A spatial cue preservation method based on ANC earphones

Abstract

The application discloses a space clue reservation method based on an ANC earphone, and comprises the following steps: constructing a space clue reservation system based on ANC; establishing an ANC filter coefficient estimation error updating formula and a step updating formula based on a CLMS algorithm; and simulating and verifying the space clue reservation effect and the noise reduction effect of the system. The space clue reservation system constructed by the CLMS algorithm improves the space clue reservation effect of binaural ANC without reducing the noise reduction amount, and is beneficial to the identification of the spatial direction of residual sound after wearing the ANC earphone.

Description

Technical Field

[0001] This invention belongs to the field of active noise cancellation and spatial cues preservation technology, and in particular, a spatial cues preservation method based on ANC headphones. Background Technology

[0002] Noise surrounds us in our daily lives and work environments, directly impacting our quality of life and work. In the military field, the increasing mechanization of various military equipment has led to growing noise problems in helicopters, ships, tanks, and other equipment, severely affecting the overall combat capability of the armed forces. Simultaneously, it poses a serious threat to the hearing of personnel, even causing permanent hearing damage. Therefore, noise control in the military field is extremely urgent. In the civilian sector, environmental noise mainly falls into three categories: residential noise, industrial noise, and traffic noise. Residential noise is commonly seen in public places with loud conversations and various street broadcasts. Industrial noise includes the noise from large machinery in work or living environments and factory equipment. Traffic noise primarily comes from vehicles such as cars, trains, and airplanes. As people increasingly value their quality of life, they also pay more attention to noise control, making the research on noise-canceling headphones increasingly important.

[0003] ANC (Audio Cancellation) utilizes the principle of sound interference cancellation. It employs a set of acquisition and processing circuits to output a sound wave signal that is opposite in phase to the noise to be eliminated but has the same frequency and energy. This signal is then combined with the original noise signal to achieve noise cancellation. ANC headphones are a successful application of ANC.

[0004] Spatial cues of sound are crucial for the auditory experience, as they are an integral part of human hearing and are essential for the ear's ability to localize sound. Spatial cues allow for better perception of the surrounding environment; for example, hearing nearby car horns allows for quick maneuvering to avoid vehicles and prevent accidents. On the battlefield, soldiers can quickly determine the direction of enemy aircraft by sound. Spatial cues also improve speech intelligibility; for instance, the "cocktail party effect" allows people to focus intently on a conversation with a specific person by locating their voice, thus minimizing the impact of other speakers and ambient noise. Therefore, spatial information from both ears is extremely important. Noise reduction methods that output only one channel are called mono-ear noise reduction, which focuses solely on noise suppression. Bilateral noise reduction involves noise reduction on both sides, but the data is not shared; both sides work independently. Neither mono-ear nor bilateral noise reduction considers preserving the spatial cues of the sound source. Binaural noise reduction, however, considers preserving the spatial information of the sound source while reducing noise, with both sides sharing data and working together. Currently, ANC headphones on the market do not consider spatial distortion of sound information.

[0005] Therefore, a major problem with existing technology is that in existing ANC headphones, the two earpieces work independently and do not consider the preservation of spatial information of the sound source. However, in actual use, we have found that after wearing the headphones, the spatial information of the sound source is distorted, making it impossible to accurately determine the spatial location of the residual sound. Summary of the Invention

[0006] The purpose of this invention is to address the problems existing in the prior art by providing a spatial cue preservation method based on ANC headphones, which further improves the spatial cue preservation performance of sound while ensuring the noise reduction level.

[0007] The technical solution to achieve the purpose of this invention is: a spatial cue retention method based on ANC headphones, the method comprising the following steps:

[0008] Step 1: Construct a spatial cue retention system based on ANC headphones;

[0009] Step 2: Establish the ANC filter step size update formula based on the CLMS algorithm;

[0010] Step 3: Simulate and verify the spatial cues preservation and noise reduction effects of the system constructed in Step 1.

