EP3945733A1 - Method for directional signal processing for a hearing aid - Google Patents

Abstract

The invention relates to a method for directional signal processing for a hearing device (2), with a first input converter (24) of the hearing device (2) generating a first input signal (E1) from a sound signal (28) from the environment, with a second input converter (26) of the hearing aid (2) a second input signal (E2) is generated from the sound signal (28) from the environment, with a first directional signal (Xr1) and a second Directional signal (Xr2) are formed, with the second directional signal (Xr2) exhibiting a relative attenuation in the direction of a first useful signal source (14), with the first directional signal (Xr1) exhibiting a relative attenuation in the direction of a second useful signal source (18), with a first Gain parameter (G1) for gaining a first useful signal (S1) of the first useful signal source (14) and a second gain parameter (G2) for gain g of a second useful signal (S2) from the second useful signal source (18) are determined, with a reference directional characteristic (62, 63, 64) being defined for a reference directional signal (Xref), with the first amplification parameter (G1) and/or of the second amplification parameter (G2) depending on the reference directional characteristic (62, 63, 64), a corrected first amplification parameter and a corrected second amplification parameter (G2') are determined in such a way that an output directional signal (Xout), which is the sum of the the first corrected first gain parameter weighted first directional signal (Xr1) and the corrected second gain parameter (G2') weighted second directional signal (Xr2) is formed, transitions into a linearly scaled reference directional signal (Xref) when the first gain parameter (G1) is equal to the second gain parameter (G2).

Description

The invention relates to a method for directional signal processing for a hearing device, with a first input converter of the hearing device generating a first input signal from a sound signal from the environment, with a second input converter of the hearing device generating a second input signal from the sound signal of the environment, with of the first input signal and the second input signal, a first directional signal and a second directional signal are formed, the second directional signal exhibiting a relative attenuation in the direction of a first useful signal source, the first directional signal exhibiting a relative attenuation in the direction of a second useful signal source, and a first Gain parameters for gaining a first useful signal from the first useful signal source and a second gain parameter for gaining a second useful signal from the second useful signal source are determined.

In a hearing device, ambient sound is converted by at least one input converter into an input signal which, depending on a hearing impairment of the wearer to be corrected, is processed in a frequency band-specific manner and in particular individually tailored to the wearer and is also amplified. The processed signal is converted via an output transducer of the hearing aid into an output sound signal which is routed to the wearer's hearing. As part of the signal processing, an automatic volume control ("automatic gain control", AGC) and often also a dynamic compression are applied to the input signal or to an already pre-processed intermediate signal, in which the input signal is usually only linearly amplified up to a certain limit value becomes, and above the limit a lower gain is applied, thereby smoothing out peak levels in the input signal. This is intended in particular to prevent sudden, loud sound events from leading to an output sound signal that is too loud for the wearer due to the additional amplification in the hearing device.

However, such an AGC with integrated dynamic compression initially reacts to sound events regardless of their direction. If the wearer of a hearing aid is in a complex hearing situation, e.g. in a conversation with several conversation partners, one conversation partner can trigger the compression, e.g may suffer.

The invention is based on the object of specifying a method for directional signal processing in a hearing device which, in particular in connection with AGC and/or dynamic compression, is also suitable for complex hearing situations with more than one useful signal source.

The stated object is achieved according to the invention by a method for directional signal processing for a hearing device, a first input signal being generated from a sound signal of the environment by a first input converter of the hearing device, a second input signal being generated from the sound signal of the environment by a second input converter of the hearing device a first directional signal and a second directional signal are formed on the basis of the first input signal and the second input signal, the second directional signal having a relative attenuation in the direction of a first useful signal source, the first directional signal having a relative attenuation in the direction of a second useful signal source, and wherein a first amplification parameter for an amplification of a first useful signal from the first useful signal source and a second amplification parameter for an amplification of a second useful signal from the second useful signal source are determined the. According to the invention, it is provided that a reference directional characteristic for a reference directional signal, which can be represented in particular as a superimposition of the first directional signal and the second directional signal, is defined, with the first gain parameter and/or the second gain parameter depending on the reference Directional characteristics, a corrected first gain parameter and a corrected second gain parameter are determined in such a way that an output directional signal, which is the sum of the first directional signal weighted with the corrected first gain parameter and the second directional signal weighted with the corrected second gain parameter, is formed into a linearly scaled reference -directional signal transitions when the first gain parameter is equal to the second gain parameter, and wherein at least one of said two corrected gain parameters differs from the corresponding underlying n gain parameter is different. Advantageous and partly inventive configurations are the subject matter of the subclaims and the following description.

The directional output signal with the required property can be formed either as the correspondingly described superimposition, or using at least one suitable intermediate signal, the directional output signal being formed in such a way that the required property is met when the two amplification parameters are equal.

The first directional signal can be formed for further signal processing, so that in particular the signal components of the first directional signal are included in the output directional signal. In the event that the output directional signal is formed from the signal components of at least one suitable intermediate signal - e.g. a formation of the first and second directional signal as well as the output directional signal based on forward and backward cardioid signals - the first directional signal is used in particular formed to determine the corrected first and / or corrected second gain parameter, since the directional information in particular about the second useful signal source, which can be removed from the first directional signal, is of decisive advantage for this.

An output signal is preferably generated on the basis of the output directional signal, which output signal is converted into an output sound signal by an output converter of the hearing device. In this case, interference noise and/or feedback and/or further signal processing steps can also be suppressed, in particular as a function of the frequency band, when the output signal is generated from the output directional signal. In particular, the method can be carried out by frequency band, so that the first and the second gain parameter, the first and the second directional signal and the reference directional signal and finally the corrected first and second gain parameter and the output directional signal are determined separately for each frequency band or for groups of individual frequency bands or be defined.

In this case, an input transducer includes in particular an electroacoustic transducer which is set up to generate a corresponding electrical signal from a sound signal. In particular, when the first or second input signal is generated by the respective input converter, pre-processing can also take place, e.g. in the form of linear pre-amplification and/or A/D conversion. The correspondingly generated input signal is given in particular by an electrical signal whose current and/or voltage fluctuations essentially represent the sound pressure fluctuations of the air.

In this case, the direction of the first useful signal source is preferably directed into the front hemisphere with respect to a frontal direction of a user of the hearing device defined by the intended use of the hearing device. In this case, the first useful signal source is particularly preferably located at least approximately in the frontal direction, so that in particular corresponding approximations of a frontal source can be made for the signal processing. The direction of the second useful signal source is preferably directed outside an angular range of +/-45°, particularly preferably +/-60°, around the frontal direction. In particular, the direction of the second useful signal source is directed towards the rear hemisphere.

A relative weakening of the first directional signal is to be understood here in particular as meaning that the directional characteristic in question has a sensitivity in the direction of the second useful signal which is reduced compared to the sensitivity averaged over all directions, and in particular has a local, preferably a global minimum. The first and the second directional signal preferably have as complete an attenuation as possible in the direction of the second or the first useful signal source.

The present invention solves the following problem in particular: If a directional output signal is formed using two directional signals, each of which has a relative, preferably complete, attenuation in the direction of another useful signal source, the amplification of the second useful signal, for example, does not depend on the resulting directional output signal depends only on the corresponding amplification parameter with which the second directional signal is weighted, but also on the direction of the second useful signal source, since the second directional signal has a non-trivial directional dependency in this regard. If the first directional signal then has a complete or almost complete weakening in the direction of the second useful signal source, said directional dependency cannot be compensated for by a correction term of the first directional signal. The same can apply to an amplification of the first useful signal by the first directional signal.

A corresponding correction must therefore be made via the contributions of the respective directional signal itself. In the context of the invention, this is solved by correcting the "scaling" of the respective directional signal, ie adjusting the respective gain parameter, so that the corrected gain parameter allows this directional dependency of the corresponding directional signal to be taken into account.

On the other hand, the reference directional characteristic is specified as the "normal state" which should be achieved if both useful signals are to be amplified with the same amplification parameters (the assumption here is that this applies in particular to identical useful signals, which only arrive from different directions). In this case, the amplification of the individual, "equally loud" useful signals should lead to the reference state, ie, for example, to an omnidirectional directional characteristic or a directional characteristic that models filtering by a pinna.

