Active noise reduction earphone and howling suppression method, device and storage medium thereof

Abstract

This application discloses an active noise-canceling headset and its feedback suppression method, device, and storage medium. For an active noise-canceling headset including a feedforward microphone, a feedback microphone, a call microphone, and a speaker, when feedback is detected, the target feedback path is first located in the feedforward and feedback paths. That is, it is determined whether the feedback is occurring in the feedforward path, the feedback path, or both paths. Based on this, targeted feedback suppression is performed according to the target feedback path to ensure the effectiveness of feedback suppression. Simultaneously, after feedback suppression, the acoustic state of the headset can be detected at a set period until the detected acoustic state indicates that there is no feedback. The release time of feedback suppression is accurately determined, and the feedback suppression on the target feedback path can be released, preventing intermittent feedback from the headset and ensuring normal user operation.

Description

Technical Field

[0001] This application relates to the field of active noise cancellation technology, and more specifically, to an active noise cancellation headphone and its feedback suppression method, device and storage medium. Background Technology

[0002] Active noise-canceling headphones can reduce ambient noise heard by the human ear to a certain extent, thus protecting hearing and improving playback quality. Common active noise-canceling headphones typically have three microphones per earbud, and their usual installation location is as follows... Figure 1 As shown. The feedforward (FF) microphone and the call microphone are located on the headphone stem, while the feedback (FB) microphone is located inside the headphone near the sound outlet. The feedforward and feedback microphones together constitute the headphone's active noise cancellation microphone, used to achieve feedforward noise cancellation, feedback noise cancellation, and pass-through functions.

[0003] Under normal wearing conditions, neither the active noise cancellation nor the pass-through function of the headphones will produce feedback because there is no obvious positive feedback loop between the feedforward microphone and the headphone speaker; it can be considered an open-loop path. The feedback microphone and the headphone speaker operate in a negative feedback loop. Neither of these loops will form an unstable positive feedback loop, thus preventing feedback. However, when the headphones are blocked, squeezed, or tightly held, the feedforward / feedback loop will become a positive feedback loop at certain frequencies. That is, the microphone picks up the signal, the filter amplifies it, and the headphone speaker outputs it back to the microphone, ultimately creating self-excitation and producing feedback. When feedback occurs, the headphones' audio functions cannot work properly, severely affecting normal use. Therefore, it is necessary to research a suppression scheme for feedback in active noise-canceling headphones. Summary of the Invention

[0004] In view of the above problems, this application is made to provide an active noise-canceling headphone and a method, device, and storage medium for suppressing feedback from the same, so as to detect and suppress feedback from the active noise-canceling headphone and ensure normal use by the user. The specific solution is as follows:

[0005] In a first aspect, a method for suppressing feedback in active noise-canceling headphones is provided, wherein the active noise-canceling headphones include a feedforward microphone, a feedback microphone, a call microphone, and a speaker, and the method includes:

[0006] After detecting feedback from the active noise-canceling headphones, the target feedback path is located. The target feedback path includes a feedforward path composed of the feedforward microphone and the speaker, and / or a feedback path composed of the feedback microphone and the speaker.

[0007] Suppress howling based on the target howling pathway;

[0008] The acoustic state of the active noise-canceling headphones is detected at a set period until the acoustic state indicates that there is no feedback from the active noise-canceling headphones, and then feedback suppression on the target feedback path is released.

[0009] Secondly, an active noise-canceling headphone feedback suppression device is provided, wherein the active noise-canceling headphone includes a feedforward microphone, a feedback microphone, a call microphone, and a speaker, and the device includes:

[0010] The feedback detection unit is used to detect whether feedback occurs in active noise-canceling headphones;

[0011] The feedback path localization unit is used to locate the target feedback path generated by the current feedback after detecting feedback from the active noise-canceling headphones. The target feedback path includes a feedforward path composed of the feedforward microphone and the speaker, and / or a feedback path composed of the feedback microphone and the speaker.

[0012] A howling suppression unit is used to suppress howling based on the target howling path;

[0013] The feedback suppression release unit is used to detect the acoustic state of the active noise-canceling headphones at a set period, and release feedback suppression on the target feedback path after the acoustic state indicates that there is no feedback in the active noise-canceling headphones.

[0014] Thirdly, an active noise-canceling headphone is provided, comprising: a feedforward microphone, a feedback microphone, a call microphone, a speaker, a feedforward filter, a feedback filter, and a processor;

[0015] The processor is used to implement the various steps of the active noise-canceling headphone feedback suppression method as described above.

[0016] Fourthly, a storage medium is provided on which a computer program is stored, which, when executed by a processor, implements the various steps of the active noise-canceling headphone feedback suppression method as described above.

