Abstract
The application relates to a hearing aid comprising a forward path comprising a) a multitude of input units for providing a multitude of electric input signals IN.sub.i, i=1, . . . , M, representative of sound, b) a multi input beam former filtering unit for providing a beam formed signal Y.sub.BF from said multitude of electric input signals, c) a gain unit for applying a hearing aid gain G.sub.HA to said beam formed signal Y.sub.BF, and providing a processed signal, and d) an output unit for providing stimuli perceivable by a user as sound based on said processed signal or a signal derived therefrom. The hearing aid further comprises e) a gain control unit for limiting said hearing aid gain G.sub.HA to a modified full-on gain value G′.sub.FOG. The multi input beam former filtering unit is configured to apply a current frequency dependent directional gain G.sub.DIR,i to each of said multitude of electric input signals IN.sub.i, and the gain control unit is configured to determine the modified full-on gain value G′.sub.FOG in dependence of said current directional gains G.sub.DIR,i, i=1, . . . , M, and a previously determined full-on gain value G.sub.FOG. Thereby an improved hearing aid is provided. The invention may e.g. be used for hearing instruments, headsets, or active ear protection systems.
Claims
1. A hearing aid comprising a forward path comprising a multitude of input units for providing a multitude of electric input signals IN.sub.i, i=1, . . . , M, representative of sound, a multi input beam former filtering unit for providing a beam formed signal Y.sub.BF from said multitude of electric input signals, a gain unit for applying a hearing aid gain G.sub.HA to said beam formed signal Y.sub.BF, and providing a processed signal, and an output unit for providing stimuli perceivable by a user as sound based on said processed signal or a signal derived therefrom, the hearing aid further comprising a gain control unit for limiting said hearing aid gain G.sub.HA to a modified full-on gain value G′.sub.FOG, wherein the multi input beam former filtering unit is configured to apply a current frequency dependent directional gain G.sub.DIR,i to each of said multitude of electric input signals IN.sub.i, and wherein the gain control unit is configured to determine the modified full-on gain value G′.sub.FOG in dependence of said current directional gains G.sub.DIR,i, i=1, . . . , M, and a previously determined full-on gain value G.sub.FOG.
2. A hearing aid according to claim 1 wherein the gain control unit is configured to determine a current modified full-on gain value G′.sub.FOG in dependence of a maximum value G.sub.DIR,max of said current directional gains G.sub.DIR,i, i=1, . . . , M.
3. A hearing aid according to claim 1 wherein the gain control unit is configured to determine a current modified full-on gain value G′.sub.FOG in dependence of a maximum value G.sub.DIR,max of said current directional gains G.sub.DIR,i, i=1, . . . , M, and the previously determined full-on gain value G.sub.FOG.
4. A hearing aid according to claim 1 wherein the gain control unit is configured to determine the modified full-on gain value G′.sub.FOG as a difference between the previously determined full-on gain value G.sub.FOG and the maximum value G.sub.DIR,max of the current directional gains multiplied by a positive constant α, G′.sub.FOG=G.sub.FOG−α G.sub.DIR,max.
5. A hearing aid according to claim 1 wherein the gain control unit comprises a configurable smoothing unit configured to determine a smoothed value <G.sub.DIR,max> of the maximum value G.sub.DIR,max of the current directional gains, and to use the smoothed value <G.sub.DIR,max> in the determination of the modified full-on gain value G′.sub.FOG, e.g. G′.sub.FOG=G.sub.FOG−<G.sub.DIR,max>.
6. A hearing aid according to claim 5 wherein the gain control unit is configured to control a release time and/or an attack time of the smoothing unit in dependence of a current full on gain margin ΔG.sub.FOGm, ΔG.sub.FOGm being a difference between the previously determined full-on gain value G.sub.FOG and the sum of the current hearing aid gain G.sub.HA and the maximum value G.sub.DIR,max of the current directional gains ΔG.sub.FOGm=G.sub.FOG−(G.sub.HA+G.sub.DIR,max).
7. A hearing aid according to claim 6 wherein the gain control unit is configured to set a release time constant involved in determining the smoothed value <G.sub.DIR,max> to a value smaller than or equal to a first value τ.sub.rel,FAST, in case the current full-on gain margin ΔG.sub.FOGm is below a first threshold value ΔG.sub.LIM,fast, i.e. for ΔG.sub.FOGm<ΔG.sub.th,LIM, where ΔG.sub.LIM,fast is larger than zero.
8. A hearing aid according to claim 1 wherein the gain control unit is configured to control the beam former filtering unit in dependence of the maximum value G.sub.DIR,max of the current directional gains.
9. A hearing aid according to claim 1 wherein the gain control unit is configured to control the beam former filtering unit in dependence of the previously determined full-on gain value G.sub.FOG, the current hearing aid gain G.sub.HA and the maximum value G.sub.DIR,max of the current directional gains.
10. A hearing aid according to claim 1 wherein the gain control unit is configured to control the beam former filtering unit in dependence of the current full on gain margin ΔG.sub.FOGm, ΔG.sub.FOGm being a difference between the previously determined full-on gain value G.sub.FOG and the sum of the current hearing aid gain G.sub.HA and the maximum value G.sub.DIR,max of the current directional gains ΔG.sub.FOGm=G.sub.FOG−(G.sub.HA+G.sub.DIR,max).
11. A hearing aid according to claim 10 wherein the gain control unit is configured to control the beam former filtering unit to reduce said maximum value G.sub.DIR,max of the current directional gains, in case the current full on gain margin ΔG.sub.FOGm, is smaller than a threshold value.