It can thus be achieved that not only useful signals from frontal useful signal sources, but also, e.g. An equivalent formulation at the level of the intermediate signals possibly forming the directional signals is also possible.

For this purpose, the reference directional signal is preferably represented on the basis of the two directional signals, so that the corrected gain parameters can be determined using the corresponding coefficients in this representation. It is particularly possible here that, for example, only the second corrected gain parameter has a real non-trivial correction to the second gain parameter, while the first corrected gain parameter is identical to the first gain parameter. However, both corrected gain parameters, that is to say the first and the second, can each differ from their underlying first or second gain parameter. This can be the case in particular if neither of the two useful signal sources is occupied in a preferred direction (e.g. the frontal direction) with respect to the hearing aid.

The corrected second amplification parameter is advantageously determined in such a way that the second useful signal is amplified by the second amplification parameter compared to the reference directional characteristic by the output directional signal, and/or the corrected first amplification parameter is determined in such a way that that the first useful signal is amplified by the output directional signal compared to the reference directional characteristic by the first amplification parameter. The result of this is that the output directional signal amplifies each of the two useful signals with the "correct" amplification parameter for the individual useful signal, depending on the direction of the respective useful signal source.

This is now only possible because those signal contributions in the output directional signal which originate, for example, from the second useful signal - e.g. a corresponding second directional signal - can be weighted using the corrected second gain parameter in such a way that a directional dependency of the second directional signal (or a comparable , based on at least one suitable intermediate signal formed signal with a relative attenuation in the direction of the first useful signal source) via the correction in the corrected second gain parameter can be compensated. The same applies to the signal contributions of the first useful signal in the output directional signal. In particular, the first corrected gain parameter can be identical to the first gain parameter.

The corrected second gain parameter is expediently formed as a product of the second gain factor and a correction factor, the correction factor corresponding to a linear coefficient of the second directional signal in a representation of the reference directional signal as a linear combination of the first directional signal and the second directional signal.

A mathematical formulation can be achieved as follows: If the input signals E1 and E2 are combined to form a vector E = [E1, E2] ^T , a first directional signal Xr1 formed using the two input signals can be represented at least approximately using the weights w = [w1, w2] ^T of the two input signals, i.e. Xr1 = E ^T · w (provided the acoustic transit time between the two input transducers is less than one sample, which is usually the case). In a comparable way, the second directional signal Xr2 can be represented as Xr2 = E ^T · u with the weights u = [u1, u2] ^T .

mit dem (skalaren!) korrigierten ersten und zweiten Verstärkungsparameter G1', G2'. Stellt man das Referenz-Richtsignal Xref ebenfalls in der Basis E1, E2 dar, so erhält man auch hier einen entsprechenden Gewichtungsvektor wref = [wr1, wr2]^T: Xref = E^T · wref. Anhand des ersten und des zweiten Richtsignals Xr1, Xr2 lässt sich nun infolge deren linearer Unabhängigkeit (da die relative Abschwächung jeweils auf eine andere Nutzsignalquelle gerichtet ist) das Referenz-Richtsignal darstellen zu $$\mathrm{Xref}=\mathrm{a}\cdot \mathrm{Xr}\mathrm{1}+\mathrm{b}\cdot \mathrm{Xr}\mathrm{2},$$

und somit $$\mathrm{wref}=\mathrm{a}\cdot \mathrm{w}+\mathrm{b}\cdot \mathrm{u}\left(\mathrm{f}\ddot{\mathrm{u}}\mathrm{r\; beide\; Vektorkomponenten}\right)$$

The output directional signal Xout can then be represented as $$\mathrm{Xout}={\mathrm{E}}^{\mathrm{T}}\cdot \left(\mathrm{G}\mathrm{1}\mathrm{ʹ}\cdot \mathrm{w}+\mathrm{<figure-callout\; id="2"\; label="G"\; filenames="imgf0001.png,imgf0002.png"\; state="\{\{state\}\}">G</figure-callout>}\mathrm{2}\mathrm{ʹ}\cdot \mathrm{and}\right)$$

with the (scalar!) corrected first and second gain parameters G1', G2'. If the reference directional signal Xref is also represented in the basis E1, E2, a corresponding weighting vector wref=[wr1, wr2] ^T : Xref=E ^T wref is also obtained here. The reference directional signal can now be represented on the basis of the first and the second directional signal Xr1, Xr2 as a result of their linear independence (since the relative attenuation is in each case directed to a different useful signal source). $$\mathrm{xref}=\mathrm{a}\cdot \mathrm{Xr}\mathrm{1}+\mathrm{b}\cdot \mathrm{<figure-callout\; id="2"\; label="Xr"\; filenames="imgf0001.png,imgf0002.png"\; state="\{\{state\}\}">Xr</figure-callout>}\mathrm{2},$$

and thus $$\mathrm{wref}=\mathrm{a}\cdot \mathrm{w}+\mathrm{b}\cdot \mathrm{and}\left(\mathrm{f}\ddot{\mathrm{and}}\mathrm{r\; both\; vector\; components}\right)$$

Equation (ii') is now multiplied to the left by h ^T (α), where h(α) = [h1(α), h2(α)] ^T represents a (measured, estimated or modeled) transfer function on the hearing aid, which, for example, takes into account a propagation of a sound from a sound source to the first input transducer or to the second input transducer, and for α = 0° in a relative transfer function from the first input transducer to the second input transducer with respect to a frontal sound source, and α is the angle for which preferably h ^T (α) · u = 0 applies (ie the angle of that direction for which the second directional signal preferably has an at least approximately total attenuation, namely the direction of the first useful signal source). Depending on the knowledge of h ^T (α) · wref and h ^T( α) · w, one can determine a value of a from this (the two scalar products can be tabulated in particular for different values of w1, w2 and α). The value of a can also be used to determine a value for b in one of the two components of (ii').

die gewünschte Eigenschaft erfüllen, im Ausgangs-Richtsignal Xout nach Gleichung (i) für G1 = G2 in das mit G1 skalierte Referenz-Richtsignal Xref überzugehen.It can then be shown that the corrected gain parameters $$\begin{array}{ccc}\mathrm{G}\mathrm{1}\mathrm{ʹ}& =& \mathrm{a}\cdot \mathrm{G}\mathrm{1},\\ \mathrm{<figure-callout\; id="2"\; label="G"\; filenames="imgf0001.png,imgf0002.png"\; state="\{\{state\}\}">G</figure-callout>}\mathrm{2}\mathrm{ʹ}& =& \mathrm{b}\cdot \mathrm{<figure-callout\; id="2"\; label="G"\; filenames="imgf0001.png,imgf0002.png"\; state="\{\{state\}\}">G</figure-callout>}\mathrm{2}\end{array}$$

fulfill the desired property of merging into the reference directional signal Xref scaled with G1 in the output directional signal Xout according to equation (i) for G1=G2.

It has proven to be advantageous if the corrected first gain parameter is determined as the first gain parameter when the second directional signal has its minimum sensitivity in the direction of the first useful signal source and preferably has as complete an attenuation as possible in this direction, and the first directional signal particularly preferably has its maximum sensitivity having. If the direction of maximum sensitivity for the first directional signal Xr1 is in the direction of the first useful signal, in which the second directional signal has its minimum sensitivity, this can be used in particular for normalizing the first directional signal so that the sensitivity in this direction is 1 is set, and thus h ^T (α) · w = 1 applies. If, in addition, the attenuation by the second directional signal is at least approximately complete in this direction, i.e. h ^T (α) u ≈ 0, then a = 1 can be set in equation (ii), which means that in equation (iii) G1' = G1 , i.e. the identity of the first and corrected first amplification parameter.

