The present disclosure relates to an audio system and to a signal processing method, each for an ear mountable playback device, e.g. a headphone, comprising a speaker.
Nowadays a significant number of headphones, including earphones, employ techniques that enhance the sound experience of a user, such as noise cancellation techniques. For example, such noise cancellation techniques are referred to as active noise control or ambient noise cancellation, both abbreviated with ANC. ANC generally makes use of recording ambient noise that is processed for generating an anti-noise signal, which is then combined with a useful audio signal to be played over a speaker of the headphone. ANC can also be employed in other audio devices like handsets or mobile phones.
Various ANC approaches make use of feedback, FB, microphones, feedforward, FF, microphones or a combination of feedback and feedforward microphones. Efficient FF and FB ANC is achieved by tuning a filter or by adjusting an audio signal, e.g. via an equalizer, based on given acoustics of a system.
Hybrid noise cancellation headphones are generally known. For instance, a microphone is placed inside a volume that is directly acoustically coupled to the ear drum, conventionally close to the front of the headphones driver. This is referred to as the feedback (FB) microphone. A second microphone, the feedforward (FF) microphone, may be placed on the outside of the headphone, such that it is acoustically decoupled from the headphones driver.
For each system to work effectively, the headphone preferably makes a near perfect seal to the ear/head of the user which does not change whilst the device is worn and that is consistent for any user. Any change in this seal as a result of a poor fit will change the acoustics and ultimately the ANC performance. This seal is typically between the ear cushion and the user's head, or between an earphone's rubber tip and the ear canal wall.
For most noise cancellation headphones and earphones, effort is put into maintaining a consistent fit when being worn and from user to user to ensure that the headphone acoustics do not change and always have a good match to the noise filters. However, “leaky” earphones and headphones, which do not make a seal between the ear cushion/tips and the ear, have a large variation in the acoustics when worn by different people. Furthermore the acoustics can vary for the user whilst the earphone moves in their ear as a result of typical everyday head movements. Therefore, for any headphones or earphones which are leaky, some adaptation is required to ensure that the filters always match the acoustics.
An objective to be achieved is to provide an improved concept for adjusting active noise control algorithms to an acoustic leakage condition of an ear mountable playback device like a headphone, earphone or mobile handset.
This object is achieved with the subject matter of the independent claims. Embodiments and developments of the improved concept are defined in the dependent claims. The improved concept is based on the idea of estimating an acoustic leakage condition in terms of its extent, i.e. determining a degree of acoustic leakage between an ear mountable playback device and the ear canal of the user, during regular usage of said ear mountable playback device. This leakage condition is consequently used to enhance the sound experience of the user, e.g. by removing unwanted portions of a sound signal transmitted to the ear canal of the user via noise control algorithms. Said unwanted portions may be ambient noise with a degree that is dependent on the extent of the acoustic leakage, for instance. In order to achieve a sufficient noise control without attenuating the wanted signal, e.g. an audio signal such as a music signal, the improved concept further employs a compensation filter that matches a driver to FB microphone response transfer function of the ear mountable playback device such that efficient signal subtraction may be performed for optimal noise control results.
In contrast, at present tuning of noise control filters, such as feedforward and feedback filters, for conventional earphones and headsets is only performed once during or at the end of production of the ANC devices, for example by measuring acoustic properties of the device. In particular, tuning is performed during a calibration process with some measurement fixture like an artificial head with a microphone in the ear canal of the artificial head. The measurement, including the playing of some test sound, is coordinated from some kind of processing device which can be a personal computer or the like. To achieve an optimum ANC performance for each ANC device produced, a dedicated measurement has to be performed for each of the ANC devices under control of the processing device, which is time-consuming, especially if larger volumes of ANC devices are to be calibrated.
In the following, the improved concept will be explained, sometimes referring to a headphone or earphone as an example of the playback device. However, it shall be appreciated that this example is not limiting and will also be understood by a skilled person for other kinds of playback devices where different acoustic leakage conditions can occur during usage by a user. In general the term playback device should include all types of audio reproducing devices.
