METHOD, DEVICE, HEADPHONES AND COMPUTER PROGRAM FOR ACTIVE INTERFERENCE NOISE REDUCTION

Description

[0001] The present invention relates to a method for active noise cancellation. The present invention further relates to a device for implementing the method. Furthermore, the invention relates to a headset configured to carry out a method according to the invention or having a device according to the invention, as well as to a computer program with instructions that cause a computer to execute the steps of the method.

[0002] The high levels of noise pollution caused by airplanes, trains, or cars, for example, and perceived as ambient noise by people outside or inside these vehicles, can lead to stress and even serious mental and physical illnesses in those affected. For this reason, active noise cancellation (ANC) technologies are known to reduce such disruptive ambient noise, making them an important feature for headphones or so-called hearables.

[0003] This involves artificially generating an additional sound signal that matches the disturbing sound as closely as possible, but with opposite polarity. The two sound signals are then superimposed using destructive interference to cancel out the background noise as much as possible. For this purpose, headphones with active noise cancellation are measured using one or more microphones integrated into the headphones. The acoustic transfer function of the headphones is then used to calculate the amount of ambient noise that would remain at the ear. To compensate for this amount, a signal of the opposite polarity is generated in the headphones and output via a loudspeaker, which also reproduces the desired sound.Modern ANC headphones typically use fixed feedforward and feedback filters, allowing attenuation of up to 30 dB at low frequencies. However, the filter performance depends sensitively on the fit of the headphones and the user's ear shape. Adaptive algorithms can also be considered to improve the level of noise cancellation. However, such adaptive algorithms require high processing power and are therefore currently unsuitable for headphones, hearables, or hearing aids.

[0004] Most commercially available ANC headphones are equipped with a built-in speaker and two microphones. One of the microphones is directed towards the headphones' surroundings to measure a reference signal in the form of ambient noise and is often referred to as the reference microphone. The other microphone is directed towards the user's ear canal or eardrum to detect an internal error signal and is also referred to as the error microphone. The acoustic transmission from the external reference microphone to the internal error microphone is referred to as the primary path, and the transmission from the speaker to the error microphone is referred to as the secondary path.

[0005] Measuring these primary and secondary paths enables a customized design and thus significantly improves the performance and robustness of an ANC system. The secondary path can be measured using the speaker and the internal microphone, with the signal-to-noise ratio at the internal microphone being quite high due to the passive isolation of the headphones. Measuring the primary path, on the other hand, requires an additional external speaker setup and a suitable measurement environment, making it complex and difficult for end users to perform.

[0006] Against this background, it is an object of the invention to provide an improved method and an improved device for active noise suppression, in particular for suppressing disturbing ambient noise in headphones, as well as a corresponding headphone and a computer program for carrying out the method.

[0007] This object is achieved by a method having the features of claim 1, a corresponding device according to claim 8, a corresponding headset according to claim 10 and a computer program according to claim 11. Preferred embodiments of the invention are the subject of the dependent claims.

[0008] The invention makes use of the knowledge that, particularly in the case of in-ear headphones, but also in headphones with other designs, there can be a significant correlation between the frequency spectra of primary and secondary paths and that this can be exploited to achieve an optimization of the noise suppression even without measuring the primary path.

[0009] According to the inventors' knowledge, in the inventive method for active noise suppression, a transfer function for a secondary path between a loudspeaker and an error microphone is measured. Based on the measured transfer function for the secondary path, a transfer function for a primary path between a reference microphone and the error microphone is estimated. Then, based on the estimated transfer function for the primary path, filter coefficients are determined for filtering to generate the cancellation signal.

[0010] In particular, at least one reference microphone detects noise signals, a loudspeaker outputs a cancellation signal and an error microphone detects the remaining residual signal after the cancellation signal is superimposed with the noise signal.

[0011] According to one embodiment of the invention, the active noise suppression is carried out during the reproduction of a useful audio signal by means of a headphone, wherein one or more reference microphones are located on the outside of the headphone and the error microphone is located on the inside of the headphone.

[0012] Preferably, the transfer function for the secondary path is individually measured for a user and an individual transfer function for the primary path is estimated for the user based on the individually measured transfer function for the secondary path.