und können insbesondere adaptiv ermittelt werden. In einem wichtigen Spezialfall ist a2 = 0, d.h., das zweite Richtsignal Xr2 ist durch das rückwärtsgerichtete Kardioid-Signal Xa, und damit durch das zweite Zwischensignal Z2 gegeben. Dies ist z.B. der Fall, wenn die erste Nutzsignalquelle, für welche das zweite Richtsignal eine relative, bevorzugt maximale, und besonders bevorzugt totale Abschwächung aufweist, im Bereich der Kerbe des rückwärtsgerichteten Kardioid-Signals liegt oder dort angenommen wird.A first intermediate signal and a second intermediate signal are preferably formed on the basis of the first input signal and the second input signal, with the first directional signal being formed as a superposition of the first intermediate signal and the second intermediate signal, and an associated first superimposition parameter being determined and/or the second directional signal is formed as a superimposition of the second intermediate signal with the first intermediate signal, and an associated second superimposition parameter is determined in the process. In this case, in particular a forward-pointing and a backward-pointing cardioid signal Xc, Xa are used as the first intermediate signal and the second intermediate signal. The first overlay parameter a1 and the second overlay parameter a2 then result from the representation $$\begin{array}{cc}\mathrm{Xr}\mathrm{1}& =\mathrm{Xc}+\mathrm{a}\mathrm{1}\cdot \mathrm{yes},\\ \mathrm{<figure-callout\; id="2"\; label="Xr"\; filenames="imgf0001.png,imgf0002.png"\; state="\{\{state\}\}">Xr</figure-callout>}\mathrm{2}& =\mathrm{a}\mathrm{2}\cdot \mathrm{Xc}+\mathrm{yes},\end{array}$$

and can in particular be determined adaptively. In an important special case, a2=0, ie the second directional signal Xr2 is given by the backward-directed cardioid signal Xa, and thus by the second intermediate signal Z2. This is the case, for example, when the first useful signal source, for which the second directional signal has a relative, preferably maximum, and particularly preferably total attenuation, is in the region of the notch of the backward-directed cardioid signal or is accepted there.

In an advantageous embodiment, the corrected first gain parameter G1' is formed as a product of the first gain factor G1 and a first correction factor a, and the corrected second gain parameter G2' is formed as a product of the second gain factor G2 and a second correction factor b.

$$=\mathrm{a}\cdot \mathrm{G}\mathrm{1}\cdot \mathrm{Xr}\mathrm{1}+\mathrm{b}\cdot \mathrm{G}\mathrm{2}\cdot \mathrm{Xr}\mathrm{2.}$$

A first reference superimposition parameter aref1 and a second reference superimposition parameter aref2 are preferably defined for a superimposition of the first intermediate signal Z1 and the second intermediate signal Z2, which forms the reference directional signal Xref, the first correction factor a being based on a product of the second superimposition parameter a2 with the second reference overlay parameter aref2 and in particular based on a deviation of the product from the first reference overlay parameter aref1, and/or wherein the second correction factor b is formed based on a deviation of a product of the first overlay parameter a1 with the first reference overlay parameter aref1 from the second reference -Overlay parameter aref2 is formed. The output directional signal Xout is preferably formed using the first directional signal Xr1 weighted with the corrected first gain parameter G1′ and using the second directional signal Xr2 weighted with the corrected second gain parameter G2′ according to equation (iii) as $$\mathrm{Xout}=\mathrm{G}\mathrm{1}\mathrm{ʹ}\cdot \mathrm{Xr}\mathrm{1}+\mathrm{<figure-callout\; id="2"\; label="G"\; filenames="imgf0001.png,imgf0002.png"\; state="\{\{state\}\}">G</figure-callout>}\mathrm{2}\mathrm{ʹ}\cdot \mathrm{<figure-callout\; id="2"\; label="Xr"\; filenames="imgf0001.png,imgf0002.png"\; state="\{\{state\}\}">Xr</figure-callout>}\mathrm{2}$$

$$=\mathrm{a}\cdot \mathrm{G}\mathrm{1}\cdot \mathrm{Xr}\mathrm{1}+\mathrm{b}\cdot \mathrm{<figure-callout\; id="2"\; label="G"\; filenames="imgf0001.png,imgf0002.png"\; state="\{\{state\}\}">G</figure-callout>}\mathrm{2}\cdot \mathrm{<figure-callout\; id="2"\; label="Xr"\; filenames="imgf0001.png,imgf0002.png"\; state="\{\{state\}\}">Xr</figure-callout>}\mathrm{2.}$$

Using equations (iv) and (ii), Xref can be calculated as a function of the forward and specify the reverse cardioid signal Xc or Xa (i.e. depending on the two intermediate signals Z1 and Z2 and in this case with Z1 = Xc and Z2 = Xa): $$\begin{array}{cc}\mathrm{xref}& =\left(\mathrm{a}+\mathrm{b}\cdot \mathrm{a}\mathrm{2}\right)\cdot \mathrm{Xc}\cdot \left(\mathrm{a}\cdot \mathrm{a}\mathrm{1}+\mathrm{b}\right)\cdot \mathrm{yes},\hfill \\ \mathrm{xref}& =\mathrm{aref}\mathrm{1}\cdot \mathrm{<figure-callout\; id="2"\; label="Xc\; +\; aref"\; filenames="imgf0001.png,imgf0002.png"\; state="\{\{state\}\}">Xc</figure-callout>}+\mathrm{aref}\mathrm{}\mathrm{2}\cdot \mathrm{yes}\hfill \end{array}$$

The linear factors in equation (ii"), which are associated with the forward and backward cardioid signal Xc and Xa, respectively, can be interpreted here as a first and second reference overlay parameter aref1 and aref2, ie $$\begin{array}{cc}\mathrm{aref}\mathrm{1}& =\mathrm{a}+\mathrm{b}\cdot \mathrm{a}\mathrm{2},\\ \mathrm{<figure-callout\; id="2"\; label="aref"\; filenames="imgf0001.png,imgf0002.png"\; state="\{\{state\}\}">aref</figure-callout>}\mathrm{2}& =\mathrm{a}\cdot \mathrm{a}\mathrm{1}+\mathrm{b}\mathrm{.}\end{array}$$

The correction factors a, b in equation (v') then result as a function of the two reference overlay parameters aref1, aref2 $$\begin{array}{cc}\mathrm{a}& =\left(\mathrm{aref}\mathrm{1}-\mathrm{a}\mathrm{2}\cdot \mathrm{<figure-callout\; id="2"\; label="aref"\; filenames="imgf0001.png,imgf0002.png"\; state="\{\{state\}\}">aref</figure-callout>}\mathrm{2}\right)/\left(\mathrm{a}-\mathrm{a}\mathrm{1}\cdot \mathrm{a}\mathrm{2}\right),\hfill \\ \mathrm{b}& =\left(\mathrm{<figure-callout\; id="2"\; label="aref"\; filenames="imgf0001.png,imgf0002.png"\; state="\{\{state\}\}">aref</figure-callout>}\mathrm{2}-\mathrm{a}\mathrm{1}\cdot \mathrm{aref}\mathrm{1}\right)/\left(\mathrm{1}-\mathrm{a}\mathrm{1}\cdot \mathrm{a}\mathrm{2}\right)\mathrm{.}\hfill \end{array}$$

For the special case aref1 =1, Equation (vii) turns into $$\begin{array}{cc}\mathrm{a}& =\left(\mathrm{1}-\mathrm{aref}\mathrm{2}\cdot \mathrm{a}\mathrm{2}\right)/\left(\mathrm{1}-\mathrm{a}\mathrm{1}\cdot \mathrm{a}\mathrm{2}\right),\hfill \\ \mathrm{b}& =\left(\mathrm{<figure-callout\; id="2"\; label="aref"\; filenames="imgf0001.png,imgf0002.png"\; state="\{\{state\}\}">aref</figure-callout>}\mathrm{2}-\mathrm{a}\mathrm{1}\right)/\left(\mathrm{1}-\mathrm{a}\mathrm{1}\cdot \mathrm{a}\mathrm{2}\right)\mathrm{.}\hfill \end{array}$$

In a further advantageous embodiment, based on the first and the second superimposition parameter a1, a2, based on the first and the second reference superimposed parameter aref1, aref2 and based on the first amplification parameter G1 and based on the second gain parameter G2, an effective first superimposition parameter aeff1 and an effective second superimposition parameter aeff2 are determined, with the output directional signal Xout based on a superimposition of the first intermediate signal Z1 weighted with the first effective superimposition parameter aeff1 and the second intermediate signal weighted with the second effective superimposition parameter aeff2 Intermediate signal Z2 is formed, ie in particular as Xout ∝ aeff1 · Z1 + aeff2 · Z2.