[0017] Using the above technical solution, this application provides a feedback suppression method for active noise-canceling headphones, including a feedforward microphone, a feedback microphone, a call microphone, and a speaker. When feedback is detected in the headphones, the target feedback path is first located in the feedforward and feedback paths. That is, it is determined whether the feedback is occurring in the feedforward path, the feedback path, or both paths. Based on this, targeted feedback suppression is performed according to the target feedback path to ensure the effectiveness of feedback suppression. Simultaneously, after feedback suppression, the acoustic state of the headphones can be detected at a set period until the detected acoustic state indicates that there is no feedback. The release time of feedback suppression is then accurately determined, and the feedback suppression on the target feedback path can be released, preventing intermittent feedback from the headphones and ensuring normal user operation. Attached Figure Description

[0018] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the scope of this application. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings:

[0019] Figure 1 An example is a schematic diagram of an active noise-canceling headphone structure;

[0020] Figure 2 An example is shown in the block diagram of an active noise cancellation / pass-through function for headphones;

[0021] Figure 3 This is a schematic flowchart of an active noise-canceling headphone feedback suppression method disclosed in an embodiment of this application;

[0022] Figure 4 An example of a transfer function H s2b (w) Schematic diagram of amplitude versus frequency under different wearing conditions;

[0023] Figure 5 An example of a transfer function H s2f (w) Schematic diagram of amplitude versus frequency under different wearing conditions;

[0024] Figure 6 An example of spectrograms of headphone microphone and speaker signals during feedback howling and feedforward howling is provided;

[0025] Figure 7 This is a schematic diagram of an active noise-canceling headphone feedback suppression device provided in an embodiment of this application. Detailed Implementation

[0026] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0027] This application provides a method for suppressing howling in active noise-canceling headphones. When applied to active noise-canceling headphones, it can detect and suppress howling, reducing the impact of howling on the normal use of the headphones.

[0028] This application's solution can be applied to suppress feedback in active noise-canceling headphones. Active noise-canceling headphones can include headphones with active noise cancellation / pass-through functionality, typically comprising a feedforward microphone (FFmic), a feedback microphone (FBmic), a talk microphone (TALKmic), and a speaker (SPK). Figure 1 As shown. Furthermore, to ensure the proper functioning of the headphones' active noise cancellation / pass-through capabilities, the headphones may also include a feedforward filter and a feedback filter. For example... Figure 2 It illustrates a block diagram of the principle of active noise cancellation / pass-through function in headphones.

[0029] x(n) represents the ambient noise collected by the feedforward microphone, n is the number of digital signal sampling points, s(n) represents the digital signal output by the speaker, and H s2f (w) represents the transfer function from the loudspeaker to the feedforward microphone, H s2b (w) represents the transfer function from the speaker to the feedback microphone, H f2b (w) represents the transfer function from the feedforward microphone to the feedback microphone, H ff (w) represents the feedforward noise reduction filter in active noise reduction mode, K ff (w) represents the feedforward pass-through filter in pass-through mode, H fb (w) represents the feedback noise reduction filter in active noise reduction mode, K fb (w) represents the feedback pass-through filter in pass-through mode.

[0030] Next, combined Figure 3 The active noise-canceling headphone feedback suppression method of this application may include the following steps:

[0031] Step S100: After detecting feedback from the active noise-canceling headphones, locate the target feedback path that is currently causing the feedback.

[0032] The target howling path includes a feedforward path consisting of the feedforward microphone and the speaker, and / or a feedback path consisting of the feedback microphone and the speaker.

[0033] In different situations, the headphone may only have feedback noise in the feedforward path, or only in the feedback path, or both the feedforward and feedback paths may have feedback noise simultaneously.

[0034] In this step, after detecting howling from the headphones, the first step is to locate the howling path that is generating the howling, which is defined as the target howling path. The target howling path can be one or both of the feedforward path and the feedback path.

[0035] Step S110: Suppress howling according to the target howling path.

[0036] Specifically, this application can perform targeted howling suppression based on the target howling path where the current howling occurs. For example, if the target howling path includes a feedforward path, howling suppression can be performed on the feedforward path; if the target howling path includes a feedback path, howling suppression can be performed on the feedback path, thus ensuring the effectiveness of howling suppression.

[0037] In one optional feedback suppression method, the gain of the filter corresponding to the target feedback path can be reduced by a set step size. After reducing the gain, it is continued to detect whether feedback still occurs in the active noise-canceling headphones. If so, the gain of the filter corresponding to the target feedback path can be further reduced until feedback is detected to be absent from the active noise-canceling headphones.

[0038] The filter corresponding to the target howling path may include a feedforward filter corresponding to the feedforward path and / or a feedback filter corresponding to the feedback path.

[0039] The gain adjustment step size can be 6dB or other values.

[0040] Of course, other methods of suppressing howling can also be used, which are not listed one by one in the embodiments of this application.

[0041] Step S120: Detect the acoustic state of the active noise-canceling headphones according to a set period until the acoustic state indicates that the active noise-canceling headphones are free of howling, and then release howling suppression on the target howling path.

[0042] It should be noted that after the above steps of howling suppression, it can only ensure that the headphones will not howl at that moment. However, in order to ensure the normal audio function of the headphones, it is also necessary to accurately determine the release timing of howling suppression to ensure that howling suppression will not recur intermittently after it is released. Therefore, in this embodiment, after howling suppression is performed, the acoustic state of the active noise-canceling headphones is detected at a set period to determine whether howling still exists in the active noise-canceling headphones. When howling is detected to be gone, it is considered that the release timing of howling suppression is met, and howling suppression on the target howling path can be released.

