents a preliminary estimate of the maximum allowable injection level in the masked region without considering the fine structure of the masking.

[0069] S602, based on the initial injection boundary, the feasibility of injection is determined by the audible probability constraint, and the adjusted boundary threshold is obtained.

[0070] After obtaining the initial injection boundary, it is used as the test injection level and substituted into the preset auditory model to calculate the audible probability of the detected signal at this level and compare it with the preset threshold. If the audible probability exceeds the threshold, the injection level is gradually reduced and re-evaluated until the audible probability meets the requirement of not exceeding the threshold. The corresponding level at this time is used as the adjusted boundary threshold. If the audible probability at the initial injection boundary is already lower than the threshold, it is directly used as the adjusted boundary threshold to ensure that the injection feasibility meets the auditory masking constraint.

[0071] S603, within the adjusted boundary threshold, combined with the spectral masking threshold, the injection level of the potential masking region is dynamically improved by the inter-band energy balance method, wherein the spectral masking threshold is extracted from the audio spectrum by the psychoacoustic model.

[0072] After obtaining the adjusted boundary thresholds, a psychoacoustic model is used to extract the spectral masking thresholds for each frequency band from the current audio spectrum, serving as a perceptual safety boundary that cannot be exceeded during injection. Under the premise of not exceeding this boundary threshold, an inter-band energy balancing method is employed to dynamically allocate injection energy based on the differences in masking margins in each potential masking region. Specifically, the injection level is moderately increased for regions with large masking margins, while a lower injection intensity is maintained for regions with limited masking margins. This maximizes the overall signal-to-noise ratio improvement effect in each masking region while ensuring that the injection intensity of all frequency bands remains below the audible masking threshold, avoiding the introduction of audible distortion.

[0073] S604, based on the enhanced injection level, calculate the logarithmic ratio of the signal-to-noise ratio after injection for each frequency band to obtain the enhanced signal-to-noise ratio for each frequency band.

[0074] After determining the actual injection level of each frequency band after boosting based on the spectral masking threshold and energy balance method, the logarithmic ratio of the signal-to-noise ratio after injection for each frequency band is calculated using the following formula:

[0075] In the formula, This represents the original program signal power. To inject detection signal power, Let be the noise power, which is the power of the remaining non-injected component in the differential spectrum. The calculated logarithmic ratio is the signal-to-noise ratio (SNR) of each frequency band after injection enhancement. This enhanced SNR typically shows a significant improvement over the initial SNR in potentially obscured regions, providing a higher-quality data foundation for reliable reconstruction of subsequent impedance curves.

[0076] This embodiment obtains the initial injection boundary by weighted summing of the energy gap amplitude and the energy difference between adjacent frequency bands in the potential masking region. A safe boundary threshold is then obtained through audible probability constraints. Within this threshold, a spectral masking threshold extracted from a psychoacoustic model is used, and an inter-band energy balancing method is employed to dynamically increase the injection level. Finally, the enhanced signal-to-noise ratio (SNR) is calculated. This series of operations ensures that the increase in injection volume fully utilizes the spectral dip and local spectral slope information of the energy gap while strictly adhering to audible masking characteristics and audible probability constraints. This avoids the problems of audible distortion caused by excessive injection or limited SNR improvement due to insufficient injection. Thus, it maximizes the SNR gain of the potential masking region while maintaining concealment, significantly improving the identification capability and measurement accuracy of abnormal frequency bands in loudspeaker line impedance detection.

[0077] In one embodiment, the impedance curve is reconstructed using the enhanced signal-to-noise ratio to identify abnormal locations in the line impedance, thus completing remote network debugging of the audio, including steps S701 to S704, each step of which is as follows: S701, based on the enhanced signal-to-noise ratio, the noise compensation coefficient is obtained by calculating the ratio of noise level to signal strength in each frequency band, and the impedance curve is reconstructed.

[0078] Based on the enhanced signal-to-noise ratio of each frequency band, the noise compensation coefficient is first calculated for each frequency band:

[0079] In the formula, The original signal power, This represents noise power. For frequency bands where the signal-to-noise ratio is below the reliable reconstruction threshold (typically 15 dB), a compensation coefficient is used to enhance the signal power:

[0080] In the formula, k is the noise compensation coefficient. The compensated signal power is combined with the voltage and current response data of the current frequency band to calculate the impedance magnitude at each frequency point, that is, the ratio of voltage amplitude to current amplitude. Through frequency domain interpolation and smoothing, a complete speaker line impedance curve is reconstructed, which reflects the true impedance characteristics of the line at different frequencies.

