Active noise reduction method, device, circuit, equipment and system

Abstract

The present invention discloses a noise reduction method, device, circuit, equipment and system for anti-low-frequency saturation. Among them, the method includes: Step S100, obtaining an environmental audio signal collected by a noise microphone; Step S200, inputting the environmental audio signal into a DC estimation module and a low-frequency amplitude estimation module respectively, so that the DC estimation module performs DC estimation on the environmental audio signal to obtain a DC signal, and the low-frequency amplitude estimation module performs low-frequency amplitude estimation on the environmental audio signal to obtain a low-frequency amplitude estimation signal; Step S300, determining a low-frequency signal according to the magnitude of the signal difference, where the signal difference is the magnitude of the difference between the low-frequency amplitude estimation signal and the DC signal; Step S400, generating a noise reduction signal according to the adjusted low-frequency signal; Step S500, performing environmental noise reduction based on the noise reduction signal. The low frequency can be dynamically determined, so that the DC estimation and the low frequency are dynamically combined. It can control the dynamic to ensure that the path is not saturated while keeping the noise reduction amount basically unchanged.

Description

Technical Field

[0001] The present invention relates to the technical field of audio signal processing, and particularly to an active noise cancellation method, device, circuit, equipment and system. Background Art

[0002] Noise not only affects people's normal life, but is even harmful to hearing. With the increasing severity of noise pollution, technologies for suppressing noise have attracted more attention from researchers. Active Noise Canceller (ANC) is based on the principle of waveform interference, and generates a waveform with the same amplitude and opposite phase as the noise through a secondary path (such as a speaker) for cancellation. In existing active noise cancellation solutions, it is easily affected by low-frequency noise in the environment, resulting in data path saturation, and thus generating special abnormal sounds (such as clicking and puffing sounds). For example, when wearing noise-canceling headphones and shaking the head rapidly, opening or closing the car door while wearing noise-canceling headphones during driving, or passing through a bumpy road while wearing noise-canceling headphones during driving, etc., these scenarios are prone to large-amplitude low-frequency signals with relatively large sound pressure amplitudes. When the noise cancellation filter performs gain, it will cause the data to be unable to be expressed, resulting in path saturation.

[0003] In the prior art, dynamic range control (DRC), a limiter, or reducing gain is used for control. These methods do not distinguish frequency bands, but compress the entire signal amplitude, resulting in a loss of noise cancellation performance within the entire bandwidth. DRC, limiting amplitude, reducing gain, etc. all monitor the amplitude (or power) of the input signal in real time. When the amplitude exceeds the set threshold, DRC controls the gain through the configured attack time / release time, the limiter directly compresses the amplitude, and reducing gain directly reduces the gain on the path to prevent saturation. At the same time, the DRC and limiter solutions will cause the total harmonic distortion (THD) to deteriorate, and abnormal sounds may also be introduced if the parameters are not adjusted well. There are also some methods that switch the noise cancellation filter, resulting in increased complexity and increased memory. It can be seen that in the prior art, the amplitude of the entire bandwidth is compressed, resulting in a decrease in noise cancellation performance across the entire frequency band.

[0004] Therefore, how to ensure that the data path is not saturated while minimizing the impact on the noise cancellation amount has become an urgent technical problem to be solved. Summary of the Invention

[0005] Based on the above situation, the main purpose of the present invention is to provide an active noise cancellation method, device, equipment and system to minimize the impact on the noise cancellation amount while ensuring that the data path is not saturated.

[0006] To achieve the above object, the technical solution adopted by the present invention is as follows:

[0007] In a first aspect, an embodiment of the present invention discloses an active noise reduction method, including:

[0008] Step S100: Obtain an environmental audio signal collected by a noise microphone;

[0009] Step S200: Input the environmental audio signal into a DC estimation module and a low-frequency amplitude estimation module respectively, so that the DC estimation module performs DC estimation on the environmental audio signal to obtain a DC signal, and the low-frequency amplitude estimation module performs low-frequency amplitude estimation on the environmental audio signal to obtain a low-frequency amplitude estimation signal; wherein, the bandwidth of the low-frequency amplitude estimation module is greater than the bandwidth of the DC estimation module;

[0010] Step S300: Dynamically adjust the low-frequency signal according to the magnitude of the signal difference, wherein the bandwidth of the low-frequency signal is positively correlated with the magnitude of the signal difference, and the signal difference is the difference between the low-frequency amplitude estimation signal and the DC signal;

[0011] Step S400: Generate a noise reduction signal according to the adjusted low-frequency signal, and the noise reduction signal is a signal with the same amplitude as the low-frequency signal;

[0012] Step S500: Perform environmental noise reduction based on the noise reduction signal.

[0013] Optionally, in step S300, a weighted sum of the DC signal and the low-frequency amplitude estimation signal is obtained to get the low-frequency signal, wherein the greater the signal difference, the greater the weighting coefficient of the low-frequency amplitude estimation signal.

[0014] Optionally, in step S300, the following method is used to weight to obtain the low-frequency signal:

[0015] dat = (1 - α)*dat0 + α*dat1

[0016]

[0017] αp = abs(dat1 - dat0) / R

[0018] wherein, dat is the low-frequency signal, α and (1 - α) are weighting coefficients, dat0 is the DC signal, dat1 is the low-frequency amplitude estimation signal; αp is the signal difference; R is a constant coefficient used to control the weight configuration value.

[0019] Optionally, in step S300, when the magnitude of the signal difference exceeds a threshold, the DC signal is compensated to obtain the low-frequency signal; when the magnitude of the signal difference does not exceed the threshold, the DC signal is used as the low-frequency signal.