[0013] Advantageously, the filtering is carried out using a forward-looking FIR filter or IIR filter.

[0014] According to a further embodiment of the invention, an estimation function for the primary path is determined by measuring and analyzing both the transfer function for the secondary path and the transfer function for the primary path in advance in a training process for different persons and / or fits of the headphones.

[0015] It is advantageous if For measured values ​​in frequency ranges of the transfer functions where deterministic changes for the primary path and the secondary path are present, a principal component analysis is carried out with subsequent dimensionality reduction of the measured values ​​obtained in the training process; based on principal components and mean values ​​determined by the principal component analysis, complex gain vectors for the primary paths and the secondary paths are determined; and a linear mapping is determined which minimizes the error between the determined and the estimated gain vectors of the primary paths.

[0016] Accordingly, a device according to the invention for active noise suppression comprises at least one reference microphone; a loudspeaker; an error microphone; a digital filter for generating a cancellation signal; a digital signal processor configured to generate a measurement signal that can be output via the loudspeaker and to evaluate a signal detected by the error microphone, in order to measure a transfer function for a secondary path between the loudspeaker and the error microphone; to estimate a transfer function for a primary path between the reference microphone and the error microphone based on the measured transfer function for the secondary path; and to adapt filter coefficients for the digital filter based on the estimated transfer function for the primary path.

[0017] According to one embodiment of the invention, the digital filter is designed as an FIR filter or IIR filter.

[0018] The invention also relates to a headset configured to carry out the method according to the invention or having a device according to the invention, as well as to a computer program with instructions that cause a computer to carry out the steps of the method according to the invention.

[0019] Further features of the present invention will become apparent from the following description and claims in conjunction with the figures. Fig. 1 schematically shows an in-ear headphone with an acoustic primary and secondary path; Fig. 2 shows a flow diagram of the inventive method for active noise cancellation; Fig. 3 shows a block diagram of a headphone according to the invention; Fig. 4 shows spectra of measured primary paths (a) and secondary paths (b); Fig. 5 shows a) spectra measured based on the individual secondary paths and the average primary path and b) spectra measured based on the individual secondary paths and the respective estimated primary path of the active transfer function from the reference microphone to the error microphone; Fig. 6 shows the median of the primary path | P ( z )| and the spectrum | H ( z )| for different primary path estimates; Fig. 7 shows a box plot for the energy ratio for different primary path estimates; and Fig. 8 schematically shows the use of a headset in conjunction with an external computing device.

[0020] To better understand the principles of the present invention, embodiments of the invention are explained in more detail below with reference to the figures. It is understood that the invention is not limited to these embodiments and that the described features may also be combined or modified without departing from the scope of the invention as defined in the claims.

[0021] The method according to the invention can be used in particular for active noise suppression in in-ear headphones, as in Figure 1schematically shown. The in-ear headphones 10 are positioned on the user's ear, with an earpiece 14 of the in-ear headphones inserted into the external auditory canal 15 to hold it in place. Depending on the individual fit in the auditory canal, the earpiece can partially shield external noise, so that this noise then reaches the user's eardrum 16 only at a reduced level.

[0022] A background noise signal x(t) arriving at the headphones from the environment is recorded by a reference microphone 11 directed away from the ear canal. The in-ear headphones 10 also have an error microphone 12 directed toward the ear canal 15 and a loudspeaker 13 located near the error microphone 12. A cancellation signal ŷ(t) can be output by means of the loudspeaker 13. The error microphone 12 records the remaining residual signal e(t) after the cancellation signal ŷ(t) has been superimposed on the background noise signal x(t). The primary acoustic path P a (s) describes the transfer function from the reference microphone 11 to the error microphone 12, while the secondary acoustic path S a (s) describes the transfer function from the loudspeaker 13 to the error microphone 12.The in-ear headphones shown have only one reference microphone, but several reference microphones can also be used, each of which then has a separate primary path.