In an embodiment of an audio system according to the improved concept, which is to be used for an ear mountable playback device like a headphone, earphone, mobile phone, handset or the like, this system comprises a speaker that is configured to generate a speaker signal on the basis of an audio output signal. The system further comprises an error microphone that is configured to generate a disturbed audio signal on the basis of ambient noise and the speaker signal. A further microphone of the audio system is configured to generate a noise signal on the basis of the ambient noise. The audio system further comprises a first noise filter that is configured to generate a first compensation signal by applying filter operations to the noise signal, and to be adapted based on a compensated error signal.
The audio system according to the improved concept further comprises a first mixer that is configured to generate the audio output signal by superimposing an audio signal, the first compensation signal and a second compensation signal. A compensation filter of the audio system is configured to generate a third compensation signal by applying filter operations to the audio signal, and to be adapted based on an acoustic leakage condition. A second noise filter is configured to generate the second compensation signal by applying filter operations to an intermediate compensation signal that is generated by subtracting the third compensation signal from the disturbed audio signal.
The audio system further comprises an error compensation unit that is configured to generate the compensated error signal on the basis of the disturbed audio signal and the third compensation signal. Furthermore, the audio system further comprises a detection unit that is configured to estimate the acoustic leakage condition on the basis of a response of the first noise filter or of the disturbed audio signal and the audio output signal.
For example, the speaker of the audio system is arranged in a housing of the playback device such that a first volume is arranged on the preferential side for sound emission of the speaker. The housing may have an opening for coupling the first volume to the ear canal volume of the user. The housing may further comprise a front vent that is covered with an acoustic resistor and couples the first volume to the ambient environment. The front volume will also be coupled to the ambient environment via an acoustic leakage due to an imperfect fit of the earphone to the ear of the user. This acoustic leakage varies from person to person and depends on how the earphone sits in the ear at a specific time. The error microphone is for example a feedback error microphone that is arranged within the first volume such that it detects sound output from the speaker as well as ambient sound, i.e. ambient noise. For example, it is arranged close to the opening.
In addition, a second volume is arranged within the housing on the side of the speaker facing away from the preferential side for sound emission. The second volume is acoustically coupled to the ambient environment via a rear vent of the housing which may also be covered with an acoustic resistor. The further microphone may for example be a feedforward microphone which is for example arranged outside of the rear volume, i.e. at the outside of the housing, in order to predominantly sense ambient noise.
The first noise filter is a feedforward noise filter, for instance, and is configured to generate the first compensation signal by filtering the noise signal from the feedforward microphone. The feedforward active noise control, FF ANC, algorithm detects the ambient noise outside the headphone via the feedforward microphone, processes it via the first noise filter and provides an anti-noise signal, the first compensation signal, to the speaker such that superposition of the anti-noise signal and the noise signal occurs at the location of the ear in order to produce noise cancellation. In detail, the residual noise ERR at the location of the ear and/or the error microphone can be characterized by
Err=AE−AFFM·F·DE,
wherein AE is the ambient to ear acoustic transfer function, AFFM is the ambient to FF microphone transfer function, F is the FF filter and DE is the driver to ear acoustic transfer function.
In order to minimize the residual noise, an efficient FF ANC requires matching the first noise filter F to a target acoustic response:
The second noise filter is a feedback noise filter, for instance, and is configured to generate the second compensation signal by filtering an intermediate compensation signal that may correspond to the disturbed audio signal from which the wanted signal, e.g. the audio signal, or a signal derived from the wanted signal, i.e. the third compensation signal, is subtracted. In other words, the intermediate compensation signal is composed primarily, if not exclusively, of the portion of the disturbed audio signal that is generated by the ambient noise by means of the error microphone, in the following referred to as the noise portion.
The first mixer is configured to generate the audio output signal that is provided to the speaker by means of superimposing the audio signal, the first compensation signal and the second compensation signal. In this context, the first and the second compensation signal correspond to signals that destructively interfere with the ambient noise between the speaker and the error microphone and/or the ear canal of a user of the ear mountable playback device.
The compensation filter is for example a filter analogous to that described in US 2017/0140746 A1. According to the improved concept, the compensation filter in this disclosure serves a twofold of purposes. For both these purposes, the compensation filter applies filter operations to the audio signal for generating the third compensation signal in such a manner that the third compensation signal is primarily, or exclusively, composed of the portion of the disturbed audio signal that is generated from the speaker signal and detected by the error microphone, in the following referred to as the speaker portion.
