Audio pre-stage switching transient noise suppression method

Abstract

This invention discloses a method for suppressing transient noise during audio preamplifier switching, relating to the field of audio signal processing technology. The method involves acquiring the output signal after the target channel is switched on, filtering out in-band audio components, extracting out-of-band background noise energy, and obtaining the background noise fluctuation amplitude. This amplitude is compared with the in-band level fluctuation within an observation window over time periods to determine the source-side silence period. The background noise fluctuation amplitude is used as a readiness criterion to exclude samples taken during the source-side silence period. The starting point and rise rate of the background noise fluctuation amplitude recovery are tracked to determine the actual readiness time and gain recovery slope of the channel. From the actual readiness time, the preamplifier gain is gradually increased, and gain stepping is paused when the background noise falls back, until the preamplifier gain recovers to its nominal value. This invention avoids premature fulfillment of the readiness condition due to source-side silence, matches gain recovery with background noise recovery, and suppresses audible pops generated during switching and the release of source-side silence.

Description

Technical Field

[0001] This invention relates to the field of audio signal processing technology, and in particular to a method for suppressing transient noise during audio preamp switching. Background Technology

[0002] When an audio preamplifier switches between multiple signal sources, the charging and discharging of the coupling stage, the re-establishment of the DC operating point of the gain stage, and sudden changes in input impedance can create steps and impulses at the output, manifesting as audible pops or noise. Conventional suppression methods involve lowering the gain or muting the output before the switching action, and then restoring the gain to its nominal value after a delay following the switching, according to a preset recovery slope. To avoid overly conservative delay values ​​that could lead to lost call content, engineering practices typically introduce a readiness criterion in addition to the delay: continuously sampling the output of the target channel after connection and statistically analyzing its level fluctuations within an observation window. When the level fluctuations are below a set threshold and the output is sufficiently stable, the preamplifier path is considered to have entered a steady state, and gain recovery is immediately initiated. The implicit premise of this criterion is that the output stability stems from the completion of the analog path's own establishment process.

[0003] When the current signal source is a receiver with source-side functionality, this premise does not hold. There is an inverse correlation between the observed parameters upon which the criterion is based and the actual readiness of the channel. When the RF input level drops below the source-side threshold, the source-side circuit activates, the audio output is cut off, leaving only a nearly constant static level. The randomly fluctuating noise components in the demodulation path are removed. The level fluctuation within the observation window is not only small but also smaller than the level fluctuation after the path is truly established. The smoother the output, the more it indicates that the channel is still in a source-side muted state rather than ready. The readiness condition is thus met prematurely, and the gain rises to a high level along the recovery slope. The timing of the RF level recovery and source-side deactivation is independent of the switching sequence. Once it occurs after the gain has already increased, the removed background noise and the DC operating point jump caused by the source-side action simultaneously enter the path and are amplified by high gain, forming a popping sound with an amplitude greater than the original switching transient. Conventional methods to improve decision reliability by extending the stabilization wait time have the opposite effect under these conditions: the longer the wait, the higher the probability that the decision time falls within the source-side silence period; simply tightening the fluctuation threshold also makes it easier for muted channels to pass the decision. The key issue is that the statistics used for decision-making are located within the audio frequency band, and the output of this band is precisely determined by the source-side circuit. The two states of a channel being muted by the source side and the channel itself being quiet are almost indistinguishable from each other in terms of observation, and the decision lacks a basis to reflect the establishment status of the path itself. In addition, if the gain recovery process and the channel level change process overlap in time and the rate of change do not match, even if the recovery timing is correctly determined, new audible noise will still be generated during the recovery period.

[0004] Therefore, how to distinguish whether the stable output performance after preamp switching is due to the completion of the path establishment or the source side mute, and thus determine the actual ready time of the channel, and ensure that the gain recovery process does not conflict with the channel's own level changes, is a technical problem that needs to be solved in this field. Summary of the Invention

[0005] To address the issues of premature fulfillment of conventional readiness criteria due to source-side muting and amplification of background noise and DC operating point jumps by high gain after source-side deactivation, this invention provides a method for suppressing transient noise during audio preamplifier switching. This method includes the following steps: S1. Acquire the output signal after the target channel is turned on, filter out the audio band components, extract the background noise energy outside the audio band, track the level fluctuation of the background noise energy, and obtain the background noise fluctuation amplitude. S2. Statistically measure the in-band level fluctuation of the output signal within the observation window, compare the noise floor fluctuation amplitude with the in-band level fluctuation by time period, and determine the time period when both are below their respective thresholds as the source-side silent period. S3. Change the ready determination quantity from the in-band level fluctuation quantity to the noise floor fluctuation amplitude, and exclude the sampling that falls into the source-side silent period from the ready determination, to obtain the ready determination benchmark. S4. Based on the readiness determination criteria, continue to monitor the output signal during the excluded sampling period, track the starting point of the noise floor fluctuation amplitude recovery, and obtain the noise floor recovery rate after the starting point. S5. Based on the noise floor recovery rate, match the rise rate of the preamplifier gain during the recovery period with the noise floor recovery rate to obtain the gain recovery slope. S6. Take the time corresponding to the starting point as the actual ready time of the channel, and gradually increase the preamplifier gain of the target channel according to the gain recovery slope from that time to obtain the target channel output state after the preamplifier gain is restored to the nominal value.