[0020] Optionally, when compensating the DC signal to obtain the low-frequency signal, the following formula is used for compensation to obtain the low-frequency signal:

[0021] dat = dat0 + cmp

[0022] cmp = dat1 - dat0 - sign(dat1 - dat0) * thd

[0023] Wherein, dat is a low - frequency signal, cmp is a compensation value, dat0 is a DC signal, dat1 is a low - frequency amplitude estimation signal, thd is a threshold value, and sign() is a sign - taking function.

[0024] Optionally, in step S300, determine the bandwidth of the DC estimation module according to the magnitude of the signal difference to dynamically adjust the bandwidth of the DC estimation module; use the DC signal as the low - frequency signal, wherein the larger the signal difference, the larger the bandwidth of the DC estimation module.

[0025] Optionally, calculate the bandwidth of the DC estimation module using the following formula:

[0026] BW0 = f(dat1, dat0)

[0027] Wherein, f() is a function that is positively correlated with the signal difference.

[0028] In a second aspect, an embodiment of the present invention discloses an active noise - reduction device, including:

[0029] An audio signal acquisition module, configured to acquire an environmental audio signal collected by a noise microphone;

[0030] A signal estimation module, configured to input the environmental audio signal into a DC estimation module and a low - frequency amplitude estimation module respectively, so that the DC estimation module performs DC estimation on the environmental audio signal to obtain a DC signal, and the low - frequency amplitude estimation module performs low - frequency amplitude estimation on the environmental audio signal to obtain a low - frequency amplitude estimation signal; wherein, the bandwidth of the low - frequency amplitude estimation module is greater than the bandwidth of the DC estimation module;

[0031] A low - frequency signal determination module, configured to dynamically adjust the low - frequency signal according to the magnitude of the signal difference, wherein the bandwidth of the low - frequency signal is positively correlated with the magnitude of the signal difference, and the signal difference is the difference between the low - frequency amplitude estimation signal and the DC signal;

[0032] A noise - reduction signal generation module, configured to generate a noise - reduction signal according to the adjusted low - frequency signal, and the noise - reduction signal is a signal with the same amplitude as the low - frequency signal;

[0033] A noise - reduction module, configured to perform environmental noise reduction based on the noise - reduction signal.

[0034] Optionally, in the low - frequency signal determination module, perform weighted summation on the DC signal and the low - frequency amplitude estimation signal to obtain the low - frequency signal, wherein the larger the signal difference, the larger the weighting coefficient of the low - frequency amplitude estimation signal.

[0035] Optionally, in the low-frequency signal determination module, the low-frequency signal is obtained by weighting in the following manner:

[0036] dat = (1 - α)*dat0 + α*dat1

[0037] αp = abs(dat1 - dat0) / R

[0038]

[0039] where dat is the low-frequency signal, α and (1 - α) are weighting coefficients, dat0 is the DC signal, dat1 is the low-frequency amplitude estimation signal; αp is the signal difference; and R is a constant coefficient used to control the weight configuration value.

[0040] Optionally, in the low-frequency signal determination module, when the magnitude of the signal difference exceeds the threshold, the DC signal is compensated to obtain the low-frequency signal; when the magnitude of the signal difference does not exceed the threshold, the DC signal is used as the low-frequency signal.

[0041] Optionally, when compensating the DC signal to obtain the low-frequency signal, the following formula is used for compensation to obtain the low-frequency signal:

[0042] dat = dat0 + cmp

[0043] cmp = dat1 - dat0 - sign(dat1 - dat0)*thd

[0044] where dat is the low-frequency signal, cmp is the compensation value, dat0 is the DC signal, dat1 is the low-frequency amplitude estimation signal, thd is the threshold, and sign() is the sign function.

[0045] Optionally, in the low-frequency signal determination module, the bandwidth of the DC estimation module is determined according to the magnitude of the signal difference to dynamically adjust the bandwidth of the DC estimation module; the DC signal is used as the low-frequency signal, where the larger the signal difference, the larger the bandwidth of the DC estimation module.

[0046] Optionally, the following formula is used to calculate the bandwidth of the DC estimation module:

[0047] BW0 = f(dat1, dat0)

[0048] where f() is a function that is positively correlated with the signal difference.

[0049] In a third aspect, an embodiment of the present invention discloses an active noise reduction circuit, including:

[0050] A noise microphone for collecting ambient audio signals;

[0051] A DC estimation module, whose input end is connected to the output end of a noise microphone, and the DC estimation module is used to perform DC estimation on an environmental audio signal to obtain a DC signal;

[0052] A low-frequency amplitude estimation module, whose input end is connected to the output end of a noise microphone, and the low-frequency amplitude estimation module is used to perform low-frequency amplitude estimation on an environmental audio signal to obtain a low-frequency amplitude estimation signal;

[0053] A noise reduction signal generation module, whose input ends are respectively connected to the output end of the DC estimation module and the output end of the low-frequency amplitude estimation module, and the low-frequency signal calculation module is used to dynamically adjust a low-frequency signal according to the magnitude of a signal difference and generate a noise reduction signal, wherein the bandwidth of the low-frequency signal is positively correlated with the magnitude of the signal difference, and the signal difference is the difference between the low-frequency amplitude estimation signal and the DC signal.

[0054] Optionally, the noise microphone is a reference microphone or an error microphone.

[0055] In a fourth aspect, an embodiment of the present invention discloses a noise reduction circuit, including: a first noise reduction circuit and a second noise reduction circuit;

[0056] The first noise reduction circuit and the second noise reduction circuit are respectively implemented by using the active noise reduction circuit disclosed in the third aspect; the noise reduction signals output by the noise reduction signal generation modules in the first noise reduction circuit and the second noise reduction circuit are superimposed through an adder; wherein:

[0057] The noise microphone in the first noise reduction circuit is a reference microphone;

[0058] The noise microphone in the second noise reduction circuit is an error microphone.

[0059] In a fifth aspect, an embodiment of the present invention discloses an audio playback device, including:

[0060] A processor, configured to implement the method disclosed in the first aspect.

