PhoneAbility

7. Panel Presentations

7a. Telephony and the hearing impaired user - Technical opportunities - John Barnes, Telecommunications Committee, Hearing Concern

Most people with normal hearing take for granted their ability to use the telephone system with very few problems, particularly when using the fixed or hard-wired network. The connection has to be very bad indeed before any significant difficulty is experienced. International standards recommend the end-to-end losses [1] from mouth to ear and most national networks set their networks up to meet these standards. These losses are specified in terms of Overall Loudness Rating (OLR) [2] which is a comparison measurement with a standardised end-to-end connection called the Intermediate Reference System (IRS) [3] having typical frequency characteristics. Thus the internationally agreed OLR for an international connection is 10 dB OLR. In other words 10dB quieter than the IRS end-to-end.

ERP = Mouth Reference Position. ERP = Ear Reference Position

Fig. 1 Speech path, mouth to ear losses, for a typical telephone connection having an OLR of 10 dB and a metre air path

A typical telephone connection
Figure 1 shows and end-to-end connection having an OLR of 10dB and the corresponding sound levels at different points in the acoustic path showing that the path loss for speech over the typical telephone connection is about 5.5 dB. A hypothetical 1 metre air path is also represented and loss over this path is about 30 dB, slightly less if one includes the obstacle effect of the head and room reverberation. Thus it is deduced that the typical telephone connection has a gain of about 25 dB over that of a typical face-to-face conversation. This makes it a useful "hearing aid" in its own right and explains why many hard of hearing folk who find it difficult to have a face-to-face conversation can cope quite well on the telephone without a hearing aid.

So the first characteristic of a telephone set that might be mentioned as being useful to hearing impaired users is that it should have transmission levels that enable the connection to meet international standards in respect of OLR. If additional amplification [4, 5] is provided in the receive path to the earphone this will give even better reception for hearing impaired users.

Coupling a hearing aid to a telephone set
For users who normally wear a hearing aid and also need to use it to converse over a telephone connection some form of coupling is necessary between the telephone earphone and the aid. The simplest form of coupling is acoustic whereby the user places the telephone earphone close to the microphone of the aid. However on the majority of handsets, which are designed to work into a sealed ear cavity, the inevitable acoustic leak can give rise to serious changes in the quality of the received speech depending on the acoustic characteristics of the receiver. Figure 2 shows the effect of a loosely coupled earphone compared with the sealed condition. This seriously impaired frequency characteristic is carried over to the microphone of the hearing aid. Figure 3 shows the result of acoustic coupling a BE51 BTE (Behind The Ear) hearing aid to a typical telephone set and underlines the advantage of the use of a low acoustic impedance earphone which does not suffer the loss of the low frequencies to the same extent. (These measurements were made using KEMAR a head and torso simulator, similar to that in [10] and were part of a study carried out by the Working Group on telephony for Hearing Impaired people [6])

Figure 4 gives an idea of the sort of characteristic that can be expected using inductive coupling [5, 7] in this case a BE31 BTE aid, little wonder that hearing aid users are great champions of inductive coupling! Calculations have shown that this connection has a loudness rating of -27dB OLR that is 59 dB louder than a 1 metre air path. With the additional receiving amplification the user should be capable of having a sensible telephone conversation over the longest/quietest connections.

Clearly there are tolerances on all these calculations and the result of any coupling configuration is going to be very dependent on the positioning of the handset with respect to the hearing aid microphone or its inductive pick-up coil. With practice most hearing aid users achieve good results. For In The Ear aids (ITE) inductive coupling is not usually so successful as the pick-up coil is usually smaller and placed for optimum use within an inductive room loop rather than a telephone inductive field. For acoustic coupling however better success is often reported, (assuming no acoustic feedback), particularly for those handsets that have earcap geometric properties that provide a good seal to the human ear. [8] A good handset seal, in addition to preserving the low frequency performance of the receiver, also helps to exclude local environmental noise.

