Determining the acoustic characteristics of a hearing instrument

Abstract

The present application proposes to determine the acoustic characteristics of a hearing instrument. A method for determining the characteristics of a hearing instrument is disclosed herein. The hearing instrument includes at least one input transducer operable to provide an input audio signal in response to a sound sensed in the environment of the hearing instrument, a signal processing unit, and at least one output transducer. The method includes: transmitting an acoustic probe signal through the output transducer, receiving an input audio signal from a microphone, analyzing the received input audio signal to determine the characteristics of the hearing instrument based on the input transducer response to the transmitted acoustic probe signal, wherein the method further includes filtering the received input audio signal to selectively attenuate one or more signal components corresponding to the acoustic probe signal, and wherein transmitting the acoustic probe signal includes transmitting a combined acoustic output signal, the combined acoustic output signal including the acoustic probe signal and an acoustic hearing instrument signal obtained from the filtered input audio signal.

Description

Technical Field

[0001] The present invention relates to hearing instruments and methods for determining the acoustic characteristics of hearing instruments. Background Art

[0002] Different types of hearing instruments are known in the art. Examples of hearing instruments include hearing aids for hearing-impaired users, hearing enhancement devices for enhancing the hearing ability of people with normal hearing, and hearing protection devices designed to prevent noise-induced hearing loss. Hearing instruments typically include an input transducer, a signal processing unit, and an output transducer. During use, the input transducer provides an audio input signal in response to sounds sensed by the hearing instrument in the environment. The signal processing unit processes the audio input signal to generate a hearing instrument audio signal, and the output transducer transmits an acoustic output representing the hearing instrument audio signal generated by the signal processing unit.

[0003] The acoustic output of the hearing instrument perceived by the user depends on the hearing instrument and its environment, in particular on the characteristics of the feedback path between the output transducer and the input transducer of the hearing instrument.

[0004] Feedback is a well-known problem in hearing instruments, and there are several systems in the art for suppressing and eliminating feedback.

[0005] Feedback can occur along external and / or internal feedback paths. Feedback along the external feedback path includes the transmission of sound between the output transducer and the input transducer of the hearing aid along a path outside the hearing instrument. This problem (also known as acoustic feedback) occurs, for example, when the hearing instrument mold does not fit the wearer's ear perfectly, or when the ear mold includes ducts or openings for ventilation purposes, for example. In both examples, sound "leaks" from the output transducer to the microphone and thus causes feedback.

[0006] Feedback in hearing instruments can also occur along the internal feedback path, since sound can be transmitted from the output transducer to the input transducer via a path inside the hearing instrument housing. This transmission can be brought about by air, or caused by mechanical vibrations in the hearing instrument housing or some components within the hearing instrument.

[0007] With the development of very small digital signal processing (DSP) units, it has become possible to execute advanced algorithms for suppressing feedback in microdevices such as hearing instruments. In order to provide effective feedback cancellation, it is highly desirable to know about the characteristics of the hearing instrument, in particular about the characteristics of the feedback path.

[0008] For this purpose, it is known to measure such characteristics, in particular in-situ measurements with the hearing instrument in an operating position. This typically involves positioning the hearing instrument or at least one of its components in the user's ear canal.

[0009] Methods for measuring the characteristics of hearing instruments are known, such as the acoustic impulse response for feedback path identification. Generally, such measurements are performed as open-loop measurements. In hearing instruments, open-loop measurements are typically used in the initial fitting process to identify the feedback path from the output transducer to the input transducer. For this purpose, an acoustic probe signal is sent by the output transducer, and the microphone response is recorded and analyzed to determine the desired characteristics, for example by fitting a model of the feedback path to the recorded response data.

[0010] Compared to closed-loop identification, the advantage of open-loop identification is that it provides highly accurate and unbiased results. Closed-loop identification typically tends to be less efficient and is affected by bias when the feedback and external signals are correlated. Decoupling techniques provide some help, but cannot achieve the same guaranteed performance as open-loop identification. The disadvantage of prior art open-loop identification is that other sounds cannot be played during the identification process.

[0011] There is a desire for a method of performing open-loop identification without disturbing the normal operation of the hearing instrument or at least reducing such interference.

[0012] The object of the present invention is to provide a hearing instrument and a method for determining the characteristics of a hearing instrument that overcomes or at least reduces one or more of the above disadvantages of the prior art methods and / or solves other problems of the prior art solutions, or at least can serve as an alternative to the prior art solutions. Summary of the Invention

[0013] Embodiments of a method for determining the characteristics (especially acoustic characteristics) of a hearing instrument are disclosed herein. The hearing instrument includes at least one input transducer, a signal processing unit, and at least one output transducer. The input transducer is operable to provide an input audio signal in response to sound sensed in the environment of the hearing instrument. The method includes:

[0014] - Sending an acoustic probe signal by the output transducer,

[0015] - Receiving an input audio signal from a microphone,

[0016] - Analyzing the received input audio signal to determine the characteristics of the hearing instrument based on the input transducer response to the sent acoustic probe signal,

[0017] wherein the method further includes filtering the received input audio signal to selectively attenuate one or more signal components corresponding to the acoustic probe signal, and wherein sending the acoustic probe signal includes sending a combined acoustic output signal, the combined acoustic output signal including the acoustic probe signal and an acoustic hearing instrument signal obtained from the filtered input audio signal.

