Insitu and standard audiogram in Genie2

Hearing Aid Self-Fitting and Adjusting [DIY]
1

Hi, i’m wearing oticon more1 - and am quite challenging to get the most out of any technical device. Investigating in some “unexplicable phenomena” i came to a startling finding:
It seems that
Genie2 does not take into account the impedance of the ear canal!

Let me start from the beginning:
My first fitting, based on a headphone-audiogram by my provider, is still the best i can get. When i came home with my new more1s, my wife watched TV (turned very low) and i sat on a sofa about 3m away and could understand her (even lower speaking), understand her TV, hear our mechanical clock and “knew” that our neigbours were talking in their garden.
This became my standard test setup for any new setting.
Next session my provider tried to optimize the setting with InSitu - and the next day i phoned him and begged to remotely switch back to the original, for hearing was so much worse.

As soon as i had payed my mores, i started to experiment myself.
No matter how i tried, any fitting based on InSitu was much worse than that based on the original audiogram.

For decades i made my living with finding the flaws of other developers and tricking around their bungle that their equipment would work. So it was natural for me to get curious:

A standard audiogram documents the hearing loss from the start of the ear canal to the brain. Good for doctors and health insurances but not optimal to fit hearing aids that whisper directly on the eardrum.
An InSitu-audiogram documents the hearing loss from the eardrum to the brain without any influence of the ear canal. Depending on the individual ear canal this is quite different to the standard audiogram.
If the InSitu was done with (nearly) closed earmoulds, the difference between the two audigrams should be the impedance of the ear canal.

As i wear PowerBass-Domes that are by design quite open, naturally the low frequencies should be much “worse” than the higher frequencies and those be “better” than standard. At the middle frequencies the difference was up to 15db.
I found, that the more i shifted the InSitu to the classic, the more the fitting was like the classic fitting.
At last i faked an InSitu-curve, that was identical to the classic audiogram. After recomputing the fitting - the insertion gain was also identical. This could not be true, for it would mean, that i had to enter InSitu-points that were more than 15db above the perceivable level - really loud!
So i compared the two fittings - normal audiogram and identical InSitu-audiogram db by db as insertion gain and as input/output-curve at all available frequencies.
They were identical!

What this means:
Genie2 does not take into account the ear-canal impedance and just feeds the audiograms - regardless that they have a totally different origin and meaning - into the fitting formulas whithout any correction.
And this means, that a fitting based on the standard audiogram will always be “sharper”, therefore better understandable but also more sensitive to unwanted noise. While the InSitu will always be “softer” and therefore have worse speech-performance.
Of course both paths can be tweaked with the fine-tuning assistant, but does it make sense to compute an only roughly estimated fitting and then manually fine-tune until it works?

Hey you out there! Providers and Audiologists! Did anyone from oticon tell you this?
And how do you cope with that behaviour of the software?

(I use a mean between what i hear in InSitu and the original - it works quite good)

2

I don’t know anything about Oticon, but in general my approach as a user is to start with the fit proposed by the fitting software based on the audiogram and use in-situ measurements to help me fine tune things.

3

I can’t make any sense of what your saying, have you heard of REM,thats how the audiologists get around some of these issues.

1 Like
4

The part you said in bold above is not quite true. Your assumption is that the receiver is all the way inside the ear canal, almost touching the ear drum. This is not necessarily a correct assumption. The receiver probably just sits a little bit right into the entrance of the ear canal. I contend that it doesn’t go all the way deep inside the ear canal to the point of almost touching your ear drum. That would not be a safe thing to do in the first place.

Also, the In-Situ audiometry is not done with closed ear molds. It’s done with whatever fitting the user uses, whether it be an open dome or a close dome (with vent or not), or a power dome, or a custom mold.

Then you forgot to account for the differences between the (supposedly) calibrated (to some kind of standard) headphones speakers used in the silent booth where the HCP does your audiogram test, compared to the performance of the actual hearing aid, which is not necessarily calibrated to the same way as the headphone speakers in the silent booth. The hearing aid may under-amplify or over-amplify, or be right on, so it may not produce the same loudness level compared to the calibrated standard that the silent booth headphones are supposedly calibrated to.

In sort, in-situ audiometry is simply a way to generate an audiogram that takes into account the variations caused by your hearing aid’s actual performance (whether it be under or over performance or right on the dot), in addition to also take into account how the fitting your HCP chooses for you to wear may alter the deliver of what the hearing aid sends out. By this virtue, the audiogram that you get from an in-situ audiometry session is more real (and therefore accurate) compared to the audiogram your get from the HCP in the silent booth. That’s because the in-situ audiogram is base-lined on the actual amplifying device that you’ll be wearing in real life, instead of the HCP’s audiogram which can only be base-lined on a calibrated standard. Calibrated standard may sound official and better, but real-life in-situ is actually more accurate because it takes into account real-life factors that will affect what you hear as delivered by the device that you actually wear.

For example, if you go by the HCP’s audiogram, but your hearing aid under-amplifies for some reason, and the fitting impedes the delivery some more (due to open domes or not snug enough fitting or some kind of leak or whatever), then you won’t be happy because the delivery based on the HCP’s audiogram will underperform. But the in-situ will automatically factor in this underperformance because you’ll have to turn up the tone volume in question up a little louder to be able to hear the same threshold that you heard in the silent booth. So you should be more happy with the delivery based on the in-situ audiogram because it knows where your actual tones’ threshold levels is more accurately based on your own feedback using your own hearing devices that you wear everyday and not some silent booth’s device that you only wear once.

REM kind of does all this “factoring in” as well, but it does it better than in-situ because in-situ is only for the basic threshold tone test for the audiogram, while REM measures the performance of the whole system via actual real life (well simulated) sound tests over the airwaves all the way from an external speaker through your ear canal to where the mic is placed inside to ensure that whatever is delivered up to that point where the mic is placed factors in all the deficiencies (or surplus) of the whole system and adjusted accordingly to meet whatever the target is.

1 Like
5

By the way, I doubt if any registered HCP on this forum frequents the DIY section here anyway. I may be wrong, it’s just that I never saw personally any forum-registered HCP make any comment in this DIY section.

6

Thank you. This is great explanation.