[0043] Here, taking the suppression of the target howling path as reducing the gain of the filter corresponding to the target howling path as an example, releasing howling suppression can include: increasing the gain of the filter corresponding to the target howling path to the normal level before howling suppression according to a set step size.

[0044] In this embodiment, it is considered that feedback generally does not occur when the headphones are worn normally. However, when the headphones are blocked or held in the hand, the acoustic state of the headphones will change significantly, and the transfer function H from the headphone speaker to the feedback microphone will change. s2b (w) Speaker to feedforward microphone transfer function H s2f (w) and the transfer function H from the feedforward microphone to the feedback microphone. f2b The amplitude of (w) will increase significantly. Figure 4 and Figure 5 Only H is shown as an example. s2b (w) and H s2f The curve change of (w), H f2b The curve change of (w) and H s2f (w) Similarly, it is not shown in this article.

[0045] Depend on Figure 4 , Figure 5 It can be seen that, compared to the normal wearing state and the loose wearing state, the headphone jack blocking / pressing state causes the feedforward or feedback path to become an unstable positive feedback, resulting in headphone howling. s2b (w) and H s2f (w) shows a significant increase in amplitude in certain frequency bands.

[0046] Therefore, in this step, the acoustic state of the active noise-canceling headphones is periodically detected to determine whether the howling of the active noise-canceling headphones has stopped, and howling suppression on the target howling path is released once it stops.

[0047] The set period for acoustic state detection in this step can be adjusted. If you want to accurately determine when the feedback suppression release is needed, you can adjust it to real-time detection or set a shorter interval period.

[0048] The feedback suppression method provided in this application, upon detecting feedback in the headphones, first locates the target feedback path in the feedforward and feedback paths, determining whether the feedback is occurring in the feedforward path, the feedback path, or both. Based on this, targeted feedback suppression is performed on the target feedback path to ensure its effectiveness. Simultaneously, after feedback suppression, the acoustic state of the headphones is monitored at set intervals until the detected acoustic state indicates no feedback occurs. The release time for feedback suppression is then accurately determined, releasing the feedback suppression on the target feedback path to prevent intermittent feedback from the headphones and ensuring normal user operation.

[0049] In some embodiments of this application, an implementation method for detecting feedback in active noise-canceling headphones is provided.

[0050] Traditional solutions typically determine whether headphone feedback is occurring by checking for signal clipping in the speaker. If clipping is detected, the speaker is immediately shut down. However, this method is inaccurate. Clipping can occur when the speaker input signal is high, even without feedback. Current solutions might misinterpret this as feedback and shut down the speaker, preventing the user from hearing the normal signal.

[0051] To address the shortcomings of the existing technology, this application conducted relevant research and discovered that the true cause of howling is: signals at certain frequencies change from negative feedback to unstable positive feedback, causing the signal amplitude to continuously increase and become much higher than signals at other frequencies, ultimately resulting in howling. Therefore, this embodiment designs a howling detection method from this perspective, which specifically includes the following steps:

[0052] S1. Obtain the loudspeaker signal s(n) and calculate the power spectrum S(l, ω) of the loudspeaker signal frame by frame.

[0053] Where l is the time frame and ω is the angular frequency.

[0054] S2. Based on the power spectrum of the speaker signal in each frame, calculate the average power of the speaker signal in each frame, and determine the frequency point with the highest power in the power spectrum of the speaker signal in each frame as the candidate frequency point for howling in the corresponding frame.

[0055] Specifically, the average power of each frame of the loudspeaker signal can be expressed as: In the power spectrum of each frame of the loudspeaker signal, the frequency point with the highest power is determined as the candidate frequency point for howling in the corresponding frame, and the power of the candidate frequency point is denoted as S. max (l,ω h ).

[0056] S3. Determine whether the difference between the power of the candidate frequency point of feedback and the average power of the speaker signal exceeds the set power threshold for a continuous L frames; if so, determine that feedback has occurred in the active noise-canceling headphones.

[0057]

[0058] The value of L can be between 5 and 10, for example, L can be 8.

[0059] According to the feedback detection method provided in this embodiment, if the difference between the power of the feedback candidate frequency points and the average power of the signal in L consecutive frames of the signal output by the speaker exceeds a set power threshold, it indicates that the signal of the feedback candidate frequency points has changed from negative feedback to unstable positive feedback, causing the amplitude of the signal output by the speaker to continuously increase and become much higher than that of other frequency points, thus ultimately producing feedback. Using the solution provided in this embodiment for feedback detection is more consistent with the root cause of feedback, and therefore the obtained feedback detection results are more accurate.

[0060] In some embodiments of this application, the process of locating the target howling path generated by the current howling in the aforementioned step S100 is further described.

[0061] As described above regarding existing technical solutions, these solutions merely determine the amplitude of the speaker output signal and then shut off the speaker signal. They lack true feedback detection and the process of locating the feedback path. However, as explained in the working principle of active noise-canceling headphones, there are two distinct paths: a feedforward path and a feedback path. When feedback occurs in active noise-canceling headphones, it could be caused by one or both of these paths. Different feedback paths require different suppression methods to effectively suppress the feedback. Therefore, it is necessary to provide a solution that accurately locates the target feedback path.

