WO1999004487A1

WO1999004487A1 - Signal-processing device

Info

Publication number: WO1999004487A1
Application number: PCT/IB1997/000902
Authority: WO
Inventors: Joseph Renier Gerardus Maria Leenen; Joseph Johannes Godefridus Jansen; Henry Cloetens; Zong Liang Wu; Stefan Marcel Maria Note
Original assignee: Koninklijke Philips Electronics N.V.; Philips Ab
Priority date: 1997-07-18
Filing date: 1997-07-18
Publication date: 1999-01-28
Also published as: JPH1168495A

Abstract

The signal-processing device (220) comprises an AGC-device (224), a programmable processor (222) and a memory means (226). The AGC-device (224) forms part of the programmable processor (222). The memory means (226) is embodied so as to comprise a set of instructions which can be executed by the programmable processor (222). An AGC-instruction forms part of this set of instructions. The AGC-device (224) can be activated by executing the AGC-instruction by means of the programmable processor (222).

Description

Signal-processing device

Description

The invention relates to a signal-processing device comprising an AGC- device, a programmable processor and a memory means. The invention also relates to a hearing aid comprising such a device.

The invention further relates to a memory means and to a hearing aid comprising such a memory means.

Such a device is disclosed in DE-A 4 407 032. The device known from this German patent application is intended specially to process speech signals. In this device, an AGC-device is used to amplify these speech signals. In general, such an AGC-device is an automatic amplifying circuit which multiplies an input signal by a specific amplification factor so as to generate an output signal. The amplification factor generally depends on a specific value derived from the input signal or the output signal. For example, the amplification factor may depend on the amplitude of the output signal. In the known device, the amplification of a speech signal by the AGC-device is controlled by a table stored in a memory. From this table, the amplification factor for the AGC-device is determined by means of the output signal. The known device further comprises a coding circuit. An input of this coding circuit is connected to the output of the AGC-device. After the speech signals have been amplified by the AGC-device, they are coded in a specific way in the coding circuit, whereafter they are stored in a speech memory. The known device does not enable the AGC-device to be used in a flexible manner. For example, in the known device a signal- processing sequence is fixed. This means that the signal is always first processed by the AGC-device and then by other signal-processing components. Another signal-processing sequence is impossible.

It is an object of the invention to provide a device of the type mentioned in the opening paragraph, in which the AGC-device can be used in a flexible manner. To achieve this, the device in accordance with the invention is characterized in that said AGC- device forms part of the programmable processor, the memory means being embodied so as to comprise a set of instructions which can be executed by the programmable processor, with an AGC-instruction forming part of the set of instructions which can be executed by the programmable processor, which AGC-device can be activated by execution of the AGC- instruction by the programmable processor.

By incorporating the AGC-device as a kind of co-processor in the programmable processor, with the execution of an AGC-processing operation by the AGC- device corresponding to the execution of the AGC-instruction by the programmable processor, the AGC-device can indeed be used in a flexible manner. This enables, on the one hand, a signal to be amplified a number of times by the AGC-device. This can be achieved, for example, by incorporating the AGC-instruction a number of times in the set of instructions which can be executed by the programmable processor, or by incorporating, in the instruction set, the AGC-instruction in a loop which is traversed several times.

On the other hand, in the device in accordance with the invention there are no limitations regarding the signal-processing sequence. By programming all signal- processing operations in the set of instructions which can be executed by the programmable processor, the sequence of these signal-processing operations can be adapted at will. In this set-up, a signal-processing operation corresponds to a number of instructions which can be executed by the programmable processor. The signal-processing sequence can now be changed by changing the sequence of the instructions in the instruction set corresponding to these signal-processing operations.

The memory means in accordance with the invention is characterized in that the memory means is embodied so as to comprise a set of instructions which can be executed by a programmable processor, which set of instructions which can be executed by the programmable processor includes at least one AGC-instruction, and, during execution of the set of instructions, the AGC-instruction can be executed at least a first time and a second time, and the set of instructions which can be executed by the programmable processor further includes a combination step for combining a first result of the first time that the AGC-instruction has been executed with a second result of the second time that the AGC- instruction has been executed.

