JPH06506322A - Bimodal audio processing device - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

[Detailed description of the invention] Bimodal audio processing device Technical field The present invention can be used to improve the hearing ability of people with disabilities by using acoustic hearing aids or ox-shell implants. processing of sound for the purpose of providing an information signal to a hearing aid device or both. Regarding improvements.

Look at the skin Throughout this specification, acoustic hearing aids refer to acoustic hearing aids that are placed in or around the ear of a disabled person. and output a sound suitable for at least partially compensating for the hearing loss of the disabled person. A type of hearing aid that is equipped with 1M cowhide implant hearing aids throughout this specification. The device is installed inside the body of a disabled person and electrically stimulates the disabled person's nervous system. 1 Devices containing components designed to partially compensate for the hearing loss of persons with normal disabilities. says.

There is a tendency for people with disabilities who have some residual hearing to have a genuine cow's shell implant installed in the contralateral ear. be. Many people with disabilities choose to use traditional acoustic hearing aids with formal cowhide implants. However, the combined It is difficult to accept. These people with disabilities receive acoustic hearing aids or cowhide implants. Do not try to use either hearing device or both together.

The purpose of the present invention is to provide a package that can drive both acoustic hearing aids and cowhide implant hearing aids. Equipped with an imodal hearing aid device, thereby making binaural information acceptable to persons with disabilities. The aim is to improve the quality of information.

There are two additional problems experienced with prior art hearing aids.

(1) Quickly and easily determine the nature and degree of hearing loss for people with disabilities in order to provide appropriate hearing aids. (2) is to measure the quality of hearing aids that are appropriate for the specific needs of the user. It is difficult to match the abilities of the two. Recently, the gain characteristics of the device at different frequencies There are a few hearing aid devices on the market that allow on-line control. Of course, this These devices may be equipped with formant extraction and similar feature extraction circuits. It does not have sufficient audio processing ability.

A further object of certain embodiments of the present invention is to combine acoustic hearing aids that address these issues. The goal is to have a signal processing device for use in the field.

Moon W Accordingly, in one broad form of the invention, audio information received from a microphone is processing means for receiving and processing the audio signal; An implant hearing aid in which the optical signal is implanted in the first ear of a disabled person and in the second ear of a disabled person. Bimodal information can be delivered to an acoustic hearing aid that is attached to or adjacent to the A pimodal hearing aid for the hearing-impaired is provided.

In a further broad form of the invention, the An electric wire that is electrically connected to the hearing aid transducer and placed in the ear of a person with a disability. includes a sound/audio processing device electrically connected to an air signal transducer; , the audio processing device receives and processes audio input information to transmit the audio input information to the hearing aid transducer. generating an acoustic signal from the transducer and an electrical signal from the electrical signal transducer; A bimodal hearing aid for people with hearing loss that allows interference bimodal information to be provided to people with disabilities. Be prepared.

In a broader form, electronically configurable sound/speech processing devices for the hearing impaired. The processing device includes a configuration means and a signal processing means. the processing device receives audio information and processes the audio by the signal processing means. processing according to the parameters set by the configuration means; , generates an output signal to excite the hearing aid transducer, and means for providing one or more electronic signal inputs or inputs for the purpose of changing said parameters. is adapted to receive software input.

In a further broad form of the invention, the sound/speech processing device provides a hearing aid and Methods and means are provided for controlling buoyant beef shell implants.

In the specified format, the sound/speech processing device has audio features FO, Fl, F2. ．． Al, A2. A3. A4. Hearing aid transcription using A5 and voiced/unvoiced decision generating the output signal that excites the diffuser. The above features are as follows. specifically defined.

FO is the fundamental frequency.

Fl is the first formant frequency.

F2 is the second formant frequency.

AO is the amplitude of FO.

A1 is the amplitude of Fl.

A2 is the amplitude of F2.

A3 is the amplitude of band 3 from 2000 to 2800 Hz (7).

A4 is the amplitude of band 4 from 2800 to 4000 Hz (7).

A5 is the amplitude of band 5 of 4000 Hz or higher.

According to a further particular form of said sound/speech processing device, said signal processing means Word features Al, A2. A3. Language spectrum defined as a function of A4 and A5 includes means for dynamically varying the amplitude of different frequency bands, and The magnitude of is between the threshold level and the optimal level at frequencies Fl and F2 and in higher frequency bands. It rises appropriately between bells.

In a further particular type of said sound/audio processing device, said signal processing means filter means, the settings of which are based on the language parameters extracted from said signal processing means; the output signal is dynamically changed according to the filtering means to remove the influence of noise. and/or dynamically adapted to overcome the user's particular hearing deficiencies.

Further, in the specified form of said sound/audio processing device, said processing device includes: The signal processing means includes means for reconstructing the signal in real time, thereby The amplitude and/or frequency characteristics of the output signal are optimally controlled to enhance the user's speech recognition. Melt.

In a first operating mode of the acoustic/audio processing device, the configuration The first filter means set by the first filter means is set by the acoustic engineer during the hearing aid installation process. provided based on the dimensions set by. Filter settings remain constant after installation is held at and remains constant thereafter.

In a variant of the operating mode of said sound/audio processing device, said processing device The meter includes dynamically changing filter means and the processing device is The writing method signal is extracted from the above configuration by extracting language parameters that act on the audio information. providing to said hearing aid transducer according to information provided in the ulation means; Ru.

In a further specified mode of operation of said sound/speech language processing device, said The output signals are integrated by said signal processing means using only language parameters.

's mouth Next, embodiments of the present invention will be described with reference to the drawings.

FIG. 1 is a schematic diagram of a bimodal hearing aid according to a first embodiment of the present invention; Figure 2 is a schematic diagram of the main functional components including the entire bimodal hearing aid.

FIG. 3 is a schematic diagram of the circuit configuration of both the acoustic processing circuit and the imbrand processing circuit.

Figure 4 shows the im- brant electronics for various stationary phonemes using the multi-peak code strategy. a chart showing examples of patterns of electrical excitation; Figure 5 is a graph showing the general volume expansion function of the audio processing section when driving the imbrandt. , Figure 6 is a schematic diagram of the circuit configuration of the functional components of the bimodal processing device that drives the acoustic hearing aid. , FIG. 7 shows the sound processing of an audio signal to drive an acoustic hearing aid according to a particular form of the invention. A schematic diagram of a circuit configuration of components including a sound processing device.