[0006] Compared with the prior art, the present invention uses the noise floor fluctuation amplitude outside the audio band to reflect the establishment state of the target channel itself, excludes the source-side silent period from the ready determination, and matches the preamp gain recovery rate with the noise floor recovery rate, which can avoid premature gain increase and the superposition of the two level change processes, thereby suppressing the audible popping sound generated by preamp switching and source-side release. Attached Figure Description

[0007] Figure 1 This is a flowchart illustrating the transient noise suppression method for audio preamplifier switching according to the present invention.

[0008] Figure 2 This is a schematic diagram illustrating the process of determining the source-side silent period and forming the readiness judgment criteria of the present invention.

[0009] Figure 3 This is a schematic diagram of the pre-stage gain step recovery and pause continuation execution process of the present invention. Detailed Implementation

[0010] The embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0011] The following specific examples illustrate the implementation of this disclosure. Those skilled in the art can easily understand other advantages and effects of this disclosure from the content disclosed in this specification. Obviously, the described embodiments are only a part of the embodiments of this disclosure, and not all of them. This disclosure can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of this disclosure. It should be noted that, in the absence of conflict, the following embodiments and features in the embodiments can be combined with each other. Based on the embodiments in this disclosure, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this disclosure.

[0012] like Figures 1 to 3 As shown, the audio preamp switching transient noise suppression method includes: S1. Acquire the output signal after the target channel is turned on, filter out the audio band components, extract the background noise energy outside the audio band, and track the level fluctuation of the background noise energy to obtain the background noise fluctuation amplitude.

[0013] After the target channel is connected, its output signal is acquired. The acquired output signal is filtered to remove in-band audio components, retaining only out-of-band noise components. The noise floor energy is extracted from the retained out-of-band noise components, and the level fluctuation of the noise floor energy is tracked point by point according to the sampling time sequence. The amplitude of the noise floor energy level fluctuation within the continuous sampling interval is the noise floor fluctuation amplitude.

[0014] In scenarios where the signal source at the front end has mute control, such as when a tuner with mute function activates its mute circuit and cuts off the audio output when the received signal level drops below the mute threshold, or when a Bluetooth audio module cuts off its analog output via mute control during the initial link establishment phase, the audio output is cut off when the source-side mute is activated, leaving only a nearly constant static level within the audio band. The originally randomly fluctuating broadband noise components in the source output are also removed. At this time, the output within the audio band is determined by the source-side mute control. Regardless of whether the channel itself has been established, the in-band observations show extremely small fluctuations, and the two states of the source being mute and the channel itself being quiet are almost indistinguishable within the band. The noise floor outside the audio band fluctuates with the path establishment process. There is almost no fluctuation when the source-side mute is activated, and random fluctuations reappear after the source-side mute is released, which can reflect the establishment status of the path itself. Therefore, this step shifts the observation object to outside the audio band, using the level fluctuation of the out-of-band noise floor energy as the basis for subsequent judgments.

[0015] Specifically, the filtering and extraction are performed as follows: The output signal after the target channel is turned on is filtered by a high-pass filter. The cutoff frequency of this high-pass filter is higher than the upper limit of the audio frequency band, resulting in out-of-band noise. The out-of-band noise signal is integrated over continuous sampling intervals to obtain the noise floor energy of each sampling interval. The noise floor energy of each sampling interval is converted into a level value according to the sampling time sequence, and the level change between adjacent sampling points is recorded point by point to obtain the noise floor level fluctuation sequence. The difference between the upper and lower envelopes of the noise floor level fluctuation sequence within continuous sampling intervals is taken to obtain the noise floor fluctuation amplitude. The output signal acquisition point is located on the pre-stage switching path, before the pre-stage gain stage. The acquired signal is the output of the target channel itself and is not affected by the pre-stage gain being in a low-gain state. The high-pass filter cutoff frequency is set at 22kHz, leaving a 2kHz transition band above the upper limit of the audio frequency band (20kHz). The lower limit of the audio frequency band (25Hz) and the upper limit (20kHz) are determined based on the rated bandwidth of the target audio system. The acquisition path is a monitoring acquisition path set in the pre-amplifier stage, with a sampling rate of 96kHz. The upper limit of the anti-aliasing filter passband is approximately 40kHz, and an out-of-band bandwidth of approximately 18kHz is reserved above the high-pass cutoff frequency. The out-of-band noise floor energy is sufficient, and the out-of-band observation is not affected by the program content within the audio band. The transition band width of 2kHz is determined according to the stopband attenuation achieved by the 4th-order Butterworth high-pass filter at the upper limit of the audio frequency band, so that the components within the audio band, along with the effect of the source-side mute control on this frequency band, are completely isolated outside the passband. Its adjustable range is between the upper limit of the audio frequency band and the upper limit of the passband of the acquisition path anti-aliasing filter, adjusted according to the measured noise floor spectrum distribution of the target channel.

[0016] In one embodiment, the sampling interval duration is 10 ms, which is one-tenth of the path establishment time of 100 ms. The path establishment time is calculated based on the time constant of a coupling capacitor of 10 μF and an input impedance of 10 kΩ and confirmed by actual measurement. The sampling interval shortens as the sampling rate increases and lengthens as the noise floor power decreases. After the noise floor energy is converted to a level value in decibels, the order of magnitude difference in noise floor energy before and after the source-side muting action can be used for subsequent comparisons on a similar scale.

