But there’s nearly no case in which I would choose to use a proprietary rationale. If a hearing aid manufacturer locks certain features into their proprietary rationale, which is rare now, I’d still just select that in the software but then fit the gain to an independent target. For example, to turn off fast compression for certain patients I may choose Phonak’s adaptive digital contrast rationale, but then I’m still going to adjust it to DSL or NL2.
1 Like
So to confirm my understanding here, even though @aliakhtar 's audi told him that he can only do REM against NL1 or NL2, @Neville is saying that if the user insists on fitting the REM against the mfg’s proprietary rationale because the proprietary rationale’s target curve is made available to the HCP, it’s entirely possible and he can do it, although he still prefers to do it against NL2 for some reason.
If some feature is locked and only available with the proprietary rationale (although very rare according to @Neville), he can still just choose that proprietary rationale to unlock that feature, but still ignore the proprietary target curve and adjust REM against the NL2 target curve.
Please correct me if my understanding is wrong in any of the above.
My question is why so resistant on fitting against the proprietary rationale’s target curve in the first place? After all, the mfgs know better than the audis, and if they go through great length to make their own proprietary rationale that they think is best fitted for their HAs, why would an audi resist doing REM against the proprietary target curve so much?
If doing REM against either takes the same effort because the target gain curves are readily available for both, then why such a big resistance from the audi? What if the patient insists that he/she likes the mfg rationale better then NL2 because the proprietary rationale sound more natural and less sharp?
1 Like
I’d like to know this as well, as I’ve always been led to believe that REM cannot be used with the manufacturer’s proprietary formula.
1 Like
What if the manufacturer formula doesn’t offer the same gain vs the same input depending on the conditions for a window of time before the sample?
Serious question?
Floating point linearity shifts the amount of gain applied depending on long and short term averaging of noise and speech.
How are you going to plot that?
Edit: technically any dynamic gain system could cause issues in this regard.
1 Like
@Um_bongo
Let me get this straight: proprietary rationale implies longer processing time? Is this generally the case or only with certain manufacturers?
To expand on @tenkan 's point, my understanding is that if a REM stand-alone system/equipment is used, then it’s natural to understand that this REM stand-alone equipment can only have access to the standard fitting rationales, hence without access to the proprietary fitting rationale of a HA mfg, it cannot perform REM against a proprietary target curve because it wouldn’t know how to generate one.
But I notice that the Oticon Genie 2 software as something called REM Autofit (see screenshot below). I assume that as long as the HCP has REM equipment that is compatible with the Oticon REM Autofit, then REM Autofit can deliver the proprietary rationale’s target gain curve to the REM equipment so that adjustment can be done against the proprietary target curve. This would solve the issue of a stand-alone REM system that can only be fitted against a open standard fitting rationale.
This is just my assumption because it seems to make sense, but I don’t know enough about REM equipment to know if this assumption is true or not.
No, why?
Proprietary rationales may not necessarily give the same amount of gain for two identical signals if they are preceded by different average amounts of speech, noise or speech in noise.
REM measures the application of gain over an input signal - the target at a given frequency is ‘fixed’ by a fitting formula like NAL 2. What if Otiwhizzo 5 doesn’t yield a repeatable result as it’s generating a variable amount of linear gain, depending a window of previous measured sound.
How do you include that level of complexity in your REM screen?
Why not either measure against a good target like DSL/NAL2 or just retain a proportional speech window and make sure the average output gets to that?
Ok, I get it. It’s the unforseeable variation of gain generated by proprietary rationales.
Thanks for your explanation!
1 Like
Sorry, let me be more clear. When he asked, my brain provided me with a few options for hacking it, all semi-arduous, and some of them possibly specifically dependent on my having the particular brand of real-ear equipment that I have. I did not mean to imply that it was straight-forward. Sure, I’d give it a try if a patient asked for it. I wouldn’t choose to do it. But you guys on this forum are a bunch of weirdoes. I love you, but not once has one of my patients EVER asked for a specific rationale.
What makes you think that this is the case? What do you think the manufacturer’s purpose/goal is behind making their own rationales? (Relatedly, who do you think your clinician is talking to when they “call the manufacturer”?)
4 Likes
Then why would the HA mfgs bother to come up with their own proprietary fitting rationales then? Just for fun? There has to be a good reason why they would bother to come up with their own fitting rationale that makes sense.
I remember watching an education video on Audiology Online by Donald Schum who was then a VP at Oticon. He was talking about the challenges of fitting unconventional hearing loss type like the cookie bite type, and how many HA mfgs would use the conventional approach and compensate for the loss in the area where the hearing loss is present, which is the mid range frequencies. But the point he was making in that video was that through the research they did at Oticon, they find that doing that doesn’t seem to help and actually makes it worse. He said that their research showed that focusing on increasing in the mid-highs compensation gives a better result in terms of better speech clarity, despite the fact that it would compensate where the audiogram says compensation (or what little of it is needed) is not needed there to get a flat response. The point of the video was that they took what they learn (about how to treat the cookie bite loss to yield better speech understanding results) and applied it to their proprietary VAC+ fitting rationale.