Figure 8 is a circuit diagram of the sound processing device; FIG. 9 is a circuit diagram of a picad filter used in the acoustic processing device of FIG. 8.

FIG. 10 is a circuit diagram of an output driver for an acoustic hearing aid.

FIG. 11 is a schematic diagram illustrating a preferred mode of operation of the audio processing device for audio input.

Figure 12 shows the sound intensity versus frequency for an acoustic hearing aid operating according to mode 1. Figure.

Figure 13 is related to an example of the operation of modes 1 and 3 of the sound processing device for one frequency. Figure 14 is a schematic diagram of how to install a Baydal hearing aid.

Figure 15 is a flowchart of how to attach the acoustic hearing aid part of the device in the first mode. to, Figure 16 shows the diagnostic and programming unit utility for use with bimodal hearing aids. It is a schematic diagram of the circuit configuration for.

beautiful Referring to Figure 1, the pimodal hearing aid is a pot beef shell implant in one ear of a disabled person. Having the ability to have information via a sound processing board in the other ear via an acoustic hearing aid.

Both im-brand hearing aids and acoustic hearing aids deliver live audio from a single microphone. Controlled by the same language processor that derives the input.

The speech processing device extracts several parameters from the input acoustic sound waves suitable for speech recognition. Some audio parameters extracted by the audio processing device are converted into electrical signals. and sent to the purchasing cowhide imbrandt. These characteristics are the basics of changing the audio waveform. The signal can then be amplified and sent to an acoustic hearing aid. disabled person Some people have already received implants and some have implants in their ears that have not yet been implanted. Even people who wear hearing aids may still have some hearing loss. These obstacles The error is that the sound produced by conventional acoustic hearing aids is not compatible with the sound produced by implants. It is often reported that it is not possible. People with such disabilities tend to rely on one or the other. This is a bimodal option for people with disabilities who do not make full use of their limited hearing ability. A hearing aid is desired. Contains M midshell implant hearing aids and speech processing acoustic hearing aids Such devices allow these disabled people to access any independent hearing aids currently available. This makes it possible to identify voices rather than instruments.

In general, if a disabled person has some hearing left, 1M cattle tend to have low frequencies. The shell implant is located at a position corresponding to higher frequencies (usually above 700Hz). generates excitation.

Therefore, by combining two channels, either channel can be Useful information can be provided over a wider range of frequencies than can be provided. Change The temporal decomposition of frequency and residual hearing is achieved through the use of a pimodal hearing aid. better than that obtained with pulsating electrical signals.

In addition to the above-mentioned use of "1'-5 Imodal", the driving part of the acoustic hearing aid of this device is The shell can be used as a speech processing hearing aid independently of the output of the implant hearing aid. this When used in this way, it has advantages over traditional acoustic hearing aids. Conventional hearing aids use frequency/ Adjustment of gain characteristics is limited to a small number of selections, and many users are unable to receive proper hearing aids. There were restrictions on implementation due to the with more flexible frequency/gain There is a need for a hearing aid and this is achieved with this hearing aid. In addition, scales and beef shell in bran The feature extraction circuit, which is the basis of hearing aids, allows hardware to extract important characteristics of the audio signal. can be measured in quiet conditions and under moderate noise. These properties are selectively amplified , a new signal that is enhanced in relation to the remaining acoustic signal or exclusively carries the same information. It is used to synthesize different speech waveforms. This is the acoustic signal processing that outputs to acoustic hearing aids. The process is executed by the management device (12).

The synthesized waveforms are used to overcome special problems.

For example, a high frequency sound above the user's hearing limit may be heard at a lower frequency within the user's hearing range. It appears as a high frequency. This is the broad peak in the language spectrum. provides excellent frequency resolution, allowing one peak to be masked by other adjacent peaks. Facilitate the decline. There is no other attachable device that can perform these processes.

A sound/audio processing device takes the audio signal from the microphone and measuring selected features of the voice (including frequency and amplitude of formants), and In the case of the first example, the second example In this example, the output is controlled only to the acoustic hearing aid.

1.0 1 - Bimodal Figure 2 is a schematic circuit diagram of the operation of this device. The auditory part is protected by existing patents and patent applications by the same applicant and is Blunt hearing aids use a similar strategy to that already developed for implant users to act on one ear of the disabled person. Additionally, users of bimodal devices Receive audio signals through an acoustic hearing aid in the ear that is not connected. Bimodal hearing aid capabilities This signal is specially conditioned to send correction information to the implant hearing aid. is possible and makes the most of the remaining hearing ability of the disabled.

Figure 2 discloses in detail the body-worn parts of the bimodal device, including the microphone 13, a language processing device 11 closely coupled to the speech processing device 12, and an acoustic hearing aid. The implant hearing aid 15 includes an inner coil hearing aid 14 and an implant hearing aid 15. 19 and an outer coil 20 to communicate wirelessly with the audio processing device 11. 8 includes an electronic array 16 electrically connected by a harness 17.

Furthermore, FIG. 2 shows the diagnostic and programming unit 21 and the diagnostic programming input. 2 shows auxiliary features of the interface 22.

Currently, the diagnostic and programming unit 21 is on a personal computer. It is embedded as a scanning program, while the diagnostic programming interface Base 22 is a communication card connected to the PC bus for 1 diagnostics and programming. Unit 21 is a clinic with a sound system that optimizes the interview of people with disabilities according to established standards. Testing device operating parameters for voice processing device 11 and/or acoustic hearing aid processing device 12 and control.

These parameters can be accessed via the diagnostic programming interface 22 for voice processing. map memory storage 23 within the management device 11. implant hearing aid 1 5 when driving the audio processing device 11, and when driving the acoustic hearing aid 14, the acoustic hearing aid How the processing device 12 processes the audio signals received from the microphones 13 on both sides. is determined by referring to the parameters stored in the map memory storage mechanism 23. Depends on.

acoustic hearing aid processor 12 and acoustic hearing aid 14 and diagnostic and programming unit 21 The components shown in Figure 2 other than the computer program that controls the functions of the The 1M cowhide implant hearing aid is mentioned somewhere in the requested patents and patent applications. As far as the operation of the vessels is concerned, they are the same.

The exact method of electrically exciting the audio processing device 11 and the implant hearing aid is now explained. For example, when placed in the ears of a person with a disability, Excitation of the excitation electrode can be done digitally or analogously. Until now, One of the Applicants, Kokriya Proprietary Limited, uses pulsating electrical signals. Digital electronic excitation warfare using so-called pulsating signals given to the transducer I have been pursuing strategy.