[0018] Accordingly, since the signal components corresponding to the probe signal are attenuated in the input audio signal of the hearing instrument, or even removed from the input audio signal, it is not necessary to interrupt the normal operation of the hearing instrument in order to determine the hearing instrument characteristics. However, the determination of the hearing instrument characteristics is substantially not affected by the normal operation of the hearing aid.

[0019] In particular, a high-quality, unbiased feedback path estimation can be achieved without temporarily depriving the user of the acoustic sensory input from the environment. For example, when the determination of the feedback path characteristics is performed as part of the fitting process for adjusting the hearing instrument, the risk that the user misses important information during the adjustment process is greatly reduced. Various embodiments of the methods disclosed herein even enable the hearing instrument to transmit the probe signal over an extended period of time, for example at a low level, even possibly to a level where it is substantially inaudible to the user of the hearing instrument, or at least less distracting or annoying. Even in the case of a lower probe signal level, an accurate determination of the characteristics can still be achieved because the measurement can be extended over a longer period of time with no or at least only a small perceivable interference with the normal operation of the hearing instrument.

[0020] For the purposes described herein, filtering to selectively attenuate one or more signal components corresponding to the acoustic probe signal will also be referred to as probe-blocking filtering.

[0021] The characteristics of the hearing instrument can be transfer characteristics, such as a transfer function or an impulse response, in particular the transfer characteristics of the feedback path between the output transducer and the input transducer of the hearing instrument.

[0022] Accordingly, analyzing the received input audio signal to determine the hearing instrument characteristics can include system identification processes known in the art. For example, the analysis can include calculating an impulse response to, for example, determine the filter coefficients of a filter (such as a linear filter) for modeling the determined impulse response.

[0023] The probe-blocking filtering can be implemented in series with additional digital signal processing of the hearing instrument, or as part of the whole thereof. The additional signal processing can include hearing loss compensation and / or other conventional signal processing of the hearing instrument known in the art. Thus, in some embodiments, the acoustic hearing instrument signal is obtained by means of said filtering and the additional signal processing of the received input audio signal.

[0024] A detection blocking filter that performs detection blocking filtering can be placed on the input side of a signal processing unit that performs additional signal processing, on the output side of the signal processing unit, or somewhere in between. For example, some signal processing can be performed before detection blocking filtering, while other signal processing can be performed after detection blocking filtering. Thus, in some embodiments, additional signal processing is performed before and / or after the filtering. In some embodiments, the detection blocking filter is integrated into a signal processing unit that performs additional processing.

[0025] Placing the detection blocking filter on the input side of the signal processing unit, or on the input side of some additional signal processing, can reduce storage requirements by sharing a cyclic summation buffer with the filter and minimize potential interactions with other algorithms.

[0026] Placing the detection blocking filter on the output side of the signal processing unit, or on the output side of some additional signal processing, can potentially assist other identification methods that can run simultaneously, such as fast adaptive feedback cancellation. Additionally, placing the detection blocking filter on the output side of the signal processing unit or at least some of the signal processing ensures a detection signal that is as clean as possible, regardless of other potentially non-linear processing options implemented by the signal processing unit. Furthermore, this arrangement only requires a single instance of the detection blocking filter, regardless of the number of input transducers.

[0027] The signal component corresponding to the acoustic detection signal can be a frequency component of the detection signal, particularly one or more main frequency components. In some embodiments, the detection signal has a spectrum that includes only a set of discrete detection frequencies, thereby facilitating selective attenuation of one or more signal components corresponding to the acoustic detection signal. To this end, the filtering can include selectively attenuating frequency components at the discrete and spaced-apart detection frequencies. In particular, the filtering can divide the audio spectrum into a set of passbands separated by notches at the detection frequencies. Additionally, this type of detection signal has been found to facilitate the accurate determination of the characteristics of a hearing instrument, particularly the accurate characterization of the feedback path.

[0028] In particular, in some embodiments, the detection signal is a pseudo-random sequence of sound samples that repeats after a predetermined number L of samples. Preferably, the detection signal implements a maximum length sequence (MLS).

[0029] In some embodiments, filtering is performed by a comb filter, particularly a recursive comb filter, or by other suitable filters for selectively blocking multiple frequencies or narrow frequency bands. This filtering can be achieved by polyphase decomposing the signal into L phases, where each phase is independently filtered with the same prototype response, resulting in a frequency response where the prototype response shape is repeated L times. It has been found that a first-order prototype response shape provides an efficient and effective implementation, but it should be understood that second-order or even higher-order implementations can also be used.

[0030] In some embodiments, the filter is an adaptive filter, particularly a filter including a gain that depends on the signal level, thereby providing improved echo cancellation and suppression of unwanted reflections for various acoustic environments and detection signal types.