[0011] Furthermore, the construction of the spatial cue retention system based on ANC headphones described in step 1 specifically includes:

[0012] Step 1-1: Construct a single-channel composite adaptive ANC system;

[0013] Steps 1-2: Establish the ANC filter step size update formula based on the FxLMS algorithm;

[0014] Steps 1-3 involve conducting an impact analysis on spatial cues using a single-channel composite adaptive ANC system.

[0015] Steps 1-4: Construct a dual-channel composite adaptive ANC system based on the CLMS algorithm.

[0016] Furthermore, the construction of the single-channel composite adaptive ANC system described in step 1-1 specifically includes:

[0017] Step 1-1-1: Perform offline identification of the secondary channels;

[0018] Step 1-1-2: The feedforward filter filters the input reference signal;

[0019] Step 1-1-3: Perform feedback reference signal estimation;

[0020] Step 1-1-4: The feedback filter filters the feedback reference signal;

[0021] Step 1-1-5: The output signal of the feedforward filter and the output signal of the feedback filter are superimposed and then passed through the secondary channel;

[0022] Step 1-1-6: The secondary channel output signal is superimposed with the desired signal to obtain the residual signal.

[0023] Furthermore, the offline identification of the secondary channel described in step 1-1-1 specifically includes: the input signal is a random noise signal, the secondary channel is trained and estimated before the single-channel composite adaptive ANC system is run, and finally the fixed training result is saved as the secondary channel model, and the model is introduced into the single-channel composite adaptive ANC system.

[0024] Furthermore, the feedforward filter described in step 1-1-2 filters the input reference signal, specifically including: the input reference signal is a historical sequence of the signal at the current time, and after feedforward filtering, the feedforward signal to be passed through the secondary channel is calculated.

[0025] Furthermore, the feedback reference signal estimation mentioned in step 1-1-3 specifically refers to the sum of the error signal and the speaker signal after passing through the secondary channel;

[0026] The feedback filter described in step 1-1-4 filters the feedback reference signal, specifically by: based on the feedback reference signal calculated in step 1-1-3, performing feedback filtering, and calculating the signal to be superimposed on the output of the feedforward filter.

[0027] Furthermore, the ANC filter step size update formula established in steps 1-2 based on the FxLMS algorithm is as follows:

[0028] w(n+1)=w(n)+μe(n)x(n)

[0029] In the formula, μ is the step size, n is the time, e(n) is the signal residual at the current time, x(n) is the input of the filter at the current time, w(n) is the weight vector of the filter at the current time, w(n+1) is the weight vector of the filter at the next time, and w(n) = [w0(n)w1(n)...w N-1 (n)] T N is the order of the filter, and T represents the matrix transpose.

[0030] Furthermore, the influence analysis of spatial cues using a single-channel composite adaptive ANC system described in steps 1-3 specifically includes: comparing and analyzing the IPD information of the binaural signals before and after ANC, wherein the IPD is obtained by calculating the cross-power spectral density of the binaural signals and then taking the phase information, and its formula is expressed as:

[0031]

[0032] In the formula, Γ l,r Γ represents the cross-power spectral density of the left and right ears. l,l Indicates the self-power spectral density;

[0033] For input signals at different angles, IPD information is calculated in the above manner, and then the IPD error before and after ANC is calculated to obtain the relationship between IPD error and input angle and signal frequency.

[0034] Furthermore, the construction of the dual-channel composite adaptive ANC system based on the CLMS algorithm described in steps 1-4 requires combining the left and right ear signals into a complex signal. The filter coefficients are updated using the complex LMS algorithm. After passing through feedforward and feedback filters, the real and imaginary parts of the complex signal are extracted and passed through the left and right ear secondary paths, respectively. Specifically, this includes:

[0035] Step 1-4-1: Perform offline identification of the secondary channels;

[0036] Step 1-4-2: Combine the left and right ear signals into a complex signal as a complex reference signal;

[0037] Step 1-4-3: The feedforward filter filters the input complex reference signal;

[0038] Step 1-4-4: Perform feedback complex reference signal estimation;

[0039] Steps 1-4-5: The feedback filter filters the feedback complex reference signal;

[0040] Steps 1-4-6: The output signals of the feedforward filter and the feedback filter are superimposed, and then the left and right ear signals are separated according to the real and imaginary parts, and passed through the left and right secondary channels respectively.