[0062] Reference Figure 6 It provides an example of spectrograms of headphone microphone and speaker signals during feedback howling and feedforward howling.

[0063] Figure 6 The left half of the image shows the spectrograms of each signal during feedback howling, while the right half shows the spectrograms of each signal during feedforward howling.

[0064] Signals from three microphones are acquired, where the feedforward microphone corresponds to the feedforward time-domain signal x. ff (n), the feedback microphone corresponds to the feedback time-domain signal x fb (n), the time-domain signal x corresponding to the microphone. talk (n).

[0065] Depend on Figure 6 It can be seen that when feedback occurs, the speaker signal s(n) is in a saturated state. Since the feedback microphone is very close to the speaker, the feedback time-domain signal x corresponding to the feedback microphone is... fb (n) will also be truncated. When the feedback path experiences howling while the feedforward path does not, the feedback time-domain signal x fb (n) will be truncated, but the feedforward time-domain signal x ff (n) and the time-domain signal x talk (n) will not be clipped; in this case, the transfer function H from the speaker to the feedback microphone... s2b (w) Low frequencies will rise. If the feedback is occurring in the feedforward path, then x fb (n), x ff (n) and x talk (n) will all be truncated.

[0066] Therefore, in this embodiment, the feedforward time-domain signal x corresponding to the feedforward microphone is used. ff (n) The time-domain signal x corresponding to the microphone talk (n) and the transfer function H from the speaker to the feedback microphone s2b (w) is used to locate the target howling path.

[0067] Specifically, the process of locating the target howling path generated by the current howling may include:

[0068] S1. Acquire the feedforward time-domain signal x of the current frame captured by the feedforward microphone. ff (n) and the call time-domain signal x of the current frame acquired by the call microphone. talk (n).

[0069] S2, Based on feedforward time-domain signal x ff (n) Calculate the average energy P of the feedforward microphone in the current frame. ff (l), and, based on the call time-domain signal x talk (n) Calculate the average energy P of the microphone in the current frame. talk (l).

[0070]

[0071]

[0072] Where N is the number of sampling points in a frame.

[0073] S3, if the average energy P of the feedforward microphone in the current frame ff (l) Less than the first energy threshold T ff (l), and the average energy P of the call microphone in the current frame talk(l) Less than the second energy threshold T talk (l) indicates that only the feedback path is currently experiencing a whistling sound.

[0074] Among them, the first energy threshold T ff (l) and the second energy threshold T talk (l) can take the value of 30dB or other values.

[0075] Furthermore, the process of locating the target howling path generated by the current howling may also include:

[0076] S4. Calculate the transfer function H from the speaker to the feedback microphone. s2b (w) The average amplitude in the set low-frequency band.

[0077] Combination Figure 4 As shown, the transfer function H s2b (w) The amplitude variation covers almost the entire frequency band. In the low frequency band: 50Hz to 1000Hz, the three states of loose wearing, normal wearing and headphone jack blockage can be distinguished. Therefore, in this step, the low frequency band can be set to 50Hz to 1000Hz.

[0078] Transfer function H s2b (w) The average amplitude of the low-frequency band in the current frame can be expressed as |H s2b (l)|, its calculation method can be:

[0079] S41. Obtain the feedback frequency domain signal X of the current frame captured by the feedback microphone. fb (l, ω), and the frequency domain signal X of the current frame of the loudspeaker. spk (l, ω);

[0080] S42. Calculate the feedback frequency domain signal X in the set low frequency band. fb (l, ω) and the frequency domain signal X of the current frame of the loudspeaker spk The ratio of (l, ω) is used to obtain the average amplitude based on the ratio.

[0081]

[0082] Ratio H s2b (l, ω) are complex numbers, and the average amplitude |H can be obtained after taking the modulus. s2b (l)|.

[0083] S5. If the average energy P of the feedforward microphone in the current frame ff (l) Not less than the first energy threshold T ff (l), and the average energy P of the call microphone in the current frame talk (l) Not less than the second energy threshold Ttalk (l), and the average amplitude |H s2b (l) | Less than the set first amplitude threshold T s2b (l) indicates that only the feedforward path is currently experiencing a whistling sound.

[0084] Specifically, as explained above, when feedback feedback occurs, the transfer function H from the speaker to the feedback microphone... s2b (w) will rise in the low-frequency band, therefore, if the transfer function H is determined in this step... s2b (w) Average amplitude |H in the set low-frequency band s2b (l) If the value is less than the set first amplitude threshold, it indicates that no howling has occurred in the current feedback path, and the only possibility is that the howling has occurred in the feedforward path.

[0085] Among them, the first amplitude threshold T s2b (l) can take the value -10dB or other values.

[0086] S6. If the average energy P of the feedforward microphone in the current frame ff (l) Not less than the first energy threshold T ff (l), and the average energy P of the call microphone in the current frame talk (l) Not less than the second energy threshold T talk (l), and the average amplitude |H s2b (l)|Not less than the set first amplitude threshold T s2b (l) indicates that both the current feedforward path and the feedback path are experiencing howling.

