Audiophile hearing aids

There is something that I am not seeing addressed in any of the posts so far. I am an engineer and have done some testing on my own. I noticed that all audiograms test for threshold where we can no longer hear a tone. However, I am finding that at normal listening volumes my hearing profile is much different than the threshold (audiogram) shows. My hearing does not drop off nearly as badly as the audiogram shows and the difference between my ears is much smaller. Instead of 20 db drop offs at higher frequencies, it is more like a 3 or 4 db. So for audiophiles who are listening to music at normal volumes it would be important to have a hearing test for frequency response to see what adjustments are actually needed and not only for hearing threshold that is of little importance to music listening.

How did you test this?

Have you explored whether your thresholds are different if the audiologist uses a more distinctive tone?

Loudness growth measures and MCL tests have been around for years - they can measure how you perceive sound growth with your particular level of recruitment. Somebody even used to make a ‘smiley face tool’ to allow the client to respond with how they felt about different loudness presentations.

However, they’re massively subjective, don’t deal with competing sounds in real environments (Lombard Effect) and don’t categorise the levels relative to all the sound energy across the sound spectrum. Meaning that the results will vary by the context, mood and sound stimulus used.

Like you noted, your preference for music probably means you’d express way more loudness growth ‘in music’ but if i played you speech babble with a noise sample over it, you’d hit the decline button much faster. Which loudness result is appropriate for you, at home, in your car, in a bar with friends and music?

I have provided a brief description of how I did my own testing in a response to Neville – just a short time ago. I can provide a more detailed description to those who are interested. Just let me know.

I had two audiograms done and the results were somewhat comparable. The first was at Cosco and it showed some hard-to-believe variations from frequency to frequency although the overall results were roughly similar. The second one was at a business that specializes in hearing aids and made more sense. They used the three repeating beeps that were easier to hear. I am not questioning the validity of the threshold test results. What did not make sense to me was how to correlate those to my real-world experience at more normal volumes. At normal and low listening sound levels I am getting quite a different result in how my ears respond – basically flat out to 5 kHz and then only a 10 dB drop at 8 kHz. And yes I can hear 10 kHz if I turn up the volume even more. That is way different than the 30 dB drop off from the threshold test. As I mentioned in my response to Neville, that may be something that could be important to those of us trying to get the most realistic sound when listening to music.

I am curious about the testing you did in more detail. I found you posted something general but more details would be helpful if anyone wants to try to replicate it. Possibly it would be better to post more details in the thread about this?

Comparing threshold to low level hearing test

As an engineer I am able to do my own frequency scan with high quality headphones. What I observe is that at low volume levels (a bit above threshold) my hearing response is much smoother and does not drop off as much. Between 250 and 2,000 it is flat and also my ears do not differ by 10 db. Has anyone done any work to measure hearing response at other that threshold levels?

Have you been able to use this information to improve your hearing or aid adjustments for day to day living?

Here are some details of my test setup. I used a BK Precision 4017 signal generator that produces a constant output at all frequencies. It was set for a 100 mV peak-to-peak output. This waveform was sent to a 10 Kohm 10 turn potentiometer as a voltage divider to provide an accurate adjustable voltage signal to the earphones. The amplitude of the signal at the earphone connection was then measured on a Tektronix 7904 oscilloscope with a plug in that has 5 mV/div sensitivity. I should mention that any signal generator or scope that has this capability will work just fine. The attenuated signal from the potentiometer was then fed to my Sony MDR-V6 over-the-ear headphones. I did also measure the sound level at the earphones with a dBA sound level meter and for most of my testing used low volume tones of about 35 dBA. I did also do frequency scans at higher volumes to see if the results were any different. They were not. However, the results are very different with a much better (smoother) frequency response than my audiogram that only measures hearing threshold. Let me know if I have missed something.

How do you account for the acoustic coupling from the headphone to your ear-drum?

Audiometers are calibrated to ensure the correct SPL at the drum, do you have a 2cc 0.707 or Zwislocki coupler on a calibrated Bruel and Kjaer reference mic to a calibrated sound acquisition system?

“…approximately 7cc” I had no idea that my ear canal has such a large volume? Did you by any chance mean 0.7 cc? Sounds like the precise positioning of the hearing aid earpiece is very important to how well it will work.

No, the encapsulated volume under a headphone can is deemed to be approximately 5cc though larger Western male ears can be bigger. This includes the volume of the concha as well as the canal.

You’d normally have about 0.7cc residual volume behind an IIC aid.

What an interesting observation, Electroniker. For starters, I think all tone threshold audiograms should be conducted with the beep-beep option rather than steady - because that can submerge into tinnitus. And, yes, like you, I have noticed although not yet measured, the non-linear behaviour within real music: it’s almost as though a ‘gate’ is active within certain frequency bands, but not equally across each of them. Turning up the volume generally obscures it, but it would indeed not be revealed by standard tone tests.
Then there’s the matter of ‘recruitment’ - i.e. how much spl we can tolerate.
Perhaps that one could better be assessed by using a steady-level multi-tone track, say 15 to 20 simultaneous frequencies, more representative of music?
But, hey ho - everyone’s hearing is different!

Indeed there must be an opportunity for a manufacturer to devote the kind of musical insight and care that goes into the best mikes and loudspeakers - along with detailed performance specifications and (of course) an option to self-adjust if we feel that brave.
Probably, because of a smaller market, they’d arrive with a ‘boutique’ price tag. …And while on that topic, I have to say that the most musical BTE aid I have heard in 30 years is a ‘Cyber Sonic’.
Sweet, warm, no artefacts, not deafeningly loud, though not enough high end for my ski-slope ears. Possibly it is analogue. Price on eBay: five English pounds, delivered! There are several such to choose from, most quite clunky to look at, but for enjoying music at home who cares?
I think one of the deafness charities should gather up some such examples of truly affordable instruments and publish listening results. It’s unfortunate that hearing aids are a ‘distress purchase’ - which opens the door to inflated prices.

Oh, that ‘dismal whirl’. Well said, classicalmusic. Decades went past until I happened upon an audiologist, bless her kindly young soul, who went the extra mile to make the best of some very basic UK national health BTEs. Gave me the lows, down to 40 if you please, and sharply low-passed the highs to match my hearing cutoff and minimise a pointless challenge to processing. She set volume compression across the range, rather than band-by-band.
Speech is less helped, of course, but my current priority is enjoying (on hi fi speakers) a wonderful gift of 150 classical CDs from a neighbour!
Prior to my Phonak M70s I was tried with a pair of Signias. Oh, the overshoots (that’s mis-design) and general clog-and-crud whenever all the violins played, or the whole choir sang. And, the receivers had a sharp mechanical resonance at different points in the octave, which caused soloists to jump left and right, phrase by phrase.
Just to be fair, maybe more recent Signias, or the more costly ones, are better.

Our wonderful hearing aids are “medical equipment”

I wish that was a reason to sell hearing aids that allow us to play and enjy music, rather that “Just Good Enough” to understand words…

I understand your point and I think overall we are in pretty good agreement. However, at normal and soft sound levels my ears are almost the same with at most a few percent less sensitivity in my right ear. It is not even close to perceiving tones as “half as loud”. And frequency response is also significantly better than the audiogram would indicate. So maybe we have to work with the “gooey level” as you suggest.