[0031] In some embodiments, the filter defines multiple notches at the detection frequencies, each notch having a width, i.e., the filter attenuates frequencies within a specific narrow range around each detection frequency while preferably passing all other frequencies substantially unchanged or with little change. In some embodiments, the width of the notches can be predetermined. In other embodiments, when the filter is an adaptive filter, it can be configured to adaptively adjust the width of the notches, e.g., in response to a change in the signal level of the received input audio signal. To this end, the adaptive filter can include a level-dependent gain, or it can otherwise adaptively control the notch bandwidth. In one embodiment, the adaptive filter includes a first level tracker and a second level tracker, where the first level tracker is configured to track the input audio signal at a first rate and the second level tracker is configured to track the input audio signal at a second rate slower than the first rate. For stationary conditions, i.e., when the levels tracked by the two level trackers are substantially equal, a baseline value of the level-dependent gain can be used, or the notch width can otherwise be controlled to have a baseline width. The baseline bandwidth can be chosen small enough such that the reflections are sufficiently masked by the received audio signal while still allowing sufficient attenuation of the response tail and flexibility to accommodate changes in the feedback path. The change relative to the baseline bandwidth can be made proportional to the difference between the fast and slow level estimates. When the signal level suddenly drops (indicating that a previously masked long reflection tail may become apparent), the gain is temporarily increased, or the notch is temporarily widened. When the signal level suddenly increases (which may become apparent as an echo), the level-dependent gain is temporarily decreased, which results in a narrower notch, or the notch bandwidth is temporarily reduced.

[0032] In some embodiments, the bandwidth of the notches can be uniform across all detection frequencies. To this end, the gain can be a scalar gain. In other embodiments, the notch bandwidth can be made frequency-dependent, e.g., by implementing the gain as a linear-phase FIR filter.

[0033] The present invention relates to different aspects including the above methods, and hereinafter relates to corresponding apparatuses, systems, methods, and / or products, each of which produces one or more of the benefits and advantages described in combination with one or more of the other aspects, and each of which has one or more embodiments, one or more embodiments corresponding to the embodiments described in combination with one or more other aspects and / or disclosed in the appended claims.

[0034] In particular, according to one aspect, an embodiment of a hearing instrument is disclosed herein, comprising:

[0035] - at least one input transducer operable to provide an input audio signal in response to sensing sound in the environment of the hearing instrument,

[0036] - a signal processing unit,

[0037] - at least one output transducer,

[0038] - a signal generator for generating a probe signal, the probe signal being configured to cause the output transducer to transmit an acoustic probe signal,

[0039] - a response analysis circuit configured to analyze the input audio signal from the input transducer to determine a characteristic of the hearing instrument based on the input transducer response to the transmitted acoustic probe signal,

[0040] wherein the hearing instrument further comprises:

[0041] - a probe blocking filter configured to filter the received input audio signal to selectively attenuate one or more signal components corresponding to the acoustic probe signal, and

[0042] - a combiner configured to combine the probe signal and a hearing instrument signal obtained from the filtered input audio signal and feed the combined signal to the output transducer to transmit a combined acoustic signal.

[0043] For the purposes described herein, the terms "signal processing unit" and "response analysis circuit" include any suitably configured circuit or device configured to perform the processing described herein to be performed by the respective processing units. For example, the signal processing unit and / or the response analysis circuit may be or include an ASIC processor, an FPGA processor, a suitably programmed general-purpose processor, a microprocessor, a circuit component, or an integrated circuit.

[0044] A hearing instrument can be a hearing aid for hearing-impaired users, a hearing enhancement device for enhancing the hearing ability of people with normal hearing, a hearing protection device for preventing noise-induced hearing loss, etc. For example, the hearing instrument can be or include hearing instruments of types such as BTE, RIE, ITE, ITC, CIC, etc. Description of the Drawings

[0045] Figure 1 A block diagram schematically showing an embodiment of a hearing instrument.

[0046] Figure 2 A schematic illustration of an open-loop measurement of the characteristics of a hearing instrument.

[0047] Figure 3 A schematic illustration of an implementation of a hearing instrument as described herein.

[0048] Figure 4 A schematic illustration of an embodiment of a detection-blocking filter of an implementation of a hearing instrument as described herein.

[0049] Figure 5A and 5B shows Figure 4 the detection-blocking amplitude response of the filter.

[0050] Figure 6 A schematic illustration of another embodiment of a detection-blocking filter of an implementation of a hearing instrument as described herein.

[0051] Figures 7A - 7D shows Figure 6 the response characteristics of the detection-blocking filter.

[0052] Figure 8 A schematic illustration of yet another embodiment of a detection-blocking filter of an implementation of a hearing instrument as described herein.

[0053] Figures 9A - 9D shows Figure 8 the response characteristics of the detection-blocking filter.

[0054] Figure 10 A schematic illustration of yet another embodiment of a detection-blocking filter of an implementation of a hearing instrument as described herein.

[0055] Figure 11 A schematic illustration of yet another embodiment of a detection-blocking filter of an implementation of a hearing instrument as described herein.

[0056] Figures 12A - 12D shows Figure 11 the response characteristics of the detection-blocking filter.

[0057] Figures 13 - 15Shows yet another embodiment of a detection blocking filter for an embodiment of a hearing instrument as described herein.

