WO2009153718A1

WO2009153718A1 - Earphone arrangement and method of operation therefor

Info

Publication number: WO2009153718A1
Application number: PCT/IB2009/052511
Authority: WO
Inventors: Sriram Srinivasan
Original assignee: Koninklijke Philips Electronics N.V.
Priority date: 2008-06-17
Filing date: 2009-06-12
Publication date: 2009-12-23

Abstract

An earphone arrangement comprises a first and second earphone arranged to operate in a binaural arrangement. An audio data exchange between the earphones supports a binaural processing of the earphone arrangement. The earphones may specifically be hearing aids each of which comprises one or more microphones. An audio environment processor (119) of the hearing aids determines an audio environment characteristic in response to signals from at least one microphone (101, 103) of the first earphone. A binaural controller (117) controls a data rate of an audio data exchange between the first and second earphone in response to the audio environment characteristic. Specifically, the data rate of the audio data exchange may be varied in dependence on a signal to interference ratio, a signal to noise ratio and/or an angle of arrival estimate for an interferer. The approach may typically reduce power consumption and increase battery life.

Description

Earphone arrangement and method of operation therefor

FIELD OF THE INVENTION

The invention relates to an earphone arrangement and a method of operation therefor and in particular, but not exclusively, to binaural processing for hearing aids.

BACKGROUND OF THE INVENTION

Advanced processing of audio signals has become increasingly important in many areas including e.g. telecommunication, content distribution etc. A particular important area for audio signal processing is in the field of earphones and in particular hearing aids. In recent years, hearing aids have increasingly applied complex audio processing algorithms to provide an improved user experience and assistance to the user. For example, audio processing algorithms have been used to provide an improved signal to noise ratio between a desired sound source and an interfering sound source resulting in a clearer and more perceptible signal being provided to the user. In particular, hearing aids have been developed which include more than one microphone with the audio signals of the microphones being dynamically combined to provide directivity for the microphone arrangement. Such directivity may be achieved by beam forming algorithms which in some cases may be adaptive such that they are dynamically directed towards a desired sound source. As another example, noise cancelling algorithms may be applied to reduce the interference caused by undesired sound sources and background noise.

A monaural hearing aid solution consists of a single hearing aid which fits on one ear of the user and typically provides functions such as compression, amplification and feedback cancellation. Most modern hearing aids have two or more microphones mounted on them and provide directional processing through spatial filtering techniques. A system wherein two independent monaural hearing aids are used simultaneously is still considered a bilateral monaural hearing aid solution.

A binaural system is one where hearing aids on two ears collaborate with one another. Such an approach promises at least three advantages compared to a monaural system: improved speech intelligibility in noise (especially when the speech and noise signals originate from different directions), improved localization resulting in a better perception of the auditory scene, and improved noise reduction due to a larger microphone spacing and a higher number of microphones.

In binaural operation the signals from both hearing aids are combined to generate the audio output signals from the hearing aids. Specifically, binaural beam forming may be achieved by combining the microphone signals from both left and right hearing aids and this binaural beam forming may provide a substantially improved beam forming relative to e.g. beam forming based only on the microphones from one hearing aid.

Binaural beam forming in hearing aids requires exchange of microphone signals between the hearing aids worn on the left and right ears. This exchange occurs over a wireless link as it is not generally desired to have a wired link for aesthetic reasons. However, wireless transmission introduces power and computational constraints and in particular introduces significant additional power consumption by the hearing aids resulting in a reduced battery life, which is an important concern for hearing aid users. Hence, an improved ear phone/ hearing aid system would be advantageous and in particular a system allowing increased flexibility, facilitated implementation, reduced resource usage, practical implementation, reduced power consumption, increased battery life and/or improved performance would be advantageous.

SUMMARY OF THE INVENTION

Accordingly, the Invention seeks to preferably mitigate, alleviate or eliminate one or more of the above mentioned disadvantages singly or in any combination. According to an aspect of the invention there is provided earphone arrangement comprising a first and second earphone arranged to operate in a binaural arrangement, the first earphone comprising: determining means for determining an audio environment characteristic in response to a signal from at least one microphone of the first earphone; and control means for controlling a data rate of an audio data exchange between the first and second earphone in response to the audio environment characteristic, the audio data exchange supporting a binaural processing of the earphone arrangement. The invention may allow improved performance of an ear phone arrangement which specifically may be a hearing aid arrangement. In particular, the invention may allow reduced resource consumption and in many embodiments a reduced power consumption and thus increased battery life can be achieved. Specifically, in many embodiments, the invention may allow performance close to a fully binaural operation but with a substantially reduced resource usage.

The system may specifically allow a flexible and dynamic system wherein the resource expended by the binaural operation may be continuously adapted to reflect the performance improvement achievable by the binaural operation in the current conditions.

The first and/or second earphone may be a hearing aid. The arrangement may comprise means for performing a binaural processing in response to the exchanged audio data. In particular, a binaural combination of signals from the first and second earphone may be performed using the exchanged audio data. The binaural operation may for example be a binaural beam forming.

The determining means may be comprised in the first earphone. The term monaural refers to a system wherein each ear is considered independently and the word binaural refers to a system wherein audio fed to one ear is determined in response to signals obtained (directly or indirectly) from microphones associated with both ears. Specifically, for hearing aids, the term monaural may be used to refer to a hearing aid for one ear which operates independently of any other hearing aid whereas the term binaural may refer to the use of interworking hearing aids for each ear. Each earphone or hearing aid may specifically comprise one or more microphones.

The exchanged audio data may specifically correspond to one or more microphone signals.

In some embodiments, the arrangement may further comprise second determining means for determining a second audio environment characteristic in response to signals from at least one microphone of the second earphone; and second control means for controlling a data rate of a second audio data exchange between the first and second earphone in response to the second audio environment characteristic, the second audio data exchange supporting a binaural processing of the earphone arrangement. The audio data exchange may include the second audio data exchange i.e. a data rate of audio exchanged between the two headphones may depend on an audio environment characteristic determined from both at least one microphone of the first earphone and at least one microphone of the second earphone.

In accordance with an optional feature of the invention, the control means is arranged to reduce the data rate if the audio environment characteristic does not meet a criterion. This may allow improved performance, facilitated operation and/or reduced resource usage. In particularly, it may in many embodiments allow an efficient and low complexity trade-off between e.g. audio quality and battery life.

The reduction of the data rate may specifically be a switching off of the exchange of all or part of one or more microphone signals for which audio data is exchanged from the first earphone to the second earphone or from the second earphone to the first earphone. Specifically, exchanged audio data supporting binaural processing for the first earphone and/or the second earphone may be switched off if the criterion is not met. Thus, binaural operation may only be supported if the criterion is met. In accordance with an optional feature of the invention, the audio environment characteristic comprises a Signal to Interference Ratio, SIR, indication, a Signal to Noise Ratio, SNR, indication and an angle of arrival estimate for an interferer; and the criterion comprises: a requirement that the SIR indication is below a first threshold; a requirement that the SNR indication is above a second threshold; and a requirement that the angle of arrival estimate is within an angle interval.