[0017] Furthermore, the noise floor fluctuation amplitude is taken as the difference between the upper and lower envelopes of the noise floor level fluctuation sequence over several consecutive sampling intervals, reflecting the total amplitude of the out-of-band noise floor fluctuations during that period. The typical setting for the number of consecutive sampling intervals used to take the envelope is to make its total duration the same as the number of intervals used in the observation window for subsequent statistical analysis of in-band level fluctuations. Specifically, the number of intervals is 50, with a total duration of 500ms, consistent with the observation window used in step S2 for statistical analysis of in-band level fluctuations. This value ensures that the envelope difference spans a sufficiently long time, is not affected by the accidental values ​​of a single interval, and is aligned with the observation window for segment-by-segment comparison; it is adjusted synchronously with the length of the observation window.

[0018] In one possible implementation, this step is performed by a high-pass filter and a noise floor detection unit set in the front-end switching path. The noise floor detection unit updates the noise floor level fluctuation sequence at the end of each sampling interval and outputs the noise floor fluctuation amplitude according to the number of intervals mentioned above.

[0019] S2. Compare the noise floor fluctuation amplitude with the in-band level fluctuation of the output signal within the observation window by time period, and determine the time period when both are lower than their respective thresholds as the source-side silence period.

[0020] The noise floor fluctuation amplitude is used as input, and the in-band level fluctuation of the output signal within the observation window is statistically analyzed. The noise floor fluctuation amplitude and the in-band level fluctuation are compared segment by segment over time. The time segment in which both are below their respective thresholds is determined as the source-side silent period.

[0021] In the scenario of source-side mute, the output within the audio band is cut off. The in-band level fluctuation within the observation window is smaller than the level fluctuation after the path is truly established, and the out-of-band noise floor is almost non-volatile. When the channel itself is quiet but ready, the in-band level fluctuation is also very small, but the out-of-band noise floor still fluctuates randomly. The two states are indistinguishable in terms of in-band observations, but there is a clear difference in the amplitude of the noise floor fluctuation. Based on this, this step examines both simultaneously within the same time period, and only the time period in which both are below their respective thresholds is determined as the source-side mute period.

[0022] Specifically, the determination of the source-side silence period is performed as follows: A sampling sequence is extracted from the output signal according to an observation window; the level fluctuations of the audio components within the observation window are statistically analyzed to obtain the in-band level fluctuation amount; the in-band level fluctuation amount and the noise floor fluctuation amplitude are time-aligned and compared segment by segment within the same observation window to obtain segment-by-segment comparison results; in the segment-by-segment comparison results, time periods where the noise floor fluctuation amplitude is lower than a preset fluctuation threshold and the in-band level fluctuation amount is lower than a preset reference fluctuation amount are judged as periods where both are below their respective thresholds, thus obtaining the initial silence period; adjacent initial silence periods are merged according to the sampling time sequence to obtain the source-side silence period. The in-band level fluctuation amount is statistically analyzed on the output signal without high-pass filtering, and the difference between the maximum and minimum values ​​of the level sequence measured in decibels within the observation window is taken. The typical setting for the observation window length is to cover multiple sampling intervals and be shorter than the shortest possible interval between the source-side mute action and release, typically 500ms, covering 50 sampling intervals and being shorter than the shortest possible interval between the source-side mute action and release (1s). This shortest interval is determined based on the mute response characteristics of the signal source, such as the tuner's mute response time or actual measured statistics of the mute action and release interval of the source device. The observation window length is adjusted according to the multiple relationship between the mute response time and the sampling interval duration. The duration of the observation window and the noise floor fluctuation amplitude are taken to be the same as the total duration of the continuous sampling intervals spanned by the envelope, and the two correspond one-to-one according to the window.

[0023] In one embodiment, a preset reference fluctuation is pre-calibrated: A switching operation is performed under the condition that the signal source is outputting normally and the source-side mute is not activated. After this switching, the level fluctuation of the audio components within the established path interval of the output signal is statistically analyzed, and the statistical result is stored as the preset reference fluctuation, which is directly called upon during subsequent switching. The typical setting of the established path interval is a period of time after the switching operation, following the charging and discharging of the coupling element and the re-establishment of the DC operating point of the gain stage. The setting is based on the path establishment time calculated according to the time constant of the coupling element and the input impedance and confirmed by actual measurement. Its adjustable range is adjusted according to the time constant of the coupling element and the establishment time of the gain stage. During judgment, the fluctuation deviation is obtained by subtracting the in-band level fluctuation from the preset reference fluctuation. A negative fluctuation deviation indicates that the current in-band level fluctuation is less than the normal level when the channel is ready. Within the time period where the fluctuation deviation is negative, the noise floor fluctuation amplitude is compared with a preset fluctuation threshold to obtain the noise floor comparison result. The time period where the fluctuation deviation is negative and the noise floor fluctuation amplitude is lower than the preset fluctuation threshold is the initial mute period.

[0024] Preferably, the preset fluctuation threshold is typically set to be lower than the normal level of the noise fluctuation amplitude after the path is established and higher than the residual level of the noise fluctuation amplitude during the source-side quiet period. The setting is based on the statistical distribution calibration of the noise fluctuation amplitude measured in the target channel under the two states of source-side quiet and path establishment. Its adjustable range is between the residual level and the normal level during the quiet period.

