Earphone audio output method and system, intelligent earphone and storage medium

Abstract

The invention relates to the technical field of audio processing, in particular to an earphone audio output method and system, an intelligent earphone and a storage medium. The method comprises the following steps: acquiring an original audio signal, extracting a low-frequency signal in the original audio signal, and acquiring frequency information and intensity information of the low-frequency signal; obtaining a reconstruction signal corresponding to the low-frequency signal based on the frequency information, dynamically adjusting the gain of the low-frequency signal based on the intensity information, and obtaining an enhanced signal corresponding to the low-frequency signal; fusing the reconstructed signal and the enhanced signal into a newly-added low-frequency signal, and fusing the newly-added low-frequency signal and a non-low-frequency signal in the original audio signal to obtain a fused audio signal; performing balance adjustment on gains of a newly added low-frequency signal and a non-low-frequency signal in the fused audio signal to obtain a full-band signal; and obtaining a hearing demand and an acoustic parameter of the intelligent earphone, and adjusting the full-band signal according to the hearing demand and the acoustic parameter to obtain a target output signal. According to the invention, innate attenuation of low-frequency sound of the intelligent earphone can be complemented.

Description

A headphone audio output method, system, smart headphone, and storage medium Technical Field

[0001] This invention relates to the field of audio processing technology, and in particular to a headphone audio output method, system, smart headphone, and storage medium. Background Technology

[0002] In the field of audio equipment, OWS (open-back headphones) / TWS (true wireless earbuds) semi-open headphones have become one of the mainstream products in the consumer electronics market due to their advantages such as comfortable wearing, wide soundstage, and no need to close the ear canal. However, due to the limitations of the semi-open acoustic structure, there is a natural sound wave leakage channel between the headphones and the ear canal. For low-frequency sound waves with longer wavelengths below 250Hz, their energy cannot form effective reflection and superposition in the ear canal as in in-ear headphones, resulting in a significant attenuation of audio signals in this frequency band, with an attenuation of about 40%. This directly manifests as thin bass and insufficient depth, making it difficult to meet users' needs for a full bass listening experience. Summary of the Invention

[0003] This invention provides a headphone audio output method, system, smart headphone, and storage medium to solve the problem of significant attenuation of low-frequency sound waves.

[0004] This invention discloses a headphone audio output method applicable to semi-open or fully open smart headphones. The headphone audio output method includes: acquiring an original audio signal; extracting a low-frequency signal from the original audio signal; acquiring frequency information and intensity information of the low-frequency signal; acquiring a reconstructed signal corresponding to the low-frequency signal based on the frequency information; dynamically adjusting the gain of the low-frequency signal based on the intensity information; acquiring an enhanced signal corresponding to the low-frequency signal; fusing the reconstructed signal and the enhanced signal into a new low-frequency signal; fusing the new low-frequency signal with non-low-frequency signals from the original audio signal to obtain a fused audio signal; equalizing the gain of the new low-frequency signal and the non-low-frequency signals in the fused audio signal to obtain a full-frequency signal; acquiring the listening requirements and acoustic parameters of the smart headphones; and adjusting the full-frequency signal according to the listening requirements and acoustic parameters to obtain a target output signal.

[0005] Optionally, the step of adjusting the full-band signal to obtain the target output signal according to the listening requirements and the acoustic parameters includes: dividing the full-band signal into multiple independent audio segments according to frequency; obtaining a compression threshold and compression ratio set for each audio segment; when the signal amplitude of the audio segment exceeds the compression threshold, compressing the signal amplitude using the compression ratio to obtain a frequency segment control signal.

[0006] Optionally, the step of obtaining the compression threshold and compression ratio set for the low-frequency audio band includes: obtaining the dynamic range parameters of the smart headphones, the dynamic range parameters including frequency response curves, acoustic structure and sound unit performance, and obtaining the compression threshold and compression ratio based on the dynamic range parameters and the listening requirements.

[0007] Optionally, after the step of acquiring the frequency band control signal, the method includes: acquiring a global limiting value, constructing a limiter based on the global limiting value, smoothly limiting the peak value of the frequency band control signal through the limiter, and acquiring the target output signal.

[0008] Optionally, the step of obtaining the global limiting value includes: obtaining the limiting parameters and safety specification parameters of the smart headphones, wherein the limiting parameters include a digital domain clipping threshold, a reference power amplifier output upper limit, and a speaker unit displacement threshold; and obtaining the global limiting value based on the limiting parameters and the listening requirements.

[0009] Optionally, the step of equalizing the gain of the newly added low-frequency signal and the non-low-frequency signal in the fused audio signal includes: obtaining the frequency response curve and acoustic structure parameters of the smart headphones; obtaining a global equalization curve based on the frequency response curve, the acoustic structure parameters and the listening requirements; and performing gain equalization adjustment on the non-low-frequency signal and the newly added low-frequency signal based on the global equalization curve.

[0010] Optionally, the step of dynamically adjusting the gain of the low-frequency signal based on the intensity information includes: increasing the gain when the low-frequency signal intensity is below standard, and decreasing the gain or not increasing the gain when the low-frequency signal intensity meets standard.