[0081] S702, based on the impedance curve, the frequency bands whose amplitude in the impedance curve exceeds a preset abnormal threshold are marked as abnormal frequency bands.

[0082] After reconstructing the complete impedance curve, a preset abnormal threshold is set. The typical value of the preset abnormal threshold is 1.5 times the average normal impedance, which can be adjusted within the range of 1.2 to 2.0 times depending on the speaker model and specifications. Then, each frequency band on the impedance curve is traversed, and frequency bands with impedance amplitudes exceeding the threshold are marked as abnormal frequency bands. Their frequency range and corresponding impedance deviation values ​​are recorded, thereby initially identifying the location of line faults such as open circuits, short circuits, or aging.

[0083] S703, based on the abnormal frequency band, calculate the amplitude difference between the abnormal frequency band and its adjacent frequency bands to obtain the adjacent frequency band comparison value.

[0084] After identifying the abnormal frequency band, the impedance amplitude value of the abnormal frequency band and the average impedance amplitude of several adjacent normal frequency bands are extracted. By calculating the difference between the impedance amplitude value and the average impedance amplitude, a comparison value between adjacent frequency bands is obtained. This comparison value quantifies the degree of deviation of the abnormal frequency band from the surrounding normal area, which is used to help determine the fault type, such as positive deviation usually corresponding to an open circuit, negative deviation corresponding to a short circuit, etc., and to assess the severity of the abnormality, providing a basis for finally accurately locating the impedance abnormality.

[0085] S704, the position where the comparison value of the adjacent frequency band exceeds the preset comparison threshold is identified as the abnormal position of the line impedance, and the remote network debugging of the audio is completed.

[0086] The process involves obtaining comparison values ​​for adjacent frequency bands and then comparing them with a preset comparison threshold. A typical value for this threshold is ±20% of the normal impedance mean or a deviation tolerance calibrated based on historical data. If the comparison value for an abnormal frequency band exceeds this threshold, the location is confirmed as an abnormal position in the speaker line impedance. The specific physical distance coordinates are then output using a time-domain reflection method or a frequency-domain localization algorithm. By integrating all confirmed abnormal positions and their corresponding impedance curve characteristics, the complete online speaker line impedance detection results are presented to the user through a remote debugging interface. This enables remote network debugging of the audio processor and facilitates equipment status monitoring and fault diagnosis under live broadcast conditions.

[0087] This embodiment calculates the noise-to-signal strength ratio of each frequency band using the enhanced signal-to-noise ratio to obtain the noise compensation coefficient. This coefficient is then used to reconstruct the impedance curve. Frequency bands whose amplitude exceeds a preset abnormal threshold are marked as abnormal bands. The amplitude difference between the abnormal band and its adjacent band is calculated to obtain a comparison value. Finally, the location where the comparison value exceeds the preset threshold is confirmed as the abnormal position of the line impedance. This process overcomes the problem of unstable impedance calculation when the signal-to-noise ratio is insufficient by utilizing noise compensation. The dual determination of the abnormal threshold and adjacent comparison values ​​reduces false alarms and missed alarms, making anomaly localization more accurate and reliable. This provides an effective means for audio processors to remotely diagnose and troubleshoot speaker line faults online, in real-time, and covertly.

[0088] Based on the same inventive concept, this application also provides a system for implementing the aforementioned remote network debugging system for audio processors. The solution provided by this system is similar to the solution described in the above method; therefore, the specific limitations in one or more embodiments of the remote network debugging device for audio processors provided below can be found in the limitations of the remote network debugging method for audio processors described above, and will not be repeated here.

[0089] In one exemplary embodiment, such as Figure 5 As shown, a remote network debugging system for audio processors is provided, comprising: The acquisition module 801 is used to acquire the instantaneous spectral energy distribution of the original program signal and identify the energy gap frequency band that can be used to inject the detection signal; The determination module 802 is used to analyze the masking energy deficiency in the energy gap frequency band through a psychoacoustic model and determine the injection level of the detection signal under the audible probability constraint. The mixing module 803 is used to inject the detection signal into the energy gap frequency band based on the injection level to obtain a mixed signal spectrum; The enhancement module 804 is used to acquire the original program signal spectrum, perform differential processing on the mixed signal spectrum and the original program signal spectrum, calculate the signal-to-noise ratio of each frequency band, and enhance the signal-to-noise ratio of each frequency band to obtain the enhanced signal-to-noise ratio of each frequency band. The debugging module 805 is used to reconstruct the impedance curve through the enhanced signal-to-noise ratio, confirm the abnormal location of the line impedance, and complete the remote network debugging of the audio.

