EP2244491B2 - Method for operating a hearing aid with feedback suppression and hearing aid with a diplexer - Google Patents

Abstract

The method involves segmenting an input signal (100) of the hearing device into a low frequency and a high frequency signal components (101,102) by a frequency separating filter (1). The high frequency signal components are distorted. The low frequency signal component and the distorted high frequency signal component are overlaid to an output signal. A predetermined threshold frequency (GF) of the frequency separating filter is positioned through an analysis of the input signal such that artifacts in an output signal (105) of the hearing device are reduced. An independent claim is also included for a hearing device with a frequency distortion module.

Description

The invention relates to a specified in claim 1 method for operating a hearing and a specified in claim 6 hearing with a crossover.

In hearing devices, in particular in hearing aids, frequency-distorting algorithms are used for different purposes and at different points of signal processing. For example, is from the DE 699 22 940 T2 a hearing aid with a combination of audio compression and feedback suppression known. Common to all frequency-distorting algorithms is that they should usually only act from a so-called cut-off frequency, because distortions of low frequencies disturb the sound impression, but distortions of high frequencies are less critical.

FIG. 1 shows a block diagram of an exemplary implementation of a frequency distortion in a hearing aid. An input signal 100 is divided by a crossover filter 1 ("split-band filter") with a predefinable limit frequency GF ("split frequency") into a low-frequency and a high-frequency signal component 101, 102. The high-frequency signal component 102 is then distorted in a frequency equalizer 2. The distorted output signal 103 is supplied to an input of an adder 3. The low-frequency signal component 101 passes through an all-pass filter 4, which rotates the phase of the signal component 101 in such a way that signal deletion in the region of the limit frequency GF does not occur during a subsequent signal addition in the adder 3. The phase-rotated low-frequency signal component 104 is supplied to a further input of the adder 3. At the output of the adder 3, the sum of the two signal components 103, 104 is available as an output signal 105.

Crossovers are not ideal and have a finite frequency overlap at their cutoff frequency GF. In FIG. 2 is shown as an example of the frequency response of a crossover in a hearing aid with the limit frequency GF 1800 Hz. The curves K1, K2 show the damping D in dB as a function of the frequency F in Hz in the range 0 to 4000 Hz. The curve K1 shows a low-pass characteristic and the curve K2 a high-pass characteristic.

If a low-pass K1 filtered signal component is not distorted and a high-pass K2 filtered signal component is distorted, superimposing of both signal components occurs in addition to the signal components K1, K2, above all in the region of the limit frequency GF, which results in an output signal of the hearing aid as modulation or strong roughness is perceived. Both effects are very disturbing and usually noticeably more noticeable in the perception by a hearing aid wearer than the frequency distortion itself.

In addition to this "electrical" signal superposition, it can also lead to an acoustic superposition of a frequency-distorted and a non-frequency-distorted signal. Especially in the low frequency ranges and with an open hearing aid supply, an acoustic overlay is not negligible. Direct sound is superimposed on the frequency-distorted signal component that the hearing aid outputs, which in turn leads to the artefacts already described.

The WO 93/20669 A1 discloses a hearing apparatus with a variable high-pass filter which attenuates a low-frequency range of a microphone signal of the hearing device for the suppression of noise.

It is an object of the invention to overcome the disadvantages described and to provide a method for operating a hearing device and a hearing device with crossover, which reduce the occurrence of artifacts in a superposition of undistorted and frequency-distorted signal components.

According to the invention, the stated object is achieved with the method for operating a hearing device of independent claim 1 and with the hearing device of independent claim 5.

The invention claims a method for operating a hearing device by dividing an input signal into a low-frequency and a high-frequency signal component with a crossover and setting a cutoff frequency of the crossover based on an analysis of the input signal such that artifacts in an output signal of the hearing device are reduced. The invention offers the advantage that artefacts resulting from signal superimposition can be avoided.

A further development of the method may comprise a distortion, for example compression or displacement, of the high-frequency signal component and superposition of the low-frequency signal component and the distorted high-frequency signal component relative to the output signal, whereby artifacts can be formed.

Another embodiment of the method may include determining the cutoff frequency by analyzing the signal level or the tonality of the input signal.

The method may preferably include setting the cutoff frequency to a frequency at which the input signal has the lowest possible signal level and / or the lowest possible tonality. This effectively minimizes artifacts caused by overlapping bands of the crossover network.

Furthermore, the method may include detecting feedback, wherein the cutoff frequency is temporarily lowered upon detection of feedback. The advantage of this is that feedback whistles is avoided as a result of an unfavorable cut-off frequency.

The invention also claims a hearing device with a crossover characterized by a frequency difference, which divides an input signal into a low-frequency and a high-frequency signal component. The hearing device also comprises a first means by which the cutoff frequency of the crossover can be adjusted based on an analysis of the input signal such that artefacts in an output signal of the hearing aid are reduced. The first means may be a switching logic.