12. A hearing aid according to claim 10 wherein the gain control unit is configured to determine a beam former control signal DIRctr for controlling the beam former filtering unit between an un-restrained ON-state, when said current full on gain margin ΔG.sub.FOGm is above a first threshold value ΔG.sub.DIR,ON, and an OFF-state, when said current full on gain margin ΔG.sub.FOGm is below a second threshold value ΔG.sub.DIR,OFF.
13. A hearing aid according to claim 12 wherein the gain control unit is configured to determine a smoothed value <ΔG.sub.FOGm> of said current full on gain margin ΔG.sub.FOGm, and to use said smoothed value <ΔG.sub.FOGm> in the determination of the beam former control signal DIRctr instead of said current full on gain margin ΔG.sub.FOGm.
14. A hearing aid according to claim 1 comprising a multitude M of analysis filter banks each for providing a time-frequency representation IN.sub.i(k,m) of a respective different one of the multitude of electric input signals IN.sub.i, i=1, . . . , M, k being a frequency index and m being a time index.
15. A hearing aid according to claim 1 comprising a hearing instrument or an active ear-protection device or other audio processing device, which is adapted to improve, augment and/or protect the hearing capability of a user by receiving acoustic signals from the user's surroundings, generating corresponding audio signals, possibly modifying the audio signals and providing the possibly modified audio signals as audible signals to at least one of the user's ears.
16. A hearing aid according to claim 1 wherein the beamformer filtering unit comprises minimum variance distortionless response (MVDR) beamformer.
17. A hearing aid according to claim 1 wherein the beamformer filtering unit comprises generalized sidelobe canceller (GSC) structure.
18. Use of a hearing aid as claimed in claim 1.
19. A method of operating a hearing aid comprising a forward path comprising a multitude of input units for providing a multitude of electric input signals IN.sub.i, i=1, . . . , M, representative of sound, an output unit for providing stimuli perceivable by a user as sound based on a processed signal or a signal derived therefrom, the method comprising providing a beam formed signal Y.sub.BF from said multitude of electric input signals, applying a current frequency dependent directional gain G.sub.DIR,i to each of said multitude of electric input signals IN.sub.i, applying a hearing aid gain G.sub.HA to said beam formed signal Y.sub.BF, and providing a processed signal, and providing a previously determined full-on gain value G.sub.FOG, limiting said hearing aid gain G.sub.HA to a modified full-on gain value G′.sub.FOG, determining said modified full-on gain value G′.sub.FOG in dependence of said current directional gains G.sub.DIR,i, i=1, . . . , M, and said previously determined full-on gain value G.sub.FOG.
20. A data processing system comprising a processor and program code means for causing the processor to perform the steps of the method of claim 19.
21. A computer program comprising instructions which, when the program is executed by a computer, cause the computer to carry out the method of claim 19.
Description
BRIEF DESCRIPTION OF DRAWINGS
[0092] The aspects of the disclosure may be best understood from the following detailed description taken in conjunction with the accompanying figures. The figures are schematic and simplified for clarity, and they just show details to improve the understanding of the claims, while other details are left out. Throughout, the same reference numerals are used for identical or corresponding parts. The individual features of each aspect may each be combined with any or all features of the other aspects. These and other aspects, features and/or technical effect will be apparent from and elucidated with reference to the illustrations described hereinafter in which:
[0093] FIG. 1 shows an exemplary first embodiment a hearing aid comprising control unit for implementing a Full-On Gain limitation system connected to a beam former filtering unit and an amplification unit according to the present disclosure,
[0094] FIG. 2 shows an embodiment of control unit for implementing a Full-On Gain limitation system according to the present disclosure,
[0095] FIG. 3A shows an illustration of an exemplary scheme for operating a gain control unit of a hearing aid according to the present disclosure from start time t.sub.0 to an end time t.sub.11, and in the left part an exemplary functional relationship between the current full on gain margin ΔG.sub.FOG, and the beam former control signal DIRctr for controlling the beam former filtering unit,
[0096] FIG. 3B illustrates an exemplary functional relationship between the current full on gain margin ΔG.sub.FOGm and the attack τ.sub.att and release τ.sub.rel time constants involved in determining the smoothed value <G.sub.DIR,max> in a first time interval from a start time t.sub.0 to an intermediate time t.sub.6 during increasing desired hearing aid gain G.sub.HA, (i.e. during decreasing full on gain margin ΔG.sub.FOGm), and
[0097] FIG. 3C illustrates an exemplary functional relationship between the current full on gain margin ΔG.sub.FOGm and the attack τ.sub.att and release τ.sub.rel time constants involved in determining the smoothed value <G.sub.DIR,max> in a second time interval from an intermediate time t.sub.6 to an end time t.sub.11 during decreasing desired hearing aid gain G.sub.HA, (i.e. during increasing full on gain margin ΔG.sub.FOGm), and
[0098] FIG. 4 shows a flow diagram of an embodiment of a method of operating a hearing aid according to the present disclosure.
[0099] The figures are schematic and simplified for clarity, and they just show details which are essential to the understanding of the disclosure, while other details are left out. Throughout, the same reference signs are used for identical or corresponding parts.
[0100] Further scope of applicability of the present disclosure will become apparent from the detailed description given hereinafter. However, it should be understood that the detailed description and specific examples, while indicating preferred embodiments of the disclosure, are given by way of illustration only. Other embodiments may become apparent to those skilled in the art from the following detailed description.
DETAILED DESCRIPTION OF EMBODIMENTS
[0101] The detailed description set forth below in connection with the appended drawings is intended as a description of various configurations. The detailed description includes specific details for the purpose of providing a thorough understanding of various concepts. However, it will be apparent to those skilled in the art that these concepts may be practised without these specific details. Several aspects of the apparatus and methods are described by various blocks, functional units, modules, components, circuits, steps, processes, algorithms, etc. (collectively referred to as “elements”). Depending upon particular application, design constraints or other reasons, these elements may be implemented using electronic hardware, computer program, or any combination thereof.