This may allow improved performance, facilitated operation and/or reduced resource usage. In particular, it has been found that the requirements may allow an efficient detection of audio environments wherein binaural operation is particularly desirable and likely to justify the additional resource usage. The angle interval may specifically comprise an angle corresponding to a frontal direction for the earphone arrangement when in use.

Particularly advantageous performance may typically be achieved for a first threshold between 0 and 20 dB (and in particular for a first threshold of substantially 10 dB) and/or a second (SNR) threshold between 0 and 2OdB (and in particular for a second threshold of substantially 10 dB) and an angle interval of between 100-140 degrees. In accordance with an optional feature of the invention, the audio data exchange is an exchange of audio data from the first earphone to the second earphone.

This may allow improved performance, facilitated operation and/or reduced resource usage. The audio environment characteristic determined from the at least one microphone of the first earphone may control the transmission of audio data to the second earphone thereby allowing the resource usage expended by binaural operation to be flexibly adjusted based on the conditions experienced at the first earphone.

The arrangement may additionally exchange audio data from the second to the first earphone. This reverse exchange of audio data may be controlled in response to the audio environment characteristic independently of or correlated with the exchange of audio data from the first earphone to the second earphone.

In accordance with an optional feature of the invention, the audio data exchange is an exchange of audio data from the second earphone to the first earphone. This may allow improved performance, facilitated operation and/or reduced resource usage. The audio environment characteristic determined from the at least one microphone of the first earphone may control the transmission of audio data from the second earphone thereby allowing the resource usage expended by binaural operation to be flexibly adjusted based on the conditions experienced at the first earphone. The arrangement may additionally exchange audio data from the first to the second earphone. This reverse exchange of audio data may be controlled in response to the audio environment characteristic independently of or correlated with the exchange of audio data from the second earphone to the first earphone.

In accordance with an optional feature of the invention, the audio environment characteristic is indicative of a performance gain of the binaural processing of the earphone arrangement relative to a monaural operation of the earphone arrangement.

This may allow improved performance, facilitated operation and/or reduced resource usage.

In accordance with an optional feature of the invention, the audio environment characteristic comprises a Signal to Interference Ratio, SIR, indication.

This may allow improved performance, facilitated operation and/or reduced resource usage. The inventor has realized that a SIR measure provides particularly useful information for dynamically adjusting the binaural processing of an earphone arrangement and of the data exchange resource used by this. In accordance with an optional feature of the invention, the audio environment characteristic comprises a Signal to Noise Ratio, SNR, indication.

This may allow improved performance, facilitated operation and/or reduced resource usage. The inventor has realized that a SNR measure provides particularly useful information for dynamically adjusting the binaural processing of an earphone arrangement and of the data exchange resource used by this.

In accordance with an optional feature of the invention, the audio environment characteristic comprises an angle of arrival estimate for an interferer.

This may allow improved performance, facilitated operation and/or reduced resource usage. The inventor has realized that an angle of arrival for an interferer provides particularly useful information for dynamically adjusting the binaural processing of an earphone arrangement and of the data exchange resource used by this. The interferer may for example be a dominant interferer and/or may be an interferer which is received at a signal level that exceeds a given threshold (which may be dynamically determined in response to audio characteristics).

In accordance with an optional feature of the invention, the control means is arranged to reduce the data rate by reducing a frequency bandwidth represented by exchanged audio data if the angle of arrival estimate is within a first angle interval.

This may allow improved performance, facilitated operation and/or reduced resource usage. This may in particular allow reduced resource usage while allowing any quality degradation to be reduced by a partial binaural operation being enabled. Furthermore, for a given reduced data rate this may allow improved benefit of the performed binaural processing.

The angle interval may specifically comprise an angle corresponding to a frontal direction for the earphone arrangement when in use. Particularly advantageous performance may typically be achieved for a first angle interval of between 100-140 degrees.

In accordance with an optional feature of the invention, the control means is arranged to reduce a data rate for audio data transmitted from the second earphone to the first earphone if the angle of arrival is in an angle interval corresponding to a head shadow for the second earphone when in use.

The reduction may furthermore be performed only if a SNR indication is below a threshold. The first angle interval may have a size of between 50 and 70 degrees and may specifically include an angle of 30 degrees relative to the frontal direction when in use and in the direction of the ear associated with the first earphone when in use.

In accordance with an optional feature of the invention, the control means is arranged to set a frequency bandwidth for audio signals exchanged by the audio data in response to the audio environment characteristic. This may allow improved performance, facilitated operation and/or reduced resource usage. In particular it may allow reduced resource usage while allowing any quality degradation to be reduced by a partial binaural operation being enabled. Furthermore, for a given reduced data rate this may allow improved benefit of the performed binaural processing. In accordance with an optional feature of the invention, the audio data exchange comprises a first audio data exchange from the first earphone to the second earphone and a second audio data exchange from the second earphone to the first earphone; and the control means is arranged to control the data rate differently in response to the audio environment characteristic for the first audio data exchange and the second audio exchange.

This may allow improved performance, facilitated operation and/or reduced resource usage. In particular, an asymmetric data rate may allow an asymmetric binaural operation/ processing such that resource usage is dynamically adapted to the potential benefit of binaural processing at each individual earphone/ear. In accordance with an optional feature of the invention, at least one earphone of the first and second earphones comprises a binaural beam former for performing a binaural beam forming using at least a first signal from a first microphone of the at least one earphone and at least a second signal from a second microphone of an other earphone of the first and second earphones, the second signal being communicated from the other earphone to the at least one earphone by the audio data exchange.

This may allow improved performance, facilitated operation and/or reduced resource usage. In particular, improved beam forming and/or reduced power consumption may be achieved.

In accordance with an optional feature of the invention, the other earphone is arranged to low pass filter the second signal prior to communication to the at least one earphone.

This may allow improved performance, facilitated operation and/or reduced resource usage. In particular, this may allow an improved trade-off between audio quality and power consumption in many embodiments and scenarios. In accordance with an optional feature of the invention, the at least one earphone is arranged to low pass filter the first signal prior to performing the binaural beam forming.

This may allow improved performance, facilitated operation and/or reduced resource usage. In particular, it may allow an improved binaural beam forming from a closer correspondence between the characteristics of the audio signals from the two earphones.

In accordance with an optional feature of the invention, the at least one earphone further comprises a monaural beam former arranged to perform a monaural beam forming based only on signals from microphones of the at least one earphone; and means for combining an output signal of the binaural beam former and an output signal of the monaural beam former to generate an output signal of the at least one earphone.