[0061] In a sixth aspect, an embodiment of the present invention discloses an audio playback device, including:

[0062] The circuit disclosed in the third aspect or the fourth aspect.

[0063] In a seventh aspect, an embodiment of the present invention discloses an audio signal processing system, including: a first audio playback device and a second audio playback device; the first audio playback device and the second audio playback device form a pair of audio playback device pairs,

[0064] The first audio playback device has the device disclosed in the second aspect; the second audio playback device has the device disclosed in the second aspect; or,

[0065] The first audio playback device has the circuit disclosed in the third aspect or the fourth aspect; the second audio playback device has the circuit disclosed in the third aspect or the fourth aspect.

[0066] Optionally, it further includes:

[0067] A sound source device for performing audio data interaction with the first audio playback device and the second audio playback device.

[0068] In an eighth aspect, an embodiment of the present invention discloses a computer-readable storage medium, on which a computer program is stored. The computer program stored in the storage medium is used to be executed to implement the method disclosed in the first aspect.

[0069] In a ninth aspect, an embodiment of the present invention discloses a chip of an audio device, on which an integrated circuit is provided. The integrated circuit is designed to implement the method disclosed in the first aspect, or integrates the circuit disclosed in the third aspect or the fourth aspect.

[0070]

Beneficial effects

[0071] According to an active noise reduction method, device, circuit, equipment and system disclosed in an embodiment of the present invention, after obtaining an environmental audio signal collected by a noise microphone, the environmental audio signal is respectively input into a DC estimation module and a low-frequency amplitude estimation module, so that the DC estimation module performs DC estimation on the environmental audio signal to obtain a DC signal, and the low-frequency amplitude estimation module performs low-frequency amplitude estimation on the environmental audio signal to obtain a low-frequency amplitude estimation signal; wherein, the bandwidth of the low-frequency amplitude estimation module is greater than the bandwidth of the DC estimation module; then, the low-frequency signal is dynamically adjusted according to the magnitude of the signal difference, and the bandwidth of the low-frequency signal is positively correlated with the magnitude of the signal difference, and the signal difference is the difference between the low-frequency amplitude estimation signal and the DC signal; a noise reduction signal is generated according to the low-frequency signal to reduce environmental noise. Since on the one hand, the DC estimation module performs DC estimation to obtain a DC signal, and on the other hand, the low-frequency amplitude estimation module performs low-frequency amplitude estimation signal, and the bandwidth of the low-frequency amplitude estimation module is greater than the bandwidth of the DC estimation module, the low-frequency can be dynamically determined, so that the DC estimation and the low-frequency are dynamically combined together, and it can be ensured that the dynamic control does not saturate the path while keeping the noise reduction amount basically unchanged.

[0072] Other beneficial effects of the present invention will be described in the specific implementation manner through the introduction of specific technical features and technical solutions. Those skilled in the art should be able to understand the beneficial technical effects brought by the technical features and technical solutions through these introductions. Description of the drawings

[0073] The following will describe embodiments according to the present invention with reference to the drawings. In the drawings:

[0074] Figure 1Flowchart of a noise reduction method for anti-low-frequency saturation disclosed in this embodiment;

[0075] Figure 2 Schematic diagram of the circuit structure of a noise reduction circuit for anti-low-frequency saturation disclosed in this embodiment;

[0076] Figure 3 Schematic diagram of the structure of a DC removal module disclosed in this embodiment;

[0077] Figure 4 Schematic diagram of an example of the output of DC estimation and low-frequency amplitude estimation in this embodiment;

[0078] Figure 5 Schematic diagram of an example of signal input and output in the first embodiment disclosed in this embodiment;

[0079] Figure 6 Schematic diagram of an example of the output signal of the DC removal module in the first embodiment disclosed in this embodiment;

[0080] Figure 7 Schematic diagram of the time-domain comparison between the first embodiment and the traditional limiter scheme;

[0081] Figure 8 Schematic diagram of the frequency-domain comparison between the first embodiment and the traditional limiter scheme;

[0082] Figure 9 Schematic diagram of an example of the low-frequency output of the second embodiment;

[0083] Figure 10 Schematic diagram of the structure of a noise reduction device for anti-low-frequency saturation disclosed in this embodiment. Detailed implementation manners

[0084] In order to minimize the impact on the noise reduction amount while ensuring that the data path is not saturated, this embodiment discloses an active noise reduction method. Please refer to Figure 1 , which is a flowchart of an active noise reduction method disclosed in this embodiment. The active noise reduction method includes: step S100, step S200, step S300, and step S400, where:

[0085] Step S100: Obtain the environmental audio signal collected by the noise microphone. Please refer to Figure 2, which is a schematic diagram of an active noise reduction circuit structure disclosed in this embodiment. In this embodiment, the noise microphone 1 can be an error microphone or a reference microphone. That is, the execution object of the active noise reduction method in this embodiment can be the ambient audio signal collected by the error microphone or the ambient audio signal collected by the reference microphone. In the specific implementation process, when the signal collected by the noise microphone 1 is an analog signal, it can be converted into a digital signal through an AD converter. In a specific embodiment, after obtaining the ambient audio signal, the ambient audio signal can be input into the DC removal module 2 so that the DC removal module 2 determines the low-frequency signal, and then, a noise reduction signal is generated based on the low-frequency signal. In the specific implementation process, the noise microphone is exposed to a complex environment. For example, when wearing headphones with a noise reduction system and passing through a bumpy road surface, or at the moment of opening or closing a car door, it will cause the noise microphone to output a large-amplitude low-frequency signal. When using an AD converter to convert it into a digital signal, it is required that the AD converter can receive large dynamic signals in common environments to ensure that the signal does not saturate in the AD module.

