Microphone audio compensation method and speech recognition translation earphone

Abstract

This invention discloses a microphone audio compensation method and a speech recognition translation headset, belonging to the field of microphone audio compensation technology. The method includes: placing the headset to be calibrated in a pre-built standardized test environment, playing test audio, and acquiring the microphone's reception data; determining the compensation gain parameter of the secondary microphone based on the reception data; the compensation gain parameter is used to limit the frequency response curve deviation between the secondary microphone and the primary microphone within a preset deviation range; and writing the compensation gain parameter into the headset to be calibrated to obtain a calibrated headset. This invention effectively improves the reception consistency of the primary and secondary microphones by accurately selecting a calibration benchmark, building a standardized test environment, and specifically adjusting the compensation gain parameter of the secondary microphone, ensuring the performance uniformity of mass-produced speech recognition translation headsets, improving the user experience, and can be widely applied to the production calibration process of dual-microphone speech recognition translation headsets.

Description

Technical Field

[0001] This invention belongs to the technical field of smart headphone devices, specifically relating to a microphone audio compensation method and a voice recognition translation headphone. Background Technology

[0002] As an intelligent translation terminal, voice recognition translation headphones rely on the coordinated sound pickup of primary and secondary microphones to achieve accurate voice signal acquisition, which is the core foundation for subsequent voice recognition and translation. The consistency and balance of the frequency response curves of the primary and secondary microphones directly determine the quality of voice acquisition, thereby affecting the accuracy of voice recognition and translation effect of the headphones. Therefore, ensuring the matching of the frequency response curves of the primary and secondary microphones is a key link in the manufacturing of voice recognition translation headphones.

[0003] Currently, in the production and calibration of voice recognition translation headphones, the industry typically selects standard headphones as a reference for product performance testing. At the same time, it mainly ensures the consistency of microphone sound pickup through hardware structure-level control measures, such as controlling the uniformity of microphone device specifications and improving headphone assembly precision. In an attempt to keep the frequency response curve of the microphone consistent through such methods, so as to achieve the expected frequency response curve balance requirements.

[0004] However, existing technical solutions have significant technical flaws in practical applications, making it difficult to meet the consistency requirements of mass production. Relying solely on the consistency control of hardware assembly cannot eliminate the subtle individual differences that arise during microphone manufacturing and earphone assembly. Even for voice recognition and translation earphones produced in the same batch, the frequency response curves of the main and secondary microphones are still prone to significant differences. The frequency response curve balance deviates from the expected value, directly affecting the accuracy of voice acquisition, and ultimately leading to inconsistent voice recognition and translation effects for the same product.

[0005] The aforementioned issues ultimately resulted in significant differences in the actual performance of the same voice recognition translation earphone product. Some products suffered from poor voice recognition translation, leading to a poor user experience, decreased product quality stability, and increased after-sales costs. Summary of the Invention

[0006] The technical problem to be solved by the present invention is the defect of inconsistent microphone sound pickup in existing speech recognition translation headphones, which leads to unstable speech recognition translation results. Therefore, the present invention provides a microphone audio compensation method and a speech recognition translation headphone.

[0007] A microphone audio compensation method, comprising: After the earphone to be calibrated is placed in a pre-built standardized test environment, test audio is played to obtain the first sound data of the main microphone and the second sound data of the secondary microphone of the earphone to be calibrated. Based on the first and second recorded audio data, a compensation gain parameter for the secondary microphone is determined; wherein, the compensation gain parameter is used to limit the frequency response curve deviation between the secondary microphone and the primary microphone to within a preset deviation range; The compensation gain parameters are written into the earphone to be calibrated to obtain the calibrated earphone.

[0008] Furthermore, the standardized testing environment is built in the following manner: The optimal earphone that meets the preset value conditions for the evaluation indicators is selected from the standard earphones; wherein, the frequency response curve balance of the main microphone and the secondary microphone of the standard earphone is the target value; After the optimal earphone is placed in the shielded test chamber and the environment of the shielded test chamber is adjusted, it is determined whether the frequency response curve balance of the main microphone and the secondary microphone of the optimal earphone is within the first preset balance range. If so, the currently adjusted shielded test chamber environment is determined as the standardized test environment; otherwise, the shielded test chamber is readjusted until the standardized test environment is determined.