In this context, a first AGC -operation corresponds to the first time that the AGC-instruction has been executed, said first AGC-operation being characterized, inter alia, by a first compression ratio. Accordingly, a second AGC-operation corresponds to the second time that the AGC-instruction has been executed, said second AGC-operation being characterized, inter alia, by a second compression ratio.

Here, the compression ratio represents the ratio between, on the one hand, an increase in amplitude of an input signal which must be amplified by the AGC- operation and, on the other hand, the corresponding increase in amplitude of the output signal resulting from said AGC-operation. This compression ratio becomes active from a specific level of the amplitude of the input signal. This level is also referred to as "knee- value". As long as the amplitude of the input signal remains below said knee-value, each increase in amplitude of the input signal is translated by the AGC-operation into a proportional increase in amplitude of the output signal.

By combining the results of AGC-operations, a resulting signal- processing operation can be readily achieved; otherwise, this would be impossible. For example, by counting up the first result and the second result, a signal-processing operation can be achieved which corresponds to a first resulting AGC-operation, said first resulting AGC- operation being characterized, inter alia, by a first resulting compression ratio. This first resulting compression ratio is governed by the first and the second compression ratio in such a manner that it can assume values which cannot readily be achieved with a single AGC-operation. By subtracting the first result from the second result, it is alternatively possible to obtain a signal-processing operation which corresponds to a second resulting AGC-operation, which second resulting AGC-operation is characterized, inter alia, by a second resulting compression ratio, and this second resulting compression ratio is not constant but depends on an amplitude of an input signal. Also this signal-processing operation cannot readily be achieved by means of a single AGC-operation.

The use of the device and the memory means in accordance with the invention in hearing aids has particular advantages. As the hearing deficiency of hearing- impaired persons varies substantially both in nature and degree, it must be possible to adapt a transfer function of a hearing aid to the hearing of an individual user in a flexible manner. In addition, it is known from past experience that AGC-devices in hearing aids can effectively be used to form this transfer function.

By providing the hearing aid with the device and the memory means in accordance with the invention, a hearing aid is achieved whose transfer function can be set in a flexible manner. In this connection, as described hereinabove, in particular the AGC-device can be used in many ways.

These and other aspects of the invention will be apparent from and elucidated with reference to the embodiments described hereinafter.

In the drawings:

Fig. 1 shows a block diagram of an example of a device for processing signals in accordance with the invention.

Fig. 2 shows a block diagram of an example of an AGC-device for use in a signal-processing device in accordance with the invention.

Figs. 3 and 4 show a number of input-output graphs by means of which the advantages of combining AGC-operations are described.

Fig. 5 shows a block diagram of an example of a hearing system comprising a hearing aid in accordance with the invention. Fig. 6 shows a block diagram of an example of a hearing aid in accordance with the invention.

Fig. 1 shows a block diagram of a signal-processing device 220 which comprises a programmable processor 222, an AGC-device 224 and a memory means 226. Said memory means 226 may be implemented, for example, as a RAM-memory. The AGC- device 224 forms part of the programmable processor 222. The device 220 further comprises an input 228 and an output 230.

An input signal applied to the input 228 is processed by the device 220 and, subsequently, appears as an output signal at the output 230. The relation between the input signal and the output signal is determined by a transfer function of the device 220.

The memory means 226 comprises a set of instructions which can be executed by the programmable processor 222. An AGC-instruction forms part of this set of instructions. The transfer function of the device 220 is now realized by execution of, by means of the programmable processor 222, the set of instructions stored in the memory means 226. The AGC-device 224 is activated by execution of the AGC-instruction by means of the programmable processor 222.

As the AGC-instruction forms part of the set of instructions, the AGC- device 224 can be used in a flexible manner by means of the device 220. For example, by embodying the set of instructions so that during the execution of the set of instructions by the programmable processor 222 the AGC-instruction is executed at least a first time and a second time, it becomes possible to combine a first result of the first time that the AGC- instruction has been executed with a second result of the second time that the AGC- instruction has been executed. For this purpose, the set of instructions must also comprise a combination step enabling the first and the second result to be combined. For the combination step, use can be made, for example, of an addition or a subtraction.