[0025] Understandably, the initial silence period is obtained in units of observation windows. Directly adjacent initial silence periods are merged into a continuous period according to the sampling sequence. When two initial silence periods are separated by only one observation window that does not meet the conditions, this window is mostly due to occasional fluctuations within the source-side silence period, and is also merged into the same source-side silence period. The basis for this setting is that the duration of the noise floor rise after the source side is released is longer than one observation window, and an interval of only one window does not constitute a silence release. This step is executed by the front-end control unit at the end of each observation window, and the output source-side silence period is recorded according to the start and end times of the window.

[0026] S3. Based on the source-side silence period, change the ready determination quantity from the in-band level fluctuation quantity to the noise floor fluctuation amplitude, and exclude the samples falling into the source-side silence period from the ready determination, thus obtaining the ready determination benchmark.

[0027] In this embodiment, readiness determination no longer uses in-band level fluctuation as the criterion, but instead uses noise floor fluctuation amplitude. Samples falling within the source-side silent period are excluded from readiness determination. The replaced readiness determination criterion and the exclusion process together constitute the readiness determination benchmark.

[0028] In scenarios where the readiness of a pre-amplifier path is determined using conventional readiness criteria, the statistical quantity used in the criteria lies within the audio frequency band. During the source-side operation, the in-band output is nearly constant, the readiness condition is met prematurely, and once the source side is deactivated after the gain rises to a high level, a popping sound occurs. Extending the wait time or tightening the fluctuation threshold does not improve this: the former makes the determination more likely to fall within the source-side silence period, and the latter makes it easier for muted channels to pass the determination. This step replaces the determination quantity accordingly and excludes all sampling during the source-side silence period, ensuring that the determination is only performed during the period when the channel is likely to be actually ready.

[0029] Specifically, the readiness determination criteria are obtained as follows: stop using in-band level fluctuation as the readiness determination quantity, and instead use the noise floor fluctuation amplitude as the readiness determination quantity to obtain the replaced readiness determination quantity; add an exclusion mark to the samples falling into the source-side silent period to obtain the sampling sequence with the exclusion mark; pause the readiness timing during the duration of the exclusion mark and keep the preamplifier gain in a low state to obtain the readiness determination pause state; the replaced readiness determination quantity and the readiness determination pause state together constitute the readiness determination criteria.

[0030] It should be noted that the meaning of "ready" changes after the replacement of the judgment value. When judging by in-band level fluctuation, the smaller the fluctuation, the closer to readiness; when judging by noise floor fluctuation amplitude, the noise floor fluctuation amplitude must be at the normal level after the path is established to indicate that the channel is actually ready, while a noise floor fluctuation amplitude that is almost zero indicates that the channel is still in source-side silence. Therefore, the replaced readiness judgment value does not use the original threshold and comparison direction, but uses whether the noise floor fluctuation amplitude has recovered from the residual level of the silence period to the normal level as the judgment criterion.

[0031] In one embodiment, exclusion markers are added to the sampling sequence according to sampling intervals, with the start and end times of the markers coinciding with the start and end times of the source-side silent period. The readiness timer is a delay timer from the start of the switching action until gain recovery is allowed to begin, accumulating only on samples without exclusion markers. It pauses upon the appearance of an exclusion marker, preventing the delay from expiring during the source-side silent period and triggering gain recovery. Accumulation resumes from the pause point after the exclusion marker ends. Gain recovery is initiated based on the actual readiness time of the channel and not earlier than the readiness timer's expiration. The typical setting for the readiness timer is slightly longer than the time required for the coupling element to charge / discharge and for the DC operating point to re-establish. This setting is based on the path establishment time calculated using the time constants of the coupling element and input impedance and confirmed by actual measurements. Its adjustable range is adjusted according to the path establishment time of the target channel.

[0032] Furthermore, during the period of the exclusion flag, the preamplifier gain remains at the gain value reduced before the switching action. The typical setting for the reduced gain state is output mute. This setting is based on the fact that the timing of source-side mute release is unpredictable. The reduced gain state must be able to withstand the background noise and DC operating point jump entering the path when the mute is released without producing an audible pop. Its adjustable range is higher than output mute, provided that subjective listening tests confirm no audible pop is produced, to shorten the recovery time. This step is executed by the preamplifier control unit, which maintains the exclusion flag and ready timer, and outputs a reduced gain command to the preamplifier gain control circuit. The ready determination pause state is automatically released after the exclusion flag ends.

[0033] S4. Based on the readiness criterion, continue to monitor the output signal during the excluded sampling period, track the starting point of the noise floor fluctuation amplitude recovery, and obtain the noise floor recovery rate after the starting point.

[0034] During the sampling period marked with exclusion tags, monitoring of the output signal does not cease. The noise floor fluctuation amplitude continues to be sampled according to the sampling interval, and the starting point of its recovery is tracked during continuous sampling. The recovery rate of the noise floor fluctuation amplitude is calculated from the starting point to obtain the noise floor recovery rate.

[0035] In scenarios where the source-side mute release time is independent of the switching timing, the timing of mute release is determined by the state of the signal source side. For example, the mute may be released when the tuner's received signal level rises above the mute threshold, or the Bluetooth audio module may resume audio output after the link is established. When the source-side mute is released, the removed background noise re-enters the path, and the noise floor fluctuation amplitude begins to rise from the residual level of the mute period. This starting point is the earliest observable manifestation of the channel transitioning from source-side mute to actual readiness. The rate of rise is determined by both the source-side mute release process and the path establishment process itself. Based on this, this step continuously monitors the background noise fluctuation amplitude during the excluded sampling period. After detecting the starting point, the subsequent rise process is used to calculate the background noise rise rate, providing a basis for timing and rate for gain recovery.