[0009] Furthermore, the evaluation indicators include one or more of the following: recognition accuracy, distance control rate, and anti-interference rate; The recognition accuracy rate is the accuracy rate of the audio data collected by the corresponding headphones after speech-to-text processing. The distance control rate is the percentage of sound within a preset pickup distance range in the sound data collected by the corresponding headphones; The anti-interference rate is the probability that the corresponding earphone is woken up by voice by a non-wearer while it is being worn.

[0010] Furthermore, adjusting the environment of the shielded test chamber includes adjusting the placement angle and position of the optimal earphone within the shielded test chamber.

[0011] Furthermore, the test audio is a sine wave test audio, the frequency of which changes continuously from low frequency to high frequency, with the low frequency value being greater than or equal to 100Hz and the high frequency value being less than or equal to 100KHz.

[0012] Furthermore, the first and second radio data include frequency response amplitude values ​​at multiple frequency points; Determining the compensation gain parameter of the secondary microphone based on the first and second recorded audio data includes: The average frequency response amplitude of the first radio data within a preset frequency range is calculated to obtain the average first frequency response amplitude. The average frequency response amplitude of the second radio data within the preset frequency range is calculated to obtain the average frequency response amplitude of the second radio data. The amplitude gain is determined based on the difference in frequency response amplitude between the average amplitude of the first frequency response and the average amplitude of the second frequency response. The amplitude gain is determined as the compensation gain parameter of the secondary microphone in the preset frequency range.

[0013] Furthermore, the step of writing the compensation gain parameter into the earphone to be calibrated includes: Calculate the fixed-point value based on the aforementioned compensation gain parameter; The fixed-point value is written into the target register of the headphone to be calibrated.

[0014] Furthermore, before writing the fixed-point value into the target register of the headphone to be calibrated: The fixed-point value is determined to be within a preset value range.

[0015] Furthermore, after writing the compensation gain parameter into the headphone to be calibrated, the method further includes: After the earphone to be calibrated is placed in the standardized test environment, the test audio is played to obtain the third sound data of the main microphone and the fourth sound data of the secondary microphone. Based on the third and fourth audio data, calculate the frequency response curve balance of the main microphone and the secondary microphone; If the frequency response curves of the main microphone and the secondary microphone are within the second preset balance range, then the earphone to be calibrated is determined to be a calibration earphone; otherwise, the earphone to be calibrated is recalibrated.

[0016] A voice recognition translation headset includes a main microphone, a secondary microphone, and a processor; the voice recognition translation headset has compensation gain parameters written in it, which are written according to the method described above; the main microphone and the secondary microphone are used to collect audio data; the processor is used to recognize text based on the audio data, generate translated speech based on the text, and play it.

[0017] Furthermore, when worn, the main microphone is closer to the wearer's mouth than the secondary microphone.

[0018] Beneficial Effects: This invention provides targeted compensation for headphones under test in a standardized testing environment. By collecting the pickup data from the main and secondary microphones and setting the compensation gain parameters for the secondary microphone, the difference in the frequency response curves of the main and secondary microphones is controlled within a preset threshold. This precisely improves the pickup consistency of the main and secondary microphones in a single headphone, thereby ensuring the basic performance of the headphone's voice recognition. Through a series of calibration steps, the overall performance consistency of mass-produced voice recognition and translation headphones is effectively improved, ensuring that headphones of the same batch and model achieve stable pickup and recognition effects, significantly improving the user experience. Simultaneously, it enables precise calibration of headphones before shipment, effectively intercepting defective products with substandard audio performance, reducing product return rates, decreasing product rework costs, and improving production efficiency. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a schematic block diagram illustrating the main method flow steps of the present invention. Detailed Implementation

[0021] To make the above-mentioned objectives, features, and advantages of this application more apparent and understandable, the specific embodiments of this application are described in detail below with reference to the accompanying drawings. Many specific details are set forth in the following description to provide a thorough understanding of this application. However, this application can be implemented in many other ways different from those described herein, and those skilled in the art can make similar modifications without departing from the spirit of this application. Therefore, this application is not limited to the specific embodiments disclosed below.

[0022] In the description of this application, it should be understood that the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this application, "multiple" means at least two, such as two, three, etc., unless otherwise explicitly specified. In the description of this application, the step numbers are only for the convenience of explaining the embodiments of this application and do not serve to limit the order in which the steps are performed.

[0023] In this application, unless otherwise expressly specified and limited, the terms "installation," "connection," "joining," and "fixing," etc., should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components, unless otherwise expressly limited. Those skilled in the art can understand the specific meaning of the above terms in this application according to the specific circumstances.