[0102] The electronic hardware may include microprocessors, microcontrollers, digital signal processors (DSPs), field programmable gate arrays (FPGAs), programmable logic devices (PLDs), gated logic, discrete hardware circuits, and other suitable hardware configured to perform the various functionality described throughout this disclosure. Computer program shall be construed broadly to mean instructions, instruction sets, code, code segments, program code, programs, subprograms, software modules, applications, software applications, software packages, routines, subroutines, objects, executables, threads of execution, procedures, functions, etc., whether referred to as software, firmware, middleware, microcode, hardware description language, or otherwise.
[0103] The present application relates to the field of hearing devices, e.g. hearing aids, in particular to a hearing device comprising a signal processing unit allowing the execution of a number of configurable processing algorithms, e.g. level compression algorithms, feedback estimation algorithms, etc. to modify an audio input signal, e.g. according to the needs of a user of the hearing device. More specifically the disclosure deals with Full-on Gain (FOG) Limitation (and/or a maximum output limitation) for controlling the stability of a digital hearing aid by limiting the maximum allowable gain in the hearing aid.
[0104] A solution for a FOG Limitation (the TOG limit′) is proposed, which limits the hearing aid amplification (G.sub.HA) to a value (G′.sub.FOG) that is dynamically corrected by the maximum gain (G.sub.DIR,max) that is given by the directional system (G′.sub.FOG=G.sub.FOG−G.sub.DIR,max). We will refer to the correction as the ‘FOG Correction’ (ΔG.sub.FOG=G.sub.FOG−G.sub.FOG=G.sub.DIR,max).
[0105] This is a very tractable and simple solution, but it comes at the drawback that fast gain limit changes can give unpleasant audible artefacts for the user. We therefore, in an exemplary embodiment, propose a system that is slower, such that it acts fast when the limit is reached, but retracts at a slow rate, in order to avoid gain-pumping artefacts.
[0106] This leads to another drawback for the user. In situations where the directionality system utilizes large microphone gains, the user experiences a lack of gain when the system is in continuous limitation. This comes from the fact that the gain in the microphone channels do not necessary contribute to the acoustical gain.
[0107] This disadvantage can be solved by introducing a sluggishness in the FOG Correction such that, when the device gain is not close to the FOG Limit, the correction is slowly varying. However, when the device gain becomes close to the FOG limit, the correction is quickly adapted in order to obtain the correct FOG limitation.
[0108] This allows for a second system to retract the directionality gain, also dependent on the closeness of the device gain to the FOG Limit. This means that when the device gain is getting closer to the FOG Limit, as a first step, the directionality system is forced to be less directional. The consequence for the user is that device gain is utilized for amplification prioritized over directionality. When the device gain keeps increasing, the FOG Correction will be accelerated in order to give the correct limitation when the device gain reaches the FOG Limit.
[0109] FIG. 1 shows an exemplary first embodiment a hearing aid (HD) comprising control unit (CONT) for implementing a Full-On Gain limitation system connected to a beam former filtering unit (BFUa, BFUb) and an amplification unit (HAG). The hearing aid comprises a forward path for processing an input signal representing sound and providing an enhanced signal for presentation to a user. The forward path comprises a multitude of input units (here microphones M1, M2) for providing a multitude of electric input signals IN.sub.i, i=1, . . . , M, representative of sound (here IN1, IN2, i.e. M=two). The input units preferably comprise appropriate analogue to digital conversion units to provide the electric input signals (IN1, IN2) as digital signals. Each microphone path comprises an analysis filter bank (here FB-A1, FB-A2, respectively) for providing the electric input signals (IN1, IN2) in a time-frequency representation as sub-band signals X.sub.1 and X.sub.2, respectively. The forward path further comprises a multi input beam former filtering unit (BFUa, BFUb) for providing a beam formed signal Y.sub.BF from said multitude of electric input signals IN1, IN2 (here from the sub-band signals X.sub.1, X.sub.2). The forward path further comprises a gain unit (HAG, MIN, ‘X’) for applying a (possibly limited) hearing aid gain G′.sub.HA to the beam formed signal Y.sub.BF, and providing a processed signal Y′.sub.G. The forward path further comprises synthesis filter bank (FB-S) for converting the frequency sub-band signals of the processed signal Y′.sub.G to an output signal OUT in the time domain. The forward path further comprises an output unit (here a loudspeaker SP) for providing stimuli (acoustic or mechanical stimuli) perceivable by a user as sound based on said processed signal or a signal derived therefrom (here the output signal OUT). The hearing aid further comprises a gain control unit (CONT) for limiting the hearing aid gain G.sub.HA to a modified full-on gain value G′.sub.FOG (via minimum unit (MIN) which provides the minimum of two input gain values (hearing aid gain G.sub.HA and modified fill-on gain G′.sub.FOG) in the form of gain limited hearing aid gain G′.sub.HA). The beam former filtering unit is configured to apply a current frequency dependent directional gain G.sub.DIR,i to each of the multitude of electric input signals IN.sub.i, (here gains G.sub.1, G.sub.2 applied to electric input signals IN1, IN2 (or rather to sub-band versions X.sub.1, X.sub.2 thereof)). The gain control unit (CONT) is configured to determine the modified full-on gain value G′.sub.FOG in dependence of the current directional gains G.sub.DIR,i, i=1, . . . , M, (here G.sub.1=|W.sub.1|, G.sub.2=|W.sub.2|) and a previously determined full-on gain value G.sub.FOG, which is stored in a memory (MEM) of the hearing aid (e.g. provided during manufacturing of the hearing aid or during fitting of the hearing aid to the needs of a particular user). The gain control unit (CONT) is operatively connected to the gain unit (HAG) and receives the current (requested) hearing aid gain G.sub.HA. In an embodiment, the current (requested) hearing aid gain G.sub.HA is used by the gain control unit to influence the temporal effect of changes in the modified value of the full-on gain G′.sub.FOG, see FIG. 2, 3.