Very little has been attempted in respect of electrical coupling although there are international standards [5, 9] covering the subject. It is acknowledged that the plug and socket arrangements on a hearing aid for such a purpose are likely to be very small and would probably prove to be too difficult for some users with limited finger dexterity. However the socket on the telephone could be useful in providing a connection to an individual neck loop for subsequent inductive coupling to a hearing aid. This is often done with mobile telephones to keep the handset a sufficient distance from the aid to reduce radio frequency interference.

Other factors that affect the telephone conversational quality.
So far the emphasis has been on the way the telephone handset presents the received signal. However the incoming speech signal itself may be degraded either by the acoustic environment at the sending end or by other factors along the transmission path to the receiving end. A case in point often cited of poor sending quality is the tendency of call centres to place operators rather too close together in a room that has a high level of reverberation resulting in the listener overhearing conversation of adjacent operators. Also the operators themselves may not be sufficiently disciplined in maintaining the microphone position in front of their mouths thus reducing both the outgoing speech level and the signal to room noise ratio. The use of noise canceling microphones can improve the latter but make the microphone positioning somewhat more critical. Possibly the answer lies in the call centres themselves in radical redesign of the physical layouts and room treatments. Sending levels on fixed line handsets can also be affected in the same way if their physical shape does not place the microphone in front of the mouth. Mobile handsets are notorious for this.

Interference has been mentioned in respect of mobile 'phones and is well documented. However the same type of interference can also occur from DECT phones being used in association with a hearing aid, the main difference being a change in the frequency of the interfering signals.

Summary of features that are advantageous to hard of hearing people when receiving calls.
A number of features have been identified, viz;

• Transmission levels meeting International standards
  
• Additional receiving amplification
  
• Inductive coupler
  
• Low acoustic impedance receiver/earphone
  
• Sensible earcap geometry to give a good seal at the ear and exclude room noise
  
• Electrical connection socket to internationally agreed standard

Additionally there are factors that are unhelpful and these are;

• Radio frequency interference from GSM and DECT handsets
  
• Poor sending speech quality from the far end due to microphone positioning
  
• Poor sending quality due to environmental noise

Conclusions
In this short paper some of the more important features of telephone design that hard of hearing people would find helpful have been identified. Additionally some of the factors of end-to-end telephony that are unhelpful have also had a mention. Telephones can be designed to include all of the helpful features and in many cases exclude the unhelpful ones. Hearing Concern would like to encourage the inclusion of the good features so that hard of hearing people are not excluded from making full use of the telecommunication system. Two papers covering in more detail some of the issues dealt with in this paper are referenced below [11, 12]

Bibliography - References and Standards

1. ITU-T Recommendation G.111 (1993) Loudness ratings (LRs) in an international connection. International Telecommunication Union, Geneva, Switzerland.
2. ITU-T Recommendation P.79 (1988) Calculation of Loudness Ratings. International Telecommunication Union, Geneva, Switzerland.
3. ITU-T Recommendation P.48 (1988) Specification for an Intermediate Reference System. International Telecommunication Union, Geneva, Switzerland.
4. ETSI ETS 300 488 (1995) Telephony for hearing impaired people; Characteristics of telephone sets that provide additional receiving amplification for the benefit of hearing impaired users, European Telecommunication Standards Institute, Sophia Antipolis, France.
5. ITU-T Recommendation P.370 (1996) Coupling hearing aids to telephone sets. Part I - Magnetic field strength around the earcap of telephone handsets which provide for coupling to hearing aids. Part II - Characteristics of telephone sets that provide additional amplification for the benefit of hearing impaired users. Part III - Electrical coupling of telephone sets to hearing aids International Telecommunication Union, Geneva, Switzerland.
6. WGHI Report on Phase 1 of the Oftel WGHI Research project, May 1987
7. ETSI ETS 300 381 (1994) Telephony for hearing impaired people; Inductive coupling of telephone earphones to hearing aids. European Telecommunication Standards Institute, Sophia Antipolis, France.
8. ITU-T Recommendation P.35 Handset telephones. International Telecommunication Union, Geneva, Switzerland.
9. ETSI ETS 300 679 (1996) Telephony for hearing impaired people; Electrical coupling of telephone sets to hearing aids, European Telecommunication Standards Institute, Sophia Antipolis, France.
10. ITU-T Recommendation P.58 (1993) Head and Torso Simulator for telephony (1993). International Telecommunication Union, Geneva, Switzerland.
11. Barnes G J, "What do we hear through the telephone?", paper given at British Society of Audiology meeting, Wired for Sound, April 1996
12. Barnes G J, Telephony "End to end levels and the hearing impaired user", Proc. International Human Factors Symposium, Melbourne, March 1995.

