A method for implementing a headphone transparent mode

Abstract

The application discloses a method for realizing a headphone transparent mode. The prior art method needs to use a device other than the headphone to complete required path measurement, and requires complicated hardware resources, environment and process. The method uses an external microphone and an error microphone of the headphone to complete path measurement of a primary channel, uses a loudspeaker of the headphone and the error microphone to complete path measurement of a secondary channel, establishes a first-stage filter of the transparent mode based on an N-order FIR filter, establishes a cost function, and uses a genetic algorithm to solve coefficients of the FIR filter; a high shelf filter is used to establish a second-stage filter of the transparent mode, and to complete suppression of potential abnormal high-frequency components. The two-stage filters jointly establish a complete transparent mode filter. The application uses conditions of the headphone itself to complete necessary path measurement, concentrates the FIR filter on fitting of effective frequency bands in environmental human voices, and reduces difficulty and complexity of required resources of the transparent mode design.

Description

Technical Field

[0001] This invention belongs to the field of headphone technology, and in particular to the field of active noise cancellation technology for headphones, specifically relating to a method for achieving a transparency mode in headphones. Background Technology

[0002] When headphones are worn, the ear canal is blocked, causing a sharp attenuation of mid-to-high frequency sound signals from the environment, thus affecting the ear's ability to perceive ambient sounds. Transparency mode is a common headphone technology used in both consumer headphones and hearing aid headphones. Consumer headphones prioritize hearing ambient sounds, while hearing aid headphones prioritize hearing the wearer's own voice, allowing the wearer to naturally adjust their voice for comfortable conversation. The technical purpose of transparency mode is to minimize the impact on the ear's perception of ambient sounds without removing the headphones.

[0003] Existing transparency mode design methods require measuring four sets of transfer functions. The first set is the transfer function TF between the external speaker and the microphone in the artificial ear when no headphones are worn. open (f). Second group, transfer function TF between the external speaker and the microphone in the artificial ear when wearing headphones. close (f). Group 3, with headphones worn, the transfer function TF between the external speaker and the external microphone of the headphones. wear (f). Group 4, with headphones worn, the transfer function TF between the headphone speaker and the microphone in the artificial ear. s (f). The designed transparent mode filter T(f) is generally required to satisfy:

[0004] TF open (f) = TF close (f)+TF wear (f)·T(f)·TF s (f).

[0005] This type of transparency mode design method requires complex external recording hardware resources and conditions. Furthermore, during the headphone production testing phase, when the designed transparency mode filter fails to meet requirements, the complex design process prevents timely and effective path measurement and filter adjustment. In practical use, personalized design is also impossible. Summary of the Invention

[0006] The purpose of this invention is to address the shortcomings of existing technologies by providing a method for achieving headphone transparency mode.

[0007] The method of this invention is specifically as follows:

[0008] Step (1) The external speaker plays pink noise, and the external microphone of the headphones plays pink noise.mic And error microphone FB mic Simultaneously, signals are acquired at the same sampling rate fs; the external microphone signal sig... FF Sum of error microphone signal sig FB Perform m-point Fourier transform (FFT) processing on each point. This indicates rounding up to obtain the frequency domain representation of the external microphone signal FF(f) and the frequency domain representation of the external source signal of the error microphone FB(f), and the frequency sequence. This yields the transfer function of the main path. Furthermore, the pink noise frequency is 48kHz.

[0009] Step (2) The headphone speaker plays pink noise sig at the same frequency as the external speaker. DS The error microphone acquires the playback signal from the headphone speaker at a sampling rate fs, and is denoted as sig. FB′ ; for sig DS Perform m-point Fourier transform (FFT) processing to obtain the frequency domain representation DS(f) of the headphone speaker signal; for sig FB′ Perform m-point Fourier transform (FFT) processing to obtain the frequency domain representation of the source signal of the error microphone, FB′(f); from this, the transfer function of the secondary path is obtained.

[0010] Step (3) Initialize the first-level transparent mode filter: Use an Nth-order finite impulse response (FIR) filter to perform an m-point Fourier transform (FFT) on the FIR filter coefficient sequence X to obtain the frequency domain representation V(f) of the FIR filter: V(f) = convertTF(X), where convertTF(·) represents the Fourier transform process. The FIR filter coefficient sequence X = [x(1), x(2), ..., x(N)], and the coefficients x(n) ∈ [-5, 5], n = 1, ..., N; further, N ∈ [64, 256].