In particular, the audio processing device 11 is manufactured by Kokriya Proprietary Limited (of this application). (one of the co-applicants) since approximately 1982. early unit and indeed modern units also first listen to all other sounds received from microphone 13. The aim is to improve speech recognition for. This is processed by the audio processing device 11. From the raw audio input received from the Crophon 13, how is it perceived by the human auditory system? to identify and process acoustic features of speech determined to have the best characteristic speech information. It will be done.

Early forms of speech processing devices 11 presented three speech-acoustic characteristics to implant users. did. These are the amplitude expressed as the current level of the electrical excitation, and the pulsating excitation rate. The fundamental frequency (FO) or voice pitch represented and the excitation located in the ear of the disabled person. F2 is the second formant frequency (F2) represented by the position of the vibrating electrode pair. The frequency is typically within the frequency range 800 to 2500 Hz.

Later, a second excitation electrode pair representing the first formant (Fl) of speech was added.

The F1 signal is typically within a frequency of 280 to 1000 Hz. This large (FOFIF (known as FOF2 type) improves the in-domain performance of speech recognition compared to the previous FOF2 type. Good. Recently, the information given to the audio processing device 11 and processed is of medium level. It is augmented with one specific purpose: to improve speech understanding in noisy environments.

This last coded type has all the information available in the FOFIF2 type and , with additional information from three high-frequency bandpass filters. These frames The filter operates in the following frequency ranges: 2000-2800Hz, 2800-400Hz Covers 0Hz and 4000 to 8000Hz. energies within these areas The electrical energy limits the amplitude of electrical excitation of the three fixed electrode pairs at the proximal end of the electrode array. mosquito Thus, additional information about high-frequency sounds is provided at the tonally appropriate location within the pot shell. available.

The overall excitation rate for voiced sounds remains as FO (Fundamental Frequency or Vocal Pitch), but the new In large sizes, four electrical excitation pulses occur for each glottal pulse. This, the voice beeps. Compare with the FOFIF2 strategy, which produced only two pulses per cycle. new In the chord type, for voiced sounds, there are two forms representing the first and second formants. pulses of 2000 to 2800 Hz and 2800 Hz are provided as before; Another excitation pulse is generated representing an energy in the region of 4000 Hz.

For unvoiced sounds, another pulse representing energy above 4000 Hz is provided. , the 17th ormant usually has almost no energy in this frequency range, so it cannot be excited. does not occur, the excitation is approximately 260H, about twice that used in previous strategies. occurs at an arbitrary pulse rate of z.

Modern noise suppression algorithms replace voice-activated switches like those used before. Unlike others, it operates in a continuous manner. This is due to the recognition of early systems. Eliminate switching on and off that makes it difficult. In the new algorithm, the noise The wave current is continuously applied to each frequency band over a period of 10 seconds or more. this cycle The lowest level above is presumed to be background noise, and based on the amplitude related to that frequency band. Subtracted. Thus, an increase in signal amplitude above the noise level is represented to the disabled person. At the same time, the ambient noise level itself decreases to near the threshold.

Figure 3 shows a bimodal supplement in which the basic filter and one of the latest processing types mentioned above are inserted. 3 shows the processing arrangement of the auditory coupling means;

International Patent No. 1 [PCT /AU90100407 describes the operation of these constructs in detail. Its origin The entire text of the application specification and drawings are incorporated herein by reference. This details The most relevant parts of the book are included below.

The nature of the electrode arrays used in connection with modern coding strategies and the methods of their implantation. and its properties are in the literature5. For example, G. M. Clark, G. Batrick Tong, eds. , described in "Imperial Cow Shell" published by Churchill Livingstone in 1990. .

R.L. Web, P.C. Paiman, and B.K. in Chapter 9 of this book. Written by Franz and G.M. Clark, entitled “Legal Bovine Shell Insertion Surgery” is particularly appropriate.

This chapter, text, and drawings are incorporated herein by cross-reference.

The spectral peaks of Fl and F2 from the microphone audio signal by the coding strategy , and use the extracted frequency estimates to determine the more apical and more fundamental Select the excitation electrode pair at the end. Each selected electrode receives a pulse equal to the reference frequency FO. is excited at a speed of In addition to Fl and F2, spectral information of three high frequency bands is Extracted. Band 3 (2000-2800Hz), Band 4 (2800-40Hz) 00Hz) and band 5 (above 4000Hz), amplitude calculation is performed using a fixed electrode, e.g. represented individually in the seventh, fourth and first electrodes of the polar array 16 (see FIGS. 2 and 4). (see).

1st. The fourth and seventh electrodes are selected as default electrodes in the high frequency band.

After all, they are located far apart and most people with disabilities cannot be excited in these three locations. This is because it is not possible to distinguish between them. These default roles are Note that this is programmed. The three high frequency bands are the three highest in the map. If only the proximal electrodes are applied, many people with disabilities will have good contact between adjacent proximal electrodes. This is because there are many cases where the recognition of positional spacing is not shown. Furthermore, from electrical excitation The resulting overall pitch! ! ! He seems too knowledgeable.

Table 1 below shows the frequencies of various formants used in the speech coding type of the present invention. It indicates the area.

shit-work W formant or obi 280 1000Hz Fl 800-4000Hz F2 2000-2800Hz Band 3-Electrode 72800-4000Hz Band 4-Electrode Pole 14000Hz or higher Band 5 - Electrode 1 If the human signal is voiced, it has a periodic fundamental frequency.

Fl, F2 and 3rd. The speed at which the electrode pair selected from the calculated value of the fourth zone is equal to FO is sequentially excited. The most proximal electrode pair is excited first, as shown in Figure 4. It is excited by a pair of electrodes that gradually come to the top. Band 5 is not shown in Figure 4 .

Most voiced sounds contain negligible information in this frequency band. This is because they are not.