With the definitions of equations (iv) and (iii), the following representation can also be derived from equation (v): $$\mathrm{Xout}=\mathrm{G}\mathrm{1}\cdot \left[\left(\mathrm{a}+\mathrm{b}\cdot \mathrm{a}\mathrm{2}\cdot \mathrm{<figure-callout\; id="2"\; label="G"\; filenames="imgf0001.png,imgf0002.png"\; state="\{\{state\}\}">G</figure-callout>}\mathrm{2}/\mathrm{G}\mathrm{1}\right)\cdot \mathrm{Xc}+\left(\mathrm{a}\cdot \mathrm{a}\mathrm{1}+\mathrm{b}\cdot \mathrm{<figure-callout\; id="2"\; label="G"\; filenames="imgf0001.png,imgf0002.png"\; state="\{\{state\}\}">G</figure-callout>}\mathrm{2}/\mathrm{G}\mathrm{1}\right)\cdot \mathrm{yes}\right]\mathrm{.}$$

mit a, b gegeben durch die Darstellung gemäß Gleichung (vii) (bzw. für den Spezialfall aref1 = 1 durch die Gleichung vii'), sodass Gleichung (viii) übergeht in $$\mathrm{Xout}=\mathrm{G}\mathrm{1}\cdot \left(\mathrm{aeff}\mathrm{1}\cdot \mathrm{Xc}+\mathrm{aeff}\mathrm{2}\cdot \mathrm{Xa}\right)\mathrm{.}$$

A comparable representation can be obtained by factoring out the second gain parameter G2, the representation selected in equation (viii) (excluding G1) being particularly advantageous for the case G1≧G2. The linear factors in equation (viii), which are assigned to the forward and backward cardioid signal Xc and Xa, respectively, can be interpreted here as a first and second effective overlay parameter aeff1 and aeff2, ie $$\begin{array}{cc}\mathrm{aeff}\mathrm{1}& =\mathrm{a}+\mathrm{b}\cdot \mathrm{a}\mathrm{2}\cdot \mathrm{<figure-callout\; id="2"\; label="G"\; filenames="imgf0001.png,imgf0002.png"\; state="\{\{state\}\}">G</figure-callout>}\mathrm{2}/\mathrm{G}\mathrm{1},\hfill \\ \mathrm{<figure-callout\; id="2"\; label="aeff"\; filenames="imgf0001.png,imgf0002.png"\; state="\{\{state\}\}">aeff</figure-callout>}\mathrm{2}& =\mathrm{a}\cdot \mathrm{a}\mathrm{1}+\mathrm{b}\cdot \mathrm{<figure-callout\; id="2"\; label="G"\; filenames="imgf0001.png,imgf0002.png"\; state="\{\{state\}\}">G</figure-callout>}\mathrm{2}/\mathrm{G}\mathrm{1}\hfill \end{array}$$

with a, b given by the representation according to equation (vii) (or for the special case aref1 = 1 by equation vii'), so that equation (viii) turns into $$\mathrm{Xout}=\mathrm{G}\mathrm{1}\cdot \left(\mathrm{aeff}\mathrm{1}\cdot \mathrm{<figure-callout\; id="2"\; label="Xc\; +\; aeff"\; filenames="imgf0001.png,imgf0002.png"\; state="\{\{state\}\}">Xc</figure-callout>}+\mathrm{aeff}\mathrm{}\mathrm{2}\cdot \mathrm{yes}\right)\mathrm{.}$$

For G1=G2, the first and second effective overlay parameters aeff1, aeff2 merge into the first and second reference overlay parameters aref1, aref2, respectively, according to equation (vi). This can now be used to first calculate the two correction parameters a, b, which according to equation (iii) correspond to the two amplification parameters G1, G2 are assigned, and finally the two effective overlay parameters aeff1, aeff2 as a function of the reference overlay parameters aref1, aref2.

The specific form of Xout to be selected, i.e. according to equation (v) or equation (viii'), is then preferably carried out depending on which of the two representations (ii), (ii") is actually used for Xref. In particular, for equation (ii") the value aref1 = 1 must be specified.

In the event that the second directional signal Xr2 is given by the second intermediate signal Z2, i.e. in particular by the backward-looking cardioid signal Xa (and thus a2 = 0 in equation iii), the first effective superimposition parameter aeff1 is formed from the first reference superimposition parameter aref1, i.e. aeff1 = aref1.

mit einem effektiven Verstärkungsparameter Geff = (G1' + G2' · a2) und dem Verhältnis der effektiven Überlagerungsparameter aeff2/aeff1 = (G1' · a1 + G2')/Geff. Anhand der Forderungen, dass für G1 = G2 das Ausgangssignal Xout in ein skaliertes Referenzsignal Xref, und somit die effektiven Überlagerungsparameter aeff1, aeff2 in die zugehörigen Referenz-Überlagerungsparameter aref1, aref2 übergehen sollen, sowie aus der zusätzlichen Forderung, dass in diesem Fall das Ausgangssignal Xout mit dem ersten Verstärkungsparameter G1 zu verstärken ist, also Geff = G1, ergeben sich die Gleichungen $$\begin{array}{lll}\mathrm{G}\mathrm{1}& =\mathrm{Geff}& =\mathrm{G}\mathrm{1}\cdot \mathrm{a}+\mathrm{G}\mathrm{2}\cdot \mathrm{b}\cdot \mathrm{a}\mathrm{2},\\ \mathrm{aref}\mathrm{1}& =\mathrm{aref}\mathrm{1}& =\mathrm{a}+\mathrm{b}\cdot \mathrm{a}\mathrm{2},\\ \mathrm{aref}\mathrm{2}& =\mathrm{aeff}\mathrm{1}& =\mathrm{a}\cdot \mathrm{a}\mathrm{1}+\mathrm{b}\mathrm{.}\end{array}$$

The following representation can also be derived from equation (v) for Z1 = Xc, Z2 = Xa after some transformation: $$\mathrm{Xout}=\mathrm{Geff}\cdot \left(\mathrm{Z}\mathrm{1}+\left(\mathrm{<figure-callout\; id="2"\; label="aeff"\; filenames="imgf0001.png,imgf0002.png"\; state="\{\{state\}\}">aeff</figure-callout>}\mathrm{2}/\mathrm{aeff}\mathrm{1}\right)\cdot \mathrm{<figure-callout\; id="2"\; label="Z"\; filenames="imgf0001.png,imgf0002.png"\; state="\{\{state\}\}">Z</figure-callout>}\mathrm{2}\right)$$

with an effective gain parameter Geff = (G1' + G2' x a2) and the ratio of the effective beat parameters aeff2/aeff1 = (G1' x a1 + G2')/Geff. Based on the requirements that for G1 = G2 the output signal Xout should change into a scaled reference signal Xref, and thus the effective superimposition parameters aeff1, aeff2 into the associated reference superimposition parameters aref1, aref2, as well as the additional requirement that in this case the output signal Xout is to be amplified with the first amplification parameter G1, ie Geff=G1, the equations result $$\begin{array}{lll}\mathrm{G}\mathrm{1}& =\mathrm{Geff}& =\mathrm{G}\mathrm{1}\cdot \mathrm{a}+\mathrm{<figure-callout\; id="2"\; label="G"\; filenames="imgf0001.png,imgf0002.png"\; state="\{\{state\}\}">G</figure-callout>}\mathrm{2}\cdot \mathrm{b}\cdot \mathrm{a}\mathrm{2},\\ \mathrm{aref}\mathrm{1}& =\mathrm{aref}\mathrm{1}& =\mathrm{a}+\mathrm{b}\cdot \mathrm{a}\mathrm{2},\\ \mathrm{<figure-callout\; id="2"\; label="aref"\; filenames="imgf0001.png,imgf0002.png"\; state="\{\{state\}\}">aref</figure-callout>}\mathrm{2}& =\mathrm{aeff}\mathrm{1}& =\mathrm{a}\cdot \mathrm{a}\mathrm{1}+\mathrm{b}\mathrm{.}\end{array}$$

For the above special case, in which the second directional signal Xr2 is given by the second intermediate signal Z2 (in particular in the form of a backward-directed cardioid signal Xa), and therefore a2 = 0 applies, and the first reference overlay parameter aref1 = 1 is set, the first correction factor a = 1 (cf. equation vii'). Thus, the corrected first gain parameter G1' is identical to the first gain parameter G1'. For the second correction factor b, b = aref2 - 1 applies.