[0041] Steps 1-4-7: The output signals of the left and right secondary channels are superimposed on the desired signals of the left and right ears respectively to obtain the residual signals of the left and right ears.

[0042] Furthermore, the ANC filter step size update formula established in step 2 based on the CLMS algorithm is as follows:

[0043]

[0044] In the formula, μ is the step size, n is the time, w(n) is the weight vector of the filter at the current time, e(n) is the signal residual at the current time, and x(n) is the input of the filter at the current time. Let x(n) be the conjugate of x(n), and w(n), e(n), and x(n) be all complex numbers, respectively denoted as:

[0045] e(n) = e R (n)+ie I(n),x(n)=x R (n)+ix I (n), w(n)=w R (n)+iw I (n), R and I represent the real and imaginary parts respectively, corresponding to the left and right ear signals, i represents the imaginary unit, w(n) = [w0(n)w1(n)...w N-1 (n)] T N represents the filter order, and T represents the matrix transpose.

[0046] Compared with the prior art, the significant advantages of this invention are:

[0047] 1) Compared with the traditional ANC algorithm, the present invention preserves the spatial cues of the residual sound signal after ANC better in all directions.

[0048] 2) Compared with the traditional ANC algorithm, this invention retains spatial cues while also better preserving the noise reduction effect of the traditional algorithm.

[0049] 3) This invention provides a better approach to the optimization of traditional ANC algorithms and expands the performance requirements of ANC headphones.

[0050] The present invention will now be described in further detail with reference to the accompanying drawings. Attached Figure Description

[0051] Figure 1 This is a block diagram of the composite adaptive ANC system based on the CLMS algorithm of this invention.

[0052] Figure 2 This is a block diagram of a composite adaptive ANC system based on the FxLMS algorithm. Detailed Implementation

[0053] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0054] It should be noted that if the embodiments of the present invention involve directional indicators (such as up, down, left, right, front, back, etc.), the directional indicators are only used to explain the relative positional relationship and movement of the components in a certain specific posture (as shown in the figure). If the specific posture changes, the directional indicators will also change accordingly.

[0055] Furthermore, if the embodiments of this invention involve descriptions such as "first" or "second," these descriptions are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined with "first" or "second" may explicitly or implicitly include at least one of those features. Additionally, the technical solutions of the various embodiments can be combined with each other, but this must be based on the ability of those skilled in the art to implement them. If the combination of technical solutions is contradictory or impossible to implement, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection claimed by this invention.

[0056] In one embodiment, combined Figure 1 A spatial cue preservation method based on ANC headphones is provided, the method comprising the following steps:

[0057] Step 1, construct a spatial cue retention system based on ANC headphones; specifically including:

[0058] Step 1-1: Construct a single-channel composite adaptive ANC system; combined with Figure 2 Specifically, it includes:

[0059] Step 1-1-1: Offline identification of the secondary channel is performed. The input signal is a random noise signal. The secondary channel is trained and estimated before the ANC system is run. Finally, the fixed training results are saved as the secondary channel model and the model is introduced into the ANC system.

[0060] Step 1-1-2: The feedforward filter filters the input reference signal. The input signal is the historical sequence of the current input signal. After the feedforward filter, the signal to be passed through the secondary channel is calculated.

[0061] Step 1-1-3: Feedback reference signal estimation. The reference signal is estimated as the sum of the error signal and the speaker signal passed through the secondary channel.

[0062] Step 1-1-4: The feedback filter filters the feedback reference signal. Based on the calculated feedback reference signal, the signal to be superimposed on the feedforward filter output is calculated after the feedback filter.