If the input signal is silent, the energy in the F1 band (280-1000Hz) is usually It is zero. Therefore, it is a frequency band that extracts information above 4000 Hz. available. In this state, the electrode pair selected from the calculated values of F2 and bands 3 and 4 is Receive pulsating excitation. The excitation rate is periodic and varies between 200-300Hz do. Therefore, the coding strategy extracts and codes five spectral peaks. However, in one excitation sequence, only four spectral peaks are coded. be converted into

Figure 4 shows the electrical excitation patterns for various invariant state phonemes when using this encoding strategy. The main feature of the map is the frequency of the dominant spectral peaks (Fl and F2). The idea is to change the number to electrode selection. To perform this function, the electrodes are made from a pot made of cowhide. They are numbered consecutively starting with . Electrode 1 is the most proximal electrode and Pole 22 is also at the apex of the electrode array dropout.

The excitation of different electrodes usually produces a pitch perception that reflects the tonal organization of the Tango shell. Pole 22 produces the lowest pitch recognition, or the dullest sound, electrode 1 is the highest. Place pitch recognition or bring out the dullest sounds.

When omitted, the frequency range of the spectrum peaks of Fl and F2 is distributed to the total number of electrodes. As shown in Figure 4, the total number of electrodes that can be used by the map algorithm is approximately 1:2. Divide into proportions.

Inside the beef shell processing device, a rantom access memory is stored in a map memory storage device 23. We have accumulated a number of tables that are shown together as . Maps are Fl, F2 and Determine both excitation parameters and amplitude estimates for bands 3-5. Excitation parameters The encoding of involves a series of separate steps. The process is summarized in the following key points:

1. The first formant frequency (Fl) is the main spectrum of 280-1000Hz. Change to a number based on the beak.

2. Using the number of Fl in relation to one of the map tables, calculate the vibration that becomes the first formant. Determine the electrode width. Different electrodes are determined by mode.

3. Set the second formant frequency (F2) to the main frequency in the 800-4000Hz band. Change to a number based on vector peaks.

4. Use the number F2 in conjunction with one of the map tables to create an excitation representing the second formant. Determine which electrodes should be used. Different electrodes are determined by mode.

5. Set the amplitude calculation values of bands 3, 4 and 5 to the default electrodes 7, 4 of bands 3, 4 and 5. ．． 1 for each, or such other electrodes are selected when preparing the map.

6. Change the amplitude of the acoustic signal of each frequency band to a number between 0 and 150. Amplitude level sent Bell sets the acoustic amplitude (between 0-15) of the specific electrode selected in steps 2 and 4.5 above. The excitation level is determined by referring to a set of maps related to the excitation level.

7. The data is further encoded in the audio processing device and transmitted to the receiver/exciter 18. be done. The data is then decoded and the excitation signal is sent to the appropriate electrodes. . The excitation pulse appears at the same velocity as the FO during the voiced period, and at the same speed as the FO and Fl during the unvoiced period. 6 that appears at an arbitrary periodic rate within the formant region (usually 200-300 Hz). Audio processing bag! 11 further converts the acoustic signal amplitude into electrical excitation parameters using a nonlinear volume Contains growth algorithms. The amplitude of the acoustic signal is determined by the audio processing device 11 as shown in FIG. The scale is changed to a digital uniform scale with values from 0 to 150. This uniform scale is ( (combined with the information stored in the disabled person's MAP) fed to the electrodes in the electronic array 16 Determine the actual charge.

Improvements to this assembly are included in the pending specification of Koklyaburo Priestly Limited. Are listed. In the details, 1 international outbound PCT/AU90100406 is microphone Between the lophon 13 and the audio processing device 11 and between the outer coil assembly and the audio processing device 11 Discloses an improved connection system between. The text and drawings of this application Incorporated herein by cross-reference.

A noise suppression circuit is disclosed in International Publication No. 1 [PcT/AU90100404.

The text of the specification and drawings attached to this application are hereby incorporated by reference. Ru.

FIG. 6 shows a block diagram of the processing circuit showing the functional interconnection of the components that drive the acoustic hearing aid 14. It is a lock diagram. The main components are microphone 13. Automatic gain controller 24. sound Voice parameter extractor 25, encoder 26, disabled person map memory storage device 23 , a noise generator 27 , and an acoustic hearing aid signal processing device 12 .

This allows the acoustic hearing aid 14 to be driven, and the center of the bimodal hearing aid is acoustically This is a hearing aid signal processing device 12.

The acoustic hearing aid signal processing device is a software configuration and has three Contains two-pole filters, each with bandpass, lowpass, and highpass configurations. Can be used for any of the following options. The center frequency, bandwidth and output of these filters The force amplitude is controlled by a processor. Filters can be used in series or parallel, and the input The waveform may be an audio waveform, a pulse noise signal, or an external signal. External signals are connected to other microphones. crophon, other acoustic output or other acoustic signal processing device. This makes it possible to can operate in a similar manner to acoustic hearing aids (not provided by currently available hearing aids). (with more accurate gain tonality than the other types) as well as different types of processed sounds. A versatile hearing aid is obtained that can operate as a hearing aid with voice information. These three programs Acoustic hearing aid signal processor including programmable filters on a single silicon chip It is inlaid. Each filter can be used as a high pass, band pass or low pass filter. Can be used. Each filter has 128 center frequencies between 100Hz and 16000Hz number, 128 Q values between 0.53 and 120 and 128 amplitude values between O and 64 dB is obtained (Q=center frequency/bandwidth).

This chip has the versatility to cover a wide frequency range, amplitude and spectral type. Ru.

It also includes a digital-to-analog converter (DAC) and provides an excitation waveform for the filter. used to generate A DAC is any type of DAC that is directly controlled by a processing unit. It can produce waveforms (sine or pulsed), or it can produce audio waveforms or white noise. It can be converted to provide excitation by an noise generator.

A schematic diagram of the three filter circuit is shown in Figure 7.

The functional specifications of the single chip implantation of the acoustic hearing aid signal processing device are shown in Figures 8, 9 and 10. Therefore, it is given. The details of the specifications are as follows.

blood■ Figure 8 shows the overall phase of the chip, the center frequency and bandwidth can be controlled independently. Three programmable filters are provided. The output of these filters is first attenuated or amplified and then mixed. The output of one of the three filters is required If so, it can be reversed by setting the INV bit.

Figure 9 shows the Vicad filter that together with the frequency latch and Q latch make up the three filters. 1 Frequency latch and Q latch are the parameters of the bicad filter. Determine the data.

Chip phase can be changed from series to parallel or mixed configuration by three PARn bits can.

The signal source in this configuration is selected by a four channel multiplexer (MUX). Ru. This can be a buffered output of a +5 volt audio signal, an internally generated noise signal, or an external signal. Select the number. This signal source is a 7-bit digital to analog reference voltage. converter (DAC).