This results in for the output signal $$\mathrm{Xout}=\mathrm{G}\mathrm{1}\cdot \left(\mathrm{Z}\mathrm{1}+\mathrm{a}\mathrm{1}\cdot \mathrm{<figure-callout\; id="2"\; label="Z"\; filenames="imgf0001.png,imgf0002.png"\; state="\{\{state\}\}">Z</figure-callout>}\mathrm{2}\right)+\mathrm{<figure-callout\; id="2"\; label="G"\; filenames="imgf0001.png,imgf0002.png"\; state="\{\{state\}\}">G</figure-callout>}\mathrm{2}\cdot \left(\mathrm{<figure-callout\; id="2"\; label="aref"\; filenames="imgf0001.png,imgf0002.png"\; state="\{\{state\}\}">aref</figure-callout>}\mathrm{2}-\mathrm{a}\mathrm{1}\right)\cdot \mathrm{<figure-callout\; id="2"\; label="Z"\; filenames="imgf0001.png,imgf0002.png"\; state="\{\{state\}\}">Z</figure-callout>}\mathrm{2.}$$

For G1=G2 it is obvious that (xii) merges into the G1-scaled reference directional signal Xref=Z1+aref2*Z2 with the second reference beat parameter aref2.

If the second directional signal Xr2 is given by the second intermediate signal Z2, the second effective overlay parameter is advantageously formed from the first overlay parameter a1 and a ratio of the corrected second gain parameter G2' and the first gain parameter G2'/G1.

wobei der Term (aref2 - a1) · G2 in aeff2 den korrigierten zweiten Verstärkungsparameter G2' bildet. Man beachte, dass für den genannten Fall das zweite Richtsignal Xr2 durch das zweite Zwischensignal Z2 gegeben ist, und somit der zweite Überlagerungsparameter a2 = 0 gesetzt wurde. Hierdurch wird in Gleichung (x) der effektive Verstärkungsparameter Geff zu G1 (a = 1), was in Gleichung (xiii) berücksichtigt wurde.Using the first directional signal Xr1 = Z1 + a1 Z2, a formulation equivalent to equation (xii) can then be obtained in the special case mentioned (aref1 = 1, a2 = 0 and thus aeff1 = 1 in equation ix) using equation (x). , if the second useful signal is to be boosted largely exclusively by the second intermediate signal: $$\begin{array}{cc}\mathrm{Xout}& =\mathrm{G}\mathrm{1}\cdot \left(\mathrm{Z}\mathrm{1}+\left(\mathrm{<figure-callout\; id="2"\; label="aeff"\; filenames="imgf0001.png,imgf0002.png"\; state="\{\{state\}\}">aeff</figure-callout>}\mathrm{2}/\mathrm{aeff}\mathrm{1}\right)\cdot \mathrm{<figure-callout\; id="2"\; label="Z"\; filenames="imgf0001.png,imgf0002.png"\; state="\{\{state\}\}">Z</figure-callout>}\mathrm{2}\right)\hfill \\ & =\mathrm{G}\mathrm{1}\cdot \left\{\mathrm{Z}\mathrm{1}+\left[\mathrm{a}\mathrm{1}+\left(\mathrm{<figure-callout\; id="2"\; label="aref"\; filenames="imgf0001.png,imgf0002.png"\; state="\{\{state\}\}">aref</figure-callout>}\mathrm{2}-\mathrm{a}\mathrm{1}\right)\cdot \mathrm{<figure-callout\; id="2"\; label="G"\; filenames="imgf0001.png,imgf0002.png"\; state="\{\{state\}\}">G</figure-callout>}\mathrm{2}/\mathrm{G}\mathrm{1}\right]\cdot \mathrm{<figure-callout\; id="2"\; label="Z"\; filenames="imgf0001.png,imgf0002.png"\; state="\{\{state\}\}">Z</figure-callout>}\mathrm{2}\right\},\hfill \end{array}$$

where the term (aref2 - a1)*G2 in aeff2 forms the corrected second gain parameter G2'. Note that for the case mentioned, the second directional signal Xr2 is given by the second intermediate signal Z2, and thus the second overlay parameter a2 = 0 was set. As a result, the effective gain parameter Geff becomes G1 (a=1) in equation (x), which was taken into account in equation (xiii).

It proves to be further advantageous if the reference directional characteristic of the reference directional signal is chosen as an omnidirectional directional characteristic, or is chosen in such a way that a shadowing effect of human ears is simulated.

If a forward-pointing and a backward-pointing cardioid signal Xc, Xa are used as intermediate signals of the signal processing, e.g -Directional signal Xref = aref1 Xc + aref2 Xa the values aref1 = 1, aref2 = - 1. In many situations, a hearing perception that is as omnidirectional as possible is desirable as a starting point.

In the event that the reference directional characteristic is intended to simulate a shadowing effect on human ears, the reference overlay parameters aref1, aref2 are preferably determined in advance in such a way that the reference directional signal Xref=aref1 Xc+aref2 Xa has the desired spatial sensitivity , as it is caused by the shadowing of the pinna on a human ear. In this case, aref1, aref2 can be determined on a generic ear model (e.g. a KEMAR), or can also be individually adjusted to the wearer of the hearing aid by appropriate measurements.

The first directional signal preferably exhibits a maximum attenuation in the direction of the first useful signal source and/or the second directional signal exhibits a maximum, in particular total attenuation, in the direction of the second useful signal source. As a result, the influences of the respective useful signals on the respective other directional signal and thus on the relative amplification parameter can be minimized particularly effectively.

Advantageously, the first directional signal is generated using adaptive directional microphones, in particular using a first intermediate signal and a second intermediate signal, and/or the second directional signal is generated using adaptive directional microphones, in particular using the first and second intermediate signals. This can ensure that the directional signal in question has the lowest possible, preferably minimum, sensitivity in the direction of one of the two useful signal sources, so that there is a high, preferably maximum, attenuation in this direction, and in the direction of the other useful signal source, the sensitivity is as high as possible, preferably maximum Sensitivity.

It has also proven to be advantageous if the first intermediate signal is generated using a time-delayed superimposition of the first input signal with the second input signal, implemented using a first delay parameter, and/or the second intermediate signal is generated using a time-delayed superimposition of the second input signal with the second input signal, implemented using a second delay parameter first input signal is generated. In particular, the first and second delay parameters can be chosen to be identical to one another, and in particular the first intermediate signal can be generated with respect to a preferred plane of the hearing device symmetrically to the second intermediate signal, with the preferred plane being assigned to the frontal plane of the wearer, preferably when wearing the hearing device. Alignment of the directional signals to the wearer's frontal direction facilitates signal processing, since this takes into account the wearer's natural viewing direction.

In this case, the first intermediate signal is preferably generated as a forward-directed cardioid directional signal and/or the second intermediate signal is generated as a backward-directed cardioid directional signal. A directional cardioid signal can be formed by superimposing the two input signals on one another with the acoustic propagation delay corresponding to the distance between the input transducers. As a result - depending on the sign of this propagation delay in the superimposition - the direction of the maximum attenuation is in the frontal direction (backward directional cardioid signal) or in the opposite direction (forward cardioid directional signal). The direction of maximum sensitivity is the direction of maximum opposite weakening. This facilitates further signal processing, since such an intermediate signal is particularly suitable for adaptive directional microphony.

The invention also mentions a hearing system with a hearing aid, which has a first input converter for generating a first input signal from a sound signal from the environment and a second input converter for generating a second input signal from the sound signal from the environment, and a control unit which is set up to carry out the procedures described above. In particular, the control unit can be integrated in the hearing aid. In this case, the hearing system is given directly by the hearing aid. The advantages mentioned for the method and for its further developments can be transferred analogously to the hearing system.

Fig. 1

eine Gesprächssituation eines Trägers eines Hörgerätes mit zwei Gesprächspartnern, und

Fig. 2

eine bevorzugte, direktionale Signalverarbeitung für das Hörgerät in der Gesprächssituation nach Fig. 1,

Fig. 3a

eine Richtcharakteristik eines aus der direktionalen Signalverarbeitung nach Fig. 2 resultierenden Ausgangssignals, und

Fig. 3b

eine Richtcharakteristik eines alternativen, aus der direktionalen Signalverarbeitung nach Fig. 2 resultierenden Ausgangssignals.