[0011] This invention also discloses an earphone audio output system, comprising: an acquisition module for acquiring an original audio signal, extracting a low-frequency signal from the original audio signal, and acquiring frequency information and intensity information of the low-frequency signal; a low-frequency module for acquiring a reconstructed signal corresponding to the low-frequency signal based on the frequency information, dynamically adjusting the gain of the low-frequency signal based on the intensity information, and acquiring an enhanced signal corresponding to the low-frequency signal; a fusion module for fusing the reconstructed signal and the enhanced signal into a new low-frequency signal, fusing the new low-frequency signal with non-low-frequency signals in the original audio signal, and acquiring a fused audio signal; an equalization module for equalizing the gain of the new low-frequency signal and the non-low-frequency signals in the fused audio signal, and acquiring a full-frequency signal; and a target module for acquiring the listening requirements and acoustic parameters of the smart earphone, and adjusting the full-frequency signal according to the listening requirements and the acoustic parameters to acquire a target output signal.

[0012] The present invention also discloses a computer-readable storage medium storing a computer program, which, when executed by a processor, causes the processor to perform the steps of the method described above.

[0013] The present invention also discloses a smart earphone, including a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor performs the steps of the method described above.

[0014] The beneficial effects of the headphone audio output method, system, smart headphones, and storage medium provided by this invention are as follows: Based on low-frequency signal strength information, the gain is dynamically adjusted to achieve on-demand control, compensating for weak signals and suppressing strong signals. This not only specifically compensates for the inherent low-frequency attenuation of semi-open / fully open headphones, improving the fullness and extension of low-frequency sound, but also effectively avoids sound quality degradation problems such as clipping distortion and popping sounds caused by excessive gain. Gain equalization adjustment is performed on the newly added low-frequency signal and non-low-frequency signal in the fused audio signal to construct a globally balanced curve with balanced energy distribution. This optimizes the frequency band connection between the newly added low-frequency and non-low-frequency signals, eliminates frequency band energy gaps and phase deviations, and achieves a balanced listening experience across the entire frequency band, with full and clear low frequencies, clear and unrecessed mid frequencies, and transparent and non-harsh high frequencies. Attached Figure Description

[0015] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. In the drawings: Figure 1 is a flowchart of the first embodiment of the headphone audio output method provided by the present invention; Figure 2 is a flowchart of the second embodiment of the headphone audio output method provided by the present invention; Figure 3 is a schematic diagram of acoustic curve comparison provided by the present invention; Figure 4 is a structural schematic diagram of an embodiment of the headphone audio output system provided by the present invention; Figure 5 is a structural schematic diagram of an embodiment of the smart headphone provided by the present invention; Figure 6 is a structural schematic diagram of an embodiment of the computer-readable storage medium provided by the present invention. Detailed Implementation

[0016] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. The preferred embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0017] Please refer to Figure 1, which is a flowchart illustrating the first embodiment of the headphone audio output method provided by this invention. The headphone audio output method provided by this invention is applied to semi-open or fully open smart headphones. The core design of semi-open / fully open headphones is that the ear canal is not completely closed, leaving a gap between the earcup and the ear canal, allowing sound waves to penetrate in both directions. Low-frequency sound waves (below 250Hz) have longer wavelengths (e.g., a 50Hz sound wave has a wavelength of approximately 6.8 meters), making them more prone to leakage through the gap. Unlike in-ear headphones, they cannot form reflections and superpositions within the ear canal, ultimately resulting in a low-frequency energy attenuation of up to 40%, leading to thin, weak bass that lacks the rich, textured feel of in-ear headphones.

[0018] The headphone audio output method provided by the present invention includes the following steps: S101: acquire the original audio signal, extract the low-frequency signal from the original audio signal, and acquire the frequency information and intensity information of the low-frequency signal.

[0019] In a specific implementation scenario, smart headphones receive raw audio signals from playback devices (such as mobile phones, tablets, etc.). These raw audio signals include complete full-frequency information (typically 20Hz-20kHz). A high-precision bandpass filter can be used to extract low-frequency signals from the raw audio signal (for open-back headphones, low-frequency signals are typically below 250Hz, the range with the most severe low-frequency attenuation). The Q value (bandwidth) of the bandpass filter is set to medium to ensure that only low-frequency signals are separated, without mixing in mid-to-low frequency transition signals. The low-frequency weakness of open-back headphones lies in the physical attenuation below 250Hz; separating the low-frequency signal first allows for targeted enhancement, avoiding ineffective processing of mid-to-high frequency signals.

[0020] For the extracted low-frequency signal, its frequency and intensity information are obtained. Frequency information refers to the frequency band distribution of the low-frequency signal, such as whether it's a 60Hz deep bass, a 150Hz mid-low frequency, or a low-frequency signal composed of multiple frequency bands. Intensity information is the instantaneous amplitude value of the low-frequency signal (measured in dBFS, the amplitude measurement unit for digital audio), reflecting the strength of the low-frequency signal. For example, the low-frequency intensity corresponding to a drum beat will spike instantaneously, while the low-frequency intensity of light music will remain consistently low.

[0021] The remaining mid-to-high frequency signals (signals above 250Hz) after separating the low frequencies are temporarily stored in the buffer module, maintaining their original state, and are not involved in this step.

[0022] S102: Obtain the reconstructed signal corresponding to the low-frequency signal based on frequency information, dynamically adjust the gain of the low-frequency signal based on intensity information, and obtain the enhanced signal corresponding to the low-frequency signal.