[0036] Specifically, the tracking start point is executed as follows: Based on the readiness judgment criteria, during and continuing from the source-side silent period, the noise floor fluctuation amplitude is continuously sampled according to the sampling interval to obtain a fluctuation amplitude sequence; the increment between adjacent intervals of the fluctuation amplitude sequence is calculated to obtain a fluctuation increment sequence; in the fluctuation increment sequence, positions where the increment of multiple consecutive intervals is continuously positive and the cumulative increment exceeds a preset recovery threshold are detected to obtain candidate recovery positions; the fluctuation amplitude sequence within a preset time period after the candidate recovery position is subjected to a monotonicity test, and positions that fail the test are eliminated to obtain the starting point of the noise floor fluctuation amplitude recovery. The cumulative increment is the sum of the fluctuation increments of each consecutive positive interval starting from the interval where the increment is first positive, and the candidate recovery position is the first interval in this consecutive positive interval segment. The typical setting for the number of consecutive intervals where the increment is continuously positive is multiple adjacent intervals. This setting is based on statistical analysis of the residual fluctuations in the noise floor amplitude during the source-side quiet period, taking the number of intervals where the residual fluctuations are rarely consecutively in the same direction. The adjustable range is adjusted proportionally to the sampling interval duration and the time taken for the noise floor to recover. The typical setting for the preset recovery threshold is higher than the maximum amplitude of the residual fluctuations during the source-side quiet period. This setting is based on the calibration of the statistical distribution of the residual fluctuations during the quiet period measured in the target channel. The adjustable range is adjusted between the residual fluctuations during the quiet period and the normal fluctuation amplitude.

[0037] In one embodiment, the monotonicity test examines the fluctuation amplitude sequence within a preset time period after the candidate rise position interval by interval. If the value at the end of the sequence is higher than the value at the candidate rise position, and there are no intervals in between that fall below the value at the candidate rise position, the test is passed. Otherwise, the candidate rise position is discarded, and new candidate rise positions are detected subsequently. The time when the start point is confirmed is later than the start point itself by a preset time period, during which the preceding stage gain remains suppressed. The preset time period is typically set to cover the time required for the noise floor to rise from the residual level of the muted period to the normal fluctuation amplitude. This setting is based on the audio setup time after the signal source mutes, such as the audio setup time after the tuner mutes or the setup time for Bluetooth audio output recovery. Its adjustable range is adjusted according to this setup time, but it should not be extended further to avoid delaying the start of gain recovery.

[0038] Further, the noise floor recovery rate is obtained as follows: starting from the starting point, the noise floor fluctuation amplitude is extracted according to the sampling interval to obtain the recovery segment fluctuation sequence; in the recovery segment fluctuation sequence, the fluctuation increment between adjacent sampling intervals is divided by the corresponding duration to obtain the segmented recovery rate; the segmented recovery rate is smoothed according to the time sequence, and values ​​that deviate from the mean by more than a preset deviation threshold are removed to obtain the corrected segmented recovery rate; the corrected segmented recovery rate is averaged within the recovery segment to obtain the noise floor recovery rate. Here, the corresponding duration is the time interval between the centers of two adjacent sampling intervals; the recovery segment starts from the starting point and continues until the noise floor fluctuation amplitude recovers to the normal level after the path is established.

[0039] Preferably, the smoothing process uses a moving average. The preset deviation threshold is typically set as a fixed proportion of the average of the segmented rise rate within the rise segment. This setting is based on the statistical distribution calibration of the segmented rise rate measured during multiple source-side mute release processes of the target channel. The threshold is set by taking the proportion where rate fluctuations during normal rise fall within the threshold, and peaks caused by occasional interference fall outside the threshold. Its adjustable range is adjusted according to the regularity of the noise floor rise process. This step is executed by the front-end control unit at the end of each sampling interval, and the obtained noise floor rise rate, along with the starting point, is stored for subsequent steps.

[0040] S5. Based on the noise floor recovery rate, match the rise rate of the preamplifier gain during the recovery period with the noise floor recovery rate to obtain the gain recovery slope.

[0041] The rise rate of the preamplifier gain during the recovery period is set according to the noise floor recovery rate to match the two. After the matched gain rise rate is clipped, it is used together with the recovery time to determine the curve of gain rise over time. This curve is the gain recovery slope.

[0042] In scenarios where the gain recovery process and the channel level change process overlap in time, after the source-side mute is released, the background noise, along with the DC operating point transition, is entering the path, while the channel output level itself is in the process of recovery. If the preceding stage gain rises at a fixed slope independent of this, the two changes are superimposed at the output. When the gain rises faster than the background noise recovery, the subsequently recovering background noise is amplified by the already high-level gain, resulting in an audible increase in background noise. When the gain rises slower than the background noise recovery, the channel level has already recovered to its normal level while the gain is still low, and then the gain continues to rise, resulting in a second rise in the output level independent of the channel's own changes. Even if the recovery timing is correctly determined, rate mismatch will still generate new audible noise during the recovery period. This step matches the gain rise rate with the background noise recovery rate, ensuring that the two changes occur synchronously, resulting in only one smooth level rise at the output.