An exemplary embodiment of the invention is explained in more detail below with reference to drawings. Here each show schematically:

1

a conversation situation of a wearer of a hearing aid with two conversation partners, and

2

a preferred, directional signal processing for the hearing aid in the conversation situation 1 ,

Figure 3a

a directional characteristic from the directional signal processing 2 resulting output signal, and

Figure 3b

a directional characteristic of an alternative, from the directional signal processing 2 resulting output signal.

Corresponding parts and sizes are provided with the same reference symbols in all figures.

In figure 1 a wearer 1 of a hearing aid 2 is shown schematically in a top view, which is in a conversation situation with a first conversation partner 4 and a second conversation partner 8 . The first interlocutor 4 is positioned in a first direction 6 with respect to the carrier 1, the second interlocutor 8 in a second direction 10 relative to the carrier 1. The first interlocutor 4 is the main interlocutor of the carrier 1, the second interlocutor 8 takes part in this conversation only through isolated voice contributions. The conversation situation described here is for the upper and the lower picture of figure 1 identical. The voice contributions of the first conversation partner 4 form the first useful signal S1, the voice contributions of the second conversation partner 8 the second useful signal S2.

In order to mitigate the level peaks of the first useful signal S1 and the second useful signal S2 for the wearer 1 of the hearing aid 2 in an output sound signal of the hearing aid 2, as in the upper image of figure 1 shown, initially a first directional signal Xr1 is generated by means of adaptive directional microphony in such a way that the same has a maximum and preferably complete attenuation in the second direction 10, in which the second interlocutor 8 is positioned. This means that the useful signal S2 is not detected by the first directional signal Xr1. A compression factor, which is calculated based on the first directional signal Xr1, consequently only reacts to the first and second call partners 4 and 8 with regard to the two useful signal sources 14, 18. Here, a first amplification parameter G1 is determined, which with regard to the first useful signal S1 of the first useful signal source 14 (ie the first conversation partner 4) for each moment determines the optimal signal amplification and thus implicitly also a corresponding compression ratio.

In the picture below from figure 1 1 shows, analogously to the top image, a second directional signal Xr2 in the same hearing situation, which has a maximum and preferably complete attenuation in the first direction 6, ie the direction of the first conversation partner 4. Since the first direction 6 coincides with the frontal direction of the carrier 1, the second directional signal Xr2 is as a rearward one Designed cardioid directional signal Xa. The second amplification parameter G2, which is determined based on the second directional signal Xr2 and is assigned to it, therefore represents the optimal amplification with regard to the second useful signal S2 and in particular an associated compression ratio at any moment.

In order to be able to use compression to reduce the level peaks in an output sound signal of hearing aid 2 for wearer 1 due to the conversation contributions of both the first conversation partner 4 and the second conversation partner 8 to a level that is comfortable for wearer 1, such an output sound signal could now be formed from a linear combination of the first and the second directional signal Xr1, Xr2, which are each weighted with their respective gain parameters G1, G2. Since the first directional signal Xr1 is also formed by means of adaptive directional microphony based on a forward-directed cardioid directional signal and based on the backward-directed cardioid directional signal Xa, such a linear combination would lead to an output sound signal whose directional characteristic is similar in shape to that of the first directional signal Xr1, but with the notch 22 of maximum attenuation is shifted away from the second direction 10. On the one hand, this leads to a possibly undesirable, completely "deaf" area away from the second useful signal source 18, which on the other hand can also fluctuate in its orientation due to the dependence of such a linear combination on the speech contributions of the first conversation partner 4.

In figure 2 1 is a schematic block diagram of a method for directional signal processing for the hearing aid 2 figure 1 in the situation described there, which is intended to mitigate in particular the level peaks of the two useful signals S1, S2 of the useful signal sources 14, 18 given by the respective conversation partners 4, 8. A first input transducer 24 and a second input transducer 26 are arranged in the hearing device 2 and each of these generates a first input signal E1 and a second input signal E2 from a sound signal 28 . The sound signal 28 is the ambient sound, which therefore also contains the first and the second useful signal S1, S2. A possible pre-processing such as an A/D conversion or something similar should already be in be included in the input transducers 24, 26, which also each have a preferably omnidirectional microphone.

The first input signal E1 is now superimposed with the second input signal E2, which was delayed by a first delay parameter T1, and a first intermediate signal Z1 is formed from this. Analogously, the first input signal E1, which was delayed by a second delay parameter T2, is superimposed on the second input signal E2, and a second intermediate signal Z2 is thereby formed. In the present case and without loss of generality, the first and second delay parameters T1, T2 are each identical (T1 = T2) and are also selected in such a way that the first intermediate signal Z1 is given by a forward-directed cardioid directional signal Xc, and the second intermediate signal 36 by the reverse cardioid directional signal Xa. Based on the first intermediate signal Z1 and the second intermediate signal Z2, the first directional signal Xr1=Z1+a1×Z2 is calculated according to figure 1 generated by means of an adaptive directional microphone 40 by determining a first superposition parameter a1 in such a way that the contributions of the second conversation partner 8, ie the second useful signal S2, are maximally suppressed in the first directional signal Xr1. The first amplification parameter G1 is determined for the first useful signal S1 on the basis of the first directional signal Xr1. The determined first gain parameter G1 thus represents the optimal gain and compression of the signal contributions of the first conversation partner 4 through the first directional signal Xr1.

Using adaptive directional microphones 42, the second directional signal Xr2 can be generated from the first intermediate signal Z1 and the second intermediate signal Z2, which suppresses the contributions of the first conversation partner 4, ie the second useful signal S2, to a maximum. Since the carrier 1 is in the frontal direction in the present case, the second directional signal Xr2 is, as already mentioned, given by the backward-directed cardioid directional signal Xa. On the one hand, the second directional signal Xr2 can be permanently assumed as the backward-directed cardioid directional signal Xa. On the other hand, by means of the adaptive directional microphone 42 and a change in position of the first interlocutor 4 for the formation of the second Directional signal Xr2 from the first and the second intermediate signal Z1, Z2 are taken into account.

Analogously to the first gain parameter G1, the second gain parameter G2 is also determined on the basis of the second directional signal Xr2. This represents the optimal amplification and compression of the second useful signal S2 by the second directional signal Xr2.

mit einem zugehörigen ersten Referenz-Überlagerungsparameter aref1 und einem zweiten Referenz-Überlagerungsparameter aref2, welche so gewählt werden, dass das Referenz-Richtsignal Xref die gewünschte Referenz-Richtcharakteristik 63 aufweist, also z.B. die räumliche Filterwirkung der Pinna an einem menschlichen Ohr, insbesondere frequenzbandweise, nachbildet. Auch kann für einige oder alle Frequenzbänder eine omnidirektionale Richtcharakteristik für das Referenz-Richtsignal Xref gewählt werden (welches hierdurch seine Richtwirkung verliert). Das Referenz-Richtsignal Xref dient dabei der Definition der Referenz-Richtcharakteristik 63 und der Referenz-Überlagerungsparameter aref1, aref2, und muss vorliegend nicht zwingend als eigenständiges Signal aus den beiden Zwischensignalen Z1 und Z2 erzeugt werden (entsprechen durch gestrichelte Linien dargestellt); die Referenz-Überlagerungsparameter aref1, aref2 können vielmehr vorab festgelegt werden. Insbesondere kann aref1 = 1 festgelegt werden, sodass die Referenz-Richtcharakteristik 63 des Referenz-Richtsignals Xref in Frontalrichtung keine Abschwächung aufweist.In addition, a reference directional characteristic 63 is defined for a reference directional signal Xref. The reference directional signal Xref is obtained as a superimposition of the two intermediate signals Z1, Z2 as $$\mathrm{xref}=\mathrm{aref}\mathrm{1}\cdot \mathrm{Z}\mathrm{1}+\mathrm{<figure-callout\; id="2"\; label="aref"\; filenames="imgf0001.png,imgf0002.png"\; state="\{\{state\}\}">aref</figure-callout>}\mathrm{2}\cdot \mathrm{<figure-callout\; id="2"\; label="Z"\; filenames="imgf0001.png,imgf0002.png"\; state="\{\{state\}\}">Z</figure-callout>}\mathrm{2}$$

with an associated first reference overlay parameter aref1 and a second reference overlay parameter aref2, which are selected such that the reference directional signal Xref has the desired reference directional characteristic 63, e.g. the spatial filter effect of the pinna on a human ear, in particular by frequency band, replicates. An omnidirectional directional characteristic for the reference directional signal Xref can also be selected for some or all frequency bands (which loses its directivity as a result). The reference directional signal Xref is used to define the reference directional characteristic 63 and the reference superimposition parameters aref1, aref2, and in the present case does not necessarily have to be generated as an independent signal from the two intermediate signals Z1 and Z2 (correspondingly represented by dashed lines); rather, the reference overlay parameters aref1, aref2 can be predetermined. In particular, aref1=1 can be set so that the reference directional characteristic 63 of the reference directional signal Xref has no attenuation in the frontal direction.