[0023] In a specific implementation scenario, on the one hand, based on frequency information, the reconstructed signal corresponding to the low-frequency signal is obtained. Utilizing the harmonic superposition effect in psychoacoustics, instead of directly increasing the amplitude of the low-frequency signal, the sense of low-frequency extension is simulated by enhancing harmonics. Specifically, the fundamental frequency of the acquired low-frequency signal is identified, and its frequency information is obtained (e.g., the fundamental frequency is identified as 50Hz). The nth harmonic of this fundamental frequency is extracted, where n is an integer greater than 1. For example, the 50Hz fundamental frequency corresponds to the 100Hz and 150Hz harmonics, both belonging to the mid-to-low frequency range that the human ear is sensitive to. Furthermore, the harmonic signal can be appropriately amplified using a harmonic gain algorithm. The gain amplitude is adjusted based on the fundamental frequency, with higher harmonic gain for low-frequency fundamental frequencies (e.g., 20-60Hz) and lower harmonic gain for mid-to-low frequency fundamental frequencies (e.g., 100-250Hz). The amplified harmonic signal is then superimposed on the original low-frequency fundamental signal to generate the reconstructed signal. The reconstructed signal is not a true enhancement of low frequencies, but rather uses harmonics to make the brain perceive a fuller low frequency.

[0024] Meanwhile, by using real-time amplitude monitoring and dynamic gain calculation to dynamically adjust the gain and generate an enhanced signal, excessive gain can be avoided. Specifically, the low-frequency target intensity threshold is derived by reverse engineering based on the headphone's acoustic frequency response curve. The intensity information of the collected low-frequency signal is compared with the low-frequency target intensity threshold. If the intensity information is lower than the low-frequency target intensity threshold, a compensation gain value is automatically calculated (e.g., a gain of 12dB if it is 12dB lower). If the intensity information is higher than or equal to the low-frequency target intensity threshold, the gain value is automatically reduced or even turned off to prevent the signal amplitude from exceeding the hardware processing limit. The parameters of the dynamic gain (gain value, bandwidth) are adjusted in milliseconds according to the signal intensity, ultimately generating an enhanced signal.

[0025] The reconstructed signal addresses the problem of insufficient low-frequency extension, while the enhanced signal addresses the problem of insufficient low-frequency intensity. The dual-path parallel processing balances virtual extension and real enhancement, resulting in a low-frequency listening experience that is both full and stable.

[0026] S103: Fuse the reconstructed signal and the enhanced signal into a new low-frequency signal, and fuse the new low-frequency signal with the non-low-frequency signals in the original audio signal to obtain the fused audio signal.

[0027] In a specific implementation scenario, the reconstructed signal and the enhanced signal are fused into a new low-frequency signal. This can be achieved through a weighted fusion of the reconstructed and enhanced signals, or by direct superposition and fusion of the two signals. The fusion process employs linear superposition combined with smooth transition processing to eliminate abrupt amplitude changes after the two signals are superimposed, thereby generating the new low-frequency signal.

[0028] For example, a dynamic weighting algorithm can be introduced to adjust the proportion of the two path signals based on the real-time intensity of the low-frequency signal. If the original low-frequency signal intensity is extremely low (such as light music), the reconstructed signal accounts for 60%-70%, relying on harmonics to simulate the low-frequency listening experience, avoiding excessive gain of the enhanced signal that could cause overload; if the original low-frequency signal intensity is moderate (such as pop music), the enhanced signal accounts for 60%-70%, relying on real gain to ensure the low-frequency texture, with the reconstructed signal assisting in extension; if the original low-frequency signal intensity is extremely high (such as heavy bass electronic music): reduce the proportion of the dual-path signals, with the original low-frequency signal as the main component, to avoid excessive amplitude after superposition.

[0029] The cached non-low-frequency signals (signals above 250Hz) are retrieved and merged with the newly added low-frequency signal in different frequency bands. During the merging process, a frequency band isolation filter is activated to strictly limit the frequency range of the newly added low-frequency signal (not exceeding 250Hz) to prevent the low-frequency signal from spreading into the mid-to-high frequency band. Time-domain phase calibration is performed on the merged signal to eliminate the phase difference between the low-frequency and mid-to-high frequency signals, avoiding the hollowness in the sound quality caused by phase cancellation, and finally generating a fused audio signal.

[0030] S104: Equalize the gain of the newly added low-frequency signal and non-low-frequency signal in the fused audio signal to obtain a full-band signal.

[0031] In a specific implementation scenario, a global equalization curve can be obtained by combining the frequency response curve of the smart headphones measured in a standard anechoic chamber with the target compensation curve defined by the acoustic engineer. Three core parameters can be set for key frequency points: center frequency (F), gain value (G, in dB), and quality factor (Q value, controlling the affected bandwidth). Connecting all these frequency point adjustments forms a global equalization curve uniquely customized for the smart headphones. Based on the global equalization curve, gain equalization is performed on non-low frequency signals and newly added low frequency signals, thereby organically and harmoniously integrating the low-frequency enhancement results of the pre-amplifier technology into the complete audio spectrum.

