This page has been here for some 7 years now. Over that time I have received a number of letters from people looking for hearing aids... some with comments and information that it would have been worth keeping and passing on. Sadly I did not do that.
However (!) today I got an e-mail from Richard X which I decided to make available to those interested.
Richard was an accoustical engineer and knows a great deal about the subject; so here is his contribution: Richard's letter.
I am not an audiologist. I have no training in this area whatever. But I do have a hearing problem; I have seen three audiologists and four otologists and I've had 'inexpensive' ($500 each) analog hearing aids for nine years. Now I have the very latest - Signia digitals from Siemens.
When I wanted to make an intelligent choice of new hearing aids (they can get very costly) - and also because I enjoy learning and understanding - I attempted to elicit answers to some specific questions about how hearing aids work and what their limitations are.
I have not received satisfactory answers from the people I have asked. Either these professionals don't know, or they don't understand enough science or they don't want to bother. I suspect another explanation may be that they feel that I, being a lay person wouldn't understand.
Here is a paraphrased sample conversation I had when looking for new aids:
Q: "These hearing aids I have seem to amplify loud sounds as much as lower volume ones. This means that when I am in a noisy environment or am subjected to loud sounds like a vacuum cleaner, I have to turn them off because they are uncomfortably loud. What is needed in the circuitry is what is referred to in the electronics world as AGC or 'automatic gain control'. Even a $10 transistor radio has such a circuit. Why not my hearing aids?
A few years ago... A: "It's difficult to build such advanced circuitry into small hearing aids. The technology isn't there yet."
A few years ago... (my response was... )"I don't see that as being advanced circuitry. AGC has been around since at least the 1920's. It's really very simple."
More recently.... A: "What you are talking about is called compression. It's available in the new computerized digital hearing aids that cost $2,000 and up each."
In any event; this article represents my understanding of how hearing aids ought to work and what their limitations are. I write it because there are probably others out there in cyberland who might find the information of value. I say 'information'... there may be some errors here. If you know that something I say is not factual, I'd appreciate and e-mail. Thank you.
Your Audiogram
An audiogram is a graphical representation of how sensitive your hearing is to different frequencies.
The illustration shows an audiogram of idealized normal hearing. The vertical axis is decibels below normal sensitivity. It may be interpreted like this:
The ideal ear cannot hear sounds below 16 Hz (cycles per second) nor above 16,000 Hz. These numbers are approximate. In between these extremes, sensitivity ought to be more or less linear. The range between 300 and 8,000 is especially important in understanding speech.
People with hearing difficulties show a profile different fom this ideal. This is a representation of a typical profile. Many hard of hearing people typically show lowered sensitivity at higher frequencies
The audiogram provided by your audiologist will probably not cover the entire 16 Hz to 16,000 Hz range; but rather, a more limited part of the spectrum. The reason for this is that speech recognition, which is the most important part of hearing improvement is not dependent on frequencies much above 8,000 Hz - and also no hearing aid is able to deal with sounds much below 400 Hz or so. This has to do with physical size limitations of the transducers... the microphone and the 'speaker'.
Here is my audiogram:
Amplification Curve
The perfect hearing aid should therefore presumably have an frequency/amplification curve symmetric to the sensitivity curve, as suggested here:
The graph shows little amplification at low frequencies and compensation for low sensitivity at higher frequencies by greater amplification. The result is normal hearing ability across the frequency range.
Note that this is an idealized situation. In the real world the curve of decreased sensitivity may not be so smooth and, particularly with 'analog' hearing aids, it may be difficult to to attain a perfect frequency-amplification curve. The following illustration shows such a hypothetical compromise:
Analog - read older and cheaper type - amplifiers tend to have a linear amplification-frequency curve only slightly modifiable with with 'tone controls' - and ordinarily only at the upper and lower ends of the spectrum. In this illustration, in order for the resultant to be within the range of normal sensitivity in the central range, one has to accept undesirably high amplification (loudness) at the upper end of the spectrum.
Sensitivity at Different Volumes
In the case of the ideal normal ear one would expect that the relationship between how loud a sound is and how loud it sounds is linear. Low volume sounds sound very soft. Very loud sounds are heard as being very loud. This is illustrated in this graph:
As any hearing impaired person knows, this is not true for them. There is a lower threshold below which nothing is heard at all. At the threshold sounds are audible. As sounds become louder, they soon become annoying and then painful. Hard of hearing people have zero sensitivity to very soft sounds and are disproportionately sensitive to loud sounds. Their comfortability-loudness range is much more limited. This is illustrated in the next graph:
The conclusion is that not only should some frequencies be amplified more than others, but different loudnesses of incoming sound should be amplified by different amounts. As a rule of thumb, soft sounds below the threshold should be amplified a lot, average loudness sounds moderately, and very loud sounds not at all. This is what automatic gain control (AGC) does. They call it compression - but it's only good if it is not linear - which compression circuits may not be unless they are specifically designed to be non-linear; customized to the individual user.
What is missing in this graph is the fact that the relative sensitivity/loudness curve is frequency dependent. In other words, a person may be extremely sensitive to high frequency loud sounds, while not at all sensitive to low frequency loud sounds. This is further complicated by possible sensitivity to a loud mix of frequencies as opposed to one loud frequency, such as one is exposed to in a room with background music and many people moving about and talking.