For the following calculations, it is now taken into account that Xr2 = Z2 = Xa and therefore a2 = 0, whereby aref1 = 1 is also set.

A corrected second gain parameter G2′ is now determined as using the second gain parameter G2, the first superimposition parameter a1 and the reference superimposition parameter aref2 $$\mathrm{G}$$$$\mathrm{}\mathrm{2}\mathrm{ʹ}=\mathrm{<figure-callout\; id="2"\; label="G"\; filenames="imgf0001.png,imgf0002.png"\; state="\{\{state\}\}">G</figure-callout>}\mathrm{2}\cdot \left(\mathrm{<figure-callout\; id="2"\; label="aref"\; filenames="imgf0001.png,imgf0002.png"\; state="\{\{state\}\}">aref</figure-callout>}\mathrm{2}-\mathrm{a}\mathrm{1}\right)\mathrm{.}$$

An output directional signal Xout is now based on the first directional signal Xr1 = Z1 + a1 Z2, weighted with the first gain parameter G1, and based on the second directional signal Xr2, which in this case corresponds to the second intermediate signal Z2, weighted with the corrected second gain parameter G2' formed as $$\begin{array}{cc}\mathrm{Xout}& =\mathrm{G}\mathrm{1}\cdot \mathrm{Xr}\mathrm{1}+\mathrm{<figure-callout\; id="2"\; label="G"\; filenames="imgf0001.png,imgf0002.png"\; state="\{\{state\}\}">G</figure-callout>}\mathrm{2}\mathrm{ʹ}\cdot \mathrm{<figure-callout\; id="2"\; label="Z"\; filenames="imgf0001.png,imgf0002.png"\; state="\{\{state\}\}">Z</figure-callout>}\mathrm{2}\hfill \\ & =\mathrm{G}\mathrm{1}\cdot \left(\mathrm{Z}\mathrm{1}+\mathrm{a}\mathrm{1}\cdot \mathrm{<figure-callout\; id="2"\; label="Z"\; filenames="imgf0001.png,imgf0002.png"\; state="\{\{state\}\}">Z</figure-callout>}\mathrm{2}\right)+\mathrm{<figure-callout\; id="2"\; label="G"\; filenames="imgf0001.png,imgf0002.png"\; state="\{\{state\}\}">G</figure-callout>}\mathrm{2}\cdot \left(\mathrm{<figure-callout\; id="2"\; label="aref"\; filenames="imgf0001.png,imgf0002.png"\; state="\{\{state\}\}">aref</figure-callout>}\mathrm{2}-\mathrm{a}\mathrm{1}\right)\cdot \mathrm{<figure-callout\; id="2"\; label="Z"\; filenames="imgf0001.png,imgf0002.png"\; state="\{\{state\}\}">Z</figure-callout>}\mathrm{2.}\hfill \end{array}$$

If, for example, the first useful signal S1 and the second useful signal S2 are of the same volume in a frequency range, then the same amplification parameter G1=G2 is assigned to them for this frequency range. In this case, in the output directional signal Xout, the contributions proportional to the first beat parameter a1 cancel each other out, and the output directional signal Xout goes into Xout = G1 (Z1 + aref2 Z2), as in the G1 amplified ("scaled") ) Reference directional signal Xref via.

Instead of the generation of the output directional signal Xout using the first directional signal Xr1 and the backward-directed cardioid signal Xa as the second intermediate signal Z2, the output directional signal Xout can also be generated by using the first superposition parameter a1, using the corrected second gain parameter G2' and a first effective superimposition parameter aeff1 and a second effective superimposition parameter aeff2 are formed on the basis of the first amplification parameter G1 as $$\begin{array}{cc}\mathrm{<figure-callout\; id="2"\; label="aeff"\; filenames="imgf0001.png,imgf0002.png"\; state="\{\{state\}\}">aeff</figure-callout>}\mathrm{2}& =\mathrm{a}\mathrm{1}+\mathrm{<figure-callout\; id="2"\; label="G"\; filenames="imgf0001.png,imgf0002.png"\; state="\{\{state\}\}">G</figure-callout>}\mathrm{2}\mathrm{ʹ}/\mathrm{G}\mathrm{1}\hfill \\ & =\mathrm{a}1+\left(\mathrm{<figure-callout\; id="2"\; label="aref"\; filenames="imgf0001.png,imgf0002.png"\; state="\{\{state\}\}">aref</figure-callout>}\mathrm{2}-\mathrm{a}\mathrm{1}\right)\cdot \mathrm{<figure-callout\; id="2"\; label="G"\; filenames="imgf0001.png,imgf0002.png"\; state="\{\{state\}\}">G</figure-callout>}\mathrm{2}/\mathrm{G}\mathrm{1.}\hfill \end{array}$$

mit aeff2 nach Gleichung (xiv) und aeff1 = 1 nimmt dann die in Gleichung (xii') dargelegte Form an. Das Ausgangs-Richtsignal Xout weist dabei infolge der vorliegenden Erzeugung eine Richtcharakteristik auf, welche in Richtung der ersten Nutzsignalquelle 14 (also der Richtung des ersten Nutzsignals S1) gegenüber dem Referenz-Richtsignal Xref eine Verstärkung oder Abschwächung um einen Faktor G1 aufweist, und in Richtung der zweiten Nutzsignalquelle 18 (also der Richtung des zweiten Nutzsignals S2) gegenüber dem Referenz-Richtsignal Xref eine Verstärkung oder Abschwächung um einen Faktor G2 aufweist (siehe hierzu auch Fig. 3a und Fig. 3b).The first effective overlay parameter aeff1 has the value aeff1=1 in the present special case, but can also assume non-trivial values, in particular for a2≠0. The correspondingly formed output directional signal $$\mathrm{Xout}=\mathrm{G}\mathrm{1}\cdot \left(\mathrm{aeff}\mathrm{1}\cdot \mathrm{Z}\mathrm{1}+\mathrm{<figure-callout\; id="2"\; label="aeff"\; filenames="imgf0001.png,imgf0002.png"\; state="\{\{state\}\}">aeff</figure-callout>}\mathrm{2}\cdot \mathrm{<figure-callout\; id="2"\; label="Z"\; filenames="imgf0001.png,imgf0002.png"\; state="\{\{state\}\}">Z</figure-callout>}\mathrm{2}\right)$$

with aeff2 according to equation (xiv) and aeff1 = 1 then takes the form set out in equation (xii'). As a result of the present generation, the output directional signal Xout has a directional characteristic which has an amplification or weakening by a factor G1 in the direction of the first useful signal source 14 (i.e. the direction of the first useful signal S1) compared to the reference directional signal Xref, and in the direction of the second useful signal source 18 (i.e. the direction of the second useful signal S2) has an amplification or attenuation by a factor G2 compared to the reference directional signal Xref (see also Figures 3a and 3b ).

Based on the output directional signal Xout, an output signal Yout is finally generated via signal processing steps 50, which can in particular include additional frequency band-dependent noise suppression, which output signal Yout is converted into an output sound signal 54 by an output converter 52 of the hearing device 2.

In Figure 3a is for the in 1 illustrated hearing situation of the wearer 1 has a directional characteristic 60 of the as in 2 described generated output directional signal Xout shown. The reference directional characteristic 62 (dashed line) is given here as an onmidirectional directional characteristic. For a better overview, the first gain parameter G1 is selected as 0 dB, while the second gain parameter G2 is selected as -6 dB. The resulting directional characteristic 60 of the output directional signal Xout has a noticeable deviation from the omnidirectional reference directional characteristic 62 in the second direction 10 (ie the direction of the second useful signal source 18).