[0032] The merged audio signal is divided into three main frequency bands: low frequency (<250Hz), mid frequency (250Hz-2kHz), and high frequency (>2kHz). Each main frequency band is further subdivided into 2-3 sub-bands (e.g., low frequency is subdivided into 20-60Hz and 60-250Hz). Preset equalization standards of 25%-35% (low frequency), 35%-45% (mid frequency), and 25%-35% (high frequency) are used, and a global equalization curve is set based on these standards. If a sub-band has excessive energy (e.g., a peak at 1kHz in the mid-frequency band, resulting in a sharp sound), appropriately attenuate the gain of that band. If a sub-band has insufficient energy (e.g., a weak 8kHz in the high-frequency band, resulting in a dull sound), appropriately increase the gain of that band. During adjustment, maintain a relatively stable gain for the newly added low-frequency signal to avoid the equalization adjustment negating the previous low-frequency enhancement effect.

[0033] In one implementation scenario, the frequency response curve and acoustic structural parameters of the smart headphones are collected. The frequency response curve of the smart headphones (covering the entire frequency band from 20Hz to 20kHz) can be collected using a professional audio analyzer. The frequency response curve is analyzed to obtain key feature points. For example, the attenuation amplitude of each sub-band below 250Hz (20-60Hz ultra-low frequency, 60-250Hz mid-low frequency) is identified (e.g., 40% attenuation in 20-60Hz, 25% attenuation in 60-250Hz); the natural peaks (e.g., a 3dB peak in the 1kHz vocal band, which can sound sharp) or attenuation points (e.g., a 2dB attenuation in the 8kHz high frequency, resulting in loss of detail) in the mid-frequency band (500Hz-2kHz) and high-frequency band (2kHz-20kHz) are located; and frequency response breaks or energy abrupt changes are marked at the transition points of each frequency band (e.g., the 250Hz-500Hz mid-low frequency transition section).

[0034] The acoustic structure information of the core hardware of smart headphones is extracted, and based on this information, acoustic design parameters of the cavity are obtained, including cavity volume, cavity resonant frequency (open-back headphones commonly resonate between 80-150Hz, which can easily lead to muddiness in this frequency range), coupling method between the driver and the cavity, and ear canal fit gap (affecting the degree of low-frequency leakage). Based on the cavity acoustic design parameters, safe boundaries for gain adjustment are obtained, including the frequency response range of the driver, rated power / maximum power threshold, displacement threshold (to avoid excessive diaphragm vibration and damage), sensitivity curve, amplifier output upper limit, and digital domain 0dBFS clipping threshold. For example, based on the driver power threshold and displacement threshold, the maximum gain upper limit for each frequency band is set (e.g., maximum gain in ultra-low frequencies should not exceed 15dB to avoid driver overload and burnout; maximum gain in high frequencies should not exceed 5dB to avoid uncontrolled high-frequency diaphragm vibration). For the cavity resonant frequency (e.g., 100Hz), the gain upper limit for that frequency band is set in the curve (e.g., ≤5dB) to avoid excessive gain exacerbating resonance and producing howling or muddiness.

[0035] The system collects the listening needs of the target customers for this smart earphone, such as enhancing bass, improving vocal clarity, and enhancing the transparency of high-frequency details.

[0036] Based on the measured acoustic frequency response curve, the initial curve for defect compensation is derived in reverse to specifically address the inherent acoustic shortcomings of the product. Taking a preset equalization standard of 25%-35% for low frequencies, 35%-45% for mid frequencies, and 25%-35% for high frequencies as an example, the original frequency response curve of the headphones is measured using a professional audio analyzer, covering the entire frequency range of 20Hz-20kHz, to obtain the deviation values ​​of each frequency band from the equalization standard. For example, in the low-frequency range (20-250Hz), the attenuation is 40%, with an energy share of only 10%, far below the 25% lower limit of equalization; in the high-frequency range (2kHz-20kHz), if the attenuation is 2dB, the energy share is only 15%, also below the 25% lower limit of equalization; and in the mid-frequency range, there is a 3dB peak, with an energy share exceeding 50%, exceeding the 45% upper limit of equalization.

[0037] With the goal of achieving a balanced energy ratio across the three frequency bands, the initial step gain values ​​for each band are derived in reverse. For example, a step gain of 10-15dB is set for the 20-60Hz ultra-low frequency band, and a step gain of 5-10dB is set for the 60-250Hz mid-low frequency band, with the goal of increasing the total low-frequency energy ratio to a balanced range of 25%-35%. A step attenuation of 2-3dB is set for the 500Hz-1kHz peak frequency band, with the goal of reducing the mid-frequency energy ratio to a balanced range of 35%-45%. A step gain of 2-3dB is set for the 8kHz-16kHz attenuation frequency band, with the goal of increasing the total high-frequency energy ratio to a balanced range of 25%-35%.

[0038] Acoustic structural parameters of the smart earphones are collected, and a safety boundary for gain adjustment is set to ensure that the equalization curve adjustment does not exceed the hardware's capacity. Based on the safety boundary of gain adjustment, the initial curve for defect compensation is adjusted to ensure that the enhanced low-frequency signal can be efficiently radiated through the cavity while reducing sound wave leakage, thus obtaining a constraint curve. For example, the maximum gain limit in the low-frequency band is locked at 15dB to ensure that the proportion of low-frequency energy after adjustment does not exceed 35% (the unit's safety limit); the maximum gain limit in the high-frequency band is locked at 5dB to prevent uncontrolled high-frequency vibration of the diaphragm.