[0086] Step S200, input the ambient audio signal into the DC estimation module and the low-frequency amplitude estimation module respectively. Please refer to Figure 3 , which is a schematic diagram of a DC removal module structure disclosed in this embodiment. The DC removal module 2 is a zero-phase DC remover that subtracts the DC value (or low-frequency value) from the original signal. Specifically, the DC removal module 2 includes a DC estimation module 21, a low-frequency amplitude estimation module 22, and a low-frequency signal calculation module 23. In this embodiment, after obtaining the ambient audio signal, the ambient audio signal can be divided into two paths, one path is input into the DC estimation module 21, and the other path is input into the low-frequency amplitude estimation module 22. Thus, the DC estimation module can perform DC estimation on the ambient audio signal to obtain the DC signal dat0, and the low-frequency amplitude estimation module can perform low-frequency amplitude estimation on the ambient audio signal to obtain the low-frequency amplitude estimation signal dat1. In the specific implementation process, the DC estimation module 21 performs low-pass filtering on the original signal, and optionally uses a first-order IIR for filtering. The filtering bandwidth is marked as BW0, and the output data after filtering is marked as dat0; the low-frequency amplitude estimation module 22 is a low-pass filter with a filtering bandwidth of BW1, and the output data after filtering is marked as dat1. The bandwidth BW1 of the low-frequency amplitude estimation module 22 is greater than the bandwidth BW0 of the DC estimation module 21.

[0087] Step S300, dynamically adjust the low-frequency signal dat according to the magnitude of the signal difference. Specifically, the bandwidth of the low-frequency signal dat is positively correlated with the magnitude of the signal difference. In this embodiment, the so-called signal difference is the difference between the low-frequency amplitude estimation signal dat1 and the DC signal dat0. Specifically, the magnitude of this difference is the absolute value of the difference between the low-frequency amplitude estimation signal dat1 and the DC signal dat0, that is, abs(dat1 - dat0). Specifically, please refer toFigure 3 The low-frequency signal dat can be determined by the low-frequency signal calculation module 23. In this embodiment, when the low-frequency signal is small, the difference between the DC signal dat0 and the low-frequency amplitude estimation signal dat1 is not significant, so abs(dat1 - dat0) is close to 0. When determining the low-frequency signal dat, the low-frequency signal dat can be close to the DC signal dat0; conversely, when the low-frequency signal is large, abs(dat1 - dat0) is larger, and when determining the low-frequency signal dat, the low-frequency signal dat can be close to the low-frequency amplitude estimation signal dat1. Thus, the low-frequency signal dat can be dynamically determined.

[0088] Step S400: Generate a noise reduction signal based on the adjusted low-frequency signal dat. In this embodiment, the noise reduction signal is a signal with the same amplitude as the low-frequency signal dat. It should be noted that in the specific implementation process, there may be a certain difference between the amplitude of the noise reduction signal and the amplitude of the low-frequency signal dat, that is, a certain error is allowed.

[0089] Step S500: Perform environmental noise reduction based on the noise reduction signal. Please refer to Figure 2 The environmental noise can be reduced according to the low-frequency signal dat through the noise reduction filter 3. Specifically, the environmental noise can be reduced by the active noise cancellation (ANC) method based on the waveform interference principle. Specifically, a waveform with the same amplitude and opposite phase to the noise is generated through the secondary path (such as a speaker) for cancellation.

[0090] In this embodiment, the signal output by the AD converter may contain a DC component. The common DC component may be introduced by the microphone or the AD converter circuit mismatch. The signal output by the AD converter is connected to the DC removal module 2 to remove the influence of the DC component on the subsequent modules. At the same time, the DC removal module 2 adaptively adjusts the attenuation of the low-frequency signal according to the magnitude of the low-frequency quantity to ensure that the subsequent path will not saturate and prevent abnormal noises in complex environments (such as bumpy roads, passing over speed bumps in a car, opening and closing doors).

[0091] For the convenience of those skilled in the art to understand, in the specific implementation process, the DC estimation can use a first-order IIR low-pass filter, expressed as:

[0092] dat0[n] = (1 - β0)*dat0[n - 1] + β0*dat_in[n]

[0093] where β0 is a parameter controlling the DC component bandwidth, 0 < β0 < 1, n is a discrete sequence, and dat_in is the input signal of the DC estimation;

[0094] The low-frequency amplitude estimation also uses a first-order IIR, which is a low-pass filter with a larger bandwidth than the DC filter. That is

[0095] dat1[n] = (1 - β1) * dat1[n - 1] + β1 * dat_in[n]

[0096] Among them, β1 is a parameter for controlling the DC bandwidth, and 0 < β1 < 1. In the present invention, 0 < β0 < β1 < 1 is set, so as to ensure that the dat1 data has a larger bandwidth.

[0097] Please refer to Figure 4 , which is an example schematic diagram of the output of DC estimation and low-frequency amplitude estimation disclosed in this embodiment. In the figure, the input signal is dat_in, the DC estimation output signal is dat0, and the low-frequency amplitude estimation output signal is dat1. Figure 4 In, the data sampling rate is 48k, β0 = 1 / 2048, and β1 = 1 / 128. It can be seen that dat1 outputs more low-frequency components, while dat0 is close to the DC component. Therefore, by using the difference between dat1 and dat0, it can be analyzed whether the input signal contains too many low-frequency signals. The low-frequency signal dat is a signal that adaptively adjusts the low-frequency bandwidth according to the estimated DC dat0 and the low-frequency amplitude value dat1.