Duplex D A C1iD C level can be converted into a pulse generator or audio or external Provide fine gain control on the signal or noise source.

The most significant bit (MSB) is used to transform the output.

All filter outputs are summed and passed to the push-pull earphone assembly.

The moving object is 10 volts peak to peak across the 270 ohm (nominal) earphone. This is an effective way to provide support. The chip uses a 5 volt single supply. .

The earphones have an impedance of 88 ohms with an impedance that gradually increases to 270 ohms at IKHz. Note that it has a DC resistance of ohms. The output stage is as shown in Figure 10. It consists of a bridge of P and N transistor switches. The switch is one person's power The sound by the signal derived from the comparator longitudinally driven from the angular wave and on other inputs. This is the pulse width modulated by the voice signal. The on-resistance of the switch is less than 5 oars. It should be.

(preferably lower).

In addition to type 0 output, there is a single-ended linear output.

This can gain or subtract 5mA for a 1 volt drop.

The chip is programmed by writing to a map of the audio processor. Tsk To clarify writing between maps, bits A3-Al2 are coded. Ru. Writing to block 1800-18FF in the map also writes to the chip. be done.

Two addresses, one odd and one even (Y13 and Y14) are decoded. , the outputs are logged and these outputs (R/W new read writes) can be more selectively filtered. used to write to the router chip. Select MUX for odd number writing, even number writing Set the Automatic Sensitivity Control (ASC) latch on the chip.

The fourteen lowest address bits contain programming information for the acoustic processing chip. Used to write to registers.

Registers YO to Yll contain the frequency, Q, gain (attenuation or amplification) and three filters. program the configuration of the router. Register 12 is the chip phase Set. Register Y15 is used to write the DAC.

Referring to FIG. 8, the phase latch Y12 is as follows.

Do INV Di, D2, 03 phase bit D4. D5 DACilJI D6 DIR 00+5 volts 1 1 External signal source phase bit Do-INV Convert the output of filter 1.

Send the output of Di-PARI filter 1 to Summer.

Send the output of D2-PAR2 filter 2 to Summer.

Send the output of D3-PAR3 filter 3 to Summer.

Send the output of D6-DIRDAC to Summer.

When the filter's PAR bit is set, the cascaded filter and attenuator/amplifier is sent to Summer, but the filter output itself is sent to the bus, which is used by other filters. It is used for When the PAR bit is not set, the filter and attenuator/amplifier connections are The coalescence is sent to the pass. If the filters are cascaded, the filter gain will be It can be decided by

I'm always watching the evening Configuration latch Y3. Y7. YllDo, Di, D2 filter Selection of type D3. D4 Clock selection D5. D6 Filter input 5 Phil The filter inputs selected by D6 and D5 are shown below.

input Filter o 1 2 3 1 DAC sound voice filter 2 filter 3 inverter 2 DAC sound voice filter Filter 1 Filter 33 DAC noise Filter 1 Filter 2 clock pre A scaler is provided to extend the frequency range of the filter. This calls it itself This is done by dividing the clock by two or four before feeding it to the divider.

The decoding is as follows.

D4 D3 Divider Divider input 0 0 1 5MHz 0 1 2 2.5MHz 1 2 4 1, 25 MHz l　1　1　5　MHz　However, 2 switches (HF) or more double the center frequency However, it is open to filters.

-2ヱ5 horned fireflies Referring to Figure 9, the filter is a loop with various gain feedback buses. It consists of two integrators. The input can be a swapped capacitor or a non-swap capacitor. It is often given to the first or second person. the output is taken from the first or second integrator Ru. This results in a variety of different transfer functions, as follows.

Mode D2 Di D.

O exchange 1st 1st power drop 1 Exchange 1st 2nd reduction passage 2 Exchange 2nd 1st reduction passage 3 Exchange 2nd 2nd band pass 4 Non-exchange 1st 1st high pass 5 Non-exchange 1st 21st F area passage 6 Non-switched 2nd 1st band pass 7 Non-exchange 2nd 2nd high pass In this table, bit zero selects a specific condition.

In mode 0, Do, Di, and D2 are all zero, and the power is lowered.

Filter and switch are output off.

In some cases it is desirable to close out all other functions, this can be done by external pins. The electric panel is operated in a power descending bimodal manner.

Programmable filter clock divider is compared to the contents of the frequency latch Contains a 7-bit ripple counter, the first match causes the counter to reset to 0 Frequency latches must be accompanied by necessary count correction6 e.g. If bits are needed after two counts, all launches are high except for bit D1, which is low.

The outputs of all NOR gates connected to the latch are low except for the one connected to Dl. , now the counter counts from zero and becomes high in the first period of Q2, this NOR also goes low and the 7 input NOR at the output is reset to the counter.

Filter Q has a 4-bit pin reset along with a 3-bit programmable resistor divider. is programmed using a capacitor with a capacitor. The resistor is Y1 latch Y5 and Y9 for filters 3 and 4) Number N represented by middle bits D4-06 Programmed by The capacitor is represented by bit Do-D3. Programmed by

Q is expressed by the following formula.

Q=8 (1+1.875n)/m A switch is used to connect the amplifier's feedback resistor to a component such as a buffer attenuator. configuration. i.e. the output can be converted to another attenuator or Can drive the device.

Two separate sections are used. One is 8 4's giving +/-28dB dB step, and the others are eight 0.5 dB steps with a maximum of +/-3,5 dB. That's about it. The total range is therefore +/-31.5 dB. Two 8 channel J Select the required taps of the potential divider using L/MUX.

The attenuator/amplifier is latch Y2. Y6. Can be selected by addressing YIO Ru. Attenuation and 0 intensification by bit D6=1 is provided.

Regarding gain settings: Y2 = dB gain x 2 Regarding attenuation settings: Y2 = 64 - (dB attenuation x 2) If Y is set to 0 (all bits low), attenuator/amplifier The power to the width transducer is lowered and switched off.

Note that the HF setting adds 6dB to the gain.

The attenuator is arranged so that its gain does not change except when the signal passes through zero. child This prevents clicks and pops between gain changes. positive or negative towards zero crossing The zero-crossing detector that generates the pulses switches the latch that sends the necessary gain to the attenuator MUX. Used to trove.