[0023] Figure 2shows a schematic of the basic concept for a method for active noise cancellation, as can be carried out, for example, with such in-ear headphones. In a first step 20, a transfer function for a secondary path between the loudspeaker and the error microphone is measured. In a subsequent step 21, based on the measured transfer function for the secondary path, a transfer function for a primary path between the reference microphone and the error microphone is estimated. For this purpose, the relationships between the primary path and secondary path for the headphones in question, determined in a training phase described below, are used. The estimated transfer function then allows filter coefficients for a filter for generating the cancellation signal to be determined in a further step 22.In this way, the filter can then be adjusted so that the output cancellation signal enables the best possible compensation of the interference signal. After determining the filter coefficients by measuring the secondary path and subsequently estimating the primary path, the filter can then be used unchanged until further notice to prevent or at least reduce the impairment of the user's perception due to interference when a useful audio signal is played back via in-ear headphones. Likewise, the user may find interference suppression more pleasant even without the playback of a useful audio signal, for example, when traveling by train or plane, which reduces the volume level.

[0024] Figure 3 shows a block diagram of a device according to the invention, wherein the analog unit 30 is connected to the hardware components of Figure 1is extended by an electronic backend which is connected to the microphones 11, 12 via analog-digital converters 31, 32 and to the loudspeaker 13 via the digital-analog converter 33. The electronic backend comprises a digital filter unit 34 and a processor unit 35.

[0025] The device according to the invention can be fully integrated into an ANC headset or partially integrated into an external device, such as a smartphone. For example, the processor unit 35 can be part of such an external device.

[0026] The processor unit 35 has one or more digital signal processors, but can also contain other types of processors or combinations thereof. The digital filter 34 is designed as a time-invariant FIR forward filter Ŵ(z), which receives the digitally converted interference signal x(n) and generates the cancellation signal ŷ(n). Likewise, the digital filter 34 can also be designed as an IIR filter, usually as a biquad filter. The digital signal processor 35 generates a measurement signal m(n) and evaluates the digitized error signal e(n) in order to measure the secondary path. Furthermore, the filter coefficients of the digital filter Ŵ(z) are adapted by the digital signal processor. For this purpose, instructions are stored in a memory (not shown), which is preferably integrated in the processor unit, which instructions cause the device, when executed by the processor unit, to carry out the steps according to the method according to the invention.

[0027] The overall transfer function H(s) describes the transfer function from reference microphone 11 to error microphone 12 and, unlike the primary path, includes the influence of the ANC system. The primary path P(z) and the secondary path S(z) contain the influence of the analog-to-digital converter and the digital-to-analog converter, the loudspeaker, and the microphones.

[0028] The total transmission path is then defined as H z = P z − W ^ z S z .

[0029] Here, s and z denote the complex frequency parameters of the Laplace and z-transforms, respectively, and n denotes a discrete time index.

[0030] In the following, we first derive how the filter ratios for the FIR forward filter Ŵ (z) can be selected based on the individually measured secondary path. Next, we present an estimator for the primary path that is trained based on a series of previously measured primary and secondary paths. After the training phase, this estimator can then be fed with measured values ​​from an individual secondary path to estimate the individual primary path.

[0031] It is T = p j , s j ∈ ℝ L j = 1 , … , J the set of measured impulse responses of length L. The optimal FIR forward filter ŵ minimizes the average energy of the entire transmission path, as defined by the following cost function: C w = ∑ j ∈ T p j 0 − s j w 2 with the zero-extended primary path vector p j 0 and convolution matrix s j for the secondary path.

[0032] The optimal FIR forward filter ŵ with respect to the average is given by w ^ = arg min w C w = ∑ j ∈ T s j T s j − 1 ∑ i ∈ T s i T p i 0

[0033] To use the FIR forward filter ŵ However, to optimize individually, precise knowledge of the respective primary and secondary pathways is required.

[0034] As already mentioned, the individual secondary path can be measured using the speaker and the error microphone located inside the headphone. If the individual secondary paths for all s j in the formula above and the average of the primary paths in , i.e. p ¯ = 1 J ∑ j ∈ T p j as an estimate for p is used, the optimal filter for a given individual secondary path is: w ^ avg = s T s − 1 s T p ¯ 0

[0035] Since both the primary and secondary paths depend on the fit of the headphones and the physiology of the user's ear, this correlation can be used to establish an estimator for an individual primary path based on the characteristics of a measured individual secondary path. To do this, the frequency ranges of the transfer functions affected by deterministic changes are extracted using window functions Q p (z) and Q s (z) in the z-domain.