Firstly, this provides a music compensation mechanism that compensates for the audio signal being attenuated by the feedback active noise control, FB ANC, algorithm since the second noise filter in this case primarily, or exclusively, applies filter functions to the noise portion of the disturbed audio signal. In detail, the intermediate compensation signal that is provided to the second noise filter is primarily, or exclusively, composed of the noise portion of the disturbed audio signal with, if at all, merely a negligible speaker portion.
Secondly, for a music removal mechanism, the third compensation signal is provided to the error compensation unit, which generates from the disturbed audio signal and from the third compensation signal the compensated error signal. For example, the error compensation unit adapts the third compensation signal to match it to the speaker portion of the disturbed audio signal. In detail, the error compensation unit generates the compensated error signal that comprises the noise portion of the disturbed audio signal and at most merely a negligible contribution of the speaker portion.
The compensated error signal is consequently used to adapt, for example by means of the detection unit or of a tuning unit, a response of the first noise filter in a manner that ambient noise detected by the further microphone can be removed from the disturbed audio signal, i.e. the signal detected by the error microphone, in a more efficient manner by means of a feedforward active noise control algorithm, FF ANC, as described above, for instance. To this end, an exact matching of the response of the first noise filter for efficient FF ANC hence requires a near-to-perfectly compensated error signal that only comprises the noise contribution of the disturbed audio signal.
In reality, the acoustic transfer functions can change depending on the headphones fit. For leaky earphones, which have a highly variable leak acoustically coupling the front volume to the ambient environment, the transfer functions AE, DE and the acoustic transfer function from the driver to the error microphone, DFBM, change substantially such that it is necessary to adapt at least the first noise filter and optionally also the second noise filter in response to the acoustic signals in the ear canal to minimize the error.
From the adapted response of the first noise filter or from the driver to error microphone transfer function, the acoustic leakage condition can be detected and estimated by means of the detection unit. For example, the detection unit is configured to compare the audio output signal to the disturbed audio signal and to estimate the acoustic leakage condition based on the result of the comparison, e.g. based on a deviation between the two signals.
Alternatively or in addition, the detection unit is configured to monitor a response of the first noise filter and to estimate the acoustic leakage condition based on said response. For example, the detection unit is configured to compare the response of the first noise filter to predetermined responses for estimating the acoustic leakage condition.
The acoustic leakage condition is consequently used to adapt the compensation filter. For example, a response of the compensation filter is adapted according to a current or to a changing acoustic leakage. For example, the response of the compensation filter is configured to match an acoustic leakage dependent driver response between the speaker and the error microphone. This way, efficient noise control algorithms as explained above can be realized in order to enhance the sound experience of a user of the ear mountable playback device.
In some embodiments, the error compensation unit comprises a second mixer that is configured to generate the compensated error signal by subtracting from the disturbed audio signal a removal signal that is based on the third compensation signal.
In order to match the third compensation signal as close as possible to the noise portion of the disturbed audio signal, the error compensation unit in these embodiments is configured to further adjust the third compensation signal to achieve a better match to the driver to error microphone response, for example by means of applying further filter functions.
In some embodiments, the error compensation unit further comprises a filter element that is configured to generate the removal signal from the third compensation signal. Furthermore, for generating the removal signal, the filter element can be configured to apply filter operations to the third compensation signal. Alternatively or in addition, for generating the removal signal, the error compensation unit can be configured to control an adjustable gain of the filter element depending on the third compensation signal and the compensated error signal.
For example, the error compensation unit comprises a feedback loop that is configured to control the filter element, e.g. an adjustable gain and/or a response of the filter element, based on a deviation between the compensated error signal and the third compensation signal. This enables efficient matching of the third compensation signal to the speaker portion of the disturbed audio signal. This way, the noise portion of the disturbed audio signal can be efficiently isolated as the compensated error signal.
In some embodiments, the error compensation unit is configured to control the adjustable gain by applying an error minimization algorithm, in particular a least mean squares algorithm, to the third compensation signal and the compensated error signal.
In cases, in which the determined leakage condition is inaccurate, e.g. during or before the adaption process of the audio system, the filter parameters of the compensation filter can be partially inaccurate. An error minimization algorithm can especially in these cases lead to additional accuracy and/or to a faster adaptation.