[0098] In the first embodiment, in step S300, a weighted sum of the DC signal dat0 and the low-frequency amplitude estimation signal dat1 is performed to obtain the low-frequency signal dat. Among them, the larger the signal difference, the larger the weighting coefficient of the low-frequency amplitude estimation signal dat1. Specifically, the low-frequency signal dat can be weighted in the following manner:

[0099] dat = (1 - α) * dat0 + α * dat1

[0100]

[0101] αp = abs(dat1 - dat0) / R

[0102] Among them, dat is the low-frequency signal, α and (1 - α) are weighting coefficients, dat0 is the DC signal, and dat1 is the low-frequency amplitude estimation signal; αp is the signal difference; R is a constant coefficient, which is an empirical value used to control the weight configuration value. Specifically, it adjusts the weights of dat0 and dat1. In this embodiment, when the low-frequency signal is small, the difference between the data dat0 and the data dat1 is not large, so abs(dat1 - dat0) is close to 0, and the low-frequency signal dat is close to dat0. When the low-frequency signal is large, the larger abs(dat1 - dat0) is, the closer dat is to dat1. It can be seen from the formula that when R is fixed, the larger the difference between dat1 and dat0, the larger α is, and finally the weight of the output dat1 is larger. When the difference between dat1 and dat0 is fixed, the smaller R is adjusted, the larger a is, and the weight of the output dat1 is also larger.

[0103] Please refer to Figure 5 , which is a schematic diagram of the signal input and output example of the first embodiment disclosed in this embodiment. As Figure 5 shown, the output of the low-frequency signal is configured when R = 1.6. It can be seen that when the difference between dat1 and dat0 is larger (the low-frequency jitter part), the low-frequency output signal dat is closer to dat1; when dat1 and dat0 are closer, the dat output is closer to dat0 (the part after sampling point 3500); finally, the output of the DC removal module is as Figure 6 shown, which is a schematic diagram of the output signal example of the DC removal module of the first embodiment disclosed in this embodiment. The light dashed line is the input signal, which includes the DC component. The thick dashed line is the output of the DC removal module of the traditional scheme. It can be seen that the function of DC removal is realized. The solid line is the active DC removal module proposed in the first embodiment. It can be seen that while removing the DC, a larger low-frequency component is subtracted at the moment when the low frequency is too large, and the overall amplitude is reduced, reducing the possibility of path saturation.

[0104] Please refer to Figure 7 , which is a schematic diagram of the time-domain comparison between the first embodiment and the traditional limiter scheme. The solid line is the input signal, and the light dashed line is the traditional limiter scheme. It can be seen that the limiter compresses the amplitude exceeding the threshold. Although the overall signal amplitude is controlled, the signals in all frequency bands are compressed. The dotted-dashed line is the processing effect of the first embodiment. It can be seen that the low-frequency signal is attenuated, but the high-frequency component is not attenuated.

[0105] Please refer to Figure 8 , which is a schematic diagram of the frequency-domain comparison between the first embodiment and the traditional limiter scheme. For the traditional scheme corresponding to the dotted line, both the low-frequency and high-frequency components are compressed relative to the input signal (solid line). Moreover, harmonics are generated. For example, at the position of 2k, when a sine wave signal (±1.5) is input to ±1, the attenuated components also exist in the high frequency, resulting in poor THD performance. However, the proposed scheme only attenuates the low-frequency signal, and the THD remains basically unchanged.

[0106] In the second embodiment, in step S300, when the signal difference magnitude exceeds the threshold thd, the DC signal dat0 is compensated to obtain the low-frequency signal dat; when the signal difference magnitude does not exceed the threshold thd, the DC signal dat0 is used as the low-frequency signal dat. Specifically, when compensating the DC signal dat0 to obtain the low-frequency signal dat, the following formula is used for compensation to obtain the low-frequency signal dat:

[0107] dat = dat0 + cmp

[0108] cmp = dat1 - dat0 - sign(dat1 - dat0) * thd

[0109] That is, in the second embodiment, the following formula can be used to determine the low-frequency signal dat:

[0110]

[0111] cmp = dat1 - dat0 - sign(dat1 - dat0) * thd

[0112] where dat is the low-frequency signal, cmp is the compensation value, dat0 is the DC signal, dat1 is the low-frequency amplitude estimation signal, and thd is the threshold. sign() is the sign function. That is, sign(x) represents taking the sign bit of x. When x is positive, 1 is output; when x is negative, -1 is output. Thus, through the sign operation of the difference and the threshold, a smooth transition can be ensured.

[0113] In this embodiment, the low-frequency signal calculation module calculates the compensation value by determining whether the absolute value of the difference between dat0 and dat1 exceeds the threshold thd. When the absolute value exceeds the threshold, it indicates that the current low-frequency signal is too large and needs to be attenuated; when the absolute value is less than the threshold, it indicates that the low-frequency signal is normal, and the low-frequency signal is directly equal to the DC estimation value dat0.

[0114] Please refer to Figure 9 , which is a schematic diagram of the low-frequency output example of the second embodiment. More obviously, the low-frequency output signal effectively adapts and outputs according to the relationship between dat1 and dat0, achieving the effect of effectively controlling the dynamics of the low frequency.

[0115] In the third embodiment, in step S300, the bandwidth BW0 of the DC estimation module is determined according to the magnitude of the signal difference to dynamically adjust the bandwidth BW0 of the DC estimation module; the DC signal dat0 is used as the low-frequency signal dat. Among them, the greater the signal difference, the greater the bandwidth BW0 of the DC estimation module. Specifically, the following formula can be used to calculate the bandwidth BW0 of the DC estimation module:

[0116] BW0 = f(dat1, dat0)

[0117] where f() is a function that is positively correlated with the signal difference. As an example, the f() function can adopt a function related to the absolute value of the difference between the above-mentioned dat1 and dao0. That is, the bandwidth BW0 is controlled according to the absolute value of the difference between dat1 and dao0. When the absolute value of the difference between dat1 and dat0 is larger, BW0 is larger; when the absolute value of the difference between dat1 and dat0 is smaller, BW0 is smaller.