This may allow improved audio quality and may in particular allow improved beam forming while maintaining low power consumption. According to an aspect of the invention, there is provided an earphone of an earphone arrangement comprising the earphone and a second earphone arranged to operate in a binaural arrangement, the earphone comprising: determining means for determining an audio environment characteristic in response to a signal from at least one microphone of the earphone; and control means for controlling a data rate of an audio data exchange between the earphone and the second earphone in response to the audio environment characteristic, the audio data exchange supporting a binaural processing of the earphone arrangement.

According to an aspect of the invention there is provided a method of operation for an earphone arrangement comprising a first and second earphone arranged to operate in a binaural arrangement, the method comprising the first earphone performing the steps of: determining an audio environment characteristic in response to a signal from at least one microphone of the first earphone; and controlling a data rate of an audio data exchange between the first and second earphone in response to the audio environment characteristic, the audio data exchange supporting a binaural processing of the earphone arrangement. These and other aspects, features and advantages of the invention will be apparent from and elucidated with reference to the embodiment(s) described hereinafter.

BRIEF DESCRIPTION OF THE DRAWINGS

Embodiments of the invention will be described, by way of example only, with reference to the drawings, in which Fig. 1 is an example of an illustration of a hearing aid in accordance with some embodiments of the invention;

Fig. 2 illustrates an example of a binaural hearing aid setup with three microphones on each ear;

Fig. 3 illustrates an example of performance gains from binaural and monaural processing;

Fig. 4 illustrates an example of performance gains from binaural and monaural processing;

Fig. 5 illustrates an example of performance gains from binaural and monaural processing; Fig. 6 illustrates an example of performance gains from binaural and monaural processing;

Fig. 7 illustrates an example of performance gains from binaural processing in the presence or not of head shadow; Fig. 8 is an example of an illustration of a hearing aid in accordance with some embodiments of the invention;

Fig. 9 illustrates an example of performance gains from binaural and monaural processing; and

Fig. 10 illustrates an example of performance gains from binaural and monaural processing.

DETAILED DESCRIPTION OF THE EMBODIMENTS

The following description focuses on embodiments of the invention applicable to hearing aids but it will be appreciated that the invention is not limited to this application but may be applied to other earphone arrangements.

In the specific example, a hearing impaired user may utilize a hearing aid arrangement to assist in the hearing. The hearing aid arrangement comprises a left and right hearing aid to be worn on respectively the left and right ear by the user when in use. In the example, the two hearing aids have identical functionality although the exact physical dimensions may vary to compensate for differences between the left or right ear.

Fig. 1 shows an example of a hearing aid in accordance with some embodiments of the invention. FIG. 1 specifically shows the left hearing aid and the following description will focus on this example. However, it will be appreciated that the right hearing aid may be functionally identical to the left hearing aid and that the focus on the left ear is merely for clarity and brevity of the description.

The hearing aid comprises a microphone array which in the specific example comprises two microphones 101, 103 although it will be appreciated that in other embodiments the hearing aid may comprise more microphones and indeed in some embodiments may comprise only a single microphone. The two microphones 101, 103 are coupled to an audio processor 105 which is further coupled to an audio transducer 107 which radiates the resulting audio signal into the left ear when in use.

The left hearing aid furthermore comprises a wireless receiver 109 which is arranged to receive a wireless data communication from the right hearing aid. The received data is fed to a receive data processor 111 which extracts the appropriate audio data from the communication. It will be appreciated that any suitable method of communicating data between the right and left hearing aid may be used including proprietary techniques or publicly standardized techniques such as e.g. Bluetooth™. The exchanged audio data is binaural audio data for performing a binaural processing at the left hearing aid. In particular, the audio data may comprise the microphone signals from the microphone(s) of the right hearing aid. In other embodiments, the audio signal(s) received from the right hearing aid may be signals generated by a processing of the microphone signals from the microphone(s) of the right hearing aid. For example, the microphone signals may be filtered, amplified, compressed etc prior to communication to the left hearing aid. Also, in some scenarios, microphone signals may be combined into fewer audio signals prior to the transmission to the left hearing aid.

The receive data processor 111 is coupled to the audio processor 105 which is fed the received audio data. The audio processor 105 may then process the received audio data together with the local microphone signals from microphones 101, 103 to generate a binaurally processed signal that is output to the audio transducer 107.

Specifically, the audio processor 105 may perform a binaural operation/ process that combines the local microphone signals with the received audio signals in order to improve the generated audio signal. As a specific example, the binaural operation may be a beam forming algorithm that is based on the combined microphone array comprising the microphones of the both the left and right hearing aid. Such a beam forming may substantially improve the performance of any beam forming based only on the local microphones 101, 103 since it increases (and typically doubles) the number of microphones of the array as well as significantly increases the distance between the microphones. Indeed, the distance between the left and right hearing aids are typically much larger than the distance between the microphones of a hearing aid thereby allowing a much larger effective array and thus enabling effective beam forming at much lower frequencies.

However, the wireless communication of audio data is typically very resource demanding and in particular increases both power consumption and computational resource (which further increases power consumption). Indeed, the increased power consumption may often substantially reduce the battery life thereby reducing the user benefit.

In order to support binaural operation at the right hearing aid, the left hearing aid furthermore comprises a data generator 113 which is coupled to the microphones 101, 103 and which generates suitable audio data/signals that are transmitted to the right hearing aid by a transmitter 115. The transmitter 115 may specifically be a Bluetooth™ transmitter but it will be appreciated that the transmission to the second hearing aid may be in accordance with any suitable transmission protocol or standard. Typically, the same transmission approach will be used for transmissions from the right to the left hearing aid as from the left to the right hearing aid.

The hearing aid furthermore comprises a binaural controller 117 coupled to receive data processor 111 and the data generator 113. The binaural controller 117 is arranged to control the audio data exchange between the two hearing aids. In particular, the binaural controller 117 is arranged to control a data rate of the audio data exchange between the first and second hearing aid in response to an audio environment characteristic determined from the signal from at least one of the local microphones. The controlled data rate may be of the transmission from the left to the right hearing aid, from the right to the left hearing aid or both. The binaural controller 117 is coupled to an audio environment processor 119 which is further coupled to the microphones 101, 103 and which is arranged to determine an audio environment characteristic from the microphone signals from the two microphones 101, 103. Specifically, the audio environment processor 119 is arranged to determine an audio environment characteristic comprising e.g. a Signal to Interference Ratio (SIR) indication, a Signal to Noise Ratio (SNR) indication and/or an angle of arrival estimate for an interferer. In the system, the audio environment characteristic is a characteristic which is indicative of a performance gain of a binaural processing of at least one of the left or right hearing aid relative to a monaural operation of the hearing aid.

Thus, the binaural controller 117 can specifically reduce the data rate of the exchanged audio data for the binaural operation if the audio environment characteristic has a value(s) that is indicative of the benefit of using binaural operation not being sufficiently high.