[0039] Based on listening preferences, gain optimization is performed on the constraint curve to obtain a global equalization curve. For example, if the listening preference leans towards bass, the gain of the 20-60Hz ultra-low frequency range is increased by 2-3dB within safe limits, and the gain of the 60-250Hz mid-low frequency range is increased by 1-2dB, while slightly attenuating the 1-2kHz mid-frequency range (to prevent low frequencies from masking vocals). Furthermore, the global equalization curve can be smoothed to eliminate abrupt gain changes at frequency band transitions, ensuring that the curve fluctuation range is ≤±2dB.

[0040] Based on the global equalization curve, the gain of newly added low-frequency and non-low-frequency signals in the fused audio signal is adjusted to ensure a balanced energy distribution across all frequency bands. A multi-channel high-precision bandpass filter can be used to split the fused audio signal into multiple independent sub-bands, each corresponding to a different gain range of the global equalization curve. A multi-band parametric equalizer is employed, with an independent adjustment channel configured for each sub-band, strictly adhering to the different gain values, center frequency, and Q value of the global equalization curve.

[0041] The adjusted sub-band signals are linearly weighted and fused according to the energy proportion weight set by the global equalization curve; at the same time, the phase calibration module is enabled to correct the phase deviation of each frequency band (≤15°) and eliminate the energy gap at the frequency band junction; finally, a smoothing filter is used to ensure that the energy curve of the full-band signal is smooth and continuous.

[0042] S105: Acquires the listening requirements and acoustic parameters of the smart headphones, and adjusts the full-frequency signal according to the listening requirements and acoustic parameters to obtain the target output signal.

[0043] In a specific implementation scenario, the listening requirements and acoustic parameters of the smart headphones are obtained. These acoustic parameters include the headphones' inherent hardware parameters, such as the frequency response curve, power threshold, displacement threshold, and acoustic resonance frequency of the driver unit, as well as the measured low-frequency attenuation curve of the open-back structure. Based on these acoustic parameters, the signal amplitude is adjusted to ensure it does not exceed the power / displacement limits of the driver unit (to avoid unit damage) and to avoid the cavity resonance frequency (e.g., if the cavity resonance is at 80Hz, the gain in that frequency band is reduced to avoid feedback).

[0044] Based on listening preferences, further adjustments are made. If the user prefers heavy bass, the gain of the newly added low-frequency signal is appropriately increased; if the user prefers vocals, the gain of the mid-frequency band (250Hz-2kHz) is increased to make vocals clearer. After adjustment, a limiter is used to finally control the peak values ​​of the full-frequency signal (ensuring that they do not exceed the digital audio 0dBFS clipping threshold) to generate the target output signal.

[0045] As described above, this embodiment dynamically adjusts the gain based on low-frequency signal strength information, achieving on-demand control by compensating for weak signals and suppressing strong signals. This not only specifically compensates for the inherent low-frequency attenuation of semi-open / fully open headphones, improving the fullness and extension of the low-frequency sound, but also effectively avoids sound quality degradation problems such as clipping distortion and popping sounds caused by excessive gain. Gain equalization is applied to the newly added low-frequency signal and non-low-frequency signal in the fused audio signal to construct a globally balanced energy distribution curve. This optimizes the frequency band transition between the newly added low-frequency and non-low-frequency signals, eliminates frequency band energy gaps and phase deviations, and achieves a balanced listening experience across the entire frequency range, with full and clear low frequencies, clear and unrecessed mid frequencies, and transparent and non-harsh high frequencies.

[0046] Please refer to Figure 2, which is a flowchart illustrating a second embodiment of the headphone audio output method provided by the present invention. The headphone audio output method provided by the present invention includes the following steps: S201: Acquire the original audio signal, extract the low-frequency signal from the original audio signal, and acquire the frequency information and intensity information of the low-frequency signal.

[0047] S202: Obtain the reconstructed signal corresponding to the low-frequency signal based on frequency information, dynamically adjust the gain of the low-frequency signal based on intensity information, and obtain the enhanced signal corresponding to the low-frequency signal.

[0048] S203: Fuse the reconstructed signal and the enhanced signal into a new low-frequency signal, and fuse the new low-frequency signal with the non-low-frequency signals in the original audio signal to obtain a fused audio signal.

[0049] S204: Equalize the gain of the newly added low-frequency signal and non-low-frequency signal in the fused audio signal to obtain a full-frequency signal.

[0050] In a specific implementation scenario, steps S201-S204 are basically the same as steps S101-S104 of the first embodiment of the headphone audio output method provided by the present invention, and will not be described again here.

[0051] S205: Divide the full-band signal into multiple independent audio segments according to frequency, and obtain the low-frequency audio segment from the independent audio segments.

[0052] In a specific implementation scenario, a multi-channel high-precision bandpass filter is used to split the full-band signal, after gain equalization, into three independent signal channels: low-frequency (<250Hz), mid-frequency (250Hz-2kHz), and high-frequency (>2kHz). Each channel corresponds to a set of customized compression parameters (compression threshold and compression ratio). During the splitting process, the Q value of the filter (1.5-2.5) is adjusted to ensure clear boundaries between each frequency band and avoid control failures caused by signal crosstalk.