[0058] Figures 16A - 16D Shows the response characteristics of a first-order variant with a second-order variant and different Q-factors. Detailed Description

[0059] Figure 1 Schematically shows a block diagram of an embodiment of a hearing instrument (e.g., a hearing aid). The hearing instrument includes an input transducer 101 (e.g., one or more microphones), a signal processing unit 103, and an output transducer 102 (e.g., a speaker), also referred to as a receiver, an implantable transducer, etc. The input transducer 101 receives incoming sound and converts it into an audio signal. The signal processing unit processes the audio signal and outputs a hearing instrument signal, which is fed into the output transducer 102. In particular, the signal processing unit 103 can process the audio signal to compensate for the hearing loss of the hearing aid user, i.e., the hearing instrument signal can be a hearing loss compensated audio signal, which is adapted to restore the loudness of the sound sent by the receiver to the loudness of the incoming sound perceived by a normal listener. Thus, the hearing instrument processor 103 includes elements such as an amplifier, a compressor, and a noise reduction system. The output converter 102 is configured to output an acoustic output signal based on the hearing instrument signal, where the acoustic output signal can be received by the human auditory system, so that the user hears the sound.

[0060] The feedback path 104 is shown as a dashed line between the output transducer 102 and the input transducer 101. This feedback can cause the input transducer 101 to pick up sound from the output transducer 102, which can lead to well-known feedback problems, such as whistling.

[0061] To compensate for the feedback, some hearing instruments include a feedback compensation filter 106, which can be configured to feed a compensation signal to a subtraction unit 105, so that the compensation signal is subtracted from the audio signal provided by the input transducer 101 before being processed in the signal processing unit 103. When the characteristics of the feedback path 104 are known or can be accurately determined, the filter characteristics of the compensation filter 106 can be selected or controlled such that the feedback path can be compensated.

[0062] Accordingly, for the above and / or other purposes, it is desirable to determine the characteristics of the feedback path of a hearing instrument or other characteristics of a hearing instrument. When the hearing instrument is in an operating configuration relative to the user's head, particularly when at least a part of the hearing instrument is located in the user's ear canal, the characteristics of the hearing instrument can indicate the acoustic characteristics of the hearing instrument. The acoustic characteristics can thus include the characteristics of the acoustic environment around the hearing instrument, such as the acoustic characteristics of the user's ear canal, such as the fit of the hearing instrument's mold within the ear canal. Additionally, such characteristics may change over time. Therefore, it is desirable to perform the determination of the hearing instrument characteristics in-situ, i.e., when the hearing instrument or at least one component of the hearing instrument is in an operating position, such as when a component of the hearing instrument is located in the user's ear canal.

[0063] Figure 2 Schematically shows an open-loop measurement of the characteristics of a hearing instrument (e.g., the hearing instrument as shown in conjunction with Figure 1 In the open-loop measurement, the normal signal processing path between the input transducer 101 and the output transducer 102 is interrupted. Instead, the probe signal generator 210 generates a probe signal and feeds the probe signal into the output transducer 102. The probe signal is configured such that the output transducer 102 transmits an acoustic probe signal. The signal analyzer circuit 220 analyzes the input audio signal from the input transducer 101 to determine the characteristics of the hearing instrument based on the input transducer response to the transmitted acoustic probe signal.

[0064] Figure 3 Schematically shows an embodiment of a hearing instrument. The hearing instrument includes an input transducer 101, one or more signal processing units 103, and an output transducer 102, all of which are the same as those described in conjunction with Figure 1 The hearing instrument may include other components, such as the feedback compensation filter and / or other components described in conjunction with Figure 1 The hearing instrument further includes a signal generator 210 and a response analysis circuit 220 configured to determine the characteristics of the hearing instrument, e.g., as described in conjunction with

[0065] The signal generator 210 and / or the response analysis circuit 220 can be implemented separately from the signal processing unit 103, or partially or fully integrated with the signal processing unit 103 into a single processing unit. Figure 2 The hearing instrument of Figure 3 differs from the hearing instrument of Figure 2 in that the determination of the hearing instrument characteristics is performed without disturbing the normal operation of the hearing instrument and without significantly affecting the normal operation of the hearing instrument.

[0066] To this end, the hearing instrument includes a detection blocking filter 330, which is configured to filter the received input audio signal to selectively attenuate one or more signal components corresponding to the acoustic detection signal. The detection blocking filter is selective in that it blocks or at least attenuates the signal components corresponding to the acoustic detection signal, while preferably not affecting or only minimally affecting the remaining signal components.

[0067] In Figure 3 an embodiment, the detection blocking filter 330 is shown as part of the signal processing unit 103. However, it should be understood that the detection blocking filter can also be implemented separately from the signal processing unit 103. Generally, part or all of the additional signal processing can be performed before detection blocking filtering. Similarly, part or all of the additional signal processing can be performed after detection blocking filtering. This is schematically shown by the dashed box 303 in Figure 3 which represents additional signal processing other than detection blocking filtering. It should be understood that some embodiments may include such additional signal processing only on the input side of the detection blocking filtering, while other embodiments may include such additional signal processing only on the output side of the detection blocking filter 330. Additionally, other embodiments may include some additional signal processing on the input side of the detection blocking filtering and other additional signal processing on the output side. Examples of the additional signal processing 303 may include hearing loss compensation, feedback compensation, etc. The signal processing unit 103 thus outputs a hearing instrument signal, such as a hearing loss compensation signal, in which the signal components corresponding to the detection signal have been attenuated.