[0011] Step (4) Construct the cost function In the specified frequency range F V The summation is performed on the above, and the solution is obtained using a genetic algorithm. When the number of iterations of the genetic algorithm reaches the preset maximum number of iterations MG, the coefficient sequence X′ of the first-level transparent mode filter is obtained.

[0012] Furthermore, specify the frequency range F V ∈[100,4000]Hz, the population size of the genetic algorithm PS=1000·N, the preset maximum number of iterations MG∈[30,50], and the crossover probability CF∈[0.3,0.8].

[0013] Step (5) suppresses the high-frequency part of the output of the first-stage transparent mode filter by using the second-stage transparent mode filter. The second-stage transparent mode filter is a second-order overhead filter.

[0014] Furthermore, the center frequency f of the elevated filter hs ∈[4000,6000]Hz, gain G hs ∈[-10,-30]dB, quality factor Q hs ∈[0.5,2].

[0015] Step (6) The first and second stage transparency mode filters constitute a complete transparency mode filter, and the headphone transparency mode is realized through the complete transparency mode filter.

[0016] In modern communication systems, the bandwidth for voice signal transmission is 3kHz, with a frequency range of 300–3300Hz. It has been proven that the bandwidth of current communication systems provides sufficient sound signal quality. Therefore, the upper limit of the flat frequency band in the transparency mode design reaches 4000Hz, which is sufficient for the headset wearer to perceive ambient human voices and distinguish the emotions in the speaker's voice. This invention draws on the practice of preserving the voice signal within the effective bandwidth in communication systems, utilizing two stages of filters to perform the function of a transparency mode filter. The first-stage FIR filter is used to ensure the flatness of the human voice within the effective bandwidth as much as possible, which may cause some excessive boosting of high-frequency signals. The introduction of the second-stage overhead filter will suppress the excessive boosting of high frequencies, avoiding discomfort from high-frequency signals and the occurrence of feedback.

[0017] This invention eliminates the need for additional recording equipment. It utilizes only an external microphone, an error microphone, and the headphone speaker to measure the path information required for transparency mode design and to design the transparency mode filter. This method reduces design complexity and, by leveraging the suppression capabilities of overhead filters, concentrates the compensation performance of the FIR filter within the effective bandwidth, improving the flatness of the effective bandwidth for ambient human voices. Furthermore, based on the fact that this invention does not rely on external equipment for path measurement, it has the potential for real-time measurement and updating of transparency mode performance. Attached Figure Description

[0018] Figure 1 This is a schematic diagram of the overall process of the method of the present invention;

[0019] Figure 2 This is a schematic diagram of the transparent mode deployment of the present invention. Detailed Implementation

[0020] The present invention will be further described below with reference to specific embodiments and accompanying drawings. It should be understood that the embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0021] like Figure 1 and 2 As shown, a method for implementing headphone transparency mode is as follows:

[0022] Step (1) The external speaker plays 48kHz pink noise, and the external microphone 1 and error microphone 2 of the headphones simultaneously acquire signals at the same sampling rate fs = 48kHz; the external microphone signal sig FF Sum of error microphone signal sig FB Perform 8192-point Fourier Transform (FFT) processing to obtain the frequency domain representation FF(f) of the external microphone signal and the frequency domain representation FB(f) of the external source signal from the error microphone, and the frequency sequence. This yields the transfer function of the main path.

[0023] Step (2) The headphone speaker 3 plays pink noise sig at the same frequency as the external speaker. DS The error microphone 2 acquires the playback signal from the headphone speaker at a sampling rate fs, and is denoted as sig. FB′ ; for sig DS Perform an 8192-point Fourier transform (FFT) to obtain the frequency domain representation DS(f) of the headphone speaker signal; then perform a sig... FB′ An 8192-point Fourier transform (FFT) is performed to obtain the frequency domain representation of the source signal from the error microphone, FB′(f). The transfer function of the secondary path is then derived from this.

[0024] Step (3) Initialize the first-level transparent mode filter 4: Use a 128th-order finite impulse response (FIR) filter to perform an 8192-point Fourier transform (FFT) on the FIR filter coefficient sequence X to obtain the frequency domain representation V(f) of the FIR filter: V(f) = convertTF(X), where convertTF(·) represents the Fourier transform process. The FIR filter coefficient sequence X = [x(1), x(2), ..., x(128)], and the coefficients x(n) ∈ [-5, 5], n = 1, ..., 128.