The AGC-device 224 shown in Fig. 2 is of the so-called feed-forward type. This means that an amplification of the AGC-device 224 depends directly on an input signal 240. The AGC-device 224 is controlled by five independent parameters. By means of said five parameters, influence can be exerted on an attack-time, a release-time, a knee- value, a compression-ratio and a gain.

The gain represents the overall amplification of the AGC-device 224. The compression ratio corresponds to the ratio between, on the one hand, an increase in amplitude of an input signal 240 which must be amplified by the AGC-device 224 and, on the other hand, the increase in amplitude of the corresponding output signal 268 resulting from the AGC-device 224. At a specific amplitude-level of the input signal 240, this compression ratio becomes active. This level is also referred to as the "knee-value" . As long as the amplitude of the input signal 240 remains below said knee- value, each increase in amplitude of the input signal 240 is translated by the AGC-device 224 into a proportional increase in amplitude of the output signal 268.

A sudden increase in the intensity of the input signal 240 causes the intensity of the output signal 268 to increase initially to a peak value, whereafter it decreases gradually to a stationary value. In the case of a sudden increase in the intensity of the input signal 240 by 25 dB, the attack-time is equal to the time period needed by the AGC-device 224 to regulate the intensity of the output signal 268 towards a value which is 2 dB higher than the stationary value.

A sudden decrease of the intensity of the input signal 240 causes the intensity of the output signal 268 to decrease initially to a trough-value, whereafter it increases gradually to a stationary value. In the case of a sudden decrease of the intensity of the input signal 240 by 25 dB, the release-time is equal to the time period needed by the AGC-device 224 to control the intensity of the output signal 268 to a value which is 2 dB lower than the stationary value. The input signal 240 is first rectified in a rectifier 242. This means that the absolute value of the input signal 240 is available at an output of the rectifier 242. The rectified signal forms a rough estimate of the intensity of the input signal 240.

Subsequently, in a logarithmic converter 244 a ²log(x)-operation is executed on the rectified signal. As a result, the signal is converted from the linear domain to the logarithmic domain. This has the advantage that a number of subsequent operations can be executed in a relatively easy and cheap manner.

Subsequently, the implementation of the release-time is effected in a release timer 245. Said release timer 245 comprises a block 246 for determining which one of two input signals is the largest, an adder 248 and a unit delay element 250. An output 252 of the release timer 254 is fed back to an input of block 246. In block 246, the signal 253 originating from the logarithmic converter

244 and the output signal 252 originating from the release timer 245 are compared. The largest one of these two input signals is passed on by the block 246 to the adder 248, where a value k2 is subtracted. K2 is a parameter which is connected with the release time. By subsequently delaying the resultant signal in the unit delay element 250, the output signal 252 is obtained.

By subtracting, in an adder 254, a value k3 from this output signal 252 originating from the release timer 245 and, subsequently, sending this result through a positive throughput device 256, the knee-value is implemented. K3 is a parameter which is connected with the knee-value. The positive throughput device 256 allows passage of an input signal if this input signal has a positive value. However, if the input signal has a negative value, an output signal of the positive throughput device 256 is set at zero.

The output signal of the positive throughput device 256 is subsequently multiplied by a value k4 in a multiplier 258. K4 is a parameter which is connected with the compression ratio. This multiplied signal is subsequently subtracted from a value k5 in an adder 260. K5 is a parameter which is connected with the gain. The resultant signal is subjected to a 2^X operation in an anti-logarithmic converter 262. As a result, the signal is converted from the logarithmic domain to the linear domain.

Subsequently, the attack-time of the AGC-device 224 is implemented in a first-order low-pass filter 264. The time constant of the filter 264 is determined by a value kl which is connected with the attack-time. If necessary, it is also possible to apply an overshoot limiter parallel to this filter 264, which limiter limits any peak and trough values to, respectively, a specific maximum and minimum value.

Finally, in a multiplier 266, the input signal 240 is multiplied by the signal supplied by the filter 264. The operation of the AGC-device 224 can be controlled by means of the values kl through k5. These values can be specified as arguments of the AGC-instruction.