For example, in some embodiments, the audio exchange between the left and right hearing aids may be completely switched off when the audio environment characteristic has values that correspond to an estimated benefit of the binaural processing being below a given value. Thus, as a relatively low complexity example, the binaural controller 117 may evaluate if the audio environment characteristic meets a predetermined criterion (e.g. that the values of the parameters comprised in the audio environment characteristic has acceptable values). If so, the binaural controller 117 controls the receive data processor 111 and the data generator 113 to exchange audio data with the right hearing aid. Furthermore, as the audio processor 105 receives audio data from the right hearing aid it proceeds to include this and to perform a binaural processing.

However, if the audio environment characteristic does not meet the criterion, the binaural controller 117 controls the receive data processor 111 and the data generator 113 to not exchange any audio data. Thus, the computational resource (of both the left and right hearing aid) is reduced (as no receive operation, transmit operation or binaural processing is performed). Furthermore, the power consumption (of both hearing aids) is substantially reduced, partly due to the computational resource reduction and partly due to the reduced transmit power resource usage.

The approach may allow improved performance and may in particular provide an improved trade-off between power consumption/ battery life and audio quality/ hearing assistance. Specifically, the battery life may be substantially increased while maintaining acceptable or even negligible quality degradation. It will be appreciated that in other embodiments, a more flexible control of the data rate of the audio data exchange may be implemented. For example, the data rate may be given as a given function of the audio environment characteristic.

Furthermore, it will be appreciated that any suitable method of controlling the data rate may be used. For example, if the audio signals are audio encoded in accordance with an audio encoder standard (such as e.g. MPEG 2), the encoding process may be adjusted to provide the desired data rate (e.g. the encoded audio quality may depend on the audio environment characteristic). As another example, the data rate may be controlled by characteristics or parameters of the transmitted audio signals being flexibly adjusted. For example, only low frequency signals may be transmitted with the cut-off frequency of the low pass filtering of the microphone signals being determined as a function of one or more values of the audio environment characteristic.

In the specific embodiment, the binaural controller 117 may control both the transmission of audio data from the first hearing aid to the second hearing aid as well as the transmission of audio data from the second hearing aid to the first hearing aid. However, it will be appreciated that in other embodiments, the binaural controller 117 may control only data transmissions in one direction. It will also be appreciated that the binaural controller 117 may control the data transmissions in the two different directions differently or asymmetrically. For example, if a specific criterion is met, the binaural controller 117 may terminate audio data from the left to the right hearing aid while allowing audio data transmissions from the right to the left hearing aid.

The binaural controller 117 may for example control transmissions from the right hearing aid to the left hearing aid by communicating control data to the right hearing aid using the transmitter 115. For example, if the binaural controller 117 detects that the criterion for using binaural operation at the left hearing aid is not met, it may transmit a control signal to the right hearing aid in response to which the right hearing aid ceases transmissions of audio data.

In the system where a simple binary on/off switching of binaural operation is used, the transmission of signals from one hearing aid to the other is only performed when the additional benefit offered by a binaural speech enhancement system compared to a multi- microphone monaural system is sufficiently high.

Specifically, based on the audio environment characteristic, the binaural controller 117 may identify situations where the additional benefit offered by a binaural system over a monaural system is high and situations where the improvements are only marginal. Transmission of audio data and thus binaural operation may then be allowed only when the expected benefits are high with monaural operation being used otherwise. Thus, as signals are only exchanged when the resulting benefits are sufficiently significant, an increased battery life is achieved without a significant quality reduction. In the specific example, the data rate is controlled by detecting specific values of parameters of the audio environment characteristic for which a specific data reduction is applied, and specifically specific parameter values for which binaural operation and audio data exchange is completely switched off. In the following, an analysis of the operation and performance gains of binaural processing/ operation is provided and specific examples of the data rate control by the binaural controller 117 are provided on the basis of the analysis.

Signal Model

A frequency domain model is assumed. It is assumed that there exists a desired source S(ω) with power spectral density (PSD) E{5'(ω)5'^t(ω)} = Φ,(ω) , and an interfering source /(ω) with PSD E{/(ω)/^t(ω)} = Φ_;(ω) , where f indicates complex conjugate transpose.

The signal observed at the k^th microphone on the left hearing aid can be written as X^k(ω) = H_L ^k(ω)S(ω) + G^k(ω)I(ω) + U_L ^k(ω) + V^k(ω), k = \...N, (1)

where H^k((ύ) and G^k((ύ) are the transfer functions between the k^th microphone on the left hearing aid and the desired and interfering sources respectively, £/*(ω) and V^k((ύ) correspond to uncorrelated (e.g., sensor) noise and ambient diffuse noise, respectively, at the k^th microphone on the left hearing aid, and N is the number of microphones on the left hearing aid.

It is assumed that E{C/*(ω)C/f (ω)} = δ(£ - /)Φ_H(ω)V£ . For the ambient noise, which is assumed to be diffuse, one has

= Φ_v(ω)sinc(ω<i \ k - l \ Ic)Mk , where sinc(x) = sin(x)/x , d is the distance between two adjacent microphones and c is the speed of sound in air. Further, it is assumed that S((o) , /(ω) and t/(ω) are pair wise independent.

A similar right ear model can be written as X_R ^k(ω) = H_R ^k(ω)S(ω) + G_R ^k(ω)I(ω) + U_R(ω) + V^k(ω), k = \...N, (2)

where the relevant terms are defined analogously to the left ear. The left ear quantities are defined as

n_L(ω) = [Hl(ω),H_L ²(ω),...,H?(ω)]^τ ,

G_x(CO) = [G|(ω),G_i ²(ω),..., Gf(ω)f ,

,

U^OO) = [Ul(ω),U_L ²(ω),...,U?(ω)]^τ , and

Y_L(ω) = [Vl(ω),V²(ω),...,V_L ^N(ω)f .

The right ear quantities H_Λ(ω) , G_Λ(ω) , X_R(ω) , U_Λ(ω) , and V_Λ(ω) are defined similarly.

Finally, the binaural quantities are defined as

, G(CO) = [G^ (CO) G^ (co)f ,

U(ω) = [Ui(OO) U£(ω)f and

V(ω) = [Vj(ω)Vj(ω)f .

Then, the left ear, right ear, and binaural signal model in vector notation can be written as

X_z (ω) = H_z ((D)S(OO) + G_z (oo)/(oo) + U_z (ω) + V_z (ω) X_R(ω) = U_R(ω)S(ω) + G_R(ω)I(ω) + U_R(ω) + \_R(ω) (3) X(ω) = H(OO)S(OO) + G(ω)/(ω) + U(co) + V(oo).

Head Shadow Model

Based on the spherical head shadow model described in O. Duda and W. L. Martens, "Range dependence of the response of a spherical head model," J. Acoust. Soc.

Amer., vol. 104, no. 5, pp. 3048-3058, Nov. 1998, the transfer functions H(ω) and G(ω) are determined. The transfer functions only represent the effect of head shadow, and not the effect of room reverberation. Thus the following analysis is specifically relevant in substantially anechoic environments. Fig. 2 illustrates an example of a binaural hearing aid setup with three microphones on each ear. The example will be used to describe the model.