[0053] S206: Obtain the compression threshold and compression ratio set for the low-frequency audio band. When the signal amplitude of the low-frequency audio band exceeds the compression threshold, the signal amplitude is compressed using the compression ratio to obtain the frequency band control signal.

[0054] In a specific implementation scenario, the dynamic range parameters of the smart headphones are obtained. The dynamic range parameters include frequency response curve, maximum linear displacement, sound unit performance and power distortion threshold. Based on the dynamic range parameters and listening requirements, the compression threshold and compression ratio are obtained.

[0055] For peak frequencies in the frequency response curve (such as the 1kHz mid-frequency peak), increase the compression ratio to suppress peak energy; for attenuation frequencies (such as the 8kHz high-frequency attenuation), decrease the compression ratio to ensure sufficient output in that frequency range. Considering the target listening experience, if the target is enhanced bass, set the low-frequency compression ratio to 2:1~3:1 (moderate compression to ensure impact while avoiding distortion); if the target is clear vocals, set the mid-frequency compression ratio to 1.5:1 (slight compression to preserve vocal dynamics).

[0056] Based on the performance of the sound unit (power threshold, maximum linear displacement) as the hardware safety baseline, the maximum output amplitude of each frequency band is set (e.g., the maximum amplitude of the low-frequency band should not exceed 80% of the unit's rated power). Considering the high high-frequency sensitivity of the unit, the compression threshold of the high-frequency band is increased to reduce unnecessary compression and preserve high-frequency details and dynamics. Based on the product's acoustic structure, the compression threshold of the earphone's cavity resonant frequency (e.g., 80-150Hz) is lowered (i.e., compression is initiated earlier) to avoid signal overload caused by resonance superposition. Considering the low-frequency leakage characteristics caused by the ear canal's fit gap, the compression ratio of the low-frequency band is appropriately relaxed to ensure that the enhanced low-frequency signal can be effectively radiated.

[0057] Ultimately, compression thresholds and compression ratios were determined for the low, medium, and high frequency bands to form differentiated frequency band control schemes.

[0058] Signals in three independent frequency bands are processed separately. In the low-frequency band, when the signal amplitude exceeds a set threshold, the amplitude is reduced by a compression ratio of 2:1 to 3:1 to ensure sufficient output without exceeding the low-frequency power threshold of the driver unit, thus eliminating clipping distortion and popping sounds. In the mid-frequency band, a low compression ratio of 1.5:1 is used, with only slight compression applied to extreme peak signals. This ensures the dynamic performance of vocals and instrumental melodies while avoiding the abruptness caused by excessively prominent mid-frequency peaks. In the high-frequency band, the compression ratio is set to 1:1 (almost no compression), and protection is activated only when the signal amplitude approaches the hardware limit. This maximizes the preservation of high-frequency details and soundstage openness, avoiding high-frequency interference and sound quality degradation caused by traditional full-band DRC.

[0059] S207: Obtain the global limiting value, construct a limiter based on the global limiting value, and smoothly limit the peak value of the frequency band control signal through the limiter to obtain the target output signal.

[0060] In a specific implementation scenario, a limiter performs the final precise control of the signal peak across the entire frequency band, ensuring that the audio signal is free of clipping distortion, the hardware operates safely, and complies with safety regulations. It acquires the extreme parameters and safety specification parameters of the smart headphones, and obtains a global limiting value based on these extreme parameters and listening requirements.

[0061] Limiting parameters include the power threshold of the sound unit (rated power and maximum power limit across the entire frequency range), displacement threshold (the maximum safe vibration displacement of the diaphragm is a key indicator of hardware protection), reference amplifier output limit (the maximum output voltage / current of the amplifier directly determines the maximum driveable amplitude of the signal; the limiting threshold must be 5%-10% lower than the amplifier output limit), and digital domain 0dBFS clipping threshold (the absolute clipping critical point of digital audio signals; signals exceeding this value will be forcibly truncated, resulting in irreversible clipping distortion).

[0062] Safety specifications can be obtained from EN50332 Audio Equipment Sound Pressure Safety Specification. This specification clarifies the maximum permissible sound pressure level for headphone products. The limiting threshold must ensure that the final output sound pressure does not exceed the specification limit to avoid hearing damage at high volumes. Especially for high-frequency signals, threshold control is required to prevent sound pressure from exceeding the standard.

[0063] When the peak value of the full-band signal approaches the limiting threshold (e.g., reaching 90% of the threshold), the limiter initiates soft-inflection smooth compression, gradually increasing the compression ratio from 1:1 to 5:1. This allows the signal amplitude to slowly approach the threshold, preventing steep amplitude changes near the threshold and thus avoiding threshold distortion. The limiter manages the entire full-band signal after frequency-band DRC processing, rather than processing each frequency band individually. This ensures that the peak values ​​of low, medium, and high-frequency signals are synchronously controlled, preventing overall distortion caused by peak escape in a particular frequency band.