Implementation Example
[0110] The following example shows how the FOG Limitation System (represented by the gain control unit CONT in FIG. 2) can be implemented in a multichannel sub-band system with complex valued sub-band signals (X.sub.1, X.sub.2 in FIG. 1) and complex microphone channel gains (W.sub.1, W.sub.2) in the Directionality System (BFUb in FIG. 1).
[0111] FIG. 2 shows an embodiment of control unit (CONT) for implementing a Full-On Gain limitation system according to the present disclosure.
[0112] The control unit comprises an ABS-MAX unit providing a maximum value G.sub.DIR,max of the current directional gains based on the current complex weights (W.sub.1, W.sub.2). Since the directionality gains (W.sub.1, W.sub.2) are complex-valued, they first pass an ABS operation (ABS) providing real gain values G1, G2 (G.sub.1=|W.sub.1|, G.sub.2=|W.sub.2|). Subsequently the maximum value is taken over microphone channels 1, 2, G.sub.DIR,max=MAX {G.sub.DIR,i}, i=1, 2 (MAX) (e.g. for each frequency sub-band k).
[0113] After determining the maximum value (G.sub.DIR,max) among the directional gains (G.sub.1, G.sub.2), the next step is to calculate the distance (or margin) (ΔG.sub.FOGm) between the actual device gain, i.e. Directionality Gain (G.sub.DIR,max)+Desired Amplification Gain (G.sub.HA), and the (predetermined) FOG Limit (G.sub.FOG), cf. inputs to summation unit ‘+’ in FIG. 2 providing ΔG.sub.FOGm=G.sub.FOG−(G.sub.DIR,max+G.sub.HA). If the distance measure ΔG.sub.FOGm (also in the following termed the ‘current full on gain margin’) is positive, the amplification gain G.sub.HA does not need limiting. If the value is below zero, the amplification gain needs to be limited in order to maintain the maximum allowable gain for device stability.
[0114] The lower part of FIG. 2 comprising time constant control unit TC-CT, smoothing unit FOG-SM, and combination unit ‘+’ is configured to control the FOG Correction ΔG.sub.FOG=G.sub.DIR,max dynamically. Only in the case where the FOG Limit G.sub.FOG is almost reached (ΔG.sub.FOGm, decreases towards 0 (absolute)), the time constant control block (TC-CT) speeds up the calculation of smoothing block FOG-SM (e.g. by decreasing or setting the release time to a low value τ.sub.rel,FAST, when the current full on gain margin ΔG.sub.FOGm is smaller than a threshold value ΔG.sub.LIM,fast). In other words, the time constant control unit TC-CT controls time constants of the smoothing process and provides time constant control signal TAU to the full-on gain smoothing unit FOG-SM. Based on control signal TAU and the current maximum directional gain value G.sub.DIR,max a smoothed maximum directional gain value <G.sub.DIR,max> is provided by the full-on gain smoothing unit FOG-SM. A resulting modified full-on gain value G′.sub.FOG is provided by combination unit ‘+’ as a difference (in [dB]) between the predefined full-on gain G.sub.FOG and the smoothed maximum directional gain value <G.sub.DIR,max> (i.e. G′.sub.FOG=G.sub.FOG−<G.sub.DIR,max>) Thereby a correct amplification gain limit G′.sub.FOG can be (immediately) provided (i.e. G′.sub.FOG˜G.sub.FOG−G.sub.DIR,max), when the hearing aid gain G.sub.HA is close to the FOG Limit G.sub.FOG (relatively fast or no smoothing) and a slowly varying (slowly smoothed) modified full-on gain value G′.sub.FOG can otherwise be provided. The risk of artifacts being introduced by the modification of the full-on gain can thereby be decreased.
[0115] The upper part of the drawing comprising DIR-control smoothing unit DCT-SM and mapping unit MAP is configured to determine a control parameter DIRctr (e.g. taking on values between 0 and 1), which can be used to control the directionality system (beamformer filtering unit BFU in FIG. 1. The smoothing unit DCT-SM receives the current full on gain margin ΔG.sub.FOGm from summation unit ‘+’ and provides an appropriate attack and release time to a smoothing of the full on gain margin ΔG.sub.FOGm. This is done with a view to the smoothing of the FOG Correction G.sub.DIR,max\ performed in the FOG-SM unit (the current values of attack and release times of the two smoothing processes are e.g. exchange and evaluated, cf. dashed arrow between the respective DCT-SM and TC-CT units). The smoothing unit DCT-SM provides a smoothed full on gain margin <ΔG.sub.FOGm> (signal DFOG in FIG. 2) to the mapping unit MAP. The mapping unit MAP and its control signal DIRctr implements the following scheme for controlling the directionality system (BFUa, BFUb in FIG. 1) based on the smoothed full on gain margin <ΔG.sub.FOGm>. A value of the control parameter DIRctr of “0” means that the directionality system is forced to be “off” (no directionality). If the value is “1”, the directionality system is free to operate normally (no constraints from the control unit CONT). For values between “0” and “1”, the directionality system is restrained to diminish the directional gains as DIRctr decreases from “1” to “0” and thereby to increase the current full on gain margin ΔG.sub.FOGm, thus allowing a larger hearing aid gain G.sub.HA to be applied to the beam formed signal (Y.sub.BF in FIG. 1) before the gain limit (G′.sub.FOG) is reached. In other words, gain is moved from the directionality system (by decreasing G.sub.1, G.sub.2) to the hearing aid gain (G.sub.HA), thereby prioritizing to provide gain G.sub.HA to the user at the cost of directionality. The movement of gain from the directionality system to the FOG gain limit (or vice versa) sets restrictions on the time constants for the smoothing of the FOG Correction G.sub.DIR,max in the lower part of FIG. 2 and the full on gain margin ΔG.sub.FOGm in the upper part of FIG. 2 (to avoid the introduction of artifacts), as indicated by the dashed connection between the DCT-SM and TC-CT units.