Hearing aids of the analog variety do not have this kind of signal processing capability. But digital hearings aids which contain 'microcomputers on a chip' can have very sophisticated signal processing powers:
It doesn't follow that all (or perhaps any) digital hearing aids do these things. It is noxiously difficult to nail down a particular aid's repertoire for comparison with others. In practice, one leaves the selection of brand A vs Brand B up to the dispenser.
Audiologists seem not to have the technical literature available. Brochures for the various brands may be available (not that I was given any despite asking)... but these seem to be written for the average consumer, and as far as I am concerned, are less than helpful.
Some further comments on hearing testing and aid selection/fitting.
I've had my hearing tested and an audiogram made on seven or eight different occasions. I was reasonably impressed on two of these... when they were conducted by an audiologist. One of the occasions that I was not impressed with was also conducted by an audiologist. One of the other sessions was conducted by an audiologist. The others by - you will pardon the politically incorrect language - a secretarial type woman with apparently six hours' training and no obvious scientific education.
The 'amateurs' all had this in common: the sequence of frequencies and loudnesses presented to me were all sequential - and entirely predictable. In other words I could have faked my responses and given a very different impression of what my hearing was really like. Consider the implications of this for a person applying for compensation for hearing loss suffered on the job who has my insight!
As we know there are three different styles of hearing aids. The miniature invisible in the ear canal type, the in the ear type and the behind the ear type. In physical size and price, they range from the very small and expensive to the much larger and a little less expensive.
It only stands to reason that the larger the space available to put the electronics, the easier and cheaper it is to pack in extra (and hopefully more sophisticated circuitry, accessible switches and other controls. So, with an in the ear canal type you would expect to pay more money, have less sophisticated circuitry with less options and have to change the battery more often - than with a behind the ear type.
Here is another thing. The behind-the-ear types do not have to have the circuitry mounted inside custom molded ear pieces... they are all the same shape and size and can be mass produced in their final form. Major cost saving - which is why I don't understand why the upper end ones are still hugely expensive - and not all that much less than the in the ear and in the ear canal types.
Here is something else I don't understand: Consider the behind the ear type hearing aid. They have a thin plastic tube that connects to a molded ear piece. This tube and earpiece has got to be or at least could be standard for all the different brands, and hence all the different signal processing programs, options and prices available. So, once I've had the ear piece made, why not let me try several different behind the ear types at the same time for comparison purposes? Switch back and forth by just exchanging the earpiece. This might lead to price competition - which does not see to be the case now.
Which reminds me. Suppose I make a deal with a car salesman to buy a brand new zipmobile for $20,000. Now I say to him "Ok. If I take two zipmobiles, how much for the pair?" Are we going to settle for $40,000. I don't think so. Yet with hearing aids you pay the same amount for the second one as you do for the first one. My experience is also that two hearing aids are definitely not twice as good as one.
Speaking of prices. Would you rather pay the dispensing audilogist a fee, either standardized or time based; or would you prefer they take a percentage of the price you pay for the hearing aid they sell? What is the general practise? Think about it.
The sound produced by any hearing aid can be simulated by a suitably constructed simulator and a pair of earphones. Want to know what the $700 (each) Audicon will sound like at a cocktail party? Want to know what the $3000 (each) advanced Digisound will sound like when the CD player is on? No problem. Adjust the settings on the universal simulator and there it is. You can hear precisely what the difference is with the flick of a switch. The computational power required is much less than that found in any home computer. In fact, I'll bet that circuitry required to ugrade any home computer to such a universal simulator is available off the shelf for a hundred bucks. The software might be a different matter, but sure as h... not particularly difficult to write. I have since found out that this software is available.
When I decided to got for new digitals I wrote and asked the manufacturers of all the brands under consideration for technical literature - not the same as promotional literature - on their units. The standard response was "This information is not available to consumers." No kidding. Would you buy a car from a salesman who won't tell you the displacement of the engine or whether it has overhead valves or if the steering is of the 'rack and pinion' type? I wouldn't.
Recently I have looked at hearing aid manufacturer's web sites. Only one of these gives you detailed technical information on their products. Siemens. Try starting at http://www.siemens-hearing.com.
As it turns out my Signias are adjusted by connecting them to a computer which has a set of displays each of which look just like an 'equilizer', in this case with eight different controls, one for each frequency range.
Display 1 adjusts the amplification of each frequency range to match your audiogram.
Display 2 adjusts the lower threshold amplification value for each frequency range.
Display 3 adjusts the compression applied at louder volumes.
It's a little more complicated than that; but these are the basics.
Now, I have a complaint about how the fitting sessions (I've had four so far and still isn't right) were conducted:
1. The fitting sessions were conducted in a very small room with distracting noises. A room with lousy accoustics.
2. All the adjustments were made by the audiologist on the basis of answers to questions about how I heard her talking when we were sitting right next to eachother - closer that would be the case under normal conversation.
I would prefer to have the software and interface installed on my computer at home and make the adjustments myself under 'natural' - and varied - conditions. I just isn't that difficult.
I would also think that a better way would be to expose me to a continuously varying frequency and let me adjust things so that I heard low frequency sounds at the same loudness level as high ones - and make seperate adjustments at each frequency for 'over loudness' and also for threshold setting with due regard to feedback.
I'm sure *I* could do a more satisfactory job of the adjustments in this setting than in the audiologist's cubicle.
Incidentally... the *wholesale* prices advertised for various brands of hearing aids on the internet in the United States are higher than the *retail* prices offered in audiologists offices in Canada. How do you explain that?