Firstly, define the origin to be the center of the sphere (corresponding to the head). Let a be the radius of the sphere, and r be the distance between the origin and the sound source, and define p = rla . Let θ denote the angle between a ray from the origin to the sound source and a ray from the origin to the point of observation on the surface of the sphere. The Head Related Transfer Function (HRTF) corresponding to the angle of incidence θ is then given by

77(p ,ω,θ) = -^-e ^CΨ, (4) ωα with h_m(ωpa/c)

Ψ(p,co,θ) = £(2m + l)P_ffl(cosθ) (5) m=0 h_m(ωa/c) ^'

where P_m is the Legendre polynomial of degree m, h_m is the spherical Hankel function of order m , and h_m is the derivative of h_m w.r.t its argument. Let θ_^ denote the angle between the vertical y-axis and a ray from the origin to the desired source. Let θ_; be defined similarly for the interfering source. The first microphone on the left and right ears are assumed to be located at 5π/9 and - 5π/9 respectively. It is assumed that the microphones are located on the surface of the sphere, and let d denote the length of the arc between two adjacent microphones on a single hearing aid. This yields:

H_L ^k(ω) = H(p,ω,θ_s + 5π/9 + (k - \)δ_θ )

⁽P)

H_R ^k(ω) = H(p,ω,θ, -5π/9 - (£ - l)δ_θ ), k = \... N,

where δ^, = dla is the angle subtended by the arc between two adjacent microphones. Similarly the transfer functions corresponding to the interferer are given by

G_L ^k(ω) = H(p,ω,θ_{! +} 5π/9 + (£ - l)δ_θ )

G_R ^k(ω) = H(p,ω,θ_! - 5π/9 - (^ - l)δ_θ ), k = \... N,

Multi Channel Wiener Filter

The performance of both the monaural and binaural operation will be evaluated for the example of a Multi-channel Wiener Filter (MWF) being used. An MWF is particularly suitable as it minimizes the mean squared error between the desired and estimated signal and thus serves as the theoretical optimum in terms of SNR improvement. The problem can be looked at from the perspective of the left hearing aid and one wishes to obtain a Minimum Mean Squared Error (MMSE) estimate of the desired signal observed at the first microphone of the left ear device, i/}(ω)S(ω) . The monaural MWF is given by

W^(ω) = E(Hl (ω)5(ω)X[(ω)} E(X, (ω)Xl(ω)}-¹

where

Φw = Φ,H_i(ω)H[(ω) + Φ_!G_i(ω)Gl(ω) + Φ_U\_N + Φ^fω), (9) I_N is the identity matrix of dimension N , and the (i,j)^th entry of T_M(ω) = E{Y_L (co) V^ (ω)} is given by

T_M'^J (ω) =sinc(ωd\i-j\/c) l≤iJ≤N. (10)

The binaural MWF is given by

W₅(ω) = E{H^(ω)5'(ω)X^t(ω)}E{X(ω)X^t(ω)}^"1

= Φ^(ω)H^τ(ω)Φ; -1 (H)

where

Φ_x = Φ,H(ω)H^t(ω) + Φ_!G(ω)G^t(ω) + Φ_HI_2Λ, + Φ_vr_B(co), (12)

I_2N is the identity matrix of dimension 2N , and the (i,j)^Λ entry of F₅(CO)E(V(CO)V¹XcO)) is given by

sinc(coJ \i-j\lc) l≤iJ≤N ovN≤i,j≤2N,

H(ω) = (13) sinc(ω(2α) \i- j\ Ic) otherwise.

It is assumed in (13) that d = 2a so that the distance between any pair of microphones on the left and right hearing aids can be assumed to be 2a . The corresponding mean squared error (MSE) can be written as

Using the expression for the MSE given in (14), one can compute the improvement in the signal-to-interference-plus-noise ratio (SINR) in the monaural and binaural cases as

where 101og₁₀ — p_j — is the input SINR (before processing) at the first

G_L ^ι \^z Φ_{1 +} Φ_u + Φ_v microphone on the left hearing aid.

Based on the derived improvement measures for the SINR, the benefit at the left ear due to monaural and binaural processing as a function of the location of the interferer, the number of interferers, the signal-to-interference noise ratio (SIR), and the signal-to-noise ratio (SNR) can be assessed. In additional the impact of the head shadow effect on the performance of the binaural system can be considered.

The SIR and SNR are defined as lOtogJ — and 10tog₁₀ — ,

respectively.

In the following specific performance measures will be illustrated for typical parameter values. Specifically, the level of the uncorrelated sensor noise is set to - 40 dB (i.e., Φ_u = I^"4). Also, The other relevant parameters are a = 0.0875 m, d = 0.008 m, N = 2 , and c = 343 m/s. In the example, a desired source is considered to be located straight ahead of the user when the hearing aid arrangement is in use (corresponding to 0°). For this situation and for different locations of the interferer, Fig. 3 illustrates ^M and ^B for four different frequencies 2, 4, 6, and 8 kHz corresponding to an input SIR of 0 dB and input SNR of 30 dB. Both the desired and interfering sources are in the example located at a distance of ^{r =} 1 -^ m. Thus, the example illustrates the improvement in SINR after monaural (dotted) and binaural (solid) processing for SIR = 0 dB and SNR = 30 dB.

It can be seen from Fig. 3 that for all frequencies, the differences between the monaural and binaural systems are most pronounced for interferers located approx. in the angle of arrival interval of [-60°, 60°]. This is likely to be caused by the main lobe generated for the small microphone array in the monaural case being quite broad whereas the larger microphone spacing in the binaural setup enables better separation of closely located desired and interfering sources as a narrower beam can be created. At low frequencies, the monaural array is generally not able to perform as well as the binaural processing due to the smaller effective microphone array and accordingly the benefit due to the binaural systems is larger.

Fig. 4 illustrates ^M and ^B averaged over all frequencies.

Thus, in the system, the binaural processing may be made dependent on an estimated angle of arrival of an interferer. For example, if the estimated angle of arrival falls within a given angle interval including the angle of the desired signal (which is typically, or may be assumed to be, straight ahead of the user when in user (0°). For example, binaural processing may be used if the estimate falls within the interval of [-60°, 60°]. Thus, in the example, the binaural audio data exchange may be increased if the estimate falls within the given interval and may be reduced if it is outside the interval. It will be appreciated that any suitable method or algorithm of estimating the angle of arrival of an interferer may be used and that several different approaches will be known by the person skilled in the art.

Initially, a situation was considered wherein only a single interferer is present. In a binaural system with more than one microphone at each ear, it is possible to cancel more than one interferer. In the following an example with two interferers will be considered.

In a scenario wherein two interferers are simultaneously active, a four- microphone binaural system has a clear advantage over a two -microphone monaural system.

Specifically, experiments have shown that the additional improvement ^{B m} provided by a binaural system is significant and that the binaural system provides good performance in most regions where the monaural system performs relatively poorly.

