This is the final post on transducers, during which we will talk about Uncomfortable Things.

Many of you reading this will have normal hearing. For those of you who have to put up with the full volume of fire alarms and ambulance sirens (sorry), the things that determine your limits of hearing are straightforward. On the low end (for soft sounds), you're bounded by the sensitivity of your ears to signals after they come bouncing through your ear canal. On the high end, you're bounded by pain tolerance -- things get so loud they hurt. (See? Uncomfortable Things.)

Obviously, folks like me who don't have normal hearing have different limits; my cochlea's resemblance to a razed minefield has something to do with me being unable to hear high frequencies at jet-engine volumes. (But sometimes not being able to hear a sound very well doesn't keep you from feeling pain when it's super-loud! If you cut us, do we not bleed?)

There's actually another bound in real life: background noise. The world isn't conveniently quiet, so even folks with normal hearing can't hear very soft sounds (that they'd be able to hear in a sound booth) once they step outside. That's why hearing aids don't amplify to the point of "normal hearing in a soundbooth" -- it's fine to have a nonzero microphone noise floor because they're aiming for "normal hearing in the real world" instead.

Anyhow. No matter how we amplify it and what we're aiming for, I think we can all agree that Things That Are Too Loud Are Painful. One technique some digital hearing aids use to deal with this is OLC (Output Limiting Compression), a signal-processing technique that makes sure the MPO (maximum power output) to a user's ear isn't painful. This isn't just about absolute volume; if a sound is really, really loud in one particular frequency, it will reduce the volume of that sound in other frequencies so that the totality of the sound isn't too loud.

Reserve gain is the difference between what the receiver can physically do (how loud it can pump out sound) and what the software limits (like OLC) let it do. You want a couple dB (6-10, usually) of reserve gain because the software isn't magically instantaneous -- if there's a loud sound, there will be a tiny moment where it's booming at full loudness before the OLC catches it, and we don't want that tiny moment to completely suck. You might decrease reserve gain because you're trying to squeeze more performance out of a hearing aid that's too old or not quite powerful enough -- maybe you really need a new hearing aid one but can't buy it yet, so you let loose the software limits to get that little bit of extra volume. But watch out, because you're giving yourself less buffer room that way -- it's easier for the transducers to slam into their saturation limits that way, and (as previously discussed) when you push mics and/or receivers beyond their physical limits, everything sounds terrible.

The cheap-ass solution to that is scotch tape. I was totally chuffed to find out about this today, and plan on trying it at my next concert. Putting 2-3 layers of scotch tape over the microphone is supposed to decrease sensitivity by 10-12 dB. Yes, this takes a little longer than turning down the volume (gain) of your hearing aids, but turning down the volume doesn't actually help if you hit saturation limits, because you're just getting a softer version of the crappy "my microphone is hitting its limitations!" noise. What you need is a way to not hit the limitations of your transducers in the first place -- a way to make the sound hitting your microphone (and thus exiting your receiver) less intense. Therefore: scotch tape.

We don't want to overdo pumping loud sounds into our ears, so the temptation is to make the MPO as low as possible -- but that's actually problematic in and of itself, because a low MPO means a low saturation limit for a receiver, which means it'll hit its limitations (and sound awful) more easily. A too-low MPO also makes things less intelligible, because you're smooshing the signal smaller than you need -- think about listening to music through a tinny cell phone speaker instead of through a properly large-sized sound system. (Note that the "M" in MPO stands for "maximum" -- it's the highest peak in the frequency response chart for that hearing aid. Other frequencies will almost certainly have a lower output level than that maximum, so don't confuse MPO for loudness-potential in general.)

 The definition of a "low" MPO is subjective; if the user turns the gain all the way up and it's still not making loud sounds loud enough, it's low. Of course, trying to raise the MPO comes with its own tradeoffs: you can raise the MPO by making the receiver larger (harder for the diaphragm to hit stuff if it has more space), but then it's larger and might not fit in your ear. Also, it sucks more power. Also, it's easier to break now that it's larger. Tradeoffs.

One way to make hearing aids seem "louder" is to put them in deeper. This makes sense; when we walk towards speakers to hear them more clearly, we are actually trying to decrease the distance between the sound source and our eardrum, so why not deal with the low-MPO issue by just going further into the ear canal?

Well, first of all, ouch. This doesn't infinitely scale; there's only so far the thing can go. Also, it doesn't really fix the problem. It will make the hearing aid sound "louder," but it won't make it more intelligible, in the same way that turning up the gain won't help unless you actually raise the MPO of the device. (Really, you can think of sticking the hearing aid in deeper as one particular, non-electronic way of increasing the gain of the system as a whole.) While turning up the gain might make softer sounds louder (and more audible), the loudest sounds won't get any louder like they need to be because you're running into the physical limitations of the receiver. So the loud sounds are still just as (inadequately) loud, except that now more loud sounds (including noise) are getting crushed up or at least towards the maximum-loudness ceiling.

Another way of understanding this via Not Completely Accurate But Good Enough analogy: if you and I are talking in a busy cafeteria and you say "Mel, I'm having trouble understanding you, you're not loud enough," and I responded "not loud enough? Ok, sure -- HEY EVERYONE MAKING BACKGROUND NOISE, TALK LOUDER!" -- but didn't actually raise my own voice -- that would be... less than optimally helpful. In fact, you'd be rather unhappy with me, because noise is also something that we typically don't like with hearing aids.

In talking about noise, we should make a particular distinction here: we're talking about microphone noise right now. This is actual noise that the microphone picks up; it is not "circuit noise" invented somewhere inside the circuit but not actually present at the microphone. Most microphone noise in hearing aids comes from the acoustic flow -- basically, air. Air moving through the sound coupling tube, air moving to the diaphragm. A little of it comes from the air coming through the little vent hole in the diaphragm. Very, very little of it comes from electrical noise somewhere in the FET preamp circuit -- so little we can't even hear that anyway.

We'll stop here and restart next week with more about digital hearing aids.