The reciprocating DAC is a 6-bit resistance ladder type, and the digital value in latch Y15 controls the output voltage. The input REF selected by the source select (FIG. 8) is multiplied. Bit D6 is Convert the output.

The set time is 50 microseconds.

Due to its versatile and programmable configuration, the acoustic hearing aid processor 12 has many This allows the installer to experiment with sound processing strategies and ultimately find the best acoustic supplement for the installer to obtain information. The in-branch installed in the other ear of the installer can be determined to be a hearing aid. correct the information received from the hearing aid. This versatility and programmability The pimodal processor can be used to adjust the operation of acoustic hearing aids only.

In both cases, a single microphone is used to maximize the versatility of the acoustic hearing processing device 12. Combined with pre-processing capabilities of the audio processing device 11 combined with programmability. , with features and benefits not found in hearing aids to date.

Next, we will explain bimodal hearing aids when used to drive only acoustic hearing aids. . However, the mode of operation to be explained here is acoustic compensation by one wearer. In the first embodiment, in which both a hearing aid and an implant hearing aid are used, the above-mentioned correction effect is achieved. Can be used to aid as well. Therefore, in that sense, the following explanation is based on the first embodiment. It can be applied similarly.

The nature of the correction between two hearing aids is subjective and depends on a combination of repeated testing and wearing experience. Determined by The bimodal hearing aid configuration described here achieves this correction. be done. Test processing for accumulating desired disabled person parameters in the map memory storage 23 and methods are explained later in the specification.

2.0　2　-　Hibiki　only as a barrage The acoustic hearing aid signal processing device 12 has three software configurable features. It includes a filter and, due to its inherent versatility, receives audio from the audio processing device 11 and The frequency domain processed signal directed to the hearing aid 14 is provided with almost infinite flexibility. . Four specific modes of operation are desired and achievable by the acoustic hearing aid signal processing device 12. It is said to be possible.

Four basic modes of sound output operation to acoustic hearing aids are available, and these are shown in Figure 11. Schematically shown. Each mode contains numerous variations.

In mode 1, filter parameters are set by the acoustician during the iterative installation process. and has since been fixed. In mode 2-4, the audio parameter extraction circuit Audio signals used to dynamically change filter parameters during hearing aid use Provides instant information about. In modes 2 and 3, the input signal is an audio wave It is a type. The output signal is filtered to select parts of the waveform (such as formants). ) and attenuating other parts (such as background noise) operated in such a way that In mode 4, the audio signal waveform is extracted by the audio parameter extractor. The output waveform is fully synthesized using audio parameters.

The difference between the original speech waveform and the hearing aid output increases as you go from mode 1 to mode 4. This also increases control over the frequency spectrum and output signal strength.

2.1 Mode 1-゛　・　Adjustment In this mode, the output signal is adjusted to account for the hearing loss of the disabled person.

This is done precisely by a 6-pole filter (usually specified by an acoustic engineer at all frequencies). (within 2 dB of ideal gain), automatic gain control eliminates residual hearing loss. Allows use of limited dynamic range.

The acoustic signal processing device of the second embodiment formed in mode 1 has a higher operational performance than conventional hearing aids. It also has practical advantages. These advantages set both types of hearing aids in motion It is best understood by considering the process that includes.

(a) Conventional hearing aids 2 Most commercially available hearing aids simply amplify and do not turn on at 1 o'clock. Compress the coming sound. To install hearing aids, audio technicians usually use an audiometer. Measure the user's threshold using a Manually set the appropriate target gain using the AL) rule (Byrne and Dillon, 1986). Calculate. The acoustician then examines the clinic's hearing aid inventory and determines the goals. Find the one with the gain closest to your gain. For all hearing aids, the amount of control Depending on the type of hearing instrument, some modifications can be made by the audio engineer. strange Modifiable features include overall gain combination, maximum output, and level at which compression begins. Can be done. If so, the frequency-specific variation in gain is usually at “high” and “low” frequencies. There are only two frequency bands, each corresponding to a wave. Although it looks bad, the postauricular hearing aid and body Internal hearing aids generally offer more scope for modification by the acoustician than in-the-ear hearing aids. . This is in addition to controlling these types themselves. In-the-ear hearing aids are also Yamald must be made specifically for that hearing aid by the fitter. is an expensive and time-consuming task. This makes it difficult to test and compare in-ear hearing aids. It is difficult and expensive, and many clinics avoid its use. In-the-ear hearing aids are also usually Maximum output is limited and therefore may not be suitable for people with significant hearing loss .

When forming a hearing aid, it is tested on the handicapped person's side. If this is not accepted, the acoustic The technician may choose another type of hearing aid.

must be formed to find a hearing aid that is considered acceptable by the disabled repeated until.

(b) Speech processing hearing aid of the second embodiment (mode 1): The acoustic engineer should check the hearing threshold of the disabled person. value and other hearing levels required for the strategy to be tested1 e.g. Maximum Adequate Level (MCL) ) to measure. Measurements are made using a combined shaping software rather than a separate audio meter. This is done using a diagnostic and programming unit equipped with software. These values are automatically accumulated in the data file.

- A strategy is selected and the hearing aid is configured accordingly in a maximum of approximately 5 minutes. goal game In calculation and installation are done automatically.

The graphic form is quickly accessed by the audio engineer at any time. shaped hearing aid The instrument is then presented to the subject for examination. Applicable to different settings quickly and consecutively will be tried until one is found.

This has the following advantages:

1) Use the actual device, earmold and transducer to measure the threshold of people with disabilities. By using this method, it is possible to accurately measure the target gain required for the device. Traditional supplement In hearing instruments, these measurements are usually made using headphones, and earmold sounds are Acoustic effects are evaluated separately. For in-the-ear hearing aids, the mold and hearing aid are manufactured together. Therefore, it is not possible to measure earmold acoustics prior to setup.

2) Changes are programmed into software rather than hardware adjustments. , different settings processing (e.g. NAL information, burnout) without the need for hearing aid changes. and Dillon, 1986) can be quickly implanted for testing.

3) Since the gain settings are widely programmed with frequency response.

It can be more accurate than is possible with many commercially available hearing aids. These values are then stored in the data file. automatically stored in the file. - A strategy is selected and the hearing aids are adjusted accordingly up to approx. Formed in minutes. Calculation and setting of target gain is done automatically and can be You can quickly access the graphic format at any time by Due to the wide programming of the zero frequency response given to the patient for hearing aid testing, Different settings are tried in rapid succession designs of various current hearing aids.