[0024] Example 1: Reference Figure 1 As shown, this embodiment provides a microphone audio compensation method, including the following steps: Step S1: After the earphone to be calibrated is placed in the pre-built standardized test environment, play the test audio and obtain the first sound data of the main microphone and the second sound data of the secondary microphone of the earphone to be calibrated.

[0025] In this embodiment, the test audio is a sine wave test audio, with its frequency continuously varying from low to high frequencies. The low-frequency value is greater than or equal to 100Hz, and the high-frequency value is less than or equal to 100kHz. The sine wave test audio is relatively stable, and its continuous frequency variation allows for the rapid plotting of a complete frequency response curve. This determines which frequency band the microphone is sensitive to or insensitive to, and compensation is then applied to the corresponding frequency points.

[0026] In a preferred embodiment, the microphone's sensitivity to or insensitivity in a specific frequency band is determined based on the plotted frequency response curve. The response amplitude at a 1kHz frequency point on the frequency response curve is set as a 0dB reference. If the response amplitude of the frequency response curve in the corresponding frequency band is consistently higher than the reference by 0 to +3dB, it is defined as a slightly sensitive frequency band; if the response amplitude is consistently higher than the reference by +3 to +10dB, it is defined as a significantly sensitive frequency band; and if the response amplitude is consistently higher than the reference by +10dB, it is defined as an extremely sensitive frequency band. Conversely, if the response amplitude of the frequency response curve in the corresponding frequency band is consistently lower than the reference by 0 to -3dB, it is defined as a slightly insensitive frequency band; if the response amplitude is consistently lower than the reference by -3 to -10dB, it is defined as a significantly insensitive frequency band; and if the response amplitude is consistently lower than the reference by -10dB, it is defined as an extremely insensitive frequency band.

[0027] Step S2: Determine the compensation gain parameter of the secondary microphone based on the first and second audio data; wherein, the compensation gain parameter is used to limit the frequency response curve deviation between the secondary microphone and the main microphone within a preset deviation range; the first and second audio data include the frequency response amplitude values ​​of multiple frequency points.

[0028] In this embodiment, the first and second radio data include frequency response amplitude values ​​at multiple frequency points; Based on the first and second recorded audio data, determine the compensation gain parameters for the secondary microphone, including: The average frequency response amplitude of the first radio data within a preset frequency range is calculated to obtain the average frequency response amplitude of the first radio data. The average frequency response amplitude of the second radio data within a preset frequency range is calculated to obtain the average frequency response amplitude of the second radio data. The amplitude gain is determined based on the difference in frequency response amplitude between the average amplitude of the first frequency response and the average amplitude of the second frequency response. The amplitude gain is determined as the compensation gain parameter for the secondary microphone within a preset frequency range.

[0029] Specifically, it includes the following steps: Acquire the first sound reception data of the main microphone and the second sound reception data of the secondary microphone of the headphone to be calibrated; the first sound reception data includes the frequency response amplitude values ​​of multiple frequency points, and the second sound reception data includes the frequency response amplitude values ​​of multiple frequency points; Assume the first radio data is D1 and the second radio data is D2.

[0030] The average frequency response amplitude of the first and second radio data in the preset frequency range is calculated separately to obtain the average first and second frequency response amplitudes. In practical applications, the average value of the data from 150Hz to 2kHz of the first radio data D1 can be obtained by summing the data and taking the average value, and the average value of the data from 150Hz to 2kHz of the second radio data D2 can be obtained by summing the data and taking the average value.

[0031] Calculate the difference in frequency response amplitude between the mean amplitude of the first frequency response and the mean amplitude of the second frequency response; In practical applications, the frequency response amplitude difference is expressed as: I1 = G1 - G2; Calculate the amplitude gain based on the mean and difference of the frequency response amplitude of the secondary microphone; In practical applications, the amplitude gain is expressed as: J1=10^(I1 / 20).

[0032] Step S3: Write the compensation gain parameters into the earphone to be calibrated to obtain the calibrated earphone.

[0033] The amplitude gain is used as the compensation gain parameter for the secondary microphone in the preset frequency range and written into the headphone to be calibrated.

[0034] In this embodiment, writing the compensation gain parameter into the earphone to be calibrated includes: Calculate the fixed-point value based on the compensation gain parameter; Write the fixed-point value into the target register of the headphone to be calibrated.