[0116] FIG. 3A is an illustration of an exemplary scheme for operating a gain control unit of a hearing aid according to the present disclosure. FIG. 3A illustrates a situation of increasing need for gain (hearing aid gain G.sub.HA) to be provided to the user over a first period of time (Time, t), t.sub.0<t<t.sub.6, and a second period of time, t.sub.6<t<t.sub.12, where the need for gain decreases. In an intermediate time period, t.sub.5<t<t.sub.7 (overlapping with the first and second time periods), the modified full-on gain sets a limit on the hearing aid gain (G.sub.HA, providing modified gain G′.sub.HA). The target gain is indicated in dotted line (during the intermediate time period t.sub.5<t<t.sub.7). The realized gain is indicated in solid line (during t.sub.0<t<t.sub.5 and t.sub.7<t<t.sub.12). The left and right vertical axes of the gain graph are gain-axes referring to a ‘Target gain’ G.sub.HA′ comprising the sum of the requested hearing aid gain G.sub.HA and the FOG correction, G.sub.HA′=G.sub.HA+G.sub.DIR,MAX. The leftmost, reversed axis shows the full on gain margin ΔG.sub.FOGm=G.sub.FOG−(G.sub.DIR,max+G.sub.HA) having its zero where the requested hearing aid gain G.sub.HA is equal to the full-on gain limit G.sub.FOG (because G.sub.DIR,MAX=0 for target gain larger than G.sub.DIR,OFF, cf. indication on the rightmost target gain axis). Between the leftmost target gain-axis and the full on gain margin ΔG.sub.FOGm-axis, a graph illustrating an exemplary functional dependence of the beam former control signal DIRctr on full on gain margin ΔG.sub.FOGm is shown. The shown graph implements a scheme for moving gain from the directionality system to the hearing aid gain (when certain criteria are fulfilled).
[0117] In the first time period (t.sub.0<t<t.sub.6, denoted ‘Release’ in the top part of FIG. 3A), a steady increased need for gain is assumed (e.g. corresponding to a situation where a target sound source decreases slowly in signal strength, i.e. received SPL, at the user), or where a noise source is gradually introduced. A steadily increasing target gain corresponds to a steadily decreasing full-on gain margin. Consequently, the release time constant τ.sub.rel of the smoothing algorithm for the full on gain margin ΔG.sub.FOGm is the important one in the first time period (cf. FIG. 3B). The first time period is divided into sub-time periods (determined by individual points in time t.sub.0, t.sub.1, t.sub.2, t.sub.3, t.sub.4, t.sub.5, t.sub.6), where the requested hearing aid gain G.sub.HA is in different ranges. The reaction of the adaptive full-on gain modification algorithm in each gain-range is briefly discussed in the following.
[0118] Time period t.sub.0<t<t.sub.1: G.sub.HA′≦G.sub.DIR,ON (cf. Target gain scale to the right in FIG. 3A, and the left graph showing DIRctr(ΔG.sub.FOGm)): In this gain range, the adaptive full-on gain modification algorithm is slowly reacting and the directional system is unrestrained (by the present algorithm). DIRctr=“1”.
[0119] When the requested hearing aid gain G.sub.HA approaches the FOG Limit G.sub.FOG, from below (G.sub.HA′<G.sub.FOG), the attack/release smoothing and mapping algorithm (c.f. upper part of FIG. 2, units DCT-SM and MAP) controls how fast the directionality system is forced to go from a (normal, unrestrained) mode of operation (G.sub.HA′≦G.sub.DIR,ON in FIG. 3A) to the “off” state (G.sub.HA′≧G.sub.DIR,OFF in FIG. 3A). It is important to note that these settings have to be set carefully since they are parameters of a recursive system (as mentioned above in connection with FIG. 2). If this system acts too fast, it will result in undesired on/off oscillation of the directionality system.
[0120] Time period t.sub.1<t<t.sub.2: G.sub.DIR,ON≦G.sub.HA′≦G.sub.DIR,OFF (cf. scale to the right in FIG. 3A, and the left graph showing DIRctr(ΔG.sub.FOGm)): The directionality system is in a restrained mode of operation (denoted ‘Transition’ in the left DIRctr(ΔG.sub.FOGm)-graph in FIG. 3A) controlled by signal DIRctr, “0”<DIRctr<“1”, where directionality gains G1, G2 are decreased with increasing G.sub.HA′ (cf. downwards pointing bold arrow denoted Increasing retraction of DIR-gain in FIG. 3A). The transition from DIRctr=“1” to “0” occurs between times t.sub.1 and t.sub.2. When the target gain is larger than G.sub.DIR,OFF (where the directional system is off), directional gains (G1, G2) are 1 (0 dB), and thus G.sub.DIR,MAX=1 (0 dB) as indicated on the rightmost target gain axis. This mechanism is important to maintain hearing aid gain (as long as possible, at the cost of DIR-gain).