4) Acoustic designers may find that the target gain based on the threshold measurement of the disabled person is not optimal for the disabled person. You can change the target gain function at will. Many traditional hearing aids , this changes the gain entirely to a "high" or "low" frequency band or They had to choose a different hearing aid with specifications that closely matched their needs.

5) Setting information is available to the acoustician at any stage and is not available for any commercially available supplements. Gives more “online” control than hearing devices.

6) Calculation of target gain is done automatically in most setting processes (insertion gain (This is done by hand), the new equipment saves time. and eliminate potential sources of error.

In short, the device of the second embodiment is as accurate as many conventional hearing aids, but more accurate. can be formed into Instrument setup and testing is faster, more efficient, and less prone to errors. There are few things to do.

Figure 12 is achieved using the acoustic hearing aid processing device 12 of the Paimomol speech processing device. This example includes a setting example for mode 1.

2.2 Mode 2 - Amplitude Mapping This is used to map the level of any specific frequency. The gears of the three filters are adjusted according to the amplitude and wavenumber variations in which the force is measured by the processing device. Non-linear mating on a frequency specific basis by dynamically changing the in-parameters This is the same as mode 1, except that the This means that sound engineers are at the threshold of disability. In addition to the value, measure the optimal level at which the maximum amplitude can be mapped. Advantages of this mode is to enable more accurate dynamic range mapping.

Therefore, when used properly, the relative loudness of the spectral components is preserved. It will be done. This is useful for users whose dynamic range changes allocation as a function of frequency. This loudness control method, which is better than code 1, uses peak limiting and nonlinear compression. There are many undesirable spectral distortions associated with the more commonly used larger sizes such as Avoid.

2.3 Mode Spec Le Enhancement In this mode, there are three The frequency parameters of the filter are dynamically changed (in modes 1 and 2 they are constant), the valence is determined in a way that depends on the audio parameters measured by the processing device. When changed, salient vocal features are strengthened. This mode allows a wide range of audio One example of generating a processing strategy.

Using the center frequencies of the two bandpass filters, Fl and F2 (the first and and second formant) peaks. This is a form of noise cancellation. both as a filter and as a removal of signal parts that mask information in the tracked peaks. act. The filtered signal is used on a frequency specific basis similar to mode 2. Amplify the sound to a level appropriate for the user.

This is most useful for users with impaired frequency analysis skills and those with high thresholds. It is. The equipment used in this mode modulates the signal amplitude at the fundamental frequency (FO). It could be another way to enhance this parameter. to this Nearby commercially available devices include “noise canceling” chips that include “Zeta Noise Blocker” chips. ``Hearing aids.'' These devices calculate an average long-term spectrum that represents the background noise. First, this noise is filtered out from the audio signal of the same frequency. This mode type 3 type is an effect of the speech signal at the measured formant frequency rather than noise cancellation. It is based on enhancement. This means that audio information whose frequency is close to noise is , further noise from formants is reduced but not lost. means. This large size is selected not only in noisy conditions but also in quiet conditions. Enhances voice characteristics.

FIG. 13 shows that the selected peak sensitization is performed by the acoustic hearing aid signal processing device 12. This is an example of mode 3.

2.4 Mode 4-'' This mode is similar to the other operating modes of the second embodiment and the user can use a modified version of the input signal. configuration using the parameters extracted by the audio processing device without receiving the The difference is that the fully synthesized signal is received. The signal depends on the user's hearing loss. and can be reconfigured in many different ways. The entire signal expressed by this reconstruction is provided with tight control, thus allowing for highly accurate mapping of the user's residual hearing. enable. This is due to the fact that hearing is very limited.1 Normal amplification does not allow for clear speech recognition. It is most advantageous for users who cannot obtain

A second use of this mode is for frequency conversion. Frequency that cannot be heard by the user The sounds normally generated in a computer are represented by a composite signal within the audible range of the user. this Large-scale models have been attempted in the past, but as in this case, completely resynthesized waveforms were used. However, the resynthetic type works for electrical stimulation of pot beef shell implant users. This is useful for hearing aid users who have lost most of their hearing.

Each mode of operation allows for a wide range of strategies. Each mode is not individual; Several strategies are given for combining elements from 1, e.g. The reconstructed signal (mode 4) can be added to the filtered audio signal (mode 3) .

3.0 Bimodal Referring to FIG. 2, the bimodal hearing aid communicates with an audio processing device 11 and further includes a diagnostic program. The diagnostic system communicates with the audio hearing aid processing device 12 via the logging interface 22. 1 diagnostic program programmed using diagnostic programming unit 21 The programming unit 21 is provided as a personal computer program. stomach The interface 22 is a communication card connected to the PC bus.

The software is designed for acoustic engineers to use for the frequency response adjustment operating modes described above. written to find suitable filter settings that yield specified frequency/gain characteristics. ing. Software to program other operating modes is also programmed.

Has been tested.

Referring to Sections 14.15 and 16, the basic usage process of the pimodal device is as follows: It is as follows.

When using bimodal, the setup process is outlined in flowchart form in Figure 14. has been done. Bimodal maps can be used to repeatedly test the subjective performance of bimodal hearing aids The map is created on a personal computer after several trial setups.

FIG. 15 shows the flowchart for obtaining the optimal settings of the acoustic hearing aid 14, especially in mode 1. - This is a schematic representation of the chart process.

Figure 16 shows the control program in the diagnostic program unit and the It outlines the basic interactions between the configuration and mapping process performed .

The foregoing describes only some embodiments of the present invention and is intended to be used by those skilled in the art. Obvious modifications may be made by those skilled in the art without departing from the scope and spirit of the invention.

Special Table 16-506322 (12) J 1 /+AI /mouth Il ] Around the time Handling Shifu Shin's thickness has grown by 2! number of ils F/G, 5 Special table 16-506322 (13) Ten fu F8:/fLM(HX) Bimodal processing system! Schematic diagram of the configuration process F/G, I4 cmgv＼Uman cover requested.