[0090] In one embodiment, a computer program product is provided, including a computer program or instructions that, when executed by a device, implement the steps of a remote network debugging method for an audio processor as described above.

[0091] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the steps in the above method embodiments.

[0092] This technical solution deeply integrates the spectrum energy acquisition unit, the auditory masking threshold calculation unit, and the impedance response amplitude data processing, and associates them with a psychoacoustic masking model to construct a remote network debugging and interactive system specifically for online detection of loudspeaker line impedance under live broadcast conditions. Its core advantage lies in overcoming the limitations of existing technologies that rely solely on energy gaps for signal injection or ignore auditory masking characteristics. It achieves adaptive adjustment of the injection level of the detection signal based on audible probability constraints in a dynamically changing program signal environment. Users can actively acquire abnormal frequency bands in the impedance curve and optimize the injection strategy in real time through an interactive debugging interface, thus providing a reliable decision-making basis for loudspeaker line status monitoring and achieving a leap from favorable energy domain to safe perception domain. Simultaneously, this solution fully leverages the respective advantages of spectrum energy acquisition in signal analysis, psychoacoustic models in masking assessment, and impedance response processing in anomaly localization, solving the problem in traditional methods where the detection signal is easily exposed as audible or abnormal frequency bands are masked by noise floor.

[0093] For the device embodiments, since they basically correspond to the method embodiments, the relevant details can be found in the descriptions of the method embodiments. The device embodiments described above are merely illustrative; components described as separate parts may or may not be physically separate, and components shown as units may or may not be physical units, meaning they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this disclosure according to actual needs. Those skilled in the art can understand and implement this without any inventive effort.

[0094] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. In particular, it should be noted that any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention for those skilled in the art.

Claims

1. A method for remote network debugging of an audio processor, characterized in that, include: The instantaneous spectral energy distribution of the original program signal is obtained, and the energy gap frequency bands that can be used to inject the detection signal are identified. The masking energy deficiency in the energy gap frequency band is analyzed by psychoacoustic modeling to determine the injection level of the detection signal under the audible probability constraint; Based on the injection level, the detection signal is injected into the energy gap frequency band to obtain the mixed signal spectrum; The original program signal spectrum is obtained, the mixed signal spectrum and the original program signal spectrum are differentially processed to calculate the signal-to-noise ratio of each frequency band, and the signal-to-noise ratio of each frequency band is enhanced to obtain the enhanced signal-to-noise ratio of each frequency band. By reconstructing the impedance curve using the enhanced signal-to-noise ratio, the abnormal location of the line impedance is identified, and remote network debugging of the audio is completed.

2. The method for remote network debugging of an audio processor as described in claim 1, characterized in that, The process of acquiring the instantaneous spectral energy distribution of the original program signal and identifying the energy gap frequency bands that can be used to inject the detection signal includes: Based on the instantaneous spectral energy distribution, the continuous frequency point regions in each frequency band where the spectral energy is lower than a preset threshold are identified; Extract the start frequency, cutoff frequency, lowest energy within the energy gap, and average energy around the energy gap of the continuous frequency region. Calculate the absolute value of the lowest energy within the energy gap minus the average energy around the energy gap to obtain the gap amplitude, and determine the energy gap position and energy gap amplitude for each frequency band. Based on the energy gap location and the energy gap amplitude, energy gap frequency bands that can be used to inject detection signals are selected.

3. The method for remote network debugging of an audio processor as described in claim 1, characterized in that, The step of analyzing the masking energy deficiency in the energy gap frequency band using a psychoacoustic model to determine the injection level of the detection signal under the constraint of audibility probability includes: The amount of masking energy missing in the energy gap frequency band is obtained by calculating the auditory masking threshold. Based on the amount of masking energy missing, the audibility probability of different test levels is evaluated using a psychoacoustic model; The audible probability is compared with a preset threshold, and the test level is iteratively tested from low to high. The upper limit of the level corresponding to when the audible probability does not exceed the preset threshold is determined as the injection level.