[0036] A principal component analysis (PCA) is used to determine the first K p , K s Main components U p , k , U s , k ∈ ℂ L , and the means of a set of windowed, complex frequency domain vectors of the primary path and secondary path are extracted from the set T.

[0037] The complex gain vectors g p,j and g s , jminimize the Euclidean distance between the reconstructed frequency domain vectors based on the principal components and the frequency domain vectors of the primary and secondary paths. A linear mapping a ^ ∈ ℂ K p × K s used to determine the gain vectors g p,j of the primary path to the gain vectors of the secondary path g s , j projected.

[0038] After the individual secondary path has been measured, the window function Q s (z) applied to the measured secondary path in the z-range and then the gain vector g s , j for the secondary path using the principal components and the mean of the secondary path. The gain vector is then calculated g p,j for the primary path using the linear mapping â estimated, followed by an estimate of the primary path based on the principal components as well as the mean of the primary path and the estimated gain vector g p,j for the primary path. Finally, by replacing p with the estimation of the individual primary path the individual forward filter.

[0039] The effectiveness of the proposed estimator was verified with simulations, the results of which are presented below. Measurements were performed on 25 subjects and various fits of in-ear headphones, using a sampling rate of 48 kHz. measured primary and secondary paths comprises a total of J=173 pairs of impulse responses.

[0040] Figure 4shows the spectra of the measured primary paths (a) and secondary paths (b). The shaded frequency range 40 indicates the range of the selected frequency range windows. The length of the primary path and secondary path was calculated with L =1024, the length of the forward filter is L w = 64. The set of measured primary and secondary paths was randomly split into two subsets, with a training set containing 80% and a validation set containing the other 20% of the set of measured paths. The training set was used to train the estimator as described above. Furthermore, the number of principal components K p = 1 and K s = 3. The performance of the estimator was then checked by examining the overall transfer path h j = p j 0 − s j w ^ validated, with the measurement being repeated 100 times for randomly divided subsets.

[0041] Figure 5 shows the measured magnitude spectra | H (z )|, where the filter design in a) is based on the individual secondary paths and the average primary path, and in b) on the individual secondary paths and the respective estimated primary path. In addition to the median 50, the 50% percentile 52 and 90% percentile 53 of | H ( z )| also the median 51 of the primary path | P ( z )| to indicate the passive attenuation of the headphones.

[0042] Figure 6 shows the median of the primary path | P ( z )|and the spectrum | H ( z )| for various primary path estimates. This is based on H avg ( z ) on the mean of the primary paths of the training set, H est ( z ) is based on a primary path estimate, as is H ppg ( z ), but with a perfect PCA gain vector (PPG) ) g p instead of its estimate, and finally H opt ( z ) on the actual primary path. The shaded area 60, in which | Q p ( z )| > 0, marks the frequency range in which H ( z ) is influenced by the primary path estimator. The figure shows that the median of the spectrum | H ( z )| between 250 Hz and 2.5 kHz by up to 7 dB, approaching the median based on the individual primary path.

[0043] The box plot in Figure 7 shows the energy ratio in dB for the different primary path estimates from Figure 6 (a) mean, b) estimate, c) estimate with PPG, d) optimum with knowledge of the actual primary path). Here, the energy ratio ε the windowed overall transmission path and the primary path using Q p ( z ) defined as ε = ∮ H q , j z 2 dz ∮ P q , j z 2 dz

[0044] For the various primary path estimates, the median as well as the minimum, the so-called lower whisker, and the maximum, the so-called upper whisker, are shown as horizontal lines, and the lower quartile and upper quartile are shown as rectangles surrounding the median.

[0045] As can be seen from the figure, the energy ratio is reduced ε Compared to using the mean value (a), when using the estimator (b) the median is 3.1 dB, while the difference between the maximum values, the so-called upper whiskers, is 5.0 dB.