In some embodiments, the second noise filter is further configured to be adapted based on the leakage condition.
In these embodiments, also the response of the second noise filter, i.e. the feedback filter, is adapted based on a current or on a changing acoustic leakage condition. This allows for further increasing the efficiency of the active noise control as also the performance of the FB ANC can be highly dependent on the acoustic leakage condition.
In some embodiments, the detection unit is configured to estimate the leakage condition on the basis of the disturbed audio signal and the audio output signal if a ratio between the speaker signal and the ambient noise exceeds a set threshold. Moreover, the detection unit in these embodiments is configured to estimate the leakage condition on the basis of the first noise filter, in particular of filter parameters of the first noise filter, otherwise.
Depending on the sound pressure level of the speaker signal and hence the contribution of ambient noise in the disturbed audio signal, the determination of the acoustic leakage may be more accurate in one way compared to the other. For example, if an audio signal is output from the speaker at a high sound pressure level, compared to the ambient noise level, the determination of the acoustic leakage condition via the driver response may be more accurate compared to situations at which a low level, or no, audio signal is being output from the speaker. In the latter case, the leakage determination via the response of the first filter is more accurate. The detection unit in these embodiments is therefore configured to determine a ratio between the speaker signal and the ambient noise and based on this determination estimate the acoustic leakage condition following the corresponding method.
In some embodiments, the leakage condition characterizes an acoustic leakage between an ambient of the playback device and a volume which is defined by an ear canal of a user and a cavity of the playback device. Herein, the cavity is arranged at a preferential side for sound emission of the speaker.
In some embodiments, estimating the leakage condition comprises determining a leakage value.
A convenient way of describing the acoustic leakage condition is the determination of an actual leakage value that quantifies the acoustic leakage condition currently present. For example, the leakage value is calculated as a normalized value between 0 and 1 scaling the determined acoustic leakage to a predetermined maximum and/or minimum acoustic leakage. A leakage value of 0 indicates the smallest possible acoustic leakage or no leak and a leakage value of 1 indicates the largest acceptable acoustic leakage, i.e. if the playback device has a very large leak between the front volume and the ambient environment.
In some embodiments, the compensation filter is adapted on the basis of a comparison of the leakage condition with reference leakage conditions in a lookup table.
For example, the lookup table comprises a number of predetermined acoustic leakage conditions, e.g. calibration leakage values measured at different acoustic leakage conditions that are associated to parameters of the compensation filter. The detection unit or a tuning unit may comprise a memory with said lookup table and be configured to adapt the response of the compensation filter by setting one of the associated parameters depending on the estimated acoustic leakage condition.
The lookup table can be coarse, for example it comprises five predetermined acoustic leakage conditions. The detection unit or the tuning unit can then be configured to interpolate the parameters of the compensation filter from two adjacent points of the lookup table if the estimated leakage condition is in between two predetermined acoustic leakage conditions. This process is sufficiently adequate for the music compensation mechanism.
However, for the music removal mechanism, a higher level of isolation of the noise portion of the disturbed audio signal is essential. Therefore, for the music removal mechanism an error compensation unit is employed in order to reduce the significant error between the response of the compensation filter and the driver response. This realizes a significantly improved accuracy of the music signal being removed from the disturbed audio signal during the generation of the compensated error signal.
An employment of the error compensation unit for both the music compensation and the music removal mechanism, e.g. via an adjustable gain of the compensation filter itself, for highly optimized music compensation and music removal filters seems obvious, however, is actually disadvantageous. In detail, any adaption of the music compensation filter, e.g. via adapting a gain, requires an error signal to feedback any deviation from the target response, as described above for some embodiments. If the adjustable gain is a gain of the compensation filter itself, the operation of the feedback loop that is configured to reduce the speaker portion for generating the compensated error signal can result in the desired adaptation of the gain of the compensation filter in order to match the driver response for effectively removing as much speaker signal from the disturbed audio signal as possible. However, the operation of the feedback loop can also result in the reduction of the gain of the compensation filter in order to reduce the amount of audio signal reaching the second noise filter. This is an undesired effect as the audio signal, e.g. music, would be significantly attenuated from the perspective of the user.