[0025] Step (4) Construct the cost function S(X) = ∑ f = FV {1-[convertTF(X)·P(f)]-G(f)}, within the specified frequency range F VThe summation is performed over the frequency range [100, 4000] Hz, and the solution is obtained using a genetic algorithm. When the number of iterations of the genetic algorithm reaches the preset maximum number of iterations (30), the coefficient sequence X′ of the first-level transparent mode filter is obtained. The population size of the genetic algorithm is PS = 128000, and the crossover probability CF = 0.5.

[0026] Step (5) suppresses the high-frequency component of the output of the first-stage transparent-mode filter 4 using the second-stage transparent-mode filter 5. The second-stage transparent-mode filter 5 is a second-order overhead filter. The center frequency f of the overhead filter is... hs =5000Hz, gain G hs = -20dB, quality factor Q hs =0.707.

[0027] Step (6) The first and second stage transparency mode filters constitute a complete transparency mode filter, and the headphone transparency mode is realized through the complete transparency mode filter.

[0028] It should be understood that the above-described embodiments are merely specific implementations of this application, used to illustrate the technical solutions of this application, and are not intended to limit the invention. The scope of protection of this application is not limited thereto.

Claims

1. A method for implementing headphone transparency mode, characterized in that: Step (1) Determine the transfer function G(f) of the main path; Step (2) Determine the transfer function P(f) of the secondary path; Step (3) Initialize the first-stage transparent mode filter to obtain the frequency domain representation V(f) of the FIR filter, where X represents the FIR filter coefficient sequence; use an Nth-order finite impulse response FIR filter to perform an m-point Fourier transform (FFT) on the FIR filter coefficient sequence X to obtain the frequency domain representation V(f) of the FIR filter: V(f) = convertTF(X), where convertTF(·) represents the Fourier transform process; Step (4) Construct the cost function In the specified frequency range F V Summation is performed over [100, 4000] Hz, and the solution is obtained using a genetic algorithm. When the number of iterations of the genetic algorithm reaches the preset maximum number of iterations MG, the coefficient sequence X′ of the first-level transparent mode filter is obtained. Step (5) suppresses the high-frequency part of the output of the first-stage transparent mode filter by using the second-stage transparent mode filter. The second-stage transparent mode filter is a second-order overhead filter. Step (6) The first and second stage transparency mode filters constitute a complete transparency mode filter, and the headphone transparency mode is realized through the complete transparency mode filter.

2. The method for achieving headphone transparency mode as described in claim 1, characterized in that, Step (1) specifically involves: The external speaker plays pink noise, and the external microphone of the headphones plays pink noise. mic And error microphone FB mic Simultaneously, signals are acquired at the same sampling rate fs; the external microphone signal sig... FF Sum of error microphone signal sig FB Perform m-point Fourier Transform (FFT) processing to obtain the frequency domain representation FF(f) of the external microphone signal and the frequency domain representation FB(f) of the external source signal from the error microphone, and the frequency sequence. Obtain the transfer function of the main path 3. The method for achieving headphone transparency mode as described in claim 1, characterized in that, Step (2) specifically involves: Headphone speakers playing pink noise sig DS The error microphone acquires the playback signal from the headphone speaker at a sampling rate fs, and is denoted as sig. FB ′;to sig DS Perform m-point Fourier transform (FFT) processing to obtain the frequency domain representation DS(f) of the headphone speaker signal; for sig FB Performing an m-point Fourier transform (FFT) on the signal yields the frequency domain representation FB′(f) of the source signal from the error microphone, and the frequency sequence. Obtain the transfer function of the secondary path 4. A method for implementing headphone transparency mode as described in claim 2 or 3, characterized in that: The pink noise frequency is 48kHz.

5. A method for implementing headphone transparency mode as described in claim 2 or 3, characterized in that: This indicates rounding up to the nearest integer.

6. The method for implementing headphone transparency mode as described in claim 1, characterized in that: N∈[64,256]。 7. The method for implementing headphone transparency mode as described in claim 1, characterized in that: In step (4), the population size of the genetic algorithm is PS = 1000·N, the maximum number of iterations is preset to MG∈[30,50], and the crossover probability is CF∈[0.3,0.8].

8. The method for implementing headphone transparency mode as described in claim 1, characterized in that: In step (5), the center frequency f of the overhead filter hs ∈[4000,6000]Hz, gain G hs ∈[-10,-30]dB, quality factor Q hs ∈[0.5,2].

Images