[0087] The process of determining the target howling path described above can be summarized by the following formula:

[0088]

[0089] The target feedback path determination method provided in this embodiment obtains the above determination result by analyzing the energy changes of each microphone signal, speaker signal and different transfer functions when the headphones generate feedback feedback and feedforward feedback. Based on this, by analyzing different microphone signals and speaker signals and combining them with the above determination formula, the target feedback path that is currently generating feedback can be accurately detected.

[0090] In some embodiments of this application, the process of releasing the howling suppression on the target howling path after detecting the acoustic state of the active noise-canceling headphones at a set period according to the aforementioned step S120 until the acoustic state indicates that the active noise-canceling headphones are free of howling is further described.

[0091] Normally, headphones do not produce feedback when worn properly. However, when the headphones are blocked or held in the hand, their acoustic properties change significantly, affecting the transfer function H from the speaker to the feedback microphone. s2b (w) Speaker to feedforward microphone transfer function H s2f (w) Transfer function H from feedforward microphone to feedback microphone f2b The amplitude of (w) will increase significantly in certain frequency bands, such as Figure 4 and Figure 5 (due to H) f2b (w) curve and H s2f (w) Similarly, therefore H is not shown in this application. f2b (w)) makes the feedforward or feedback path become an unstable positive feedback, thus causing howling.

[0092] As can be seen from the above-described solution, this application can detect and suppress howling, but it is still necessary to further clarify when to release howling suppression, that is, to detect when the headphones can return to normal and ensure that the headphones work normally.

[0093] In this embodiment, the process of releasing howling suppression is described separately when the target howling pathway is different:

[0094] 1. If the target howling path includes a feedback path:

[0095] After suppressing the howling, the transfer function H from the speaker to the feedback microphone is calculated according to a set period. s2b (w) Average amplitude |H in the set low-frequency band s2b (l)|, until the calculated average amplitude is less than the first set amplitude threshold T. s2b (l) when |H s2b (l)|<T s2b (l) indicates that the acoustic state of the headphones is not in the blocked / pressed state, which means that there is no howling in the feedback path, and the howling suppression of the feedback path is released.

[0096] Combination Figure 4 As shown, the transfer function H s2b (w) The increase in amplitude covers almost the entire frequency band. In the low frequency band: 50Hz to 1000Hz, the three states of loose wearing, normal wearing and headphone jack blockage can be distinguished. Therefore, in this step, the low frequency band can be set to 50Hz to 1000Hz.

[0097] One method to release the howling suppression of the feedback path is to increase the gain of the feedback filter corresponding to the feedback path to the normal level before howling suppression by adjusting the gain by a set step size.

[0098] 2. If the target howling path includes a feedforward path:

[0099] In this case, this embodiment provides several different judgment methods:

[0100] The first type

[0101] After suppressing the howling, the transfer function Hf2 from the feedforward microphone to the feedback microphone is calculated according to a set period. b (w) Average amplitude in the set mid-frequency band |H f2b (l)|, until the calculated average amplitude is less than the second set amplitude threshold T. f2b (l) when |H f2b (l)|<T f2b (l) indicates that the acoustic state of the headphones is not in the blocked / pressed state, which means that there is no howling in the feedforward path, and the howling suppression of the feedforward path is released.

[0102] The second type

[0103] Alternatively, the transfer function H from the speaker to the feedforward microphone can be calculated according to a set period. s2f (w) Average amplitude in the set mid-frequency band |H s2f (l)|, until the calculated average amplitude is less than the second set amplitude threshold T. s2f (l) when |H s2f (l)|<T s2f (l) indicates that the acoustic state of the headphones is not in the blocked / pressed state, which means that there is no howling in the feedforward path, and the howling suppression of the feedforward path is released.

[0104] Combination Figure 5 As shown, the transfer function H f2b (w) or H s2f (w) The amplitude increase is basically in the mid-frequency band of 500Hz to 4KHz. Therefore, in this embodiment, the above-mentioned mid-frequency band can be set to 500Hz to 4KHz.

[0105] Furthermore, considering the transfer function H s2f (w) It is easily affected by external environmental noise, which may lead to certain errors in the results. Therefore, the first judgment method can be preferred in this embodiment.

[0106] One way to release the squeal suppression of the feedforward path is to increase the gain of the feedforward filter corresponding to the feedforward path to the normal level before squeal suppression by adjusting the gain by a set step size.

[0107] The above steps involve calculating the transfer function H from the feedforward microphone to the feedback microphone. f2b (w) Average amplitude in the set mid-frequency band |H f2b The process of (l) can include:

[0108] With active noise cancellation or pass-through enabled, the transmission path from the feedforward microphone to the feedback microphone adds an electrical path to the existing acoustic path. Specifically, the electrical path is feedforward microphone - feedforward filter - speaker - feedback microphone. Therefore, the transfer function H can be calculated as follows: f2b (w) Average amplitude in the set mid-frequency band |H f2b (l)|:

[0109] Let's take the active noise cancellation function being enabled as an example:

[0110] With active noise cancellation enabled, the feedforward filter at time t1 can be expressed as H. ff1 The signals from the feedforward microphone and the feedback microphone are acquired, and the transfer function H between them is calculated. yx1 :

[0111] H yx1 (l, ω) = X fb1 (l,ω) / X ff1 (l, ω)

[0112] Among them, X fb1 (l, ω), X ff1 (l, ω) are the frequency domain signals of the feedback microphone and the feedforward microphone at time t1, respectively.

