This best combination resulted in an agreement between measurement and prediction of less than ±3 dB and ±20° up to 3 kHz and less than ±5 dB and ±30° up to 6–8 kHz, performing significantly better than both average transfer models and existing individualized predictions.
Seems rather nice. Although the same problem of drift from calibration that happens with the amplifiers, could also happen with the microphones. So you would still need a REM to assure that they are still functioning according to specs.
It seems like a useful development in testing technology. However, I think the current Achilles heel of REM testing and adjusting is the audiogram testing and the prescription target set based on the audiogram. While we have ever increasing numbers of channels and adjustment handles with each new generation of HA’s, we see to be stuck on about 8 frequency points for measuring hearing loss. Putting 48 channels of discrimination on 8 data points sounds like trying to make a silk purse out of a sow’s ear.
The second step of the heel is the setting of a a target prescription. The common industry practice seems to be to use industry generic formulas like the NAL-NL2. I am not sure when it was developed and revised, but it seems to be a very blunt tool in comparison to the 48 channels of resolution. My observation is that the NAL-NL2 target is heavily smoothed out. With current technology HA’s it is possible to make a much closer fitting of the gain to the loss than is accomplished with the NAL-NL2 formula. This is kind of the reverse situation where the process makes a sow’s ear out of a silk purse.
As an example here are the simulations of what the Rexton Smart Fit targets were for my HA, compared to the NAL-NL2 simulation that was actually used for the REM and fitting. They are significantly different. I appreciate it would take more time for testing, but I suspect that the overall fitting process could be significantly improved by using more test points and a more individual and instrument specific fitting target.
I’m new to all of this, but I expected HA fitting technology to be further ahead than it really is.
SmartFit Power Target
NAL-NL2 Target - Notice how much more smoothed out it is.
The assumption you’re making is that testing hearing across a huge number of pitches and providing audibility of sound at each individual pitch will yield greater success. Even if you could deliver the perfectly shaped sound to the inner ear, you would still have the problem of the spread of masking in the cochlea… where amplification at one pitch masks sound perceived at adjacent higher pitches…
From what I can tell, looking at the number of channels a hearing aid is all but useless. Phonak’s top of the line has 20, Resound’s 17, whereas Signia has 48 and Oticon 64. Other than marketing, I never see it mentioned. Having compared stereo and car specs for years, it’s hard to let go of, but I think hearing aids, if they are to be compared at all can only be done so qualitatively and not quantitatively.
During a hearing test, if you have a large difference between, say 1k and 2k, they will test 1500. But, if there is only a 10 db difference in 1k and 2k, there is no need to get any finer.
I have a question regarding thresholds between the commonly tested frequencies.
Let’s say one has a threshold of 50dB at 1KHz and 55dB at 2KHz, has there ever been a spike in sensitivity at say 1500Hz e.g. 30dB or is the threshold at 1500 always somewhere between two thresholds immediately adjacent?
Just a thought. It was more from a point of view that is it possible to have regions where there is a normal threshold despite most being a moderate hearing loss.
I seem to notice that some voices carry much better than others despite my audiogram being relatively flat at the tested thresholds.
I guess it is possible, but I don’t know how they would catch that.
I have dead spots over 4k but under 5k and those don’t get caught on a hearing test.
Wow, that’s like horseshoes and grenades “close”.
Here we go again. To me REM stands for Rapid Eye Movement which is a term used to describe deep sleep. Perhaps someone could inform uninformed people like me what REM stands for here (especially if you insist on using acronyms frequently).
Considering the forum, it is Real Ear measurement.
A system is loaded with your target from your hearing test. Microphones are placed in your ears along with hearing aids adjusted to your target.
The system plays some sounds and compares what the microphones actually hear with the expected target and recommends programming changes to the aids to get closer to hitting the target curves.