[0035] In practical applications, the target register can be a register on the chip in the earphone to be calibrated that stores the amplitude gain compensation parameters of the secondary microphone.

[0036] The fixed-point value K1 is represented as: K1=round(32768*J1). The fixed-point value is an int data type. This value is sent to the register on the chip in the earphone to be calibrated, which is used to store the amplitude gain compensation parameters of the secondary microphone. Specifically, the earphone to be calibrated can be a Bluetooth earphone, and the corresponding register is compCoeff[0].

[0037] In some embodiments of this example, compensation gain parameters for different frequency ranges are calculated and obtained, thereby simultaneously adjusting the compensation gain parameters of the secondary microphone's sound data in different frequency ranges.

[0038] In some embodiments of this example, the fixed-point value is determined to be within a preset range before the fixed-point value is written into the target register of the earphone to be calibrated.

[0039] In practical applications, before writing the fixed-point value to the target register, it is determined whether the fixed-point value exceeds the preset limit range. If it does, the writing of the fixed-point value is stopped.

[0040] Specifically, the headphone chip and the acoustic instrument program have pre-set maximum and minimum limits. If a parameter exceeds the range, the headphone chip will not burn that parameter into the chip, and the acoustic instrument program will simultaneously display an error notification for that parameter. When the calculated parameter value is within the range, the parameter value will be burned into the chip and displayed as a frequency response curve by the acoustic instrument. The secondary microphone curve will be compared with the primary microphone curve for further balance testing. If the balance exceeds the range, the acoustic instrument will ultimately determine it as NG (Not Good). If it is within the balance range, the acoustic instrument will ultimately determine it as PASS (Pass).

[0041] As a further improvement to this embodiment, after writing the compensation gain parameters into the earphone to be calibrated, the following is also included: After placing the headphones to be calibrated in a standardized test environment, test audio was played to obtain the third sound data from the main microphone and the fourth sound data from the secondary microphone. Based on the third and fourth audio data, calculate the frequency response curve balance between the main microphone and the secondary microphone; If the frequency response curves of the main microphone and the secondary microphone are within the second preset balance range, the earphone to be calibrated is determined as the calibration earphone; otherwise, the earphone to be calibrated is recalibrated.

[0042] As a specific implementation example of this embodiment, in step S1, the acoustic shielding box is closed to form a closed acoustic test environment, the CRY6151B acoustic test program is run, the test audio is played, the first sound reception data of the main microphone and the second sound reception data of the secondary microphone are obtained, and the frequency response curves of the main and secondary microphones before calibration and the balance curves of the main and secondary microphones before calibration are further collected and recorded.

[0043] In step S2, the compensation gain parameter of the secondary microphone is calculated and determined. The secondary microphone curve is then calibrated to approximate the primary microphone curve until the frequency response curves of the primary and secondary microphones are close to or coincident. Next, the calibrated frequency response curves of the primary and secondary microphones, and the calibrated balance curves of the primary and secondary microphones are collected to bring the calibrated secondary microphone curve closer to the primary microphone curve, ideally resulting in coincidence of the primary and secondary microphone frequency response curves. Finally, the difference between the primary and secondary microphone frequency response curves is measured to be minimal, meaning the parameters are close to a preset range, such as 0±1dB. In this embodiment, the difference between the primary and secondary microphone frequency response curves is defined as the difference in response amplitude for the corresponding frequency band, in dB.

[0044] The frequency response curves of the primary and secondary microphones before calibration refer to the frequency response curves of the primary and secondary microphones respectively when no gain parameter compensation has been performed on the secondary microphone. These curves reflect the original sound reception sensitivity and frequency response characteristics of the two microphones in the core audio band. Due to slight differences in hardware components and assembly, there will be significant frequency band differences between the two curves at this time. This serves as the original foundational data for subsequent algorithm compensation, used to determine the compensation direction and magnitude. The core audio band is set to cover the main frequency band of human speech; in some embodiments of this example, the core audio band is 100-4000Hz.

[0045] The pre-calibration main and secondary microphone balance curves are based on the original frequency response curves of the main and secondary microphones before calibration. The curves are calculated by measuring the difference in sound pickup across each frequency band, providing a clear and quantitative reflection of the original differences in sound pickup between the two microphones. A larger difference in the curves indicates a worse balance between the two microphones, resulting in less stable collaborative sound pickup and noise reduction. This is a crucial basis for determining whether compensation is needed for the two microphones and for identifying the core frequency band for compensation. Based on the post-calibration main and secondary microphone balance curves, the post-calibration balance can be further assessed to determine if further calibration compensation is required.

