Recruitment

Thanks for that Rick.
I do think recruitment is part of my problem, but probably not all. I liked your explanation. Very few people I have mentioned it to to had any idea what it was. The way I explained it was like a radio tuned to one frequency, but with the volume turned all the way up trying to hear a signal that was not on its frequency. When even a small signal on its frequency is detected, because the volume was at maximum, that signal is overpowering. Not nearly as complete, but simple enough that most people get the idea.

@Sierra said:

The starting point of the problem is that an audiogram is based on threshold hearing, but we live in an environment way louder that our threshold level of hearing. The whole range of gain corrections is based on one single point of threshold hearing.

Yes; and this is “wrong”; but historically the Threshold is the only measurement. (My audi picked an MOL from thin air.)

We “should” ask the client “Now turn this dial to a medium and comfortable volume” (verify it seems reasonable) “and while turning to other pitches you turn dial to make it seem a similar medium comfortable level”.

If I do that to me, below 1kHz my loudness curve is “flat” (‘normal hearing’) but above 3kHz about 42dB is scrunched into 12dB. But there is no standard test to quantify this.

I believe the big HA labs get statistical trends from mass observations, and their first-fit math includes compression based on HL levels and slopes.

@Sierra said:

recruitment concept, which is believe is more technically described as a non linear response to increasing gain

The name is bad but let that go. Actually a ‘normal’ ear is NON-linear. You can’t get 120dB through a simple linear device. The inner hairs and the auditory nerves pass a bit more than 60dB. The outer hairs are hyper-sensitive to sound, shiver when they hear their pitch, which “amplifies” the sound to the inner hairs. (If you know “regenerative radio”, same thing.) Together a 120dB range of sound fits in the 60dB range of the auditory nerves.

So there is a natural 2:1 compression from sound to brain.

UNTIL! Those outer hairs, which must be very highly tuned and high-strung, finally give-out. Now we only hear the upper 60dB of ‘normal’ sound. This seems exactly true for me with 55-60dB threshold 1.5kHz-up, and at least that much “recruitment” up there (but there is also something else, probably weak inner hairs).

I can run an equal-loudness curve on myself in a minute, but I kinda despair of trying to do this on people without years of audio experience.

I wouldn’t put much stock in this description of recruitment. This guy’s PhDs are in ancient astronomy (classics) and theology. As far as I can tell, no current hearing researcher thinks that the auditory periphery works this way–I don’t even think that the guys who originally coined this term thought it worked that way about two seconds after they coined the term, which is why it’s confusing because it’s not actually a good word for it. I hate the way audiologists throw around the word “recruitment” when confronted with a patient who is sensitive to loud noise as if it is at all meaningful. All it means is that people with (sensorinerual) hearing loss have steeper growth of loudness than people without. It is present for all people with sensorineural hearing loss, not just some.

Hyperacusis is a better model for abnormally low tolerance to loud noise.

[Also, it gets my back up when someone is like, “I have a PhD, let me explain to you how time REALLY works” and then their PhD is in French.]

@Neville said:

“I have a PhD, let me explain to you how time REALLY works” and then their PhD is in French.

It sounds better if you say it in French.

I thought Killion had a good clarification of “recruitment” but I can’t find it now.

As said, there are all sorts of New Meanings tacked onto the bad name which are not what was originally intended. Sensitivity to loud sounds is related but NOT the same. I recruit highs like McNamara yet very loud does not make me cringe. (I dislike extended loud sounds; they don’t hurt.)

https://www.etymotic.com/media/publications/erl-0133-1974.pdf is typical Villchur: thoughtful, complete, probably correct in its assertions, but not an easy read. He is making “fake recruitment” for “normal listeners” (backward from what we want). In 1974 this was a large rack of expensive equipment. The audio tracks are here: https://www.etymotic.com/media/publications/Villchur1974Soundsheet.mp31

https://www.etymotic.com/media/publications/erl-0025-1993.pdf has some nice curves. The WDRC compression he used to compensate is very micro-1990; digi-aids could fit better.

https://www.etymotic.com/media/publications/erl-0109-2008.pdf1 is another solid meal. And shows different people have different loudness curves (“recruitment”).

Further reading from Killion and peers: https://www.etymotic.com/publications2

The current Wikipedia article on recruitment is IMHO barely half right.