Copy and translation of correction band) Submission form (Article 184-8 of the Patent Act) April 30, 1993

Claims (26)

[Claims] 1. hearing, including processing means for receiving and processing audio information received from a microphone; A bimodal hearing aid for anomia, wherein the processing means is derived from the audio information. An implant hearing aid that implants processing information into the first ear of a person with a disability, and the second ear of the person with a disability. Provide binaural information for the above-mentioned disabled person by supplying acoustic hearing aids that are worn within or close to each other. A bimodal hearing aid. 2. The processing means includes an implant hearing aid signal processing device and an acoustic hearing aid signal processing device. and the implant hearing aid signal processing device operates on the audio information to transmit the audio signal to the implant. the acoustic hearing aid signal processing device electrically excites the implant hearing aid; and processing the audio information and processing information received from the user to activate the acoustic hearing aid. Hearing aids listed in scope 1. 3. When energized by the implant hearing processing device, it directly connects to the cochlea of the disabled person. 3. A bimodal hearing aid according to claim 2, which provides electrical excitation. 4. The implant hearing aid processing device converts the audio information into audio features F0, F1, F2. , A0, A1, A2, A3, A4, A5 defined in the specification. Bimodal hearing aid according to claim 2, processing according to a peaking strategy. 5. The acoustic hearing aid signal processing device uses information whose parameters are stored in the hearing aid. an electronically configurable sound including filter means that can be electronically varied according to the 3. A bimodal hearing aid according to claim 2, comprising a sound processing device. 6. Said filter means communicates said parameters with information stored in said hearing aid. 6. Hearing aid according to claim 5, comprising three filter banks that can be varied according to the information. 7. The implant hearing aid audio processing device includes a configuration means and a signal. processing means, said processing device receiving audio information and transmitting said audio information to said computer. The signal processing means according to the parameters set by the configuration means. processing by a stage to generate an output signal for exciting an acoustic hearing aid, and means for inputting one or more electronic signals to modify said parameters. 2. A bar according to claim 2 adapted to receive power or software input. Imodal hearing aid. 8. The sound/audio processing device has audio features FO, F1, F2, AO, A1, A. 2, A3, A4, A5 (as defined in the specification) and voiced/unvoiced sound determination. claim to generate an output signal that excites a hearing aid transducer. The processing device according to item 7. 9. The signal processing means is configured to process audio features A1, A2, A3, A4, and A5 parameters. Dynamically vary the amplitude of different frequency bands of a defined audio spectrum as a function of the loudness of these bands is equal to the threshold level and the frequencies of F1 and F2. and the hearing aid user's optimum level in the higher frequency bands. 8. A hearing aid according to claim 8. 10. The signal processing means has its settings determined by the audio parameters extracted from the signal processing means. filter means dynamically varying according to the output data; The signal is adapted to overcome the effects of noise and/or certain loss of hearing ability of the user. 10. The hearing aid according to claim 9, wherein the hearing aid is adapted to be able to 11. Signal processing means in the processing device reconstructs the audio signal in real time and The amplitude and/or frequency characteristics of the recorded output signal may enhance the user's speech recognition. 11. The hearing aid according to claim 10, wherein the hearing aid can be appropriately controlled. 12. In a first mode of operation of the electronically configurable sound/audio processing device, The filter means described above are based on measurements made by an acoustician during the hearing aid installation process. Claim range set by a configuration means and then kept constant Hearing aid according to paragraph 11. 13. The processing device includes filter means whose parameters change dynamically. In use, the processing device operates by means of audio parameter extraction means acting on the audio information. said hearing aid device according to information provided to said configuration means. 12. A hearing aid according to claim 11, wherein the output signal is provided in a listener. 14. The output signal is synthesized by the signal processing means using only audio parameters. A hearing aid according to any of the preceding claims. 15. A hearing aid according to any one of claims 1 to 14. How to control hearing and cochlear implants. 16. Hearing aid transducers worn next to or in the ears of people with disabilities electrically connected to a voltage signal transducer located within the disabled person's ear. a sound/audio processing device connected to the audio processing device, the audio processing device receiving audio input information; receiving and processing acoustic signals from the hearing aid transducer to the electrical transducer; The transducer generates an electrical signal to provide cochlear binaural information to the disabled person. A bimodal hearing aid for people with hearing loss. 17. An electrically configurable sound/speech processing device for persons with hearing loss, comprising: The processing device includes configuration means and signal processing means, and the processing device includes an audio processing device. receiving the voice information and implementing the parameters set by said configuration means; the audio information is processed by the signal processing means in accordance with the hearing aid transducer; said configuration. The means may include one or more electronic signal inputs or software for the purpose of modifying said parameters. A processing device adapted to receive software. 18. The sound/audio processing device has audio features FO, F1, F2, A0, A1, A2, A3, A4, A5 (as defined in the specification) and voiced/unvoiced sounds utilizing the determination to generate said output signal to excite a hearing aid transducer; 18. The processing device according to claim 17. 19. The signal processing means processes the audio features A1, A2, A3, A4, and A5 parameters. dynamically change the amplitude of different frequency bands of the audio spectrum defined as a function of the including means for adjusting the loudness of these bands to the F1 and F2 frequencies and higher frequencies. Claims that vary appropriately between the threshold level and the hearing aid user's optimal level in the frequency band 19. An electronically configurable sound/audio processing device according to claim 18. 20. the signal processing means whose settings are based on the audio parameters extracted from the signal processing means; filter means dynamically changing according to the output data; The signal is dynamically modified to overcome the effects of noise and/or specific deficiencies in the user's hearing. Adaptable electronically configurable sound/speech processing device according to claim 19 Place. 21. The signal processing means in the processing device reconstructs the audio signal in real time. the amplitude and/or frequency characteristics of the output signal are suitably controlled and used. 21. An electronically configurable sound as claimed in claim 20 for enhancing user speech recognition. voice processing device. 22. in a first mode of operation of the electronically configurable sound/audio processing device; , said filter means is configured based on measurements made by an acoustician during hearing aid installation. The second claim set by the configuration means and then kept constant. The processing device according to item 1. 23. The processing device includes filter means whose parameters change dynamically. In use, the processing device converts the output signal into an audio parameter acting on the audio information. according to the information given to said configuration means by the data extraction means. Electronically configurable according to claim 21 for providing the hearing aid transducer sound/audio processing equipment. 24. The output signal is synthesized by the signal processing means that uses only audio parameters. Electronically formed according to any one of claims 17 to 23 Possible sound/audio processing equipment. 25. Sound/sound according to any one of claims 17 to 24 Method of controlling both a hearing aid and a cochlear implant by means including a voice processing device . 26. Sound/sound according to any one of claims 17 to 24 Bimodal hearing aids that include a voice processing device.