[0063] Step 1-1-5: The output signals of the feedforward filter and the feedback filter are superimposed and then pass through the secondary channel together, which corresponds to the transmission path from the speaker to the error microphone in the actual application system.

[0064] Step 1-1-6: The secondary channel output signal is superimposed with the desired signal, that is, the reverse sound wave superposition is used to achieve ANC. The residual signal after superposition is the sound signal that the human ear can hear.

[0065] Step 1-2: Establish the ANC filter step size update formula based on the FxLMS algorithm. The specific formula is as follows:

[0066] w(n+1)=w(n)+μe(n)x(n)

[0067] In the formula, μ is the step size, n is the time, e(n) is the signal residual at the current time, x(n) is the input of the filter at the current time, w(n) is the weight vector of the filter at the current time, w(n+1) is the weight vector of the filter at the next time, and w(n) = [w0(n)w1(n)...w N-1 (n)] T N is the order of the filter, and T represents the matrix transpose.

[0068] Steps 1-3 involve analyzing the impact of spatial cues using a single-channel composite adaptive ANC system. Specifically, this includes comparing the IPD (Integrated Power Spectrum Density) of the binaural signals before and after ANC. The IPD is obtained by calculating the cross-power spectral density of the binaural signals and then extracting the phase information; the formula is as follows:

[0069]

[0070] In the formula, Γ l,r Γ represents the cross-power spectral density of the left and right ears. l,l Indicates the self-power spectral density;

[0071] For input signals at different angles, IPD information is calculated in the above manner, and then the IPD error before and after ANC is calculated to obtain the relationship between IPD error and input angle and signal frequency.

[0072] Steps 1-4 involve constructing a dual-channel composite adaptive ANC system based on the CLMS algorithm. This requires combining the left and right ear signals into a single complex signal. The filter coefficients are updated using the complex LMS algorithm. After passing through feedforward and feedback filters, the real and imaginary parts of the complex signal are extracted and processed through the left and right ear secondary paths, respectively. Specifically, this includes:

[0073] Step 1-4-1: Perform offline identification of the secondary channels;

[0074] Step 1-4-2: Combine the left and right ear signals into a complex signal as a complex reference signal;

[0075] Step 1-4-3: The feedforward filter filters the input complex reference signal;

[0076] Step 1-4-4: Perform feedback complex reference signal estimation;

[0077] Steps 1-4-5: The feedback filter filters the feedback complex reference signal;

[0078] Steps 1-4-6: The output signals of the feedforward filter and the feedback filter are superimposed, and then the left and right ear signals are separated according to the real and imaginary parts, and passed through the left and right secondary channels respectively.

[0079] Steps 1-4-7: The output signals of the left and right secondary channels are superimposed on the desired signals of the left and right ears respectively to obtain the residual signals of the left and right ears.

[0080] Step 2, establish the ANC filter step size update formula based on the CLMS algorithm as follows:

[0081]

[0082] In the formula, μ is the step size, n is the time, w(n) is the weight vector of the filter at the current time, e(n) is the signal residual at the current time, and x(n) is the input of the filter at the current time. Let x(n) be the conjugate of x(n), and w(n), e(n), and x(n) be all complex numbers, respectively representing e(n) = e R (n)+ie I (n),x(n)=x R (n)+ix I (n), w(n)=w R (n)+iw I (n), R and I represent the real and imaginary parts respectively, corresponding to the left and right ear signals, i represents the imaginary unit, w(n) = [w0(n)w1(n)...w N-1 (n)] T N represents the filter order, and T represents the matrix transpose.

[0083] Step 3: Simulation verification of the spatial cue preservation and noise reduction effects of the system constructed in Step 1. Specifically: The spatial cue preservation effect utilizes the IPD information of binaural sound to calculate the IPD of the ANC-received signal obtained by this method, and compares and analyzes it with the IPD of the ANC-received signal collected by the original headphones. Since ANC is only effective for the low-frequency part of the sound, the noise reduction effect is compared and analyzed by calculating the average noise reduction amount within 1KHz.