[0068] Thus, the detection blocking filter 330 can generally be implemented in series with other digital signal processing 303, which can include all common hearing instrument algorithms and can be placed on the input side, the output side, or somewhere in the middle. Any of these positions has advantages and disadvantages. For example, on the input side, the storage requirements can be reduced by sharing the cycle summing buffer with the filter, and the potential interaction with other algorithms is minimized. Removing the detection frequency later potentially aids other recognition methods that may be running simultaneously (e.g., it can aid in fast adaptive feedback cancellation). Removing the detection frequency only on the output side before adding the detection signal, regardless of other (possibly non-linear) processing options, ensures as clean an identification signal as possible at the detection frequency and requires only one instance of the detection blocking filter regardless of the number of microphones.

[0069] The hearing instrument further includes a combiner 340 configured to combine the probe signal from the signal generator 210 with the output from the signal processing unit 103 including the probe-blocking filter 330, i.e., to combine the probe signal with the hearing instrument signal obtained from the filtered input audio signal. The combiner 340 feeds the combined signal to the output transducer 102 for transmitting a corresponding combined acoustic output signal.

[0070] The probe signal generator 210 may include a periodic excitation circuit 212 that generates the probe signal as a pseudo-random sequence 211 that repeats every L samples. Correspondingly, the response analysis circuit 220 may include a periodic summation circuit 221 configured to record the input transducer response by performing periodic averaging in a buffer of length L and feed the buffer content into a response analyzer 222. The response analyzer may perform a system identification process to determine hearing instrument characteristics, such as determining the impulse response, e.g., as described in James M. Kates, “Room reverberation effects in hearing aid feedback cancellation”, The Journal of the Acoustical Society of America 109, 367 (2001); doi:10.1121 / 1.1332379, or by other suitable processes known in the art for system identification.

[0071] Unlike ordinary signals, the spectrum of such a periodic sequence contains only discrete frequencies:

[0072]

[0073] where n is an integer, and L is the length of the sequence, expressed as the number of signal samples. Typically, when the true transfer function is smooth enough, sampling a discrete number of frequencies can provide a good approximation, which is the case when the true impulse response vanishes sufficiently within L samples. The probe sequence used to calibrate the digital feedback cancellation system in some hearing instruments is a maximum length sequence (MLS), such as described in D. Rife and J. Vanderkooy: "Transfer-Function Measurement with Maximum-Length Sequences", Journal of the Audio Engineering Society, 37(6):419{444, June 1989. In some embodiments, the period of the maximum length sequence is 24.5 ms. The MLS method can use the effective cross-correlation between the input and output to recover the periodic impulse response (PIR) of the system under measurement.

[0074] The probe sequence can have a minimum crest factor and a flat spectrum, which facilitates decoding. Alternatively, the sequence can be shaped to, for example, increase the sensitivity at certain frequencies, or make it less obtrusive.

[0075] The discrete nature of the probe spectrum can be exploited by the various embodiments of the probe blocking filter disclosed herein to minimize the interference caused by open-loop identification. It is sufficient to block only the discrete probe frequencies, without disconnecting all the sound in the forward path between the input transducer and the output transducer. The frequencies between the probe frequencies (corresponding to non-integer values of n in Equation 1) can pass through without significantly affecting the measurement.

[0076] To this end, the probe blocking filter 330 can divide the spectrum into L / 2 unique frequency bands separated by notches at the probe frequencies f(n). For non-probe frequencies, the audible variations of the signal caused by the probe blocking filter should be minimized.

[0077] Various embodiments of the probe blocking filter 330 will be described in more detail below.

[0078] Figure 4 An example of a probe blocking filter that schematically shows an embodiment of a hearing instrument as described herein is shown. In particular, Figure 4 the probe blocking filter 330 is a recursive comb filter (which can provide an efficient implementation of the probe blocking filter).

[0079] The filter includes a gain block 331 that applies a gain w to a delayed signal, which is delayed by a delay block 332 with a delay b2. The filter also includes delay blocks 333 and 334 that apply delays b and L - b, respectively. Optionally, the filter further includes a gain adaptation block 335. Different filter characteristics can be obtained by selecting the delays b and b2, by selecting the gain w, and / or by adding or omitting the gain adaptation block 335.

[0080] Generally, the filter 330 provides complete suppression at the detection frequency, regardless of the setting of w. In some embodiments of the filter 330, the delay parameters b and b2 are selected to be equal to each other. In particular, in some embodiments, both of them are set to zero.

[0081] In a specific embodiment, the delay parameters b and b2 are set to zero, the gain adaptation block 335 is omitted, and w is a constant scalar gain, for example, in the range between 0 and 2, which defines the transient behavior around the detection frequency.

[0082] Figure 5A and 5B shows the detection blocking amplitude response of the filter for different fixed scalar values of w and for b and b2 set to zero. Figure 4 of

[0083] Specifically, Figure 5A shows the amplitude response in the frequency domain. Curve 501 shows the amplitude response for w = 1, curve 502 shows the amplitude response for w = 0.333, curve 503 shows the amplitude response for w = 0.1, and curve 504 shows the amplitude response for w = 0.033.

[0084] Figure 5B shows the reflected amplitude of the detection blocking filter in the time domain for the same w values.

[0085] From Figure 5A and 5B it can be seen that the amplitude response of the detection blocking filter has a notch at the detection frequency defined in equation (1). Small values of w provide a narrow notch with a long tail of many small reflections at multiples of the L - sample delay. The larger the value of w, the wider the notch and the fewer large reflections. For w = 1, the response is a special case, as shown by the triangular point 511 in Figure 5B with only one (unattenuated) reflection at the L - sample delay.