[0121] Time period t<t.sub.3: G.sub.HA≦G.sub.LIM,slow (cf. axis to the left in FIG. 3A, and FIG. 3B, ΔG.sub.FOGm≦ΔG.sub.LIM,slow): The requested hearing aid gain G.sub.HA is still below the threshold G.sub.LIM,slow (i.e. ΔG.sub.FOGm>ΔG.sub.LIM,slow in FIG. 3C), where the modified full-on gain is provided fast, i.e. in a mode (still) providing a Slow adaptation rate of G′.sub.FOG.
[0122] Time period t.sub.3<t<t.sub.a: G.sub.LIM,slow≦G.sub.HA′≦G.sub.LIM,fast (cf. axis to the left in FIG. 3A, and FIG. 3B, ΔG.sub.LIM,slow≧ΔG.sub.FOGm≧ΔG.sub.LIM,fast): requested hearing aid gain G.sub.HA′ is in a range where the modified full-on gain G′.sub.FOG is provided with increasing speed for increasing requested hearing aid gain G.sub.HA′ (but still below a fastest provision, i.e. in a mode providing a Changing adaptation rate of G′.sub.FOG). Looking at the ΔG.sub.FOGm axis to the left, this corresponds to a decreasing full on gain margin ΔG.sub.FOGm resulting in an increased adaptation rate (i.e. a decreasing release time constant τ.sub.rel (cf. FIG. 3B), so that the modified full-on gain value can be provided (and taken into use) with increased speed the closer we get to the ΔG.sub.LIM,fast threshold.
[0123] Time period t.sub.4<t<t.sub.5: G.sub.LIM,fast≦G.sub.HA′≦G.sub.FOG (cf. axis to the left in FIG. 3A, and FIG. 3B, ΔG.sub.LIM,fast≧ΔG.sub.FOGm): The requested hearing aid gain G.sub.HA′ is above the threshold for providing immediate (or maximum adaptation rate) of the full on gain margin ΔG.sub.FOGm and thus of the modified full-on gain G′.sub.FOG (or rather the full on gain margin ΔG.sub.FOGm) (ΔG.sub.FOG<ΔG.sub.LIM,fast).
[0124] Intermediate time period t.sub.5<t<t.sub.7: G.sub.FOG≦G.sub.HA′ (dotted part of the gain curve). In this time period, the target gain is larger than the full-on gain G.sub.FOG and hence the hearing aid gain G.sub.HA is limited to the full-on gain value G′.sub.FOG=G.sub.FOG. At time t.sub.6, the target gain starts to decrease, which prompts the release time constant τ.sub.rel (cf. FIG. 3B) to change (increase) to a value τ.sub.rel,SLOW providing a slow adaptation rate of the modified full-on gain G′.sub.FOG. (cf. vertical upwards pointing arrow on the τ.sub.rel axis in FIG. 3B).
[0125] In the second time period (t.sub.6<t<t.sub.12, denoted ‘Attack’ in the top part of FIG. 3A), a steady decreased need for gain is assumed (e.g. corresponding to a situation where a target sound source increases slowly in signal strength, i.e. received SPL, at the user), or where a noise source is gradually removed or decreased in strength. A steadily decreasing target gain corresponds to a steadily increasing full-on gain margin. Consequently, the attack time constant τ.sub.att of the smoothing algorithm for the full on gain margin ΔG.sub.FOGm is the important one in the second time period (cf. FIG. 3C). The second time period is divided into sub-time periods (determined by individual points in time t.sub.6, t.sub.7, t.sub.8, t.sub.9, t.sub.10, t.sub.11, t.sub.12), where the requested hearing aid gain G.sub.HA is in different ranges. The reaction of the adaptive full-on gain modification algorithm in each gain-range is briefly discussed in the following.
[0126] Time period t.sub.7<t<t.sub.8: G.sub.LIM,fast≦G.sub.HA′≦G.sub.FOG (cf. axis to the left in FIG. 3A, and FIG. 3C, ΔG.sub.LIM,fast≧ΔG.sub.FOGm): The attack time of the smoothing algorithm for the full on gain margin ΔG.sub.FOGm is set to a fixed relatively large value τ.sub.att,x providing relatively slow adaption of the modified full-on gain G′.sub.FOG.
[0127] Time period t.sub.8<t<t.sub.9: G.sub.LIM,slow<G.sub.HA′≦G.sub.LIM,fast (cf. axis to the left in FIG. 3A, and FIG. 3C, ΔG.sub.LIM,slow≧ΔG.sub.FOGm≧ΔG.sub.LIM,fast): The attack time of the smoothing algorithm for the full on gain margin ΔG.sub.FOGm stays fixed at the relatively large value τ.sub.att,x providing relatively slow adaption of the modified full-on gain G′.sub.FOG.
[0128] Time period t.sub.9<t: G.sub.HA′≦G.sub.LIM,slow (cf. axis to the left in FIG. 3A, and FIG. 3C, ΔG.sub.FOGm≧ΔG.sub.LIM,slow): The attack time of the smoothing algorithm for the full on gain margin ΔG.sub.FOGm stays fixed at the relatively large value τ.sub.att,x providing relatively slow adaption of the modified full-on gain G′.sub.FOG.