[0084] For signals with different input angles, the IPD error and the noise reduction error are calculated. It can be seen that the IPD error and the difference in noise reduction change with the angle information.

[0085] In one embodiment, a spatial cue retention system based on an ANC headset is provided, the system comprising:

[0086] The first module is used to build a spatial cue retention system based on ANC headphones;

[0087] The second module is used to establish the ANC filter step size update formula based on the CLMS algorithm;

[0088] The third module is used to simulate and verify the spatial cues preservation and noise reduction effects of the system.

[0089] Specific limitations regarding the spatial cue retention system based on ANC headphones can be found in the limitations of the spatial cue retention method based on ANC headphones mentioned above, and will not be repeated here. Each module in the aforementioned spatial cue retention system based on ANC headphones can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the corresponding operations of each module.

[0090] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to perform the following steps:

[0091] Step 1: Offline identification of the left and right secondary channels of the earphones;

[0092] Step 2: Combine the sound signals collected from the left and right ears into a complex signal;

[0093] Step 3: Filter the complex signal using a feedforward filter;

[0094] Step 4: Estimate the feedback reference signal;

[0095] Step 5: Filter the estimated feedback reference signal using a feedback filter;

[0096] Step 6: After filtering, separate the left and right ear signals according to the real and imaginary parts, and send them through the secondary paths on the left and right sides respectively.

[0097] For specific limitations on each step, please refer to the limitations on the spatial cue preservation method based on ANC headphones mentioned above, which will not be repeated here.

[0098] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, the computer program performing the following steps when executed by a processor:

[0099] Step 1: Offline identification of the left and right secondary channels of the earphones;

[0100] Step 2: Combine the sound signals collected from the left and right ears into a complex signal;

[0101] Step 3: Filter the complex signal using a feedforward filter;

[0102] Step 4: Estimate the feedback reference signal;

[0103] Step 5: Filter the estimated feedback reference signal using a feedback filter;

[0104] Step 6: After filtering, separate the left and right ear signals according to the real and imaginary parts, and send them through the secondary paths on the left and right sides respectively.

[0105] For specific limitations on each step, please refer to the limitations on the spatial cue preservation method based on ANC headphones mentioned above, which will not be repeated here.

[0106] This invention is highly innovative and practical. It fully integrates ANC technology, achieving ANC while preserving the spatial cues of residual sound. This greatly optimizes the problems existing in current ANC headphones, which is conducive to improving the listening experience of ANC headphones and provides a valuable reference for the iteration of new technologies.

[0107] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention without departing from its spirit and scope should be included within the protection scope of the present invention.

Claims

1. A spatial cue preservation method based on ANC headphones, characterized in that, The method includes the following steps: Step 1, construct a spatial cue retention system based on ANC headphones; including: Steps 1-4 involve constructing a dual-channel composite adaptive ANC system based on the CLMS algorithm. The left and right ear signals need to be combined into a complex signal. The filter coefficients are updated using the complex LMS algorithm. After passing through feedforward and feedback filters, the real and imaginary parts of the complex signal are extracted and processed through the left and right ear secondary paths, respectively. Specifically, this includes: Step 1-4-1: Perform offline identification of the secondary channels; Step 1-4-2: Combine the left and right ear signals into a complex signal as a complex reference signal; Step 1-4-3: The feedforward filter filters the input complex reference signal; Step 1-4-4: Perform feedback complex reference signal estimation; Steps 1-4-5: The feedback filter filters the feedback complex reference signal; Steps 1-4-6: The output signals of the feedforward filter and the feedback filter are superimposed, and then the left and right ear signals are separated according to the real and imaginary parts, and passed through the left and right secondary channels respectively. Steps 1-4-7: The output signals of the left and right secondary channels are superimposed on the desired signals of the left and right ears respectively to obtain the residual signals of the left and right ears; Step 2, establish the ANC filter step size update formula based on the CLMS algorithm: In the formula, μ is the step size, n is the time, w(n) is the weight vector of the filter at the current time, e(n) is the signal residual at the current time, and x(n) is the input of the filter at the current time. Let x(n) be the conjugate of x(n), and w(n), e(n), and x(n) be all complex numbers, respectively denoted as: , , R and I represent the real and imaginary parts, respectively, corresponding to the left and right ear signals, and i represents the imaginary unit. N represents the filter order, and T represents the matrix transpose; Step 3: Simulation verification of the spatial cues preservation and noise reduction effects of the dual-channel composite adaptive ANC system constructed in Step 1.