The proposed solution of a lookup table for the music compensation mechanism therefore removes this conflict and in addition simplifies processing since operating an adaptive process requires additional computational steps for realizing safety measures for ensuring stability. A lookup table for the music compensation mechanism with a small error of e.g. 1 dB is acceptable as there is no direct reference. That is, the user would perceive a slightly different spectrum of sound coming from the headphone relative to the driver response when noise cancellation is off, however, this difference is small enough to be barely noticeable in normal operation, and small compared to the difference in spectrum due to a changing leakage.
In contrast, if there is a similarly small error in the music removal mechanism between the noise portion of the disturbed audio signal and the third compensation signal, the attenuation, or removal, of the speaker signal for generating the compensated error signal would be substantially reduced. As the third compensation signal is to be subtracted from the disturbed audio signal, i.e. it is directly compared to it, a near-to-perfect match for good attenuation is required. Therefore, an additional adaptive stage realized by the error compensation unit is employed for the music removal mechanism.
The object is further solved by an ear mountable playback device that comprises an audio system according to one of the embodiments described above. For example, the ear mountable playback device is a headphone or an earphone. In general, the term playback device includes all types of audio reproducing devices. Where the term music is specified, it should be appreciated that this term can include any known signal e.g. a voice recording.
The object is further solved by a signal processing method for an ear mountable playback device with a speaker generating a speaker signal based on an audio output signal, with a further microphone that is configured to generate a noise signal on the basis of the ambient noise, and with an error microphone that is configured to generate a disturbed audio signal on the basis of the speaker signal and the ambient noise. The method comprises generating a first compensation signal by applying filter operations of a first noise filter to the noise signal, generating the audio output signal by superimposing an audio signal, the first compensation signal and a second compensation signal, and generating a third compensation signal by applying filter operations of a compensation filter to the audio signal. The method further comprises generating the second compensation signal by applying filter operations of a second noise filter to an intermediate compensation signal that is generated by subtracting the third compensation signal from the disturbed audio signal.
The method further comprises generating a compensated error signal based on the disturbed audio signal and the third compensation signal, estimating a leakage condition on the basis of the first noise filter or of the disturbed audio signal and the audio output signal, adapting the first noise filter based on the compensated error signal, and adapting the compensation filter based on the leakage condition.
Further embodiments of the signal processing method become apparent to a person skilled in the art from the embodiments of the audio system described above.
The improved concept will be described in more detail in the following with the aid of drawings. Elements having the same or similar function bear the same reference symbols throughout the drawings. Hence their description is not necessarily repeated in the description to the following drawings.
In the drawings:
The error microphone FB_MIC may be used according to the improved concept to provide an error signal being the basis for a determination of the wearing condition, respectively acoustic leakage condition, of the headphone HP, when the headphone HP is worn by a user.
In the embodiment of
The headphone HP in this example features a loudspeaker SP, a feedback noise microphone FB_MIC and, optionally, an ambient noise microphone FF_MIC, which e.g. is designed as a feedforward noise cancellation microphone. Internal processing details of the headphone HP are not shown here for reasons of a better overview.
In the configuration shown in
Concerning the ambient noise microphone FF_MIC, a fifth acoustic transfer function AFFM represents the acoustic sound path between the ambient sound source and the ambient noise microphone FF_MIC, and may be called an ambient-to-feedforward response function.
Response functions or transfer functions of the headphone HP, in particular between the microphones FB_MIC and FF_MIC and the speaker SP, can be used with a feedback filter function B and feedforward filter function F, which may be parameterized as noise cancellation filters during operation.
Any processing of the microphone signals or any signal transmission are left out in
In the various embodiments, the FB or error microphone FB_MIC may be located in a dedicated cavity, as for example detailed in ams application EP17208972.4.
Referring now to
In a further implementation, not shown, a headphone HP, e.g. like that shown in
In the following, several implementations of the improved concept will be described in conjunction with a specific use case. It should however be apparent to the skilled person that details described for the implementation may still be applied to other implementations.
The disturbed audio signal E is provided to the third mixer M3 which performs a music compensation process, i.e. subtracts the third compensation signal CS3 from said disturbed audio signal E and provides the resulting intermediate compensation signal to the feedback type second noise filter B for generating a further anti-noise signal, the second compensation signal CS2. For the subtraction, the third mixer M3 may be an additive mixer that comprises a signal inverter on one of its inputs, for instance. The second compensation signal CS2 is superimposed with the audio signal IN, e.g. a music signal, and the first compensation signal CS1 by means of the first mixer M1 for generating the audio output signal, which is converted to the speaker signal SPS by means of the speaker SP.