In Figs. 3 and 4, a number of graphs corresponding to AGC-operations are shown in an input-output diagram. In this diagram, the amplitude of an input signal is plotted on the horizontal axis, and the amplitude of an output signal is plotted on the vertical axis.

In Fig. 3, curve 282 represents a first AGC- operation executed by the AGC-device 224, in which the gain is equal to 0 dB, the knee- value is -40 dB and the compression ratio is 2. Curve 286 represents a second AGC-operation executed by the AGC- device 224, in which the gain is -25 dB, the knee-value is -30 dB and the compression ratio is 1.

By combining the results of two AGC-operations, a resultant AGC- operation can be readily obtained. This is illustrated in Fig. 3 by the curves 280 and 284. Curve 280 is obtained by adding a first result of the first AGC-operation to a second result of the second AGC-operation. The first resultant AGC-operation shown in curve 280 is characterized by a first resultant compression ratio of approximately 1.75. Curve 284 is obtained by subtracting the second result of the second AGC-operation from the first result of the first AGC-operation. The resultant second AGC-operation is characterized by a second resultant compression ratio which is not constant, but instead increases as the amplitude of the input signal increases.

In Fig. 4, curve 302 represents a third AGC-operation executed by the AGC-device 224, in which the gain is -20 dB, the knee-value is -50 dB and the compression ratio is 1. Curve 306 represents a fourth AGC-operation executed by the AGC-device 224, in which the gain is 0 dB, the knee-value is -50 dB and the compression ratio is 8. In Fig. 4, the combination of two AGC-operations is illustrated by the curves 300 and 304. Curve 300 is obtained by adding a third result of a third AGC-operation to a fourth result of a fourth AGC-operation. Curve 304 is obtained by subtracting the fourth result of the fourth AGC-operation from the third result of the third AGC-operation.

It is noted that it is alternatively possible to obtain a resultant AGC- operation by combining the results of more than two AGC-operations. In general, the attack- time and the release-time of the AGC-operations to be combined do not have to be equal. By varying the attack-time and/ or the release-time of the AGC-operations and, subsequently, combining the results thereof, a resultant AGC-operation having different properties can be obtained. The hearing system shown in Fig. 5 comprises a card reader 10, a computer system 12, a remote control 14 and two hearing aids 16 and 18. The computer system 12 is a device which serves to load at least one hearing algorithm into the remote control 14. A hearing algorithm comprises a set of instructions which can be executed by a programmable processor which is incorporated in the hearing aid 16, 18. By execution of the set of instructions belonging to a hearing algorithm in the hearing aid 16, 18, a desired transfer function of the hearing aid 16, 18 is realized.

The computer system 12 and the card reader 10 coupled thereto are embodied so as to be used by a hearing-aid fitter, for example an audiologist. The hearing- aid fitter has a number of smart cards on which hearing algorithms are stored. Each one of these hearing algorithms corresponds to a specific transfer function of the hearing aid 16, 18.

After the hearing-aid fitter has determined the hearing characteristics of an ear of a hearing-impaired patient, the hearing-aid fitter can select, from the available hearing algorithms, a hearing algorithm which is suitable for this ear under specific sound conditions. This means that the hearing-aid fitter selects a hearing algorithm which corresponds to a transfer function of the hearing aid 16, 18, thus enabling the hearing deficiency of the ear demonstrated by the hearing characteristics to be corrected to the extent possible under the above-mentioned sound conditions. By means of a program which can be executed by the computer system

12, the hearing-aid fitter can, subsequently, read the selected hearing algorithm from the smart card and adapt it. For this purpose, the smart card containing the selected hearing algorithm must first be introduced into the card reader 10. Subsequently, by means of the program the hearing algorithm can be read from the smart card and loaded into the computer system 12. Next, the hearing-aid fitter can adapt the selected hearing algorithm by means of the program so as to achieve a fine adjustment of the transfer function of the hearing aid 16, 18 corresponding to the hearing algorithm.