[0043] Specifically, the gain recovery slope is obtained as follows: The gain increment per unit time during the recovery period is determined based on the noise floor rise rate, resulting in a gain rise rate setting value; the deviation between the gain rise rate setting value and the noise floor rise rate is calculated, and if the deviation exceeds a preset deviation threshold, the gain rise rate setting value is adjusted until the deviation does not exceed the preset deviation threshold, resulting in an adjusted gain rise rate; the adjusted gain rise rate is compared with a preset rate upper limit, and if it exceeds the preset rate upper limit, it is limited by the preset rate upper limit, and the recovery time is extended accordingly, resulting in a limited gain rise rate; the curve of gain increase over time is determined based on the limited gain rise rate and the recovery time, resulting in the gain recovery slope. Here, the preamplifier gain is measured in decibels, and the gain rise rate and the noise floor rise rate are therefore measured in the same unit of measurement. The typical setting for the gain rise rate is the rise rate that the preamplifier gain control circuit can achieve, based on its gain resolution and control cycle, closest to the noise floor recovery rate. This setting is based on the premise that when the gain rise and noise floor recovery are synchronized, the output level rises only once. The achievable rise rate is limited by the discretization of the gain control circuit, and its adjustable range is within the allowable range of a preset deviation threshold, centered around the noise floor recovery rate. The preset deviation threshold is typically set as a fixed proportion of the noise floor recovery rate, based on subjective listening tests. It is set to the proportion where the two changes are audibly indistinguishable within this deviation range, and is not less than half the rate resolution of the gain control circuit to ensure convergence. Its adjustable range varies with the tolerance for noise during recovery; the stricter the requirement, the smaller the threshold.

[0044] In one embodiment, when the deviation exceeds a preset deviation threshold, the gain rise rate setting is adjusted in reverse according to the sign of the deviation. If the gain rises faster than the noise floor rises, it is lowered by one level according to the rate resolution of the gain control circuit; if it is slower than the noise floor rises, it is raised by one level. The deviation is recalculated after each adjustment until the deviation does not exceed the preset deviation threshold.

[0045] It should be noted that the typical setting of the preset rate limit is the maximum gain rise rate that the preamplifier gain control circuit can achieve without producing audible noise. This setting is based on two device parameters: the gain resolution and control cycle of the gain control circuit, and has been confirmed through subjective listening tests. Its adjustable range is adjusted according to the gain resolution and control cycle of the gain control circuit. When the adjusted gain rise rate exceeds the preset rate limit, the preset rate limit is taken as the gain rise rate after limiting. In this case, the gain rise is slower than the noise floor recovery. To ensure that the gain can still recover to the nominal value, the recovery time is correspondingly extended to the difference between the nominal value and the current preamplifier gain divided by the limited gain rise rate.

[0046] It is understandable that the gain recovery slope is not a single numerical value, but rather a curve showing the gain increasing over time, determined by the gain rise rate after limiting and the recovery time. The starting point is the current gain of the preamplifier in the low-voltage state, and the ending point is the nominal value. This step is executed by the preamplifier control unit after confirming the starting point, and the resulting curve is stored in the control unit for subsequent steps to write gain values ​​to the gain control circuit step by step.

[0047] S6. Take the time corresponding to the starting point as the actual ready time of the channel. From this time, gradually increase the preamplifier gain of the target channel according to the gain recovery slope to obtain the target channel output state after the preamplifier gain is restored to the nominal value.

[0048] Obtain the gain recovery slope and take the time corresponding to the starting point as the actual ready time of the channel. From the actual ready time of the channel, the gain of the preceding stage gradually increases according to the gain recovery slope. After the gain of the preceding stage is recovered to the nominal value and locked, the target channel is in the output state of complete recovery.

[0049] In scenarios where the source-side mute reactivates during gain recovery, the source-side state fluctuates around the mute control threshold. For example, if the tuner's received signal level fluctuates around the mute threshold, causing repeated mute activation, or if intermittent Bluetooth audio buffering causes intermittent output cutoff, the mute control may reactivate after being released. The noise floor fluctuation amplitude will then fall back to the level of the mute period during the recovery process. If the preamplifier gain continues to rise according to the gain recovery slope, and the mute is released again after the gain reaches a high level, the cut-off noise floor and the DC operating point jump are amplified again by the high gain, causing the popping problem to reappear. The recovery process must be adjusted according to changes in the channel state. Therefore, this step decomposes the gain recovery into several gain steps. After each step, the noise floor fluctuation amplitude is reacquired, and a pause is determined by its decline, while a continuation is determined by its rise, until all steps are completed and the preamplifier gain is locked at the nominal value.

[0050] Specifically, the gradual increase of the preamplifier gain is performed as follows: starting from the actual ready moment of the channel, the difference between the nominal value and the current preamplifier gain is divided into several gain steps according to the gain recovery slope, resulting in a step sequence; the preamplifier gain of the target channel is increased step by step according to the step sequence, and the noise floor fluctuation amplitude is acquired after each increase to obtain the recovery process monitoring result; when the recovery process monitoring result indicates that the noise floor fluctuation amplitude has fallen below the preset fall-off threshold, subsequent steps are paused and the current preamplifier gain is maintained. After the noise floor fluctuation amplitude rises again, the step sequence is continued to obtain the recovery process after continued execution; after all gain steps are completed in the recovery process after continued execution, the preamplifier gain is locked at the nominal value, resulting in the target channel output state after the preamplifier gain has recovered to the nominal value. The typical setting for the step size of each gain step is a gain amount smaller than the level step that the human ear can perceive. The setting is based on the known value of the level difference threshold in psychoacoustics and the gain resolution of the preamplifier gain control circuit, and its adjustable range is adjusted according to the gain resolution of the gain control circuit. The number of steps is determined by dividing the difference between the nominal value and the current gain of the previous stage by the amount of each step and rounding down. Any margin less than one step is added to the last step. The time interval between two adjacent steps is determined by dividing the amount of each step by the rate of increase of the gain recovery slope at that moment, so that the step sequence as a whole increases along the gain recovery slope.