[0129] Time period t.sub.10<t<t.sub.11: G.sub.DIR,ON≦G.sub.HA′≦G.sub.DIR,OFF (cf. scale to the right in FIG. 3A, and the left graph showing DIRctr(ΔG.sub.FOGm)): The directionality system is in a restrained mode of operation (denoted ‘Transition’ in the left DIRctr(ΔG.sub.FOGm)-graph in FIG. 3A) controlled by signal DIRctr, “0”<DIRctr <“1”, where directionality gains G1, G2 are allowed to increase with decreasing G.sub.HA′ (cf. upwards pointing bold arrow denoted Decreasing retraction of DIR-gain in FIG. 3A). The transition from DIRctr=“1” to “0” occurs between times t.sub.10 and t.sub.11. When the target gain is smaller than G.sub.DIR,ON (where the directional system is in a normal ON-state), directional gains (G1, G2) are allowed to vary freely (DIRctr=1) in control of the beam former filtering unit.
[0130] Time period t.sub.11<t<t.sub.12: G.sub.HA′≦G.sub.DIR,ON (cf. Target gain scale to the right in FIG. 3A, and the left graph showing DIRctr(ΔG.sub.FOGm)): In this gain range, the adaptive full-on gain modification algorithm is slowly reacting and the directional system is unrestrained (by the present full-on gain control algorithm). DIRctr=“I”.
[0131] FIG. 3B illustrates an exemplary functional relationship between the current full on gain margin ΔG.sub.FOGm and the attack τ.sub.att and release τ.sub.rel time constants involved in determining the smoothed value <G.sub.DIR,max> in a first time interval from a start time t.sub.0 to an intermediate time t.sub.6 during increasing desired hearing aid gain G.sub.HA, (i.e. during decreasing full on gain margin ΔG.sub.FOGm). FIG. 3B corresponds to an increasing target gain situation (first time period t.sub.0-t.sub.6 denoted Release in FIG. 3A). The time axis (Time, t) indicates start and end of the first time period (t.sub.0-t.sub.6) in FIG. 3A. The release time constant τ.sub.rel decreases from a larger (slow) time constant τ.sub.rel,SLOW to a smaller (fast) time constant τ.sub.rel,FAST, when current full on gain margin ΔG.sub.FOGm decreases from ΔG.sub.LIM,slow to ΔG.sub.LIM,fastIn the embodiment of FIG. 3B, the transition from τ.sub.rel,SLOW to τ.sub.rel,FAST, is shown to be liner. This need not be the case however. In another embodiment, it may be non-linear, e.g. stepwise linear or of a sigmoid form.
[0132] FIG. 3C illustrates an exemplary functional relationship between the current full on gain margin ΔG.sub.FOGm and the attack τ.sub.att and release τ.sub.rel time constants involved in determining the smoothed value <G.sub.DIR,max> in a second time interval from an intermediate time t.sub.6 to an end time t.sub.11 during decreasing desired hearing aid gain G.sub.HA, (i.e. during increasing full on gain margin ΔG.sub.FOGm) FIG. 3C corresponds to a decreasing target gain situation (second time period t.sub.7-t.sub.12 denoted Attack in FIG. 3A). The time axis (Time, t) indicates start and end of the second time period (t.sub.7-t.sub.12) in FIG. 3A. The attack and release time constants (τ.sub.att, τ.sub.rel) are set to constant relatively large values (τ.sub.att,X τ.sub.rel,SLOW), providing relatively slow smoothing (adaptation). In the embodiment of FIG. 3B, 3C, the attack time constant τ.sub.att,x is larger than the release time constant τ.sub.rel,SLOW.
[0133] Outside the transition of the release time constant from slow to fast τ.sub.rel,SLOW to τ.sub.rel,FAST (FIG. 3B), the attack and release time constants of FIGS. 3B and 3C are shown to be constant (τ.sub.att,x, τ.sub.rel,SLOW, τ.sub.rel,FAST) for varying full on gain margin ΔG.sub.FOGm. This need not be the case, however. In an embodiment, one or more of the attack and release time constants are non-linear, e.g. non-linearly, e.g. logarithmically approaching a fixed value.
[0134] In an embodiment (with reference to FIG. 3A) G.sub.DIR,OFF=G.sub.LIM,slow, (ΔG.sub.DIR,OFF=ΔG.sub.LIM,slow), so that the increase of adaptation rate of the modified full-on gain G′.sub.FOG is started when the full directional gain has been moved to the hearing aid gain (G.sub.DIR,max=0).
[0135] In a multi-channel implementation of the directionality system (BFUa, BFUb in FIG. 1) and the amplification system (HAG in FIG. 1), the FOG Limit algorithm according to the present disclosure can be implemented in independent channels. The FOG Limit (G.sub.FOG, G′.sub.FOG) is typically a frequency dependent function. The embodiments described in the present disclosure are implemented in the time frequency domain (signals of individual frequency sub-bands are treated individually). The present scheme may, however, be implemented fully or partially in the time domain.
[0136] In an embodiment, the gain control unit is configured to determine a beam former control signal DIRctr for controlling the beam former filtering unit between an un-restrained ON-state, when said current full on gain margin ΔG.sub.FOGm is above a first threshold value ΔG.sub.DIR,ON, and an OFF-state, when said current full on gain margin ΔG.sub.FOGm is below a second threshold value ΔG.sub.DIR,OFF. FIG. 3A (left side) illustrates an exemplary functional relationship between the current full on gain margin ΔG.sub.FOGm and the beam former control signal DIRctr for controlling the beam former filtering unit. The DIRctr(ΔG.sub.FOGm)-graph in FIG. 3A shows that the beam former filtering control signal DIRctr is set to “1” (corresponding to an un-restrained ON-state of the beamformer filtering unit, e.g. to operate normally), when the current full on gain margin ΔG.sub.FOGm is above a first threshold value ΔG.sub.DIR,ON. FIG. 3B further shows that the beam former filtering control signal DIRctr is set to “0” (corresponding to an OFF-state of the beamformer filtering unit), when the current full on gain margin ΔG.sub.FOGm is below a second threshold value ΔG.sub.DIR,OFF. In the OFF-state of the beam former filtering unit may e.g. be fixed to an omni-directional mode of operation. FIG. 3A further shows that when the full on gain margin ΔG.sub.FOGm is changed between the first and second threshold values ΔG.sub.DIR,OFF, ΔG.sub.DIR,ON, the DIRctr signal changes linearly between 0 and 1. In this range, the beamformer filtering unit is in a transition-state between an OFF-state and an un-restrained ON-state, where the current directional gains G.sub.DIR,i, i=1, . . . , M are influenced (limited, attenuated) by the gain control unit via beam former control signal DIRctr.