[0118] This embodiment also discloses an active noise reduction device. Please refer to Figure 10, which is a schematic structural diagram of an active noise reduction device disclosed in this embodiment. The active noise reduction device includes: an audio signal acquisition module 100, a signal estimation module 200, a low-frequency signal determination module 300, a noise reduction signal generation module 400, and a noise reduction module 500, where:

[0119] The audio signal acquisition module 100 is used to acquire the environmental audio signal collected by the noise microphone;

[0120] The signal estimation module 200 is used to input the environmental audio signal into a DC estimation module and a low-frequency amplitude estimation module respectively, so that the DC estimation module performs DC estimation on the environmental audio signal to obtain a DC signal dat0, and the low-frequency amplitude estimation module performs low-frequency amplitude estimation on the environmental audio signal to obtain a low-frequency amplitude estimation signal dat1; wherein, the bandwidth of the low-frequency amplitude estimation module is greater than the bandwidth of the DC estimation module;

[0121] The low-frequency signal determination module 300 is used to dynamically adjust the low-frequency signal dat according to the signal difference magnitude, where the bandwidth of the low-frequency signal dat is positively correlated with the signal difference magnitude, and the signal difference is the difference between the low-frequency amplitude estimation signal dat1 and the DC signal dat0;

[0122] The noise reduction signal generation module 400 is used to generate a noise reduction signal according to the adjusted low-frequency signal dat, and the noise reduction signal is a signal with the same amplitude as the low-frequency signal dat;

[0123] The noise reduction module 500 is used to reduce the environmental noise based on the noise reduction signal.

[0124] Optionally, in the low-frequency signal determination module 300, the DC signal dat0 and the low-frequency amplitude estimation signal dat1 are weighted and summed to obtain the low-frequency signal dat, where the greater the signal difference, the greater the weighting coefficient of the low-frequency amplitude estimation signal dat1.

[0125] Optionally, in the low-frequency signal determination module 300, the low-frequency signal dat is weighted in the following manner:

[0126] dat = (1 - α) * dat0 + α * dat1

[0127] αp = abs(dat1 - dat0) / R

[0128]

[0129] where, dat is the low-frequency signal, α and 1 - α are weighting coefficients, dat0 is the DC signal, dat1 is the low-frequency amplitude estimation signal; αp is the signal difference; R is a constant coefficient used to control the weight configuration value.

[0130] Optionally, in the low-frequency signal determination module 300, when the magnitude of the signal difference exceeds the threshold thd, the DC signal dat0 is compensated to obtain the low-frequency signal dat; when the magnitude of the signal difference does not exceed the threshold thd, the DC signal dat0 is used as the low-frequency signal dat.

[0131] Optionally, when compensating the DC signal dat0 to obtain the low-frequency signal dat, the following formula is used for compensation to obtain the low-frequency signal dat:

[0132] dat = dat0 + cmp

[0133] cmp = dat1 - dat0 - sign(dat1 - dat0) * thd

[0134] where dat is the low-frequency signal, cmp is the compensation value, dat0 is the DC signal, dat1 is the low-frequency amplitude estimation signal, thd is the threshold, and sign() is the sign function.

[0135] Optionally, in the low-frequency signal determination module 300, the bandwidth BW0 of the DC estimation module is determined according to the magnitude of the signal difference to dynamically adjust the bandwidth BW0 of the DC estimation module; the DC signal dat0 is used as the low-frequency signal dat, where the greater the signal difference, the greater the bandwidth BW0 of the DC estimation module.

[0136] Optionally, the following formula is used to calculate the bandwidth BW0 of the DC estimation module:

[0137] BW0 = f(dat1, dat0)

[0138] where f() is a function that is positively correlated with the signal difference.

[0139] This embodiment also discloses an active noise cancellation circuit. Please refer to Figure 2 and Figure 3 , the active noise cancellation circuit includes: a noise microphone 1 and a DC removal module 2, where the DC removal module 2 includes a DC estimation module 21, a low-frequency amplitude estimation module 22, and a noise cancellation signal generation module 23, where:

[0140] The noise microphone 1 is used to collect the environmental audio signal. The noise microphone 1 can be an error microphone or a reference microphone. It should be noted that the DC removal module 2 should perform DC removal processing on the error microphone or the reference microphone.

[0141] The input end of the DC estimation module 21 is connected to the output end of the noise microphone. The DC estimation module is used to perform DC estimation on the environmental audio signal to obtain the DC signal dat0;

[0142] The input end of the low-frequency amplitude estimation module 22 is connected to the output end of the noise microphone. The low-frequency amplitude estimation module is used to estimate the low-frequency amplitude of the environmental audio signal to obtain the signal dat1.

[0143] The input ends of the noise reduction signal generation module 23 are respectively connected to the output end of the DC estimation module and the output end of the low-frequency amplitude estimation module. The low-frequency signal calculation module is used to dynamically adjust the low-frequency signal dat according to the magnitude of the signal difference and generate a noise reduction signal. Among them, the bandwidth of the low-frequency signal dat is positively correlated with the magnitude of the signal difference, and the signal difference is the difference between the low-frequency amplitude estimation signal dat1 and the DC signal dat0.

[0144] This embodiment also discloses a noise reduction circuit. Please refer to Figure 2 which includes a first noise reduction circuit and a second noise reduction circuit;

[0145] The first noise reduction circuit and the second noise reduction circuit are respectively implemented by using the active noise reduction circuit disclosed in the above embodiment. The noise reduction signals output by the noise reduction signal generation modules in the first noise reduction circuit and the second noise reduction circuit are superimposed through an adder. Among them: the noise microphone in the first noise reduction circuit is a reference microphone; the noise microphone in the second noise reduction circuit is an error microphone. That is, in this embodiment, the two noise reduction circuits are used to perform noise reduction processing on the two noise microphones respectively.

[0146] This embodiment also discloses an audio playback device, including:

[0147] a processor for implementing the method disclosed in the above embodiment.

[0148] This embodiment also discloses an audio playback device, including the circuit disclosed in the above embodiment.

[0149] This embodiment also discloses an audio signal processing system, including a first audio playback device and a second audio playback device; the first audio playback device and the second audio playback device form a pair of audio playback device pairs.