In Figure 3b Instead of the omnidirectional reference directional characteristic 62, a reference directional signal Xref with a reference directional characteristic 64 is selected (dashed line), which models the filtering of the ambient sound by the pinna and corresponding shadowing effects. Again, the first gain parameter G1 is here chosen as 0 dB, while the second gain parameter G2 is chosen as -6 dB. The resulting directional characteristic 66 of the output directional signal Xout again shows a noticeable deviation from the omnidirectional reference directional characteristic 64 in the direction of the second useful signal source 18, with an additional weakening now taking place in this direction due to the definition of the reference directional signal Xref, which as a result of the Shading effects of the pinna go into the reference directional characteristic 64.

Although the invention has been illustrated and described in detail by the preferred embodiment, the invention is not limited by the disclosed examples and other variations can be derived therefrom by those skilled in the art without departing from the scope of the invention.

Reference List

1

carrier

2

hearing aid

4

first interlocutor

6

first direction

8th

second interlocutor

10

second direction

14

first useful signal source

18

second useful signal source

22

score

24

first input converter

26

second input converter

28

sound signal

40

adaptive directional microphone

42

adaptive directional microphone

50

signal processing steps

52

output converter

54

output sound signal

60

polar pattern

62

(omnidirectional) reference polar pattern

63

reference polar pattern

64

reference polar pattern

66

polar pattern

a1

first overlay parameter

aeff1

first effective overlay parameter

aeff2

second effective overlay parameter

aref1

first reference overlay parameter

aref2

second reference overlay parameter

E1

first input signal

E2

second input signal

G1

first gain parameter

G2

second gain parameter

G2'

corrected second gain parameter

S1

first useful signal

S2

second useful signal

T1

first delay parameter

T2

second delay parameter

yes

reverse cardioid signal

Xc

forward cardioid signal

Xout

Exit directional signal

Xr1

first directional signal

Xr2

second directional signal

yout

output signal

Z1

first intermediate signal

Z2

second intermediate signal

Claims (14)

Method for directional signal processing for a hearing aid (2), - a first input signal (E1) being generated from a sound signal (28) from the environment by a first input converter (24) of the hearing aid (2), - wherein a second input signal (E2) is generated from the sound signal (28) of the environment by a second input converter (26) of the hearing aid (2), - a first directional signal (Xr1) and a second directional signal (Xr2) being formed on the basis of the first input signal (E1) and the second input signal (E2), - wherein the second directional signal (Xr2) has a relative attenuation in the direction of a first useful signal source (14), - wherein the first directional signal (Xr1) has a relative attenuation in the direction of a second useful signal source (18), - wherein a first amplification parameter (G1) for an amplification of a first useful signal (S1) of the first useful signal source (14) and a second amplification parameter (G2) for an amplification of a second useful signal (S2) of the second useful signal source (18) are determined, - wherein a reference directional characteristic (62, 63, 64) is defined for a reference directional signal (Xref), - A corrected first gain parameter and a corrected second gain parameter (G2') being determined on the basis of the first gain parameter (G1) and/or the second gain parameter (G2) as a function of the reference directional characteristic (62, 63, 64) such that a Output directional signal (Xout), which is formed as the sum of the first directional signal (Xr1) weighted with the first corrected first gain parameter and the second directional signal (Xr2) weighted with the corrected second gain parameter (G2'), into a linearly scaled reference directional signal (Xref) transitions when the first gain parameter (G1) is equal to the second gain parameter (G2), and - wherein at least one of said two corrected gain parameters (G2') is different from the corresponding underlying gain parameter (G2). Method according to claim 1,
the corrected second amplification parameter (G2') being determined in such a way that the second useful signal (S2) is amplified by the second amplification parameter (G2) by the output directional signal (Xout) compared to the reference directional characteristic (62, 63, 64), and or
the corrected first amplification parameter being determined in such a way that the first useful signal (S1) is amplified by the output directional signal (Xout) compared to the reference directional characteristic (62, 63, 64) by the first amplification parameter (G1). A method according to claim 1 or claim 2,
wherein the corrected second gain parameter (G2') is formed as a product of the second gain factor (G2) and a correction factor, the correction factor being a linear coefficient of the second directional signal (Xr2) in a representation of the reference directional signal (Xref) as a linear combination of first directional signal (Xr1) and the second directional signal (Xr2). Method according to one of the preceding claims,
wherein the corrected first gain parameter is determined as the first gain parameter (G1) when the first directional signal (Xr1) has its minimum sensitivity in the direction of the second useful signal source (14). Method according to one of the preceding claims,
a first intermediate signal (Z1) and a second intermediate signal (Z2) being formed on the basis of the first input signal (E1) and the second input signal (E2),
wherein the first directional signal (Xr1) is formed as a superimposition of the first intermediate signal (Z1) with the second intermediate signal (Z2), and an associated first superimposition parameter (a1) is determined in the process, and/or wherein the second directional signal (Xr2) is formed as a superimposition of the second intermediate signal (Z2) with the first intermediate signal (Z1), and an associated second superimposition parameter (a2) is determined in the process. Method according to one of the preceding claims,
wherein the corrected first gain parameter (G1') is formed as a product of the first gain factor (G1) and a first correction factor, wherein the corrected second gain parameter (G2') is formed as a product of the second gain factor (G2) and a second correction factor. Method according to claim 5 with claim 6,
wherein a first reference overlay parameter (aref1) and a second reference overlay parameter (aref2) are defined for an overlay of the first intermediate signal (Z1) and the second intermediate signal (Z2), which forms the reference directional signal (Xref),
wherein the first correction factor is formed using a product of the second overlay parameter (a2) and the second reference overlay parameter (aref2), and/or
wherein the second correction factor is formed on the basis of a deviation of a product of the first overlay parameter (a1) and the first reference overlay parameter (aref1) from the second reference overlay parameter (aref2). A method according to claim 6 or claim 7,
wherein the output directional signal (Xout) is formed using the first directional signal (Xr1) weighted with the corrected first gain parameter (G1') and using the second directional signal (Xr2) weighted with the corrected second gain parameter (G2'). A method according to claim 6 or claim 7,
wherein based on the first and the second overlay parameter (a1, a2), the first and the second reference overlay parameter (aref1, aref2), and a first effective overlay parameter (aeff1) and a second effective overlay parameter (aeff2) are determined on the basis of the first and second gain parameters (G1, G2), and
wherein the output directional signal (Xout) is formed using a superposition of the first intermediate signal (Z1) weighted with the first effective superimposition parameter (aeff1) and the second intermediate signal (Z2) weighted with the second effective superimposition parameter (aeff2). Method according to claim 9,
in the event that the second directional signal (Xr2) is given by the second intermediate signal (Z2),
the first effective overlay parameter (aeff1) is formed from the first reference overlay parameter (aref1). A method according to claim 6 or claim 7,
a second effective superimposition parameter (aeff2) being determined on the basis of the first superposition parameter (a1), on the basis of the corrected first amplification parameter (G1') and on the basis of the corrected second amplification parameter (G2'), and
wherein the output directional signal (Xout) is formed using a superposition of the first intermediate signal (Z1) and the second intermediate signal (Z2) weighted with the second effective superimposition parameter (aeff2). Method according to claim 11,
in the event that the second directional signal (Xr2) is given by the second intermediate signal (Z2), the second effective superposition parameter (aeff2) is formed from the first superimposition parameter (a1) and a ratio of the corrected second amplification parameter (G2') and of the first gain parameter (G1). Method according to one of the preceding claims,
wherein the reference directional characteristic (63) of the reference directional signal (Xref) is selected as an omnidirectional directional characteristic (62), or is selected in such a way that a shadowing effect of human ears is simulated. hearing aid with - a hearing aid (2), which has a first input converter (24) for generating a first input signal (E1) from a sound signal (28) of the environment and a second input converter (26) for generating a second input signal (E2) from the sound signal (28 ) of the environment, and - a control unit which is set up to carry out the method according to any one of the preceding claims.

Images