However, the binaural system offers fewer benefits in some scenarios. For example, when both interferers are co-located, the binaural operation tends not to provide a significant additional benefit. However, this corresponds to a scenario with only a single interferer where the monaural system performs relatively well and the benefit of the binaural processing is relatively limited except when the interferers are located within the interval [- 60°, 60°] as previously described for the single interferer. Another example is when both interferers are located very close to the desired source. In this case, the binaural array is typically unable to place two spatial nulls close to the desired source and therefore cannot sufficiently attenuate the interferers.

Another example is when the interferers are located around 180°. In this example, the monaural system is able to suppress a single interferer relatively well. However, due to the front-back ambiguity in a broadside array, the binaural system is unable to sufficiently suppress an interferer located around 180° when the desired source is active at 0°.

Thus, in the system, the binaural processing may be made dependent on a number of interferers and may in particular be dependent on estimated angles of arrival for a plurality of interferers. For example, if two interferers are detected with estimated angle of arrivals that differ substantially from each other (i.e. by more than a suitable threshold) binaural processing may be switched on.

In the system, the binaural audio data exchange may furthermore be controlled in response to the SIR and/or SNR values experienced in the audio environment. For example, at high SIR values and for an interferer located at around 30°, the additional benefit of binaural processing tends to reduce to the 3 dB gain resulting from the reduction of uncorrelated noise due to the doubling of microphones.

In an example wherein the interferer is located at 120°, the monaural system is able to provide a significant improvement as the SNR increases. Since the monaural system already provides good performance, the additional benefit due to the binaural processing is not as high as when the interferer is located at 30°.

Fig. 5 illustrates ^M and ^B for four different frequencies 2, 4, 6, and 8 kHz corresponding to an input SIR of 0 dB and input SNR of 20 dB and Fig. 6 illustrates ^M and

B for four different frequencies 2, 4, 6, and 8 kHz corresponding to an input SIR of 0 dB and input SNR of 0 dB. As can be seen, there is a substantial SNR dependency of the performance gain from binaural processing.

Thus, in the system, the binaural processing may be made dependent on the SIR and/or the SNR. For example, the binaural audio data exchange may only be performed when the SIR is below a given threshold and the SNR is above a given threshold. The binaural audio data exchange may in some embodiments/ scenarios be controlled in consideration of the head shadow effect.

Specifically, to illustrate the effect of head shadow on the performance of the binaural system, results obtained using the head shadow model are compared to those obtained without applying the model. The latter case corresponds to the four microphone array mounted in free space, and the microphone signals differ from each other just by delays.

Specifically, the improvement in SINR at the left ear is considered. FIG. 7 illustrates the SINR improvement of the binaural processing with the head shadow model (dotted line) and without the head shadow model (solid line) for four different frequencies 2, 4, 6, and 8 kHz corresponding to an input SIR of 0 dB and input SNR of 30 dB. As can be seen from FIG. 7, the improvement with the free mounted array is symmetric with respect to interferer location whereas the head mounted array results in better performance when the interferer is located in the left half plane.

For interferers located in the left half plane, the input SINR is low at the left ear and higher at the right ear due to the attenuation provided by the head. In this case, the binaural system is able to benefit from the higher SINR signal observed at the right ear. The free mounted array does not have this advantage as signals at both ears have the same input SINR. This effect is more pronounced at higher frequencies. As the attenuation provided by the head (resulting in inter-aural level differences) is higher at high frequencies, the difference between the head and free mounted arrays is also higher at the high frequencies. For interferences located in the right half plane, the left ear already enjoys a high input SINR and there is not much added benefit due to the availability of the lower SINR right ear signal.

It should be noted that FIG. 7 shows the improvement in SINR due to the binaural processing, which is asymmetric with respect to interferer location in the case of the head mounted array (as only the impact on the left hearing aid is considered). The total output SINR after processing is however symmetric. A second effect especially at high frequencies is that the head mounted array suffers less from spatial aliasing compared to the free-mounted array. This is evident from the sharp nulls present in the front half plane in the free-mounted case. These nulls are less pronounced for the head mounted array.

Thus, in some embodiments, the binaural controller 117 may reduce a data rate from the right earphone to the left earphone if the angle of arrival is in an angle interval corresponding to a head shadow for the left earphone when in use by a user (i.e. when the hearing aids are worn on the left and right ears). In particular, when an interferer is located in the interval from [0°, -60°] there is not much added benefit at the left ear due to a binaural processing, and transmission of data from the right to left ear may be avoided. It will be appreciated that the symmetric situation will typically apply to the right hearing aid.

In the system of Fig. 1, the data rate of the audio data exchange is made dependent on the benefit due to binaural beam forming as assessed by the audio environment characteristic. Specifically, as illustrated the data rate may be varied depending on interferer location, frequency, angle of arrival, SNR and SINR. The resulting transmission scheme may conserve power and bandwidth and may specifically increase battery life.

It will be appreciated that many different transmission schemes modifying the data rate may be envisaged. In the following, two specific transmission schemes are presented. In the first method, which can be viewed as a hard decision method, transmission is allowed only when the audio environment characteristic meets a criterion that indicates that the benefits of the binaural processing are high, thus resulting in power savings in other situations as no signal needs to be transmitted.

The second approach is a soft decision scheme where an appropriate data-rate is selected for the audio data exchange resulting in an efficient use of the available bandwidth. The selection of the data rate takes into account the additional benefit provided by the binaural processing.

It will be appreciated that the two schemes may be combined; e.g. first, a hard decision may be made on whether or not to transmit, and in cases when transmission is performed, an appropriate bit-rate may then be calculated by the second method.

As previously shown, the additional benefit provided by the binaural processing depends on a number of parameters. In the hard decision scheme, binaural audio data is exchanged between the two hearing aids only when the different parameters of the audio environment characteristic are within a predetermined range. The exact parameter ranges may depend on the specific characteristics and preferences of the individual embodiment.

However, in many embodiments, an advantageous criterion may comprise a requirement that the SIR indication SIR is below a first threshold; a requirement that the SNR indication is above a second threshold; and a requirement that the angle of arrival estimate is within an angle interval.

As a specific example, at low SIRs (e.g., below 15 dB) and high SNRs (e.g., above 15 dB) audio data may be exchanged between the left and right hearing aids. The gain in performance resulting from a binaural system is only marginal at high SIRs and low SNRs and thus by restricting transmission to the identified range, power consumption can be minimized.

Furthermore, if only one localized interferer is present, transmission is beneficial when the interferer is located within e.g. [-60°, 60°]. Thus, the binaural audio data exchange may only be performed for interferers in this range.