[0150] The first audio playback device has the device disclosed in the above embodiment; the second audio playback device has the device disclosed in the above embodiment; or,

[0151] The first audio playback device has the circuit disclosed in the above embodiment; the second audio playback device has the circuit disclosed in the above embodiment.

[0152] Optionally, it further includes:

[0153] a sound source device for performing audio data interaction with the first audio playback device and the second audio playback device.

[0154] This embodiment also discloses a computer-readable storage medium, on which a computer program is stored. The computer program stored in the storage medium is used to be executed to implement the method disclosed in the above embodiment.

[0155] This embodiment also discloses a chip of an audio device, on which an integrated circuit is provided. The integrated circuit is designed to implement the method disclosed in the above embodiment, or integrates the circuit disclosed in the above embodiment.

[0156] According to an active noise reduction method, device, circuit, equipment and system disclosed in an embodiment of the present invention, after obtaining an environmental audio signal collected by a noise microphone, the environmental audio signal is respectively input into a DC estimation module and a low-frequency amplitude estimation module, so that the DC estimation module performs DC estimation on the environmental audio signal to obtain a DC signal, and the low-frequency amplitude estimation module performs low-frequency amplitude estimation on the environmental audio signal to obtain a low-frequency amplitude estimation signal; then, a low-frequency signal is determined according to the magnitude of the signal difference, where the signal difference is the difference between the low-frequency amplitude estimation signal and the DC signal; a noise reduction signal is generated according to the low-frequency signal to reduce environmental noise. Since on the one hand, the DC estimation module performs DC estimation to obtain a DC signal, and on the other hand, the low-frequency amplitude estimation module performs low-frequency amplitude estimation, the low frequency can be dynamically determined, so that the DC estimation and the low frequency are dynamically combined together, and it can be ensured that the dynamic control keeps the path unsaturated while keeping the noise reduction amount basically unchanged.

[0157] It should be noted that in the present invention, step numbers (letter or number numbers) are used to refer to certain specific method steps, only for the purpose of description convenience and simplicity, and by no means to limit the order of these method steps by letters or numbers. Those skilled in the art can understand that the order of relevant method steps should be determined by the technology itself and should not be unduly restricted due to the existence of step numbers.

[0158] Those skilled in the art can understand that on the premise of no conflict, the above preferred solutions can be freely combined and superimposed.

[0159] It should be understood that the above embodiments are merely exemplary and not restrictive. Without departing from the basic principles of the present invention, various obvious or equivalent modifications or substitutions made by those skilled in the art to the above details will all be included within the scope of the claims of the present invention.

Claims

1. An active noise reduction method, characterized in that, Including: Step S100: Obtain the environmental audio signal collected by the noise microphone; Step S200: Input the environmental audio signal into a DC estimation module and a low-frequency amplitude estimation module respectively, so that the DC estimation module performs DC estimation on the environmental audio signal to obtain a DC signal dat0, and the low-frequency amplitude estimation module performs low-frequency amplitude estimation on the environmental audio signal to obtain a low-frequency amplitude estimation signal dat1; wherein, the bandwidth of the low-frequency amplitude estimation module is greater than the bandwidth of the DC estimation module; Step S300: Dynamically adjust the low-frequency signal dat according to the signal difference magnitude, wherein the bandwidth of the low-frequency signal dat is positively correlated with the signal difference magnitude, and the signal difference is the difference magnitude between the low-frequency amplitude estimation signal dat1 and the DC signal dat0; Step S400: Generate a noise reduction signal according to the adjusted low-frequency signal dat, and the noise reduction signal is a signal with the same amplitude as the low-frequency signal dat; Step S500: Perform environmental noise reduction based on the noise reduction signal.

2. The active noise reduction method according to claim 1, characterized in that In the step S300, perform weighted summation on the DC signal dat0 and the low-frequency amplitude estimation signal dat1 to obtain the low-frequency signal dat, wherein the greater the signal difference, the greater the weighting coefficient of the low-frequency amplitude estimation signal dat1.

3. The active noise reduction method according to claim 2, wherein In the step S300, the low-frequency signal dat is weighted in the following manner: wherein, dat is the low-frequency signal, , 1 - are weighting coefficients, dat0 is the DC signal, and dat1 is the low-frequency amplitude estimation signal; is the signal difference; R is a constant coefficient for controlling the weight configuration value.

4. The active noise reduction method according to claim 1, characterized in that, In the step S300, when the signal difference magnitude exceeds a threshold thd, compensate the DC signal dat0 to obtain the low-frequency signal dat; when the signal difference magnitude does not exceed the threshold thd, use the DC signal dat0 as the low-frequency signal dat.

5. The active noise reduction method according to claim 4, characterized in that, When compensating the DC signal dat0 to obtain the low-frequency signal dat, the following formula is used for compensation to obtain the low-frequency signal dat: Wherein, dat is the low-frequency signal, cmp is the compensation value, dat0 is the DC signal, dat1 is the low-frequency amplitude estimation signal, thd is the threshold, and sign() is the sign bit function.

6. The active noise reduction method according to claim 1, characterized in that, In the step S300, determine the bandwidth BW0 of the DC estimation module according to the signal difference magnitude to dynamically adjust the bandwidth BW0 of the DC estimation module; use the DC signal dat0 as the low-frequency signal dat, wherein the greater the signal difference, the greater the bandwidth BW0 of the DC estimation module.

7. The active noise reduction method according to claim 6, wherein The bandwidth BW0 of the DC estimation module is calculated by the following formula: Wherein, f() is a function positively correlated with the signal difference.