[0046] Figure 8schematically shows the use of a headset 10, such as a so-called hearable, in conjunction with an external computing device 80. The external computing device 80 can, in particular, be a mobile device suitable for audio playback. For example, a smartphone, a so-called wearable, such as a smartwatch, a fitness bracelet, or data glasses, or a computer tablet can be connected to the headset.

[0047] The devices communicate wirelessly via a radio connection such as Bluetooth. Once the connection is established, audio signals can be transmitted from the external computer device 80 to the headset 10 and then played back in the conventional manner using one or more speakers integrated into the headset.

[0048] In addition, the active noise suppression according to the invention can also be carried out using the external computer device 80. For this purpose, the external computer device 80, particularly when a user uses the headphones 10 for the first time, can transmit a measurement signal to the headphones, which is then output through a loudspeaker integrated into the headphones. An error microphone integrated into the headphones 10 then records the error signal, which is transmitted to the external computer device 80. Based on this, the external computer device 80 calculates the secondary path, estimates the primary path, and subsequently determines the filter coefficients for the filter to generate the cancellation signal. The filter coefficients are then sent via the wireless connection from the external computer device 80 to the headphones 10, where the filter is adjusted accordingly so that noise is largely suppressed during playback of the audio signals.

Claims

1. Method for active noise cancellation in headphones (10), wherein in the method - a transfer function for a secondary path between a loudspeaker and an error microphone is measured (20); - a transfer function for a primary path between a reference microphone and the error microphone is estimated (21) based on the measured transfer function for the secondary path with an estimator for the primary path, wherein the estimator for the primary path is determined in advance in a training process by measuring and analyzing both the transfer function for the secondary path and the transfer function for the primary path; and - filter coefficients for filtering to generate a cancellation signal are determined (22) based on the estimated transfer function for the primary path.

2. The method of claim 1, wherein at least one reference microphone (11) detects noise signals, a loudspeaker (13) emits a cancellation signal and an error microphone (12) detects the remaining residual signal after the cancellation signal has been superimposed with the background noise signal.

3. The method according to claim 2, wherein the active noise cancellation is performed during reproduction of a useful audio signal by means of headphones (10), and one or more reference microphones (11) are located on the outside of the headphones and the error microphone (12) is located on the inside of the headphones.

4. The method according to any of the preceding claims, wherein - the transfer function for the secondary path is measured individually for a user; - an individual transfer function for the primary path is estimated based on the individually measured transfer function for the secondary path for the user.

5. The method according to of any of the preceding claims, wherein the filtering is performed by means of a forward FIR filter or IIR filter (35).

6. The method according to of any of claims 3 to 5, wherein the estimator for the primary path is determined by measuring and analyzing both the transfer function for the secondary path and the transfer function for the primary path in the training process for different people and / or fits of the headphones.

7. The method of claim 6, wherein - for measured values in frequency ranges of the transfer functions, where deterministic changes are present for the primary path and the secondary path, a principal component analysis is performed with subsequent dimension reduction of the measured values obtained in the training process; - based on principal components and mean values determined by the principal component analysis, complex gain vectors are determined for the primary paths and the secondary paths; and - a linear mapping that minimizes the error between the determined and the estimated gain vectors of the primary paths is determined.

8. Device for active noise cancellation in headphones (10), comprising - at least one reference microphone (11); - a loudspeaker (13); - an error microphone (12); - a digital filter (34) for generating a cancellation signal; - a digital signal processor (35) which is arranged - to generate a measurement signal which can be output via the loudspeaker and to evaluate a signal detected by the error microphone in order to measure a transfer function for a secondary path between the loudspeaker and the error microphone; - estimate a transfer function for a primary path between the reference microphone and the error microphone based on the measured transfer function for the secondary path with an estimator for the primary path, wherein the estimator for the primary path is determined in advance in a training process by measuring and analyzing both the transfer function for the secondary path and the transfer function for the primary path; and - adapt filter coefficients for the digital filter based on the estimated transfer function for the primary path.

9. The device according to claim 8, wherein the digital filter (34) is designed as a forward-directed FIR filter or IIR filter.

10. Headphones (10) adapted to perform a method according to any of claims 1 to 7 or a device according to claim 8 or 9.

11. A computer program comprising instructions which cause a computer to perform the steps of a method according to any of claims 1 to 7.