[0051] In one embodiment, the preset fallback threshold is typically set to be higher than the residual level of the noise floor fluctuation amplitude during the source-side silence period, lower than the normal level after the path is established, and with an interval between it and the normal level not less than the fluctuation of the normal level itself. The setting is based on the statistical distribution calibration of the measured noise floor fluctuation amplitude of the target channel, ensuring that normal fluctuations do not trigger a pause, and that the fallback can be detected promptly when the source-side silence is resumed. Its adjustable range is between the residual level and the normal level. The recovery process monitoring results are updated once after each step, taking the most recent noise floor fluctuation amplitude output by the noise floor detection unit after that step.

[0052] It should be noted that during the pause of subsequent steps, the gain of the preceding stage remains at the level at the time of the pause, neither increasing nor decreasing, while monitoring of the noise floor fluctuation amplitude continues. There is no maximum pause time limit; the transient response entering the path before the source-side mute is lifted is only amplified by the gain at the pause level. If the noise floor fluctuation amplitude rises above the preset fall-off threshold again and remains above that threshold for several consecutive sampling intervals, it is considered a rebound. The remaining steps in the step sequence are then executed from the pause point, and the time interval for subsequent steps is still determined by the gain recovery slope. The typical setting for the number of consecutive intervals remaining above the threshold is multiple adjacent intervals. This setting is based on statistical analysis of residual fluctuations during the source-side mute period, taking the number of intervals where residual fluctuations rarely exceed the threshold consecutively. The adjustable range is adjusted according to the sampling interval duration.

[0053] Furthermore, after all gain steps are executed, the preamplifier gain reaches its nominal value and is locked. Once locked, the preamplifier gain no longer adjusts with changes in the noise floor fluctuation amplitude, and the target channel enters its normal output state. This step is executed by the preamplifier control unit, which writes gain values ​​to the preamplifier gain control circuit step by step according to the step sequence, and reads the noise floor fluctuation amplitude output by the noise floor detection unit after each step is written as the monitoring result of the recovery process.

[0054] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0055] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection between apparatuses or units through some interfaces, and may be electrical, mechanical, or other forms.

[0056] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0057] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0058] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

Claims

1. A method for suppressing transient noise during audio preamp switching, characterized in that, include: S1. Acquire the output signal after the target channel is turned on, filter out the audio band components, extract the background noise energy outside the audio band, track the level fluctuation of the background noise energy, and obtain the background noise fluctuation amplitude. S2. Statistically measure the in-band level fluctuation of the output signal within the observation window, compare the noise floor fluctuation amplitude with the in-band level fluctuation by time period, and determine the time period when both are below their respective thresholds as the source-side silent period. S3. Change the ready determination quantity from the in-band level fluctuation quantity to the noise floor fluctuation amplitude, and exclude the sampling that falls into the source-side silent period from the ready determination, to obtain the ready determination benchmark. S4. Based on the readiness determination criteria, continue to monitor the output signal during the excluded sampling period, track the starting point of the noise floor fluctuation amplitude recovery, and obtain the noise floor recovery rate after the starting point. S5. Based on the noise floor recovery rate, match the rise rate of the preamplifier gain during the recovery period with the noise floor recovery rate to obtain the gain recovery slope. S6. Take the time corresponding to the starting point as the actual ready time of the channel, and gradually increase the preamplifier gain of the target channel according to the gain recovery slope from that time to obtain the target channel output state after the preamplifier gain is restored to the nominal value.

2. The method according to claim 1, characterized in that, The output signal after the target channel is turned on is sampled, the audio band components are filtered out, the background noise energy outside the audio band is extracted, and the level fluctuation of the background noise energy is tracked to obtain the background noise fluctuation amplitude, specifically including: The output signal after the target channel is turned on is filtered by a high-pass filter. The cutoff frequency of the high-pass filter is higher than the upper limit of the audio frequency band, resulting in an out-of-band noise signal. The energy of the out-of-band noise signal is integrated over a continuous sampling interval to obtain the noise floor energy of each sampling interval; The noise floor energy of each sampling interval is converted into a level value according to the sampling time sequence, and the level change between adjacent sampling points is recorded point by point to obtain the noise floor level fluctuation sequence. The noise floor level fluctuation sequence is obtained by taking the difference between the upper and lower envelopes within a continuous sampling interval.

3. The method according to claim 1, characterized in that, The method involves statistically analyzing the in-band level fluctuation of the output signal within the observation window, comparing the noise floor fluctuation amplitude with the in-band level fluctuation over time periods, and defining the time periods during which both are below their respective thresholds as source-side silent periods. Specifically, this includes: The sampling sequence is truncated according to the observation window on the output signal, and the level fluctuation of the audio band components within the observation window is statistically analyzed to obtain the level fluctuation amount within the band. The in-band level fluctuation and the noise floor fluctuation amplitude are time-aligned and compared segment by segment using the same observation window to obtain segment-by-segment comparison results. In the segment-by-segment comparison results, the time period in which the noise floor fluctuation amplitude is lower than the preset fluctuation threshold and the in-band level fluctuation is lower than the preset reference fluctuation is judged as both being lower than their respective thresholds, thus obtaining the initial judgment of the silent period. The adjacent initially determined silence periods are merged according to the sampling time sequence to obtain the source-side silence periods.