4. The method for remote network debugging of an audio processor as described in claim 1, characterized in that, The step of injecting the detection signal into the energy gap frequency band based on the injection level to obtain the mixed signal spectrum includes: Obtain impedance response amplitude data; Based on the impedance response amplitude data and the injection level, a detection signal is injected into the energy gap frequency band to obtain a preliminary mixed signal; The initial mixed signal is subjected to spectral conversion to obtain the mixed signal spectrum.

5. The method for remote network debugging of an audio processor as described in claim 2, characterized in that, The process of acquiring the original program signal spectrum, performing differential processing on the mixed signal spectrum and the original program signal spectrum to calculate the signal-to-noise ratio (SNR) of each frequency band, and enhancing the SNR of each frequency band to obtain the enhanced SNR of each frequency band includes: The spectrum of the mixed signal and the spectrum of the original program signal are differentially processed, and the signal-to-noise ratio of each frequency band is calculated based on the differential results; The frequency bands with a signal-to-noise ratio below the noise floor threshold are marked as potential masking areas; Based on each potential obscured region, and in conjunction with the corresponding energy gap amplitude and the audible probability constraint, the actual injection amount of the potential obscured region is increased without exceeding the injection level, thereby obtaining the enhanced signal-to-noise ratio of each frequency band.

6. The method for remote network debugging of an audio processor as described in claim 5, characterized in that, Based on each potential obscured region, and in conjunction with the corresponding energy gap amplitude and the audible probability constraint, the actual injection amount of the potential obscured region is increased without exceeding the injection level, to obtain the enhanced signal-to-noise ratio for each frequency band, including: The initial injection boundary is obtained by weighted summing of the energy gap amplitude corresponding to the potential shielding region and the energy difference between the potential shielding region and the adjacent frequency band; Based on the initial injection boundary, the feasibility of injection is determined by the audible probability constraint, and the adjusted boundary threshold is obtained. Within the adjusted boundary threshold, combined with the spectral masking threshold, the injection level of the potential masking region is dynamically improved through the inter-band energy balance method, wherein the spectral masking threshold is extracted from the audio spectrum by the psychoacoustic model; Based on the enhanced injection level, the logarithmic ratio of the signal-to-noise ratio after injection is calculated for each frequency band to obtain the enhanced signal-to-noise ratio for each frequency band.

7. The method for remote network debugging of an audio processor as described in claim 1, characterized in that, The process of reconstructing the impedance curve using the enhanced signal-to-noise ratio to identify abnormal locations in the line impedance and completing remote network debugging of the audio includes: Based on the enhanced signal-to-noise ratio, the noise compensation coefficient is obtained by calculating the ratio of noise level to signal strength in each frequency band, and the impedance curve is reconstructed. Based on the impedance curve, the frequency bands whose amplitude in the impedance curve exceeds a preset abnormal threshold are marked as abnormal frequency bands; Based on the abnormal frequency band, the amplitude difference between the abnormal frequency band and its adjacent frequency bands is calculated to obtain the adjacent frequency band comparison value; The location where the comparison value of the adjacent frequency bands exceeds the preset comparison threshold is identified as an abnormal location of the line impedance, thus completing the remote network debugging of the audio.

8. A remote network debugging system for an audio processor, characterized in that, The system includes: The acquisition module is used to acquire the instantaneous spectral energy distribution of the original program signal and identify the energy gap frequency bands that can be used to inject the detection signal. The determination module is used to analyze the masking energy deficiency in the energy gap frequency band through a psychoacoustic model and determine the injection level of the detection signal under the audible probability constraint. A mixing module is used to inject the detection signal into the energy gap frequency band based on the injection level to obtain a mixed signal spectrum; An enhancement module is used to acquire the original program signal spectrum, perform differential processing on the mixed signal spectrum and the original program signal spectrum, calculate the signal-to-noise ratio of each frequency band, and enhance the signal-to-noise ratio of each frequency band to obtain the enhanced signal-to-noise ratio of each frequency band. The debugging module is used to reconstruct the impedance curve through the enhanced signal-to-noise ratio, identify the abnormal location of the line impedance, and complete the remote network debugging of the audio.

9. A computer-readable storage medium, characterized in that, The storage medium stores a computer program or instructions, which, when executed by a communication device, implement the audio processor remote network debugging method as described in any one of claims 1-7.

10. A computer program product, comprising a computer program or instructions, characterized in that, When the computer program or instructions are executed by the communication device, the remote network debugging method for the audio processor as described in any one of claims 1-7 is implemented.

Images