[0137] In an embodiment, the gain control unit comprises a configurable smoothing unit configured to determine a smoothed value <G.sub.DIR,max> of the maximum value G.sub.DIR,max of the current directional gains, and to use the smoothed value <G.sub.DIR,max> in the determination of the modified full-on gain value G′.sub.FOG, e.g. G′.sub.FOG=G.sub.FOG−<G.sub.DIR,max>. The configurable smoothing unit may e.g. be configured to use different attack and release times for the smoothing. In an embodiment, the smoothing attack and/or release time are controllable in dependence of one or more parameters. FIG. 3B illustrates an exemplary functional relationship between the current full on gain margin ΔG.sub.FOGm, and the release time constant τ.sub.rel involved in determining the smoothed value <G.sub.DIR,max>.
[0138] In the exemplary scheme illustrated by FIG. 3B, the gain control unit is configured to set a release time constant τ.sub.rel involved in determining the smoothed value <G.sub.DIR,max> to a value equal to a first value τ.sub.rel,FAST, in case the current full-on gain margin ΔG.sub.FOGm is below a first threshold value ΔG.sub.LIM,fast, i.e. for ΔG.sub.FOGm<ΔG.sub.LIM,fast, where ΔG.sub.LIM,fast is larger than zero. This is advantageous to ensure a fast and immediate adaptation of the modified full-on gain value G′.sub.FOG, in case the current full on gain margin ΔG.sub.FOGm becomes small (i.e. close to zero). According to the scheme of FIG. 3B, the release time constant τ.sub.rel is increased (linearly) when the current full on gain margin ΔG.sub.FOGm is increased above the threshold value ΔG.sub.LIM,fast, but below a second threshold value ΔG.sub.LIM,slow, In FIG. 3C the release time constant τ.sub.rel is set to a second value τ.sub.rel,SLOW, when the current full on gain margin ΔG.sub.FOGm is increased above the second threshold value ΔG.sub.LIM,slow.
[0139] Typically, the currently used attack time constant τ.sub.att is set to a value larger than or equal to the currently used release time constant τ.sub.rel.
[0140] FIG. 4 shows a flow diagram of an embodiment of a method of operating a hearing aid according to the present disclosure. The hearing aid comprises a forward path comprising a multitude of input units for providing a multitude of electric input signals IN.sub.i, i=1, . . . , M, representative of sound, and an output unit for providing stimuli perceivable by a user as sound based on a processed signal or a signal derived therefrom. The method comprises [0141] S1. providing a multitude of electric input signals IN.sub.i, i=1, . . . , M, representative of sound, [0142] S2. providing a beam formed signal Y.sub.BF from said multitude of electric input signals IN.sub.i, including applying a current frequency dependent directional gain G.sub.DIR,i to each of said multitude of electric input signals IN.sub.i, [0143] S3. applying a hearing aid gain G.sub.HA to said beam formed signal Y.sub.BF, and providing a processed signal, and [0144] S4. providing a previously determined full-on gain value G.sub.FOG, [0145] S5. limiting said hearing aid gain G.sub.HA to a modified full-on gain value G′.sub.FOG, [0146] S6. determining said modified full-on gain value G′.sub.FOG in dependence of said current directional gains G.sub.DIR,i, i=1, . . . , M, and said previously determined full-on gain value G.sub.FOG.
[0147] It is intended that the structural features of the devices described above, either in the detailed description and/or in the claims, may be combined with steps of the method, when appropriately substituted by a corresponding process.
[0148] As used, the singular forms “a,” “an,” and “the” are intended to include the plural forms as well (i.e. to have the meaning “at least one”), unless expressly stated otherwise. It will be further understood that the terms “includes,” “comprises,” “including,” and/or “comprising,” when used in this specification, specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof. It will also be understood that when an element is referred to as being “connected” or “coupled” to another element, it can be directly connected or coupled to the other element but an intervening elements may also be present, unless expressly stated otherwise. Furthermore, “connected” or “coupled” as used herein may include wirelessly connected or coupled. As used herein, the term “and/or” includes any and all combinations of one or more of the associated listed items. The steps of any disclosed method is not limited to the exact order stated herein, unless expressly stated otherwise.
[0149] It should be appreciated that reference throughout this specification to “one embodiment” or “an embodiment” or “an aspect” or features included as “may” means that a particular feature, structure or characteristic described in connection with the embodiment is included in at least one embodiment of the disclosure. Furthermore, the particular features, structures or characteristics may be combined as suitable in one or more embodiments of the disclosure. The previous description is provided to enable any person skilled in the art to practice the various aspects described herein. Various modifications to these aspects will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other aspects.
[0150] The claims are not intended to be limited to the aspects shown herein, but is to be accorded the full scope consistent with the language of the claims, wherein reference to an element in the singular is not intended to mean “one and only one” unless specifically so stated, but rather “one or more.” Unless specifically stated otherwise, the term “some” refers to one or more.
[0151] Accordingly, the scope should be judged in terms of the claims that follow.