4. The method according to claim 3, characterized in that, The step of determining the time period in which the noise floor fluctuation amplitude is lower than a preset fluctuation threshold and the in-band level fluctuation is lower than a preset reference fluctuation is considered as both being lower than their respective thresholds, thus obtaining the initial judgment of a silent period, specifically includes: Within the established range of the output signal path, the level fluctuation of the components in the audio band is statistically analyzed to obtain the preset reference fluctuation amount; The difference between the in-band level fluctuation and the preset reference fluctuation is used to obtain the fluctuation deviation. During the time period when the fluctuation deviation is negative, the noise floor fluctuation amplitude is compared with the preset fluctuation threshold to obtain the noise floor comparison result; The time period in which the fluctuation deviation is negative and the noise floor fluctuation amplitude is lower than the preset fluctuation threshold is determined as both being lower than their respective thresholds, thus obtaining the initial judgment of the silent period.

5. The method according to claim 1, characterized in that, The step of changing the readiness determination quantity from the in-band level fluctuation to the noise floor fluctuation amplitude, and excluding samples falling within the source-side silent period from the readiness determination, to obtain the readiness determination criterion, specifically includes: Stop using the in-band level fluctuation as the readiness determination quantity, and instead use the noise floor fluctuation amplitude as the readiness determination quantity to obtain the replaced readiness determination quantity; An exclusion marker is added to the samples that fall within the source-side silent period to obtain a sample sequence with exclusion markers; During the duration of the exclusion flag, the ready timer is paused, and the preamplifier gain is kept low to obtain the ready determination pause state; The replacement ready determination quantity and the ready determination pause state together constitute the ready determination benchmark.

6. The method according to claim 1, characterized in that, The starting point for tracking the recovery of the noise floor fluctuation amplitude specifically includes: Based on the readiness determination criteria, the noise floor fluctuation amplitude is continuously sampled according to the sampling interval during the source-side silent period to obtain the fluctuation amplitude sequence; The increment between adjacent intervals of the fluctuation amplitude sequence is calculated for each interval to obtain the fluctuation increment sequence; In the fluctuation increment sequence, positions where the increments of multiple consecutive intervals are continuously positive and the cumulative increments exceed a preset recovery threshold are detected to obtain candidate recovery positions; A monotonicity test is performed on the fluctuation amplitude sequence within a preset time period after the candidate rebound position, and positions that fail the test are eliminated to obtain the starting point of the rebound of the noise floor fluctuation amplitude.

7. The method according to claim 1, characterized in that, The noise floor recovery rate after obtaining the starting point specifically includes: Starting from the aforementioned starting point, the noise floor fluctuation amplitude is extracted according to the sampling interval to obtain the rising segment fluctuation sequence; In the recovery phase fluctuation sequence, the fluctuation increment between adjacent sampling intervals is divided by the corresponding duration to obtain the segmented recovery rate; The segmented recovery rate is smoothed according to the time sequence, and values ​​that deviate from the mean by more than a preset deviation threshold are removed to obtain the corrected segmented recovery rate. The average value of the corrected segmented recovery rate within the recovery segment is taken to obtain the noise floor recovery rate.

8. The method according to claim 1, characterized in that, The step of matching the rise rate of the preamplifier gain during the recovery period with the noise floor rise rate to obtain the gain recovery slope specifically includes: The gain increment per unit time during the recovery period is determined based on the noise floor recovery rate, and the gain rise rate setting value is obtained. The deviation between the gain rise rate setting value and the noise floor recovery rate is calculated. When the deviation exceeds a preset deviation threshold, the gain rise rate setting value is adjusted until the deviation does not exceed the preset deviation threshold, thus obtaining the adjusted gain rise rate. The adjusted gain rise rate is compared with the preset rate upper limit. When it exceeds the preset rate upper limit, the gain rise rate is limited according to the preset rate upper limit, and the recovery time is extended accordingly to obtain the limited gain rise rate. The gain rise rate after the limiting and the recovery time are used to determine the curve of gain rise over time, and the gain recovery slope is obtained.

9. The method according to claim 1, characterized in that, The process of gradually increasing the preamplifier gain of the target channel according to the gain recovery slope from that moment onwards, to obtain the target channel output state after the preamplifier gain has recovered to its nominal value, specifically includes: Starting from the actual ready moment of the channel, the difference between the nominal value and the current previous stage gain is divided into several gain steps according to the gain recovery slope to obtain a step sequence; The gain of the target channel is increased step by step according to the step sequence, and the noise floor fluctuation amplitude is acquired after each increase to obtain the monitoring results of the recovery process. When the monitoring result of the recovery process indicates that the noise floor fluctuation amplitude has fallen below the preset fall-off threshold, the subsequent stepping is paused and the current preamplifier gain is maintained. After the noise floor fluctuation amplitude rises again, the stepping sequence is continued to be executed to obtain the recovery process after continued execution. After the recovery process following continued execution completes all the gain steps, the preamplifier gain is locked at the nominal value, resulting in the target channel output state after the preamplifier gain is restored to the nominal value.

Images