7b. What Next? - Dave Barratt - BT Age and Disability Action

We have been looking at the past, in what we mainly have talked about, amplification, inductive couplers, adjustable ringers and flashing lights. But what is coming next? What can we expect to come to help?

We have spoken about digital induction couplers. Although people have their doubts whether they will work, or have not yet seen them, so we do not know.

We have spoken about Bluetooth and then we have smarter amplification. That is something that in our research development side we have been looking at.

Fig. 1 Speech levels showing how hearing loss reduces the range available

Fig. 1 shows frequency along the bottom, with the sound level up the side. The bottom line is the normal hearing threshold below which we cannot hear anything, and above the 100 mark it is going to be painful to our ears and may well distort.

The telephone service fits quite nicely to the middle to upper range of that pattern. So, with a normal hearing person such as myself, I would hope I could hear with no problem. However, if we introduce a hearing loss, anything below that line we can no longer hear, so we are missing out parts of the audio signal.

Fig. 2 The effect of compression in making the full range of sounds available.

When you boost the signal up as in Fig. 2, if you look at the audio signal now, it has gone above the uncomfortable level. So not only have you now -- will you not be able to hear the lower tones now boosted up. The higher tones are causing problems, discomfort or maybe distortion.

We have been looking at this to see how can we get over the problem and make it so that the upper level does not cause a distortion or discomfort?

What we then do is to compress the signal. You can see from that compressed audio signal now that we have utilised what area we had left of useful hearing to actually put all of the hearing in there. So we have taken the distortion, the painful levels, down. In effect, we have amplified whatever level we required for our own individual purpose.

This is all well and good, but what happens to it? Do we put it in a telephone? Do we put it on a network? There are pros and cons for all of this. One of the main things is we are all individuals, and we have all got probably different hearing losses.

If you put it in a telephone, how would you be able to select the one that is best for you? How would you know whether your audiogram has got a high or low loss? It is a bit difficult. I am not sure how we are going to do that, but it is development.

What we have is a network-based solution. This is only a trial that we are looking at, but we are looking to dialling a number, a prefix, before you dial your main telephone number. The prefix will give you your own individual settings of amplification and compression.

The benefits are that maybe at home you have a telephone that works fine. You go to a hotel or somewhere like that, and they have not got the telephone that is suitable for you.

With this system, you would be able to dial the number that suits you. Your compression would go in and then you would be able to use their telephone with that adjustment of compression.

We are at a stage now where we want to try this. We want to see if there are any general settings. Maybe 5, 6, or 7 that fit quite a few people.

We have made the telephone so you can adjust that level. It may be you go to your phone, a big button one, or a new one. You go away, use the same setting. Because the characteristics are different, you may need to alter that. Using the buttons on the telephone, you will be able to alter the compression, the amplification and something called "slope", which is too technical for me but I am sure some of you guys will understand.

So you can adjust what is best for your own hearing needs. There is not a great deal of amplification in our network that we can put in without getting echo suppression and other problems. If we can get the signal so that the clarity is better, your hearing aid will be able to use that clarity to give you a better signal.

If you do not use a hearing aid, just getting the greater clarity will also give you a better signal, and enable communication.

That is what we are looking at within our amplification side. We are not looking at doing it linearly, which is to move the whole lot up. We are looking at specific parts of the frequency range that we can amplify. How we want to amplify, what level, and tailor it is for yourself to decide.

The only rider with this, I am afraid, is that this work is under development. There is no guarantee it is going to lead to anything. We have got a system running and we want to try it out, but we can say the learning we have taken from that we will use to provide better communication.

Back to Contents

Previous Page

Next Page