[0113] At time t2, by changing the gain of the feedforward filter, a new feedforward filter H is obtained. ff2 The transfer function H between the feedforward microphone and the feedback microphone is recalculated. yx2 :

[0114] H yx2 (l, ω) = X fb2 (l,ω) / X ff2 (l, ω)

[0115] Among them, X fb2 (l, ω), X ff2 (l, ω) are the frequency domain signals of the feedback microphone and the feedforward microphone at time t2, respectively.

[0116] Calculate the transfer function H f2b (w):

[0117]

[0118] For the above transfer function H f2b (w) Take the set average amplitude of the mid-frequency band to obtain |H f2b (l)|.

[0119] The above example, using headphones with active noise cancellation enabled, illustrates the calculation of the transfer function H. f2b (w) Average amplitude in the set mid-frequency band |H f2b (l)| process. When the headphones are in pass-through mode, it is only necessary to convert the feedforward filter from the feedforward noise reduction filter H in the aforementioned calculation process. ff Change to feedforward pass-through filter K ff That's it; the rest of the calculation process remains unchanged.

[0120] The solution described in this embodiment can detect the acoustic state of the headphones based on the existing microphone and speaker without adding any additional sensors. Once the acoustic state indicates that the active noise-canceling headphones are not experiencing feedback (e.g., the headphones are not currently blocked / pressed), feedback suppression on the target feedback path is released.

[0121] In addition, other methods can be used to detect the acoustic status of headphones, such as adding sensors to the headphones and using the parameters collected by the sensors to detect the acoustic status of the headphones.

[0122] The active noise cancellation headphone feedback suppression device provided in the embodiments of this application is described below. The active noise cancellation headphone feedback suppression device described below can be referred to in correspondence with the active noise cancellation headphone feedback suppression method described above.

[0123] See Figure 7 , Figure 7 This is a schematic diagram of an active noise-canceling headphone feedback suppression device disclosed in an embodiment of this application.

[0124] like Figure 7 As shown, the device may include:

[0125] The feedback detection unit 11 is used to detect whether feedback occurs in the active noise-canceling headphones;

[0126] The feedback path localization unit 12 is used to locate the target feedback path generated by the current feedback after detecting feedback from the active noise-canceling headphones. The target feedback path includes a feedforward path composed of the feedforward microphone and the speaker, and / or a feedback path composed of the feedback microphone and the speaker.

[0127] The howling suppression unit 13 is used to suppress howling according to the target howling path;

[0128] The howling suppression release unit 14 is used to detect the acoustic state of the active noise-canceling headphones according to a set period, and release howling suppression on the target howling path after the acoustic state indicates that the active noise-canceling headphones are free of howling.

[0129] Optionally, the process by which the above-mentioned feedback detection unit detects whether feedback occurs in the active noise-canceling headphones may include:

[0130] Acquire the speaker signal and calculate the power spectrum of the speaker signal frame by frame;

[0131] Based on the power spectrum of the loudspeaker signal in each frame, the average power of the loudspeaker signal in each frame is calculated, and the frequency point with the highest power in the power spectrum of the loudspeaker signal in each frame is determined as the candidate frequency point for howling in the corresponding frame.

[0132] Determine whether the difference between the power of the candidate frequency point of the howling and the average power of the speaker signal exceeds a set power threshold for a continuous period of L frames;

[0133] If so, it confirms that the active noise-canceling headphones are experiencing feedback.

[0134] Optionally, the process by which the howling path localization unit locates the target howling path generated by the current howling may include:

[0135] Acquire the feedforward time-domain signal of the current frame collected by the feedforward microphone and the call time-domain signal of the current frame collected by the call microphone;

[0136] The average energy of the feedforward microphone in the current frame is calculated based on the feedforward time-domain signal, and the average energy of the call microphone in the current frame is calculated based on the call time-domain signal.

[0137] If the average energy of the feedforward microphone in the current frame is less than the first energy threshold, and the average energy of the call microphone in the current frame is less than the second energy threshold, then it is determined that only the feedback path is experiencing howling.

[0138] Optionally, the process by which the howling path localization unit locates the target howling path generated by the current howling may further include:

[0139] Calculate the average amplitude of the transfer function from the loudspeaker to the feedback microphone in a set low-frequency band;

[0140] If the average energy of the feedforward microphone in the current frame is not less than the first energy threshold, the average energy of the call microphone in the current frame is not less than the second energy threshold, and the average amplitude is less than the set first amplitude threshold, then it is determined that only the feedforward path is experiencing howling.