[0046] Example 2: This embodiment provides a method for setting up a standardized testing environment, which can be used in the microphone audio compensation method described in Embodiment 1, including: Step S4: After the optimal earphone is placed in the shielded test chamber and the environment of the shielded test chamber is adjusted, determine whether the frequency response curve balance of the main microphone and the secondary microphone of the optimal earphone is within the first preset balance range; if so, determine the current adjusted shielded test chamber environment as the standardized test environment; otherwise, readjust the shielded test chamber until the standardized test environment is determined.

[0047] Adjusting the environment of the shielded test chamber also includes: adjusting the optimal placement angle and position of the headphones within the shielded test chamber.

[0048] This invention first selects standard headphones based on the balance of the frequency response curves of the main and secondary microphones as the target value. Then, based on evaluation indicators, it selects the optimal headphone as the calibration benchmark, providing a unified and reliable reference for subsequent audio compensation work and ensuring the accuracy of the calibration work from the source. By adjusting the shielded test chamber using the optimal headphone as the benchmark, a standardized test environment is built where the balance of the frequency response curves of the main and secondary microphones is within a preset range. This achieves standardization and uniformity of the calibration environment, avoiding batch calibration deviations caused by differences in test environments, ensuring consistent compensation conditions for each headphone to be calibrated, and laying an environmental foundation for the uniform performance of mass-produced headphones.

[0049] Specifically, in this embodiment, a CRY6151B electroacoustic tester is used to acquire the sound reception data of the main microphone and the secondary microphone. A matching fixture is installed inside the shielded test chamber to stably place the corresponding model of headphones. When the headphones are placed on the test fixture, the headphone stem is parallel to the speaker, and the two microphones on the stem are close to the center of the speaker.

[0050] Example 3: This embodiment provides a method for selecting the optimal headphones, which can be used in the standardized testing environment construction method described in Embodiment 2, including: Step S5: Select the optimal earphone from the standard earphones whose evaluation indicators meet the preset value conditions; wherein, the frequency response curve balance of the main microphone and the secondary microphone of the standard earphone is the target value; In this embodiment, the target value for the balance of the frequency response curves of the main and secondary microphones is 0dB.

[0051] In this embodiment, the evaluation metrics include one or more of recognition accuracy, distance control rate, and anti-interference rate; ten standard headphones are selected, and the optimal headphone is chosen from them. In this embodiment, the headphone with the best evaluation metrics is defined as the optimal headphone.

[0052] The recognition accuracy rate is the accuracy rate of the audio data collected by the corresponding headphones after speech-to-text processing; The distance control rate is the percentage of sound within a preset pickup distance range in the sound data collected by the corresponding headphones; as a preferred embodiment, the pickup distance range is 20cm.

[0053] The anti-interference rate corresponds to the probability that the earphones will be woken up by voice from a non-wearer while the earphones are being worn. By optimizing the anti-interference rate, it is ensured that even if the earphone wearer does not speak, the earphones will not be woken up to pick up sound if someone nearby speaks around the earphone wearer.

[0054] Specifically, the evaluation indicators are obtained through subjective and objective tests. The subjective test involves multiple people of different genders in different environments, and the recognition accuracy, distance control rate, and anti-interference rate of each person are statistically obtained. The objective test uses a laboratory environment to simulate human heads and torsos.

[0055] In embodiments of the present invention, the average values ​​of three indicators—recognition accuracy, distance control rate, and anti-interference rate—are calculated for each standard earphone. The earphone with the highest average value is selected as the optimal earphone. By using recognition accuracy, distance control rate, and anti-interference rate as evaluation indicators and selecting the highest average value as the preset condition, the optimal earphone with the best overall performance in speech recognition and translation can be selected. A standardized testing environment is then built based on the optimal earphone, ensuring that the earphone calibrated in the standardized testing environment has the sound reception performance suitable for speech recognition and translation work scenarios.

[0056] Example 4: This embodiment also provides a voice recognition translation headset, including a main microphone, a secondary microphone, and a processor. The voice recognition translation headset has compensation gain parameters written in it, which are written according to the method of Embodiment 1. The main microphone and the secondary microphone are used to collect audio data. The processor is used to recognize text based on the audio data, generate translated speech based on the text, and play it.