The third compensation signal CS3 is generated from the audio signal IN by means of the compensation filter C. The third compensation signal CS3 is provided to third mixer M3, as mentioned above, and in addition to the error compensation unit ECU for a music removal process. In detail, the error compensation unit ECU is configured to adjust the third compensation signal CS3 such that it matches the speaker portion of the disturbed audio signal E. The second mixer M2 of the error compensation unit ECU generates the compensated error signal EM by subtracting the adjusted compensation signal from the disturbed audio signal E such that the compensated error signal EM only, or substantially only, comprises the noise portion of the disturbed audio signal E.
The adjusted compensation signal is generated from the third compensation signal CS3 by applying filter operations of an adjustable filter element X to the third compensation signal CS3. For example, the adjustable filter element X is an adjustable gain and is adjusted by means of a feedback loop comprising a control unit CTRL that compares the third compensation signal CS3 and the compensated error signal EM and based on this comparison adjusts the gain of the adjustable filter element X. To this end, the control unit CTRL applies an error minimization algorithm, e.g. a least mean squares algorithm, for instance.
The response of the first noise filter F is adjusted depending on the compensated error signal EM such that a residual noise portion in the disturbed audio signal E is more efficiently removed by means of the first compensation signal CS1, i.e. by means of FF ANC.
The detection unit DET is configured to estimate an acoustic leakage condition from the response of the first noise filter F or from the disturbed audio signal E and the audio output signal. If a level of the audio signal IN exceeds a predetermined threshold relative to a level of the ambient noise NOISE or noise N, the detection unit DET estimates the acoustic leakage condition from the driver response, i.e. the disturbed audio signal E and the audio output signal, for instance, and otherwise from the response of the first noise filter F. In order to determine whether said threshold is exceeded, the detection unit can be configured to measure a level of the audio portion relative to the noise portion of the disturbed audio signal E, for instance.
Regarding the estimation of the acoustic leakage condition via the driver response, the detection unit can be configured to compare the audio output signal to the disturbed audio signal and to estimate the acoustic leakage condition based on the result of the comparison, e.g. based on a deviation between the two signals.
Concerning the estimation of the acoustic leakage condition via the response of the first noise filter F, the detection unit can be configured to monitor the adjustable response of the first noise filter F and to estimate the acoustic leakage condition based on said response. For example, the detection unit is configured to compare the response of the first noise filter F to predetermined responses for estimating the acoustic leakage condition
The detection engine DET can be configured to generate a leakage value for quantifying the actual leakage condition of the earphone. Consequently, the leakage value is provided to the tuning unit TU for adjusting the response of the compensation filter C such that it matches the driver response, i.e. the transfer function from the speaker SP to the error microphone FB_MIC. For example, the tuning unit TU comprises a memory with a lookup table that comprises a number of reference leakage values and respective associated filter responses. The tuning unit TU is then configured to adjust the response of the compensation filter C by setting one of the associated filter responses depending on the leakage value received from the detection unit DET. The tuning unit TU can further be configured to interpolate the adaptation of the compensation filter C if the leakage value received from the detection unit DET is between two of the reference leakage values.
In addition, the tuning unit TU can be further configured to adjust the response of the second noise filter B depending on the leakage value received from the detection unit DET, e.g. based on a second lookup table.
The tuning unit TU, the detection unit DET and the error compensation unit ECU combined essentially constitute the adaptation unit ADP illustrated in
The embodiment of the audio system AS illustrated in
Number | Date | Country | Kind |
---|---|---|---|
19200514.8 | Sep 2019 | EP | regional |
The present application is the national stage entry of International Patent Application No. PCT/EP2020/075992, filed on Sep. 17, 2020, and published as WO 2021/063692 A1 on Apr. 8, 2021, which claims the benefit of priority of European Patent Application No. 19200514.8, filed on Sep. 30, 2019, all of which are incorporated by reference herein in their entireties.
Filing Document | Filing Date | Country | Kind |
---|---|---|---|
PCT/EP2020/075992 | 9/17/2020 | WO |