[0064] Please refer to Figure 3, which is a schematic diagram comparing the acoustic curves provided by this invention. In Figure 3, the green curve is the original acoustic curve output without the method provided by this invention, and the blue curve is the acoustic curve output after using the headphone audio output method provided by this invention. Comparing the two curves, it can be seen that in the key low-frequency range below 250Hz, the blue curve achieves a comprehensive and significant amplitude improvement compared to the green curve, covering the entire low-frequency range from ultra-low frequencies (20-60Hz) to mid-low frequencies (60-250Hz), and the improvement amplitude shows a differentiated distribution. In the ultra-low frequency sub-band (such as 20-50Hz) where the original headphone acoustic curve attenuates the most, the amplitude improvement effect is the most prominent, with a maximum improvement amplitude of about 22dB. This curve comparison result directly verifies the effectiveness of this invention, which, through targeted signal enhancement processing, significantly compensates for the energy attenuation shortcomings caused by low-frequency sound wave leakage in semi-open / fully open headphones.

[0065] As described above, this embodiment, through customized parameters for each frequency band, can specifically ensure sufficient output and hardware safety in the low-frequency band while preserving the dynamics and details of the mid-to-high frequencies and avoiding volume fluctuations. Simultaneously, by combining parameter setting logic with product hardware characteristics and listening requirements, amplitude control better addresses the acoustic shortcomings of semi-open / fully open headphones, achieving the triple goals of sufficient output across all frequency bands, robust hardware safety, and stable and harmonious sound quality. Through the aforementioned precise threshold calibration and combined control logic, the limiter achieves the triple goals of hardware safety protection, complete elimination of distortion, and stable and high-quality sound, providing a reliable guarantee for the high-quality output of the final audio signal.

[0066] Please refer to Figure 4, which is a structural schematic diagram of an embodiment of the headphone audio output system provided by the present invention. The headphone audio output system 10 includes: an acquisition module 11 for acquiring the original audio signal, extracting the low-frequency signal from the original audio signal, and acquiring the frequency information and intensity information of the low-frequency signal; a low-frequency module 12 for acquiring the reconstructed signal corresponding to the low-frequency signal based on the frequency information, dynamically adjusting the gain of the low-frequency signal based on the intensity information, and acquiring the enhanced signal corresponding to the low-frequency signal; a fusion module 13 for fusing the reconstructed signal and the enhanced signal into a new low-frequency signal, fusing the new low-frequency signal and the non-low-frequency signal in the original audio signal, and acquiring a fused audio signal; an equalization module 14 for equalizing the gain of the new low-frequency signal and the non-low-frequency signal in the fused audio signal, and acquiring a full-band signal; and a target module 15 for acquiring the listening requirements and acoustic parameters of the smart headphones, adjusting the full-band signal according to the listening requirements and acoustic parameters, and acquiring a target output signal.

[0067] The target module 15 is used to divide the full-band signal into multiple independent audio segments according to frequency; obtain the compression threshold and compression ratio set for each audio segment; when the signal amplitude of the audio segment exceeds the compression threshold, the compression ratio is used to compress the signal amplitude and obtain the frequency segment control signal.

[0068] The target module 15 is used to obtain the dynamic range parameters of the smart headphones. The dynamic range parameters include the frequency response curve, acoustic structure and sound unit performance. Based on the dynamic range parameters and listening requirements, the compression threshold and compression ratio are obtained.

[0069] The target module 15 is used to obtain the global limiting value, construct a limiter based on the global limiting value, and smoothly limit the peak value of the frequency band control signal through the limiter to obtain the target output signal.

[0070] The target module 15 is used to obtain the limit parameters and safety specification parameters of the smart headphones. The limit parameters include the digital domain clipping threshold, the upper limit of the reference power amplifier output, and the speaker unit displacement threshold. Based on the limit parameters and listening requirements, the global limiting value is obtained.

[0071] The equalization module 14 is used to obtain the frequency response curve and acoustic structure parameters of the smart headphones, and obtain a global equalization curve based on the frequency response curve, acoustic structure parameters and listening requirements; and perform gain equalization adjustment on non-low frequency signals and newly added low frequency signals based on the global equalization curve.

[0072] The low-frequency module 12 is used to increase the gain when the low-frequency signal strength is insufficient, and to reduce the gain or not increase the gain when the low-frequency signal strength is sufficient.

[0073] As described above, this embodiment dynamically adjusts the gain based on low-frequency signal strength information, achieving on-demand control by compensating for weak signals and suppressing strong signals. This not only specifically compensates for the inherent low-frequency attenuation of semi-open / fully open headphones, improving the fullness and extension of the low-frequency sound, but also effectively avoids sound quality degradation problems such as clipping distortion and popping sounds caused by excessive gain. Gain equalization is applied to the newly added low-frequency signal and non-low-frequency signal in the fused audio signal to construct a globally balanced energy distribution curve. This optimizes the frequency band transition between the newly added low-frequency and non-low-frequency signals, eliminates frequency band energy gaps and phase deviations, and achieves a balanced listening experience across the entire frequency range, with full and clear low frequencies, clear and unrecessed mid frequencies, and transparent and non-harsh high frequencies.

[0074] Please refer to Figure 5, which is a structural schematic diagram of an embodiment of the smart earphone provided by the present invention. The energy storage device 20 includes a processor 21 and a memory 22. The processor 21 is coupled to the memory 22. The memory 22 stores a computer program, which the processor 21 executes during operation to implement the method described above. Detailed steps can be found above and will not be repeated here.