A further embodiment of the invention may comprise a frequency equalizer which distorts, for example compresses or shifts, the high-frequency signal component. The hearing device can further iron on an adder, which forms the output signal by superimposing the low-frequency signal component with de-distorted high-frequency signal component, wherein artifacts can be formed.

In a development, the first means can determine the cutoff frequency by analyzing the signal level or the tonality of the input signal.

Preferably, the first means can set the cutoff frequency to a frequency at which the input signal has the lowest possible signal level and / or the lowest possible tonality.

Furthermore, the hearing device may comprise a feedback suppression with a feedback detector, which temporarily lowers the cutoff frequency upon occurrence of a feedback.

Other features and advantages of the invention will become apparent from the following explanations of an embodiment with reference to schematic drawings.

Figur 1:

ein Blockschaltbild einer Anordnung mit Frequenz- weiche gemäß Stand der Technik,

Figur 2:

ein Schaubild eines Frequenzgangs einer Frequenz- weiche gemäß Stand der Technik,

Figur 3:

Schaubilder von Frequenzgängen bei einem Hörgerät gemäß Stand der Technik,

Figur 4:

Schaubilder von Frequenzgängen bei einem erfin- dungsgemäßen Hörgerät und

Figur 5:

ein Blockschaltbild einer erfindungsgemäßen Anord- nung.

Show it:

FIG. 1:

1 is a block diagram of a prior art frequency-locked device;

FIG. 2:

4 is a graph of a frequency response of a frequency sweep according to the prior art;

FIG. 3:

Charts of frequency responses in a hearing aid according to the prior art,

FIG. 4:

Charts of frequency responses in a hearing aid according to the invention and

FIG. 5:

a block diagram of an arrangement according to the invention.

FIG. 3 shows in three graphs the effect of the cutoff frequency GF of a crossover in a hearing device in a typical application. The upper diagram shows the frequency response K3 of the signal amplitude A in dB of an input signal, for example a microphone signal, as a function of the frequency F in Hz between 0 and 4000 Hz. A flute tone with a fundamental tone P1 unfavorably falls exactly to the limit frequency GF 1800 Hz of the crossover.

The middle graph of the FIG. 3 shows the frequency response of the crossover with a cut-off frequency GF at 1800 Hz. The curves K1, K2 indicate the attenuations D in dB as a function of the frequency F in Hz. The curve K1 shows a low-pass characteristic and the curve K2 a high-pass characteristic. At the outputs of the crossover are thus a low-frequency and a high-frequency signal component available. The high-frequency signal component is now shifted upwards by 100 Hz in frequency.

If the frequency-shifted by 100 Hz high-frequency signal component and the undistorted low-frequency signal component added, resulting in a frequency response K4 according to the lower graph of FIG. 3 , The curve K4 shows an output signal with a prominent double mode P2, which is perceived as a strong distortion.

According to this strong distortion is now avoided by the cutoff frequency GF of the crossover is set, for example, to 1200 Hz. In FIG. 4 this effect of the invention is shown in three graphs. The upper diagram shows as in FIG. 3 the frequency response K3 of the signal amplitude A in dB of an input signal as a function of the frequency F in Hz between 0 and 4000 Hz. A flute tone has its root P1 at 1800 Hz.

The middle graph of the FIG. 4 shows the frequency response of the crossover with a shifted to 1200 Hz limit frequency GF. The curves K1, K2 indicate the attenuations D in dB as a function of the frequency F in Hz. The curve K1 shows a low-pass characteristic and the curve K2 a high-pass characteristic. At the outputs of the crossover are thus a low-frequency and a high-frequency signal component available. The high-frequency signal component is now shifted upwards by 100 Hz in frequency.

If the frequency-shifted by 100 Hz high-frequency signal component and the undistorted low-frequency signal component added, resulting in a frequency response K5 according to the lower graph of the FIG. 4 , The curve K5 shows an output without a prominent double mode. The root note P3 of the flute is played almost distortion-free.

With the method according to the invention, therefore, a disturbing "electrical" superposition can be very effectively avoided or rendered inaudible. The solution according to the invention can be used independently of a known from the prior art on-off logic, which can complement the benefits.

If the crossover frequency GF of the crossover can be freely adjusted within a certain frequency range, overlapping artifacts can be significantly reduced by analyzing an input signal either for signal volume A or tonality and then setting the cutoff frequency accordingly. Preferably, the cutoff frequency is set to a frequency at which the input signal has a particularly low signal amplitude A or is particularly tonal.

FIG. 5 shows a block diagram of a typical embodiment of the invention. An input signal 100 of a hearing device, for example a microphone signal, is divided by a crossover network 1 with a predefinable limit frequency GF into a low-frequency and a high-frequency signal component 101, 102. The high-frequency signal portion 102 is then distorted in a frequency equalizer 2, for example, frequency-compressed or shifted the frequency. Such a distorted output signal 103 is supplied to an input of an adder 3. The low-frequency signal component 101 passes through an all-pass filter 4, which rotates the phase in such a way that signal deletions in the region of the limit frequency GF do not occur during a subsequent signal addition in the adder 3. The phase-rotated low-frequency signal component 104 is supplied to a further input of the adder 3. At the output of the adder 3 is available as a sum, an output signal 105 is available.