[0141] If the average energy of the feedforward microphone in the current frame is not less than a first energy threshold, the average energy of the call microphone in the current frame is not less than a second energy threshold, and the average amplitude is not less than a set first amplitude threshold, then it is determined that both the current feedforward path and the feedback path are experiencing howling.

[0142] Optionally, the process by which the above-mentioned feedback path localization unit calculates the average amplitude of the transfer function from the speaker to the feedback microphone in a set low-frequency band may include:

[0143] Acquire the feedback frequency domain signal of the current frame collected by the feedback microphone, and the frequency domain signal of the current frame of the speaker;

[0144] Calculate the ratio of the feedback frequency domain signal to the frequency domain signal of the current frame of the speaker in a set low frequency band, and obtain the average amplitude based on the ratio.

[0145] Optionally, the process by which the above-mentioned howling suppression unit suppresses howling according to the target howling path may include:

[0146] The gain of the filter corresponding to the target howling path is reduced by a set step size, wherein the filter corresponding to the feedforward path is a feedforward filter and the filter corresponding to the feedback path is a feedback filter.

[0147] After reducing the gain, check whether the active noise-canceling headphones still produce feedback. If so, continue to reduce the gain of the filter corresponding to the target feedback path until feedback is detected to be absent from the active noise-canceling headphones.

[0148] Optionally, the process of releasing the feedback suppression release unit, which detects the acoustic state of the active noise-canceling headphones at a set period, and releases feedback suppression on the target feedback path after the acoustic state indicates that there is no feedback from the active noise-canceling headphones, may include:

[0149] If the target howling path includes the feedback path, then after howling suppression, the average amplitude of the transfer function from the speaker to the feedback microphone in the set low frequency band is calculated according to a set period until the calculated average amplitude is less than the first set amplitude threshold, indicating that no howling occurs in the feedback path, and the howling suppression of the feedback path is released.

[0150] If the target howling path includes the feedforward path, then after howling suppression, the average amplitude of the transfer function from the feedforward microphone to the feedback microphone in the set mid-frequency band is calculated according to a set period until the calculated average amplitude is less than the second set amplitude threshold, indicating that no howling occurs in the feedforward path, and the howling suppression of the feedforward path is released.

[0151] Optionally, the process by which the above-mentioned howling suppression release unit releases howling suppression on the feedback path may include:

[0152] The gain of the feedback filter corresponding to the feedback path is increased to the normal level before the howling is suppressed according to the set step size;

[0153] Releasing the whistling suppression of the feedback path includes:

[0154] The gain of the feedforward filter corresponding to the feedforward path is increased to the normal level before howling suppression according to the set step size.

[0155] The active noise cancellation headphone feedback suppression device provided in this application embodiment can be applied to active noise cancellation headphones. The active noise cancellation headphones may include: a feedforward microphone, a feedback microphone, a call microphone, a speaker, a feedforward filter, a feedback filter, and a processor. The processor is used to implement the various steps of the active noise cancellation headphone feedback suppression method as described in the previous embodiments.

[0156] This application embodiment also provides a storage medium that can store a program suitable for execution by a processor, the program being used to implement the various steps of the active noise-canceling headphone feedback suppression method as described in the preceding embodiments.

[0157] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0158] The various embodiments in this specification are described in a progressive manner. Each embodiment focuses on the differences from other embodiments. The various embodiments can be combined as needed, and the same or similar parts can be referred to each other.

[0159] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. An active noise reducing headphone howling suppression method, characterized by, The active noise-canceling headphones include a feedforward microphone, a feedback microphone, a call microphone, and a speaker. The method includes: After detecting feedback from the active noise-canceling headphones, the feedforward time-domain signal of the current frame collected by the feedforward microphone and the call time-domain signal of the current frame collected by the call microphone are acquired. The average energy of the feedforward microphone in the current frame is calculated based on the feedforward time-domain signal, and the average energy of the call microphone in the current frame is calculated based on the call time-domain signal. Calculate the average amplitude of the transfer function from the loudspeaker to the feedback microphone in a set low-frequency band; The system identifies the target feedback path generating the current feedback, wherein the target feedback path includes a feedforward path composed of the feedforward microphone and the speaker, and / or a feedback path composed of the feedback microphone and the speaker; wherein, if the average energy of the feedforward microphone in the current frame is less than a first energy threshold, and the average energy of the call microphone in the current frame is less than a second energy threshold, then it is determined that only the feedback path is currently experiencing feedback; if the average energy of the feedforward microphone in the current frame is not less than the first energy threshold, the average energy of the call microphone in the current frame is not less than the second energy threshold, and the average amplitude is less than a set first amplitude threshold, then it is determined that only the feedforward path is currently experiencing feedback; if the average energy of the feedforward microphone in the current frame is not less than the first energy threshold, the average energy of the call microphone in the current frame is not less than the second energy threshold, and the average amplitude is not less than a set first amplitude threshold, then it is determined that both the feedforward path and the feedback path are currently experiencing feedback. Suppress howling based on the target howling pathway; The acoustic state of the active noise-canceling headphones is detected at a set period until the acoustic state indicates that there is no feedback from the active noise-canceling headphones, and then feedback suppression on the target feedback path is released.