In general, the above-described process of selecting and adapting hearing algorithms will have to be repeated a number of times by the hearing-aid fitter. Said number is equal to the product of, on the one hand, the number of ears for which the patient requires a hearing aid 16, 18 and, on the other hand, the number of different sound conditions for which an adaptation of the transfer function of the hearing aid 16, 18 is desirable. This can be explained by means of an example. Let us assume that the patient needs a hearing aid 16, 18 for both ears, and that after examination and consultation with the patient it has been decided that setting the transfer function of the hearing aid 16, 18 for two different audio conditions is desirable. This means that, in this example, the hearing-aid fitter has to select and adapt four (= two ears x two sound conditions) hearing algorithms.

The selected and adapted hearing algorithms can subsequently be loaded into the remote control 14 by means of the program. For this purpose, the remote control 14 can be coupled to the computer system 12, for example, by means of a serial connecting cable. After all hearing algorithms have been loaded into the remote control 14, the connection between the remote control 14 and the computer system 12 can be interrupted.

The patient can now control the hearing aid 16, 18 by means of the remote control 14. If necessary, one remote control 14 suffices to control two hearing aids 16, 18.

To control the hearing aid 16, 18, the remote control 14 comprises a transmitter for sending reference or control signals to the hearing aid 16, 18. To receive the reference or control signals, the hearing aid 16, 18 is provided with a suitable receiver. The reference or control signals may be in the form of infrared signals, ultrasonic sound signals or radio signals. It is alternatively possible to send the reference or control signals from the remote control 14 to the hearing aid 16, 18 via wires.

A number of different functions of the hearing aid 16, 18 can be set by the patient via the remote control 14. First, the patient can control the volume of the hearing aid 16, 18. Second, as the hearing aid 16, 18 may comprise both a microphone and a telephone coil, the patient can select a sound-reception source. In this case, the telephone coil can suitably be used as a sound-reception source in situations in which a special means for inductively transferring sound information is available. This is the case, for example, during a telephone call or in a room provided with a ring circuit. The microphone can be used as a sound-reception source in all situations. By means of the remote control 14, the patient can choose the microphone, the telephone coil or the microphone and the telephone coil as a sound-reception source.

And third, the patient can adapt the setting of the hearing aid 16, 18 for use under specific sound conditions. To this end, the patient can select a selection means of the remote control 14 which is coupled to these specific sound conditions, whereafter the associated hearing algorithm or the associated hearing algorithms are sent to the hearing aid 16, 18.

And fourth, the patient can put the hearing aid into a stand-by state. In this state, the hearing aid 16, 18 is in the off-position. In this state, the energy consumption of the hearing aid 16, 18 is minimal, while all settings of the hearing aid 16, 18 are preserved.

Fig. 6 shows a block diagram of a hearing aid 16, 18 comprising a device 220 for processing signals and a memory means 226. Said memory means 226 forms part of the device 220.

Claims

CLAIMS:

1. A signal-processing device (220) comprising an AGC-device (224), a programmable processor (222) and a memory means (226), characterized in that said AGC- device (224) forms part of the programmable processor (222), the memory means (226) being embodied so as to comprise a set of instructions which can be executed by the programmable processor (222), with an AGC-instruction forming part of the set of instructions which can be executed by the programmable processor (222), which AGC-device (224) can be activated by execution of the AGC-instruction by the programmable processor (222).

2. A memory means (226), characterized in that the memory means (226) is embodied so as to comprise a set of instructions which can be executed by a programmable processor (222), which set of instructions which can be executed by the programmable processor (222) includes at least one AGC-instruction, and, during execution of the set of instructions, the AGC-instruction can be executed at least a first time and a second time, and the set of instructions which can be executed by the programmable processor (222) further includes a combination step for combining a first result of the first time that the AGC- instruction has been executed with a second result of the second time that the AGC- instruction has been executed.

3. A memory means (226) as claimed in claim 2, characterized in that the combination step comprises an addition.

4. A memory means (226) as claimed in claim 2, characterized in that the combination step comprises a subtraction.

5. A hearing aid (16, 18) comprising a device (220) as claimed in claim 1.

6. A hearing aid (16, 18) comprising a memory means (226) as claimed in claim 2.