If two localized interferers are present, binaural operation is typically beneficial when at least one interferer is located within [60°, 300°]. If both interferers are located within a 60° neighborhood of 0° or 180°], the benefit is only marginal. Thus, the criterion for audio data exchange may be set to reflect this. As an example of a soft decision approach, the data rate may be determined by a rate distortion algorithm. Specifically, the binaural beam forming problem can be seen as a remote Wyner-Ziv due to the fact that the encoder does not have direct access to the source but only to a noisy version of the source. The right hearing aid needs to encode ^⁰⁵J such that an MMSE estimate of £v^ωJΗ^ωJ [_s obtained at the left hearing aid using the side information ^L^^ω) . Assuming that the sources are band limited and can be described as zero-mean jointly Gaussian processes, the corresponding parametric rate-distortion formula is given by

D(λ) = -L r (ξ₅(ω) + min(λ,ξ(ω))yω,

2π ^J-^∞ (16)

where the bit-rate R is expressed in bits per second and

ξ(ω) = ξ_M(ω) -ξ_B(ω)

0 < λ ≤ esssup_ω ξ(ω). ^

The quantity ^^B ^⁰ ^ represents the distortion incurred at R ⁼ °° , and ^^M ^^ω' represents the distortion incurred at R ⁼ 0 (monaural). The difference ^M \^{ω) ~} SB \^ω) _can tø_e reduced by exchanging signals at a positive rate R . Thus, as an example, the above equations may be used in a rate distortion algorithm to determine the rate allocation across different frequency bands. For example, a certain desired SINR improvement may be defined and the data rate of the audio data exchange may be adjusted accordingly. It will be appreciated that in some embodiments, the frequency dependency of the binaural processing gain may be taken into account when controlling the data rate. For example, the frequency bandwidth for the audio signals exchanged by the audio data may be modified in response to the audio environment characteristic.

For example, for most scenarios, the binaural processing gain is higher for lower frequencies than for higher frequencies. Accordingly, the signal bandwidth may be reduced for scenarios wherein the benefit at higher frequencies is negligible whereas the benefit at lower frequencies is still significant. In particular, the frequency bandwidth represented by the exchanged audio data may be reduced if the angle of arrival estimate of a single interfere is within a given angle interval wherein the benefit at higher frequencies is less significant.

For example, from the previous performance gain analysis, full bandwidth microphone signals may be exchanged between the hearing aids if the estimated angle of arrival is within [-60°, 60°] . However, for interferers located outside [-60°, 60°] only the low frequency content (e.g., up to 2 kHz) may be transmitted. This, in this case binaural processing is only performed at lower frequencies whereas at higher frequencies, the monaural system already provides good performance.

Thus, as the data rate required to represent the signal can be reduced for lower bandwidths (e.g. a reduced sample rate may be used), a substantially reduced data rate for the wireless transmission may be achieved and thus reduced power consumption can be achieved.

Fig. 8 shows an example of an implementation wherein the binaural processing is a binaural beam forming that uses one or more low pass filtered audio signals communicated from the other hearing aid. FIG. 8 specifically shows the left hearing aid and the following description will focus on this example. However, it will be appreciated that the right hearing aid may be functionally identical to the left hearing aid and that the focus on the left ear is merely for clarity and brevity of the description. The hearing aid of FIG. 8 specifically corresponds to the hearing aid of FIG. 1 with additional functionality which will be described in the following. In the example of Fig. 8, the audio processor 105 is a binaural beam former which performs a binaural beam forming using audio signals derived from the microphones of both the left hearing aid and the right hearing aid. Thus, the binaural beam forming uses both the locally generated microphone signals and the microphone signals received by the audio data exchange, i.e. transmitted from the right hearing aid to the left hearing aid.

However, in the example, the audio signal received from the right hearing aid is not a full bandwidth signal but rather is a signal that has first been low pass filtered with a suitable cut-off frequency, such as e.g. 2 kHz. Thus, in the example only the low frequency part of the microphone signal from the right hearing aid is transmitted to the left hearing aid thereby resulting in a reduced data rate of the audio data exchange and thus a reduced resource usage.

In the example, the first and second microphone signals are furthermore low pass filtered in a low pass filter 801 prior to being fed to the audio processor 105 for binaural beam forming. In the example, the low pass filter characteristics of the low pass filter 801 is selected to match the characteristics of the low pass filtering performed at the right hearing aid. Specifically, the same filter and/or cut-off frequency may be used.

Thus, the binaural beam forming is based only on lower frequency components of the microphones signals. Although, this may not provide the same quality beam forming as a beam forming based on full bandwidth signals, the degradation is often acceptable in return for the reduced power consumption and the approach may accordingly provide increased user satisfaction in many scenarios.

In some embodiments, the beam formed low frequency output signal generated by the audio processor 105 may be used directly as an audio output signal of the hearing aid. However, in many embodiments, the beam formed low frequency output signal is combined with high frequency content of the locally generated microphones signals. For example, the first and second microphone signals may be high pass filtered with a cut-off frequency corresponding to the low pass filter cut-off frequency. The high pass filtered signal of one or more of the left ear microphone signals may e.g. be added directly to the beam formed low frequency output signal to generate the hearing aid output signal. In the specific example, the local microphone signals are monaurally beam formed before being combined with the beam formed low frequency output signal from the binaural beam former. Specifically, the first and second microphone signals are fed to a high pass filter 803 which performs a high pass filtering that corresponds to (is complementary to) the low pass filtering performed by the low pass filter 801. The high pass filter 803 is coupled to a monaural beam former 805 which receives the high pass filtered microphone signals. The monaural beam former 805 then performs a monaural beam forming on these signals to generate a beam formed high frequency output signal.

The audio processor 105 and the monaural beam former 805 are coupled to a combiner 807 which is further coupled to the audio transducer 107. The combiner 807 combines the binaurally beam formed low frequency output signal and the monaurally beam formed high frequency output signal to generate the hearing aid output signal fed to the sound transducer 107. The combination may for example be a simple addition.

This approach may provide improved performance and may in particular provide improved beam forming. The approach may allow high performance while substantially reducing the power consumption of the hearing aids.

In the specific example, the left and right hearing aids have identical functionality. Accordingly, the hearing aid of FIG. 8 also comprises a low pass filter 809 for filtering the locally generated microphone signals prior to their transmission to the right hearing aid. It will be appreciated that in some embodiments, the low pass filtering may be symmetric such that the transmitted signals may be the output signals from the low pass filter 801. However, it will also be appreciated that in other embodiments, the low pass filtering may e.g. be dynamically varied asymmetrically between the first and second hearing aid (e.g. depending on the location of a dominant interferer). It will be appreciated that in many embodiments, the data rate of the audio data exchange may be adapted by adapting a characteristic of the low pass filtering. As a simple example, the cut-off frequency of the low pass filtering may be adjusted in response to angle of arrival, SIR and/or SNR measures. As a simple example, the low pass filtering may simply be switched on and off depending on whether the audio environment characteristic meets a suitable criterion.