[0086] As Figure 5AAs shown, for large values of w, the filter can add some significant gain at the center of the passband, e.g., as shown by curve 501, and to some extent, also as shown by curve 502. The maximum filter gain can be normalized by multiplying by (1 - w / 2), and the RMS filter gain can be normalized by multiplying by ; alternatively, the offset can be simply discarded when it is small enough.

[0087] As Figure 5B shown, choosing the value of w can involve a trade-off between a smaller number of larger reflections or a larger number of smaller reflections. Ideally, the reflections are all hidden by forward-time masking, the precedence effect, and ordinary masking by the direct signal. For the first two psychoacoustic effects to work, the delay should preferably be roughly in the range of 5 to 40 ms, and the subsequent attenuation should be fast enough not to cause a noticeable ringing effect (e.g., after 40 ms, the reflection can become apparent as an echo). Thus, for a moderate value of L, a fixed scalar gain w can be chosen where the perceived effects of both the first reflection and the late response tail are sufficiently minimized. Of course, for an artificial test signal, e.g., a pure tone at a probing frequency, this effect can still be demonstrated, but even then, the effect simply tapers off and is unlikely to be considered objectionable. As L gets larger, at some point, for some sounds, the first reflection or the long response tail will become apparent and even objectionable. For some sounds, a smaller value of w is better, while for other sounds, a larger value is better. To have the best of both worlds and reduce the unwanted side effects for a wide range of signal types, the implementation of an adaptive filter is therefore preferred.

[0088] Referring again to Figure 4, in one embodiment, the detection blocking filter includes a gain adaptive circuit 335 to provide an adaptive scalar gain w, which is updated by tracking the level of the primary audio signal. In one embodiment, the gain adaptive circuit may use two level trackers, one operating fast and the other operating at a slower speed than the fast one. For stationary conditions, i.e., when the two level trackers are in agreement, the gain adaptive circuit may set w to a predetermined baseline value. The baseline value can be chosen small enough so that reflections are sufficiently masked by the primary audio signal, while still allowing the response tail to be sufficiently attenuated and the flexibility to adapt to changes in the feedback path. When the fast and slow level estimates are different from each other, the gain adaptive circuit 335 may adjust w to deviate from the baseline value. For example, the gain adaptive circuit 335 may adjust w to deviate from the baseline value by a difference proportional to the difference between the slow and fast level estimates. Accordingly, when the signal level suddenly drops, indicating that a previously masked long reflection tail may become apparent, w is temporarily increased (the notch is widened). When the signal level suddenly increases and may become apparent as an echo, w is temporarily decreased (the notch is narrowed). The effect of this adaptation is to reduce the dynamic range of the update signal fed into the filter buffer, thus continuously adapting between the forms of compression / limiting and expansion.

[0089] To avoid the obvious non-linear effects introduced by adjusting w, the change in w can be made actively and / or smoothly, i.e., it is preferably to reach the new gain earlier, and to reach the new gain by applying many small steps at the sampling rate instead of applying several large steps at the block rate. To this end, to obtain a smooth transition, the gain adaptive circuit 335 may use the input from the level tracker (which may also operate at the block rate) to calculate the new target of w for each block. According to the calculated target of w, the gain adaptive circuit 335 can then derive the relative increment to be increased for each sample. The gain change can be made actively by setting positive values for b and b2, so that the filter response can adapt in advance to one or more blocks, which is particularly useful for avoiding echoes from impulse sounds.

[0090] Hearing experiments conducted by the present inventors using a wide range of audio segments show that the scalar adaptive gain w performs well for typical values of L suitable for digital feedback suppression. For example, L can be between 100 and 2000; when using a maximum length sequence, L can be chosen as L = 2 m-1, where m can be between 7 and 12, but other values can also be used. The value of L can be selected, for example, based on the desired resolution, sampling rate, and / or other factors. No deterioration in speech quality was observed, such as the fundamental frequency or harmonic frequencies that coincide with the notch frequencies. This is likely because the spread of the average speech harmonics is much wider than the notch used to suppress the detection frequency, and when this is not temporarily the case due to a rapid drop in level, the notch is temporarily widened and the forward masking effect dominates. For signals with highly concentrated spectral content, such as the sound of an ambulance / siren (where the frequency slowly moves while keeping the amplitude constant), some changes may be noticeable. However, generally, if noticeable, most of the changes in the sound are like a slight increase in room reverberation.

[0091] A non - adaptive detection blocking filter configuration may not always be able to provide the same performance as an adaptive configuration, but can still have some uses for small values of L. When L is further increased, a point may eventually be reached where even a scalar adaptive gain is no longer sufficient. When this occurs, some further improvement can be obtained by switching to a frequency - dependent gain. This can be done by implementing the gain w with a linear - phase FIR filter, for which the group delay is compensated by reducing b2, such that the combined delay still matches b. Such a filter can be designed / updated by spectral analysis of the level and calculation of the target gain for each frequency band, followed by windowed IFFT filter design. Alternatively, the calculation can be performed entirely in the time domain using linear - phase band splitting, which is effective when the number of desired frequency bands is low, or the calculation can be performed entirely in the frequency domain, which is effective when the number of desired frequency bands is high.