For adjusting or regulating the limit frequency GF according to the invention, the input signal 100 is split into frequency bands 106 with the aid of a filter bank 5. The bands 106 are then analyzed in a switching logic 6 with respect to their signal amplitude. From a predeterminable list of possible cutoff frequencies, for example 1000 Hz, 1250 Hz, 1500 Hz and 2000 Hz, that is selected in whose environment the signal level is as low as possible. With an output signal 107 of the switching logic 6, the crossover 1 is now faded to this selected limit frequency GF. As a result of the shift of the limit frequency GF artefacts caused by overlapping bands of the crossover 1, thus minimized.

In the case of a frequency distortion for improved feedback suppression in hearing devices may, depending on the ear coupling, in addition to the described "electrical" superposition also to an acoustic superposition of sound from the hearing and sound that comes directly to the eardrum acoustically come. This is due to physical reasons, especially at low frequencies. Here, too, the problem arises that undistorted signal components over the physical path overlap with intended distorted signal components via the hearing device in the same frequency band and lead to unwanted artifacts. Since this occurs preferably at low frequencies, the countermeasure would be to increase the cutoff frequency GF of the crossover 1 to such an extent that the superimposition occurring becomes insignificant. However, this increases the risk of feedback whistles in the low frequency band.

An advantageous remedy for this is the combination of the method according to the invention with a feedback detector 7 according to FIG. 5 If feedback whistles are detected - which is usually the more disturbing artifact - the cutoff frequency GF is suitably lowered by the feedback detector 7 and the feedback whistling is quickly suppressed. The superposition artifacts described above, which may occur at short notice, are the lesser evil. After successful suppression of feedback whistling and / or dominant / tonal signal components occur in the lower band, the limit frequency GF of the crossover 1 can be raised again.

LIST OF REFERENCE NUMBERS

1

Crossover / split band filter

2

Frequenzverzerrer

3

adder

4

All-pass filter

5

filter bank

6

switching logic

7

Feedback detector

100

Input signal / microphone signal

101

Low frequency signal

102

high-frequency signal

103

distorted signal

104

phase-compensated signal

105

output

106

frequency bands

107

Output signal of the switching logic

A

signal amplitude

D

damping

F

frequency

GF

cut-off frequency

K1

Frequency response lowpass

K2

Frequency response high pass

K3

Frequency response of the input signal 100

K4

Frequency response of the output signal 105

K5

Frequency response of the output signal 105 after shifting the limit frequency GF

P1

Basic tone in the input signal 100

P2

distorted fundamental in output signal 105 / mode

P3

undistorted root in the output signal 105

Claims (8)

1. Method for operating a hearing apparatus, by:

   - splitting an input signal (100) of the hearing apparatus into a low-frequency and a high-frequency signal component (101, 102) by means of a crossover filter (1),

   - frequency distortion (2) of the high-frequency signal component (102) and

   - overlaying (3) the low-frequency signal component (104) and the distorted high-frequency signal component (103) to form the output signal (105)
   characterised by:

   - setting (107) a predefinable cut-off frequency (GF) of the crossover filter (1) by analysing the input signal (100) so that artefacts in the output signal (105) are reduced.
2. Method according to claim 1, characterised by:

   - determining (6) the cut-off frequency (GF) by analysing the signal level (A) or tonality of the input signal (100).
3. Method according to one of the preceding claims, characterised by,

   - setting (107) the cut-off frequency (GF) to a frequency, at which the input signal (100) has the lowest possible signal level (A) and/or the lowest possible tonality.
4. Method according to one of claims 2 to 4, characterised by:

   - identifying (7) feedback, with the cut-off frequency (GF) being momentarily lowered when feedback is identified.
5. Hearing apparatus with

   - a crossover filter (1) characterised by a cut-off frequency (GF), which splits an input signal (100) into a low-frequency and a high-frequency signal component (101, 102),

   - a frequency distorter (2), which distorts the high-frequency signal component (102), and

   - an adding unit (3), which forms an output signal (105) by overlaying the low-frequency signal component (104) with the distorted high-frequency signal component (103), characterised by:

   - a first means (6), which can be used to set the cut-off frequency (GF) based on an analysis of the input signal (100) using the first means (6) so that artefacts in the output signal (105) are reduced.
6. Hearing apparatus according to claim 5, characterised in that the first means (6) determines the cut-off frequency (GF) by analysing the signal level (A) or tonality of the input signal (100).
7. Hearing apparatus according to claims 5 or 6, characterised in that the first means (6) sets the cut-off frequency (GF) to a frequency, at which the input signal (100) has the lowest possible signal level (A) and/or the lowest possible tonality.
8. Hearing apparatus according to one of claims 5 to 7 having feedback suppression, characterised by:

   - a feedback detector (7), which momentarily lowers the cut-off frequency (GF) when feedback is identified.

Images