2. The method of claim 1, wherein, The process of detecting whether active noise-canceling headphones experience feedback includes: Acquire the speaker signal and calculate the power spectrum of the speaker signal frame by frame; Based on the power spectrum of the loudspeaker signal in each frame, the average power of the loudspeaker signal in each frame is calculated, and the frequency point with the highest power in the power spectrum of the loudspeaker signal in each frame is determined as the candidate frequency point for howling in the corresponding frame. Determine whether the difference between the power of the candidate frequency point of the howling and the average power of the speaker signal exceeds a set power threshold for a continuous period of L frames; If so, it confirms that the active noise-canceling headphones are experiencing feedback.

3. The method of claim 1, wherein, The calculation of the average amplitude of the transfer function from the speaker to the feedback microphone in a set low-frequency band includes: Acquire the feedback frequency domain signal of the current frame collected by the feedback microphone, and the frequency domain signal of the current frame of the speaker; Calculate the ratio of the feedback frequency domain signal to the frequency domain signal of the current frame of the speaker in a set low frequency band, and obtain the average amplitude based on the ratio.

4. The method of claim 1, wherein, The method of suppressing howling based on the target howling pathway includes: The gain of the filter corresponding to the target howling path is reduced by a set step size, wherein the filter corresponding to the feedforward path is a feedforward filter and the filter corresponding to the feedback path is a feedback filter. After reducing the gain, check whether the active noise-canceling headphones still produce feedback. If so, continue to reduce the gain of the filter corresponding to the target feedback path until feedback is detected to be absent from the active noise-canceling headphones.

5. The method according to any one of claims 1 to 4, characterized in that, The step of detecting the acoustic state of the active noise-canceling headphones at a set period until the acoustic state indicates that the active noise-canceling headphones are free of feedback, and then releasing feedback suppression on the target feedback path, includes: If the target howling path includes the feedback path, then after howling suppression, the average amplitude of the transfer function from the speaker to the feedback microphone in the set low frequency band is calculated according to a set period until the calculated average amplitude is less than the first set amplitude threshold, indicating that no howling occurs in the feedback path, and the howling suppression of the feedback path is released. If the target howling path includes the feedforward path, then after howling suppression, the average amplitude of the transfer function from the feedforward microphone to the feedback microphone in the set mid-frequency band is calculated according to a set period until the calculated average amplitude is less than the second set amplitude threshold, indicating that no howling occurs in the feedforward path, and the howling suppression of the feedforward path is released.

6. The method according to claim 5, characterized in that, Releasing the whistling suppression of the feedback path includes: The gain of the feedback filter corresponding to the feedback path is increased to the normal level before the howling is suppressed according to the set step size; Releasing the whistling suppression of the feedback path includes: The gain of the feedforward filter corresponding to the feedforward path is increased to the normal level before howling suppression according to the set step size.

7. An active noise-canceling headphone feedback suppression device, characterized in that, The active noise-canceling headphones include a feedforward microphone, a feedback microphone, a call microphone, and a speaker. The device includes: The feedback detection unit is used to detect whether feedback occurs in active noise-canceling headphones; A feedback path localization unit is configured to, upon detecting feedback from active noise-canceling headphones, acquire the feedforward time-domain signal of the current frame collected by the feedforward microphone and the call time-domain signal of the current frame collected by the call microphone; calculate the average energy of the feedforward microphone in the current frame based on the feedforward time-domain signal, and calculate the average energy of the call microphone in the current frame based on the call time-domain signal; calculate the average amplitude of the transfer function from the speaker to the feedback microphone in a set low-frequency band; and locate the target feedback path generated by the current feedback, wherein the target feedback path includes a feedforward path composed of the feedforward microphone and the speaker, and / or a feedback path composed of the feedback microphone and the speaker; wherein, if If the average energy of the feedforward microphone in the current frame is less than a first energy threshold, and the average energy of the call microphone in the current frame is less than a second energy threshold, then it is determined that only the feedback path is experiencing howling. If the average energy of the feedforward microphone in the current frame is not less than the first energy threshold, the average energy of the call microphone in the current frame is not less than the second energy threshold, and the average amplitude is less than a set first amplitude threshold, then it is determined that only the feedforward path is experiencing howling. If the average energy of the feedforward microphone in the current frame is not less than the first energy threshold, the average energy of the call microphone in the current frame is not less than the second energy threshold, and the average amplitude is not less than a set first amplitude threshold, then it is determined that both the feedforward path and the feedback path are experiencing howling. A howling suppression unit is used to suppress howling based on the target howling path; The feedback suppression release unit is used to detect the acoustic state of the active noise-canceling headphones at a set period, and release feedback suppression on the target feedback path after the acoustic state indicates that there is no feedback in the active noise-canceling headphones.

8. An active noise-canceling headphone, characterized in that, include: Feedforward microphone, feedback microphone, call microphone, speaker, feedforward filter, feedback filter, and processor; The processor is used to implement each step of the active noise-canceling headphone feedback suppression method as described in any one of claims 1 to 6.

9. A storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements each step of the active noise-canceling headphone feedback suppression method as described in any one of claims 1 to 6.

Images