2. The spatial cue retention method based on ANC headphones according to claim 1, characterized in that, Step 1, which describes the construction of a spatial cue retention system based on ANC headphones, further includes: Step 1-1: Construct a single-channel composite adaptive ANC system; Steps 1-2: Establish the ANC filter step size update formula based on the FxLMS algorithm; Steps 1-3 involve conducting an impact analysis on spatial cues using a single-channel composite adaptive ANC system.

3. The spatial cue retention method based on ANC headphones according to claim 1, characterized in that, Step 1-1 describes the construction of a single-channel composite adaptive ANC system, which specifically includes: Step 1-1-1: Perform offline identification of the secondary channels; Step 1-1-2: The feedforward filter filters the input reference signal; Step 1-1-3: Perform feedback reference signal estimation; Step 1-1-4: The feedback filter filters the feedback reference signal; Step 1-1-5: The output signal of the feedforward filter and the output signal of the feedback filter are superimposed and then passed through the secondary channel; Step 1-1-6: The secondary channel output signal is superimposed with the desired signal to obtain the residual signal.

4. The spatial cue retention method based on ANC headphones according to claim 3, characterized in that, The offline identification of the secondary channel described in step 1-1-1 specifically includes: the input signal is a random noise signal, the secondary channel is trained and estimated before the single-channel composite adaptive ANC system is run, and the fixed training result is saved as the secondary channel model. The model is then introduced into the single-channel composite adaptive ANC system.

5. The spatial cue preservation method based on ANC headphones according to claim 3, characterized in that, The feedforward filter described in step 1-1-2 filters the input reference signal, specifically including: the input reference signal is the historical sequence of the signal at the current time, and after feedforward filtering, the feedforward signal to be passed through the secondary channel is calculated.

6. The spatial cue preservation method based on ANC headphones according to claim 3, characterized in that, The feedback reference signal estimation mentioned in step 1-1-3 is specifically the sum of the error signal and the speaker signal after passing through the secondary channel; The feedback filter described in step 1-1-4 filters the feedback reference signal, specifically by: based on the feedback reference signal calculated in step 1-1-3, performing feedback filtering, and calculating the signal to be superimposed on the output of the feedforward filter.

7. The spatial cue retention method based on ANC headphones according to claim 2, characterized in that, The step size update formula for the ANC filter based on the FxLMS algorithm established in steps 1-2 is as follows: In the formula, μ is the step size, n is the time, e(n) is the signal residual at the current time, and x(n) is the input of the filter at the current time. This represents the weight vector of the filter at the current time. This is the weight vector of the filter at the next time step. N is the order of the filter, and T represents the matrix transpose.

8. The spatial cue retention method based on ANC headphones according to claim 2, characterized in that, Steps 1-3 describe the impact analysis of spatial cues using a single-channel composite adaptive ANC system. Specifically, this includes comparing the IPD information of the binaural signals before and after ANC. The IPD is obtained by calculating the cross-power spectral density of the binaural signals and then extracting the phase information; the formula is as follows: In the formula, This represents the cross-power spectral density between the left and right ears. Indicates the self-power spectral density; For input signals at different angles, IPD information is calculated in the above manner, and then the IPD error before and after ANC is calculated to obtain the relationship between IPD error and input angle and signal frequency.

Images