8. An active noise reduction device, characterized in that, Including: An audio signal acquisition module (100) for obtaining the environmental audio signal collected by the noise microphone; A signal estimation module (200) is configured to input the environmental audio signal into a DC estimation module and a low-frequency amplitude estimation module respectively, so that the DC estimation module performs DC estimation on the environmental audio signal to obtain a DC signal dat0, and the low-frequency amplitude estimation module performs low-frequency amplitude estimation on the environmental audio signal to obtain a low-frequency amplitude estimation signal dat1; wherein, the bandwidth of the low-frequency amplitude estimation module is greater than the bandwidth of the DC estimation module; A low-frequency signal determination module (300) is configured to dynamically adjust a low-frequency signal dat according to the magnitude of a signal difference, wherein the bandwidth of the low-frequency signal dat is positively correlated with the magnitude of the signal difference, and the signal difference is the difference between the low-frequency amplitude estimation signal dat1 and the DC signal dat0; A noise reduction signal generation module (400) is configured to generate a noise reduction signal according to the adjusted low-frequency signal dat, and the noise reduction signal is a signal with the same amplitude as the low-frequency signal dat; A noise reduction module (500) is configured to perform environmental noise reduction based on the noise reduction signal.

9. The active noise reduction device according to claim 8, wherein, In the low-frequency signal determination module (300), the DC signal dat0 and the low-frequency amplitude estimation signal dat1 are weighted and summed to obtain the low-frequency signal dat, wherein the greater the signal difference, the greater the weighting coefficient of the low-frequency amplitude estimation signal dat1.

10. The active noise reduction device according to claim 9, characterized in that, In the low-frequency signal determination module (300), the low-frequency signal dat is obtained by weighting in the following manner: wherein, dat is the low-frequency signal, , (1 - ) is the weighting coefficient, dat0 is the DC signal, and dat1 is the low-frequency amplitude estimation signal; is the signal difference; R is a constant coefficient used to control the weight configuration value.

11. The active noise reduction device according to claim 8, wherein, In the low-frequency signal determination module (300), when the magnitude of the signal difference exceeds a threshold thd, the DC signal dat0 is compensated to obtain the low-frequency signal dat; when the magnitude of the signal difference does not exceed the threshold thd, the DC signal dat0 is used as the low-frequency signal dat.

12. The active noise reduction device according to claim 11, wherein When compensating the DC signal dat0 to obtain the low-frequency signal dat, the following formula is used for compensation to obtain the low-frequency signal dat: wherein, dat is the low-frequency signal, cmp is the compensation value, dat0 is the DC signal, dat1 is the low-frequency amplitude estimation signal, thd is the threshold, and sign() is the sign function.

13. The active noise reduction device according to claim 8, characterized in that, In the low-frequency signal determination module (300), the bandwidth BW0 of the DC estimation module is determined according to the magnitude of the signal difference to dynamically adjust the bandwidth BW0 of the DC estimation module; the DC signal dat0 is used as the low-frequency signal dat, wherein the greater the signal difference, the greater the bandwidth BW0 of the DC estimation module.

14. The active noise reduction device according to claim 13, wherein The following formula is used to calculate the bandwidth BW0 of the DC estimation module: wherein, f() is a function positively correlated with the signal difference.

15. An active noise reduction circuit, characterized in that, It includes: A noise microphone for collecting an environmental audio signal; A DC estimation module, whose input end is connected to the output end of the noise microphone, and the DC estimation module is configured to perform DC estimation on the environmental audio signal to obtain a DC signal dat0; A low-frequency amplitude estimation module, whose input end is connected to the output end of the noise microphone, and the low-frequency amplitude estimation module is used to estimate the low-frequency amplitude of the environmental audio signal to obtain a signal dat1; wherein, the bandwidth of the low-frequency amplitude estimation module is greater than the bandwidth of the DC estimation module; A noise reduction signal generation module, whose input ends are respectively connected to the output end of the DC estimation module and the output end of the low-frequency amplitude estimation module, and the noise reduction signal generation module is used to dynamically adjust the low-frequency signal dat according to the magnitude of the signal difference and generate a noise reduction signal, wherein the bandwidth of the low-frequency signal dat is positively correlated with the magnitude of the signal difference, and the signal difference is the difference between the low-frequency amplitude estimation signal dat1 and the DC signal dat0.

16. The active noise reduction circuit according to claim 15, wherein, The noise microphone is a reference microphone or an error microphone.

17. A noise reduction circuit, characterized in that, Comprising: A first noise reduction circuit and a second noise reduction circuit; The first noise reduction circuit and the second noise reduction circuit are respectively implemented by using the active noise reduction circuit as described in claim 15; The noise reduction signals output by the noise reduction signal generation module in the first noise reduction circuit and the noise reduction signals output by the noise reduction signal generation module in the second noise reduction circuit are superimposed through an adder; wherein: The noise microphone in the first noise reduction circuit is a reference microphone; The noise microphone in the second noise reduction circuit is an error microphone.

18. An audio playback device, characterized in that, Comprising: A processor for implementing the method as described in any one of claims 1-7.

19. An audio playback device, characterized in that, Comprising: The circuit as described in any one of claims 15-17.

20. An audio signal processing system, comprising: A first audio playback device and a second audio playback device; The first audio playback device and the second audio playback device form a pair of audio playback device pairs, and are characterized in that The first audio playback device has the device as described in any one of claims 8-14; the second audio playback device has the device as described in any one of claims 8-14; or, The first audio playback device has the circuit as described in any one of claims 15-17; the second audio playback device has the circuit as described in any one of claims 15-17.

21. The audio signal processing system according to claim 20, wherein, Further comprising: A sound source device for performing audio data interaction with the first audio playback device and the second audio playback device.

22. A computer-readable storage medium having a computer program stored thereon, characterized in that, The computer program stored in the storage medium is used to be executed to implement the method as described in any one of claims 1-7.

23. A chip of an audio device, on which there is an integrated circuit, characterized in that, The integrated circuit is designed to implement the method as described in any one of claims 1-7, or integrates the circuit as described in any one of claims 15-17.

Images