[0092] Figure 6 Another embodiment of a detection blocking filter schematically showing an embodiment of a hearing instrument as described herein is shown. In this embodiment, the detection blocking filter is a comb - band splitting filter. Specifically, the comb - band splitting filter includes a delay block 601 that provides an L - sample delay to the input audio signal x. The delayed signal is subtracted from the input audio signal x by a combiner 602 to provide a first response signal y1. Optionally, the filter includes another combiner 603 that adds the delayed signal to the input audio signal x to provide a second response signal y2.

[0093] The response characteristics of the comb - band splitting filter are shown in Figures 7A - 7D shown.

[0094] Figure 7A The amplitude response 701 of the response signal y1 and the amplitude response 702 of the response signal y2 are shown in Figure 7AIt can be seen that the response signal y1 has a notch at the detection frequency of the detection signal and a wide peak between the detection frequencies. In contrast, the response signal y2 has a notch between the detection frequencies of the detection signal and a wide peak at the detection frequencies. The response signal y1 can thus be referred to as a detection-blocking response signal and used as a filtered signal to be forwarded to the output transducer, optionally with additional signal processing as described herein. The response signal y2 can be referred to as a detection-passing response signal. If desired, it can be used to analyze the input audio signal x at the detection frequencies; alternatively, it can simply be discarded.

[0095] Figure 8 Another embodiment of a detection-blocking filter schematically showing an embodiment of a hearing instrument as described herein is shown. Figure 8 The filter of is a Schroeder all-pass band-splitting filter, which is similar to Figure 6 the filter of, except that the L-sample delay block is replaced by a Schroeder all-pass filter 801. The response characteristics of the Schroeder all-pass band-splitting filter are shown in Figures 9A - 9D FIG..

[0096] Figure 10 Another embodiment of a detection-blocking filter schematically showing an embodiment of a hearing instrument as described herein is shown. Figure 10 The filter of is an optimized variant of the Schroeder all-pass band-splitting filter.

[0097] Its response signals y1 and y2 can be expressed as follows:

[0098] y1[n] = (1 - w / 2) × (x[n] - x[n - L]) + (1 - w) × y1[n - L]

[0099] y2[n] = (w / 2) × (x[n] + x[n - L]) + (1 - w) × y2[n - L].

[0100] Figure 11 Another embodiment of a detection-blocking filter schematically showing an embodiment of a hearing instrument as described herein is shown. Figure 11 The filter of is Figure 10 a simplified variant of the filter of, where normalization and 2-point moving average are omitted.

[0101] The response signals y1 and y2 of this variant can be expressed as follows:

[0102] y1[n] = x[n] - x[n - L] + (1 - w) × y1[n - L]

[0103] y2[n] = w × x[n] + (1 - w) × y2[n - L].

[0104] In Figures 12A - 12D is shown Figure 11 the response characteristics of a variant.

[0105] Figures 13 - 15 Shows yet another embodiment of a detection blocking filter for an embodiment of a hearing instrument as described herein. Figures 13 - 15 The embodiment of Figure 13 shows a second-order band-splitting filter, Figure 14 shows an optimized variant of the second-order band-splitting filter, and Figure 15 shows a second-order all-pass band-splitting filter including Schroder all-pass sections 1501 and 1502.

[0106] Figures 16A - 16D Shows the response characteristics of a first-order variant ( Figures 16B - 16D ) with a second-order variant and different Q-factors and cut-off frequencies ( Figure 16A ). When the echo of the first-order filter drops linearly on the dB scale, the echo of the second-order filter drops rapidly but then increases again. The cut-off frequency that defines the notch width affects the speed of echo attenuation. For a wide notch, the initial echo is stronger, but the attenuation speed is faster than in the case of a narrow notch.

[0107] In summary, at least some aspects disclosed herein can be summarized as follows:

[0108] Embodiment 1: A method for determining characteristics of a hearing instrument, the hearing instrument including at least one input transducer, a signal processing unit, and at least one output transducer, the input transducer being operable to provide an input audio signal in response to sensing sound in the environment of the hearing instrument, the method comprising:

[0109] - Sending an acoustic detection signal by the output transducer,

[0110] - Receiving an input audio signal from a microphone,

[0111] - Analyzing the received input audio signal to determine the characteristics of the hearing instrument based on the input transducer response to the sent acoustic detection signal,

[0112] wherein the method further comprises filtering the received input audio signal to selectively attenuate one or more signal components corresponding to the acoustic detection signal, and wherein sending the acoustic detection signal includes sending a combined acoustic output signal, the combined acoustic output signal including the acoustic detection signal and an acoustic hearing instrument signal obtained from the filtered input audio signal.

[0113] Embodiment 2: The method according to Embodiment 1, wherein the acoustic hearing instrument signal is obtained by said filtering and by additional signal processing of the received input audio signal.

[0114] Embodiment 3: The method according to Embodiment 2, wherein the additional signal processing is performed before and / or after said filtering.

[0115] Embodiment 4: The method according to any one of the foregoing embodiments, wherein the detection signal has a spectrum that includes only a set of discrete detection frequencies.