[0075] Please refer to Figure 6, which is a schematic diagram of an embodiment of the computer-readable storage medium provided by the present invention. The computer-readable storage medium 30 stores at least one computer program 31, which is executed by a processor to implement the method described above. Detailed steps are described above and will not be repeated here. In one embodiment, the computer-readable storage medium may be a storage chip in a terminal, a hard disk, a portable hard disk, a USB flash drive, an optical disc, or other readable and writable storage tools, or it may be a server, etc.

[0076] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and RAMbus dynamic RAM (RDRAM), etc.

[0077] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0078] It should be understood that the above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit them. Those skilled in the art can modify the technical solutions described in the above embodiments, or make equivalent substitutions for some of the technical features; and all such modifications and substitutions should fall within the protection scope of the appended claims of the present invention.

Claims

1. A method for outputting audio from headphones, characterized in that, This method is applied to semi-open or fully open smart headphones. The headphone audio output method includes: acquiring an original audio signal; extracting a low-frequency signal from the original audio signal; acquiring frequency and intensity information of the low-frequency signal; acquiring a reconstructed signal corresponding to the low-frequency signal based on the frequency information; dynamically adjusting the gain of the low-frequency signal based on the intensity information; acquiring an enhanced signal corresponding to the low-frequency signal; fusing the reconstructed signal and the enhanced signal into a new low-frequency signal; fusing the new low-frequency signal with non-low-frequency signals from the original audio signal to obtain a fused audio signal; equalizing the gain of the new low-frequency signal and the non-low-frequency signals in the fused audio signal to obtain a full-frequency signal; acquiring the listening requirements and acoustic parameters of the smart headphones; and adjusting the full-frequency signal according to the listening requirements and acoustic parameters to obtain a target output signal.

2. The headphone audio output method according to claim 1, characterized in that, The step of adjusting the full-band signal to obtain the target output signal according to the listening requirements and the acoustic parameters includes: dividing the full-band signal into multiple independent audio segments according to frequency; obtaining a compression threshold and compression ratio set for each audio segment; when the signal amplitude of the audio segment exceeds the compression threshold, compressing the signal amplitude using the compression ratio to obtain a frequency segment control signal.

3. The headphone audio output method according to claim 2, characterized in that, The step of obtaining the compression threshold and compression ratio set for the low-frequency audio band includes: obtaining the dynamic range parameters of the smart headphones, the dynamic range parameters including frequency response curve, acoustic structure and sound unit performance, and obtaining the compression threshold and compression ratio based on the dynamic range parameters and the listening requirements.

4. The headphone audio output method according to claim 2, characterized in that, After the step of acquiring the frequency band control signal, the method includes: acquiring a global limiting value, constructing a limiter based on the global limiting value, smoothly limiting the peak value of the frequency band control signal through the limiter, and acquiring the target output signal.

5. The headphone audio output method according to claim 4, characterized in that, The step of obtaining the global limiting value includes: obtaining the limiting parameters and safety specification parameters of the smart headphones, wherein the limiting parameters include a digital domain clipping threshold, a reference power amplifier output upper limit, and a speaker unit displacement threshold; and obtaining the global limiting value based on the limiting parameters and the listening requirements.

6. The headphone audio output method according to claim 1, characterized in that, The step of equalizing the gain of the newly added low-frequency signal and the non-low-frequency signal in the fused audio signal includes: obtaining the frequency response curve and acoustic structure parameters of the smart headphones; obtaining a global equalization curve based on the frequency response curve, the acoustic structure parameters and the listening requirements; and performing gain equalization adjustment on the non-low-frequency signal and the newly added low-frequency signal based on the global equalization curve.

7. The headphone audio output method according to claim 1, characterized in that, The step of dynamically adjusting the gain of the low-frequency signal based on the intensity information includes: increasing the gain when the low-frequency signal intensity is below standard, and decreasing the gain or not increasing the gain when the low-frequency signal intensity meets standard.

8. A headphone audio output system, characterized in that, include: The acquisition module is used to acquire the original audio signal, extract the low-frequency signal from the original audio signal, and acquire the frequency information and intensity information of the low-frequency signal; The low-frequency module is used to obtain the reconstructed signal corresponding to the low-frequency signal based on the frequency information, dynamically adjust the gain of the low-frequency signal based on the intensity information, and obtain the enhanced signal corresponding to the low-frequency signal. The fusion module is used to fuse the reconstructed signal and the enhanced signal into a new low-frequency signal, and to fuse the new low-frequency signal with the non-low-frequency signals in the original audio signal to obtain a fused audio signal. An equalization module is used to equalize the gain of the newly added low-frequency signal and the non-low-frequency signal in the fused audio signal to obtain a full-frequency signal. The target module is used to acquire the listening requirements and acoustic parameters of the smart headphones, and adjust the full-frequency signal according to the listening requirements and acoustic parameters to acquire the target output signal.

9. A computer-readable storage medium, characterized in that, The device stores a computer program that, when executed by a processor, causes the processor to perform the steps of the method as described in any one of claims 1 to 7.

10. A smart earphone, characterized in that, It includes a memory and a processor, the memory storing a computer program that, when executed by the processor, causes the processor to perform the steps of the method as described in any one of claims 1 to 7.

Images