[0057] When worn, the main microphone is closer to the wearer's mouth than the secondary microphone. The main microphone is closer to the wearer's mouth, while the secondary microphone is farther from the wearer's ear.

[0058] 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.

[0059] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.

Claims

1. A microphone audio compensation method, characterized in that, include: After the earphone to be calibrated is placed in a pre-built standardized test environment, test audio is played to obtain the first sound data of the main microphone and the second sound data of the secondary microphone of the earphone to be calibrated. Based on the first and second recorded audio data, a compensation gain parameter for the secondary microphone is determined; wherein, the compensation gain parameter is used to limit the frequency response curve deviation between the secondary microphone and the primary microphone to within a preset deviation range; The compensation gain parameters are written into the earphone to be calibrated to obtain the calibrated earphone.

2. The method according to claim 1, characterized in that, The standardized testing environment was built using the following method: The optimal earphone that meets the preset value conditions for the evaluation indicators is selected from the standard earphones; wherein, the frequency response curve balance of the main microphone and the secondary microphone of the standard earphone is the target value; After the optimal earphone is placed in the shielded test chamber and the environment of the shielded test chamber is adjusted, it is determined whether the frequency response curve balance of the main microphone and the secondary microphone of the optimal earphone is within the first preset balance range. If so, the currently adjusted shielded test chamber environment is determined as the standardized test environment; otherwise, the shielded test chamber is readjusted until the standardized test environment is determined.

3. The microphone audio compensation method according to claim 2, characterized in that, The evaluation indicators include one or more of the following: recognition accuracy, distance control rate, and anti-interference rate; The recognition accuracy rate is the accuracy rate of the audio data collected by the corresponding headphones after speech-to-text processing. The distance control rate is the percentage of sound within a preset pickup distance range in the sound data collected by the corresponding headphones; The anti-interference rate is the probability that the corresponding earphone is woken up by voice by a non-wearer while it is being worn.

4. The microphone audio compensation method according to claim 2, characterized in that, The adjustment of the shielding test chamber environment includes: adjusting the placement angle and position of the optimal earphone in the shielding test chamber.

5. A microphone audio compensation method according to claim 1, characterized in that, The test audio is a sine wave test audio, and the frequency of the sine wave test audio changes continuously from low frequency to high frequency, with the low frequency value being greater than or equal to 100Hz and the high frequency value being less than or equal to 100KHz.

6. The microphone audio compensation method according to claim 1, characterized in that, The first and second radio data include frequency response amplitude values ​​at multiple frequency points; Determining the compensation gain parameter of the secondary microphone based on the first and second recorded audio data includes: The average frequency response amplitude of the first radio data within a preset frequency range is calculated to obtain the average first frequency response amplitude. The average frequency response amplitude of the second radio data within the preset frequency range is calculated to obtain the average frequency response amplitude of the second radio data. The amplitude gain is determined based on the difference in frequency response amplitude between the first average frequency response amplitude and the second average frequency response amplitude. The amplitude gain is determined as the compensation gain parameter of the secondary microphone in the preset frequency range.

7. A microphone audio compensation method according to claim 1, characterized in that, The step of writing the compensation gain parameter into the earphone to be calibrated includes: Calculate the fixed-point value based on the aforementioned compensation gain parameter; The fixed-point value is written into the target register of the headphone to be calibrated.

8. A microphone audio compensation method according to claim 7, characterized in that, Before writing the punctured value into the target register of the headphone to be calibrated: The fixed-point value is determined to be within a preset value range.

9. A microphone audio compensation method according to any one of claims 1 to 8, characterized in that, After writing the compensation gain parameter into the headphone to be calibrated, the method further includes: After the earphone to be calibrated is placed in the standardized test environment, the test audio is played to obtain the third sound data of the main microphone and the fourth sound data of the secondary microphone. Based on the third and fourth audio data, calculate the frequency response curve balance of the main microphone and the secondary microphone; If the frequency response curves of the main microphone and the secondary microphone are within the second preset balance range, then the earphone to be calibrated is determined to be a calibration earphone; otherwise, the earphone to be calibrated is recalibrated.

10. A voice recognition translation headset, characterized in that, It includes a main microphone, a secondary microphone, and a processor; the voice recognition translation headset has compensation gain parameters written in it, and the compensation gain parameters are written according to the method according to any one of claims 1 to 9; the main microphone and the secondary microphone are used to collect audio data; The processor is used to recognize text based on the radio data, generate translated speech based on the text, and play it.

Images