[0116] Embodiment 5: The method according to Embodiment 4, wherein the filtering includes selectively attenuating frequency components at said discrete detection frequencies.

[0117] Embodiment 6: The method according to any one of Embodiments 4 to 5, wherein the filtering divides the spectrum into a set of passbands separated by notches at the detection frequencies.

[0118] Embodiment 7: The method according to any one of the foregoing embodiments, wherein the detection signal is a pseudo-random sequence of sound samples repeated every L samples.

[0119] Embodiment 8: The method according to any one of the foregoing embodiments, wherein the detection signal represents a maximum length sequence.

[0120] Embodiment 9: A hearing instrument, comprising:

[0121] - at least one input transducer operable to provide an input audio signal in response to sensing sound in the environment of the hearing instrument,

[0122] - a signal processing unit,

[0123] - at least one output transducer,

[0124] - a signal generator for generating a detection signal configured to cause the output transducer to transmit an acoustic detection signal,

[0125] - a response analysis circuit configured to analyze the input audio signal from the input transducer to determine a characteristic of the hearing instrument based on the input transducer response to the transmitted acoustic detection signal,

[0126] wherein the hearing instrument further comprises:

[0127] - a detection blocking filter configured to filter the received input audio signal to selectively attenuate one or more signal components corresponding to the acoustic detection signal, and

[0128] - A combiner configured to combine the detection signal and the hearing instrument signal obtained from the filtered input audio signal and feed the combined signal to the output transducer to transmit the combined acoustic signal.

[0129] Embodiment 10: The hearing instrument according to Embodiment 9, wherein the detection blocking filter comprises a comb filter.

[0130] Embodiment 11: The hearing instrument according to Embodiment 10, wherein the comb filter is a recursive comb filter.

[0131] Embodiment 12: The hearing instrument according to any one of Embodiments 9 to 11, wherein the detection blocking filter is a first-order filter.

[0132] Embodiment 13: The hearing instrument according to any one of Embodiments 9 to 12, wherein the detection blocking filter is an adaptive filter.

[0133] Embodiment 14: The hearing instrument according to any one of Embodiments 9 to 13, wherein the detection blocking filter comprises a frequency-dependent gain.

[0134] Embodiment 15: The hearing instrument according to any one of Embodiments 9 to 14, configured to perform the steps of any one of the methods according to Embodiments 1 to 8.

[0135] Although the above embodiments have been mainly described with reference to certain specific embodiments, various modifications thereof will be apparent to those skilled in the art without departing from the spirit and scope of the present invention, as outlined in the appended claims.

Claims

1. A method for determining characteristics of a hearing instrument, the hearing instrument comprising at least one input transducer (101), a signal processing unit (103), and at least one output transducer (102), the input transducer being operable to provide an input audio signal in response to sensing sound in the environment of the hearing instrument, the method comprising: - transmitting an acoustic probe signal by the output transducer (102), - receiving the input audio signal from the input transducer (101), - analyzing the received input audio signal to determine the characteristics of the hearing instrument based on the input transducer response to the transmitted acoustic probe signal, characterized in that: the method further comprises filtering the received input audio signal to selectively attenuate one or more signal components corresponding to the acoustic probe signal, and wherein transmitting the acoustic probe signal comprises transmitting a combined acoustic output signal, the combined acoustic output signal comprising the acoustic probe signal and an acoustic hearing instrument signal obtained from the filtered input audio signal.

2. The method according to claim 1, wherein The acoustic hearing instrument signal is obtained by the filtering and by additional signal processing of the received input audio signal, wherein the additional signal processing is performed before and / or after the filtering.

3. The method according to any one of the preceding claims, wherein, The probe signal has a spectrum that includes only a set of discrete probe frequencies.

4. The method according to claim 3, wherein, Filtering comprises selectively attenuating frequency components at the discrete probe frequencies.

5. The method according to claim 1, wherein The probe signal is a pseudo-random sequence of sound samples repeated every L samples, in particular representing a maximum length sequence.

6. A hearing instrument, comprising: - at least one input transducer (101), operable to provide an input audio signal in response to sensing sound in the environment of the hearing instrument, - a signal processing unit (103), - at least one output transducer (102), - a signal generator (210) for generating a probe signal, the probe signal being configured to cause the output transducer to transmit an acoustic probe signal, - a response analysis circuit (220) configured to analyze an input audio signal from the input transducer (101) to determine the characteristics of the hearing instrument based on the input transducer response to the transmitted acoustic probe signal, characterized in that: the hearing instrument further comprises: - a probe blocking filter (330) configured to filter the received input audio signal to selectively attenuate one or more signal components corresponding to the acoustic probe signal, and - a combiner (340) configured to combine the probe signal and a hearing instrument signal obtained from the filtered input audio signal and feed the combined signal to the output transducer (102) to transmit a combined acoustic signal.

7. The hearing instrument according to claim 6, wherein, The probe blocking filter (330) comprises a comb filter, in particular a recursive comb filter.

8. The hearing instrument according to any one of claims 6 to 7, wherein, The probe blocking filter (330) is a first-order filter.

9. The hearing instrument according to any one of claims 6 to 7, wherein, The probe blocking filter (330) is an adaptive filter.

10. The hearing instrument according to any one of claims 6 to 7, wherein, The probe blocking filter (330) comprises a frequency-dependent gain.

Images