Fig. 9 illustrates an example of performance gains from binaural and monaural processing. Specifically, FIG. 9 illustrates the improvement in SINR (Signal-To-Interference- Plus-Noise Ratio) at the left hearing aid for different locations of an interfering source for three scenarios. The dashed line corresponds to the scenario when the entire frequency content is transmitted from the right hearing aid (cut-off frequency of 8 kHz). The dotted line corresponds to full monaural processing, when no signal is transmitted from the right hearing aid (corresponding to a cut-off frequency of 0 kHz). In this scenario, only the microphone signals from the left hearing aid are processed. The solid line corresponds to a particular scenario using the approach described in relation to FIG. 8. In the example, the low frequency content of the right hearing aid microphone signals up to 2 kHz are transmitted from the right hearing aid, the desired source is assumed to be located at 0 deg., the input SIR (signal-to-interference ratio) is 0 dB and the input SNR is 30 dB. The signals are sampled at 16 kHz. Fig. 10 illustrates the relative performance for the same scenario but with a cut-off frequency of 4 kHz.

As can be seen, efficient performance can be achieved with transmission of reduced bandwidth signals and thus with reduced power consumption. In particular an improved flexibility in the trade-off between power consumption and quality can be achieved. E.g. the cut-off frequency can be adjusted in response to e.g. the audio environment characteristic or e.g. in response to a current remaining battery charge.

It will be appreciated that the above description for clarity has described embodiments of the invention with reference to different functional units and processors. However, it will be apparent that any suitable distribution of functionality between different functional units or processors may be used without detracting from the invention. For example, functionality illustrated to be performed by separate processors or controllers may be performed by the same processor or controllers. Hence, references to specific functional units are only to be seen as references to suitable means for providing the described functionality rather than indicative of a strict logical or physical structure or organization. The invention can be implemented in any suitable form including hardware, software, firmware or any combination of these. The invention may optionally be implemented at least partly as computer software running on one or more data processors and/or digital signal processors. The elements and components of an embodiment of the invention may be physically, functionally and logically implemented in any suitable way. Indeed the functionality may be implemented in a single unit, in a plurality of units or as part of other functional units. As such, the invention may be implemented in a single unit or may be physically and functionally distributed between different units and processors.

Although the present invention has been described in connection with some embodiments, it is not intended to be limited to the specific form set forth herein. Rather, the scope of the present invention is limited only by the accompanying claims. Additionally, although a feature may appear to be described in connection with particular embodiments, one skilled in the art would recognize that various features of the described embodiments may be combined in accordance with the invention. In the claims, the term comprising does not exclude the presence of other elements or steps. Furthermore, although individually listed, a plurality of means, elements or method steps may be implemented by e.g. a single unit or processor. Additionally, although individual features may be included in different claims, these may possibly be advantageously combined, and the inclusion in different claims does not imply that a combination of features is not feasible and/or advantageous. Also the inclusion of a feature in one category of claims does not imply a limitation to this category but rather indicates that the feature is equally applicable to other claim categories as appropriate. Furthermore, the order of features in the claims do not imply any specific order in which the features must be worked and in particular the order of individual steps in a method claim does not imply that the steps must be performed in this order. Rather, the steps may be performed in any suitable order. In addition, singular references do not exclude a plurality. Thus references to "a", "an", "first", "second" etc do not preclude a plurality. Reference signs in the claims are provided merely as a clarifying example shall not be construed as limiting the scope of the claims in any way.

Claims

CLAIMS:

1. An earphone arrangement comprising a first and second earphone arranged to operate in a binaural arrangement, the first earphone comprising: determining means (119) for determining an audio environment characteristic in response to a signal from at least one microphone (101, 103) of the first earphone; and control means (117) for controlling a data rate of an audio data exchange between the first and second earphone in response to the audio environment characteristic, the audio data exchange supporting a binaural processing of the earphone arrangement.

2. The earphone arrangement of claim 1 wherein the control means (117) is arranged to reduce the data rate if the audio environment characteristic does not meet a criterion.

3. The earphone arrangement of claim 2 wherein the audio environment characteristic comprises a Signal to Interference Ratio, SIR, indication, a Signal to Noise Ratio, SNR, indication and an angle of arrival estimate for an interferer; and the criterion comprises: a requirement that the SIR indication is below a first threshold; a requirement that the SNR indication is above a second threshold; and a requirement that the angle of arrival estimate is within an angle interval.

4. The earphone arrangement of claim 1 wherein the audio environment characteristic is indicative of a performance gain of the binaural processing of the earphone arrangement relative to a monaural operation of the earphone arrangement.

5. The earphone arrangement of claim 1 wherein the audio environment characteristic comprises a Signal to Interference Ratio, SIR, indication.

6. The earphone arrangement of claim 1 wherein the audio environment characteristic comprises a Signal to Noise Ratio, SNR, indication.

7. The earphone arrangement of claim 1 wherein the audio environment characteristic comprises an angle of arrival estimate for an interferer.

8. The earphone arrangement of claim 1 wherein the control means (117) is arranged to set a frequency bandwidth for audio signals exchanged by the audio data in response to the audio environment characteristic.

9. The earphone arrangement of claim 1 wherein at least one earphone of the first and second earphones comprises a binaural beam former for performing a binaural beam forming using at least a first signal from a first microphone (101, 103) of the at least one earphone and at least a second signal from a second microphone (101, 103) of an other earphone of the first and second earphones, the second signal being communicated from the other earphone to the at least one earphone by the audio data exchange.

10. The earphone arrangement of claim 9 wherein the other earphone is arranged to low pass filter the second signal prior to communication to the at least one earphone.

11. The earphone arrangement of claim 9 wherein the at least one earphone is arranged to low pass filter the first signal prior to performing the binaural beam forming.

12. The earphone arrangement of claim 9 wherein the at least one earphone further comprises a monaural beam former arranged to perform a monaural beam forming based only on signals from microphones of the at least one earphone; and means for combining an output signal of the binaural beam former and an output signal of the monaural beam former to generate an output signal of the at least one earphone.

13. The earphone arrangement of claim 1 wherein the audio data exchange comprises a first audio data exchange from the first earphone to the second earphone and a second audio data exchange from the second earphone to the first earphone; and the control means (117) is arranged to control the data rate differently in response to the audio environment characteristic for the first audio data exchange and the second audio exchange.

14. An earphone of an earphone arrangement comprising the earphone and a second earphone arranged to operate in a binaural arrangement, the earphone comprising: determining means (119) for determining an audio environment characteristic in response to a signal from at least one microphone of the earphone; and control means (117) for controlling a data rate of an audio data exchange between the earphone and the second earphone in response to the audio environment characteristic, the audio data exchange supporting a binaural processing of the earphone arrangement.

15. A method of operation for an earphone arrangement comprising a first and second earphone arranged to operate in a binaural arrangement, the method comprising the first earphone performing the steps of: determining an audio environment characteristic in response to a signal from at least one microphone of the first earphone; and controlling a data rate of an audio data exchange between the first and second earphone in response to the audio environment characteristic, the audio data exchange supporting a binaural processing of the earphone arrangement.