So to answer your question, my understanding is that these are the types of things that an HA mfg like Oticon may want to promote their own fitting rationales over the standard rationales, because it would incorporate these things that they have learned through their research into it. And that’s why their fitting rationales are proprietary, to protect these knowledges that they learned through their research that are incorporated into the rationales. I’ve heard of posters on this forum (most notably @Abarsanti) who has a cookie bite loss and prefers to use the Oticon VAC+ because he thinks it works best for him compared to other standard fitting rationales.
Also, because they design their HAs, they know the strengths and weaknesses of what their HAs can do, so a rationale that is customized to the design of their HAs would be better than a standard rationale that would not have the proprietary data to know to take into account the HA brand specific types of things like performance and capabilities.
So let me respectfully ask the same question back at you, why do you as an HCP think HA mfgs come up with their own proprietary rationale for then? It’s not meant as a challenge, but more as a curiosity to understand the viewpoint of the HCPs on this.
Sorry, but I honestly don’t get the point of this question. OK, so I think my clinician who sold me Oticon HAs is talking to Oticon when they “call the manufacturer”. It’s really over my head the implication of this question.
Haha, I think it’s more like maybe most of your patients probably don’t even know what a fitting rationale is, much less that there are standard ones and proprietary ones. And I’d also venture to say that some probably don’t even know what REM is, although maybe more may know about REM by now especially if they watch YouTube videos from Dr. Cliff pounding on relentlessly about the value of REM.
Nevertheless, people like us on this forum hear and learn what these 2 things are (the rationales and REM), so it prompted it to spur a dangerous combination of questions on how these 2 things work together (or apparently not). Not really weirdos, but more like “inquiry minds want to know”, haha.
2 Likes
So based on this, the explanation given is that standard fitting rationales like NL2 are static and predictable, but a proprietary fitting rationale from an HA mfg is most likely (or can be) dynamic and therefore may not give predictable and repeatable results then?
Is there a reason why the standard ones are static but the proprietary ones can (and likely) be dynamic? What are the pluses and minuses between these 2 types of rationales?
Obviously for REM test, a fixed/static rationale that is predictable and repeatable is better because you don’t want the target gain to change all over the map then. Is that why @Neville in an earlier post was saying that mfgs typically recommend verifying against NL2 for their proprietary algorithms? Because doing REM against their (most likely) dynamic rationales is really not optimal?
Hmmm… so different manufacturers have different “rationales” for fitting that don’t equally apply with REM testing? So, attempting to fit an aid with REM won’t work across various manufacturers?
so, the whole thing is more complex than we here at this forum might understand, despite our listening in to on line tutorials about REM?
I’m not a licensed audiologist. Some here are. Even licensed audis disagree about the benefits of REM.
No, that’s not what I said; I said you couldn’t measure to a defined target if the target itself is dynamic.
You can ‘still’ measure to a defined target like NAL2 if you want verification of results, the problem is that due to the conditional programming of the hearing aid, the result in the field is potentially going to differ.
This means that the people using REM as the holy grail for fitting and making a huge song and dance about it online - ‘Don’t trust your Audiologist if they don’t use REM’ etc. aren’t necessarily correct. They might have a method that tells you the arrows are landing nearer the target, but it’s not perfect; especially with more modern dynamic aids.
How many times on here does someone write: ‘my aids sounded great in the booth/soundproof room/shop, but as soon as I went to X they performed differently’ ?
I like REM, I like the fact that it removes some of the subjectivity from ‘does that sound clearer?’, but use it as an absolute measure and you’ll come unstuck - pretty much as they did in the early REM use in the NHS where they mandated it and ended up overprescribing a whole load of hearing aids. OTOH, if you ‘can’ accurately reproduce the complex sound-field of a restaurant with real conversations mixed into the clatter in your fitting centre, then go for it. Preferably with ‘Live Speech Mapping’ to see the output rather than a strict NAL2 line.
Until then - fitting to a basic estimate (with a splash of REM) and developing a client focussed prescription from there, isn’t a bad answer.
5 Likes
Yes, exactly that, per my last answer. You end up with a series of potentially overprescribed aids which people return or don’t use; especially in noisy environments……
The Modernising NHS Programme went through quite a series of this as there were many early mandates that Hearing Aids ‘must’ be fitted to full DSL/NAL2 targets. Some manufacturers added ‘test/REM modes’ to meet the requirements of the fitting/REM, but in my mind that’s just cheating as the actual real-world performance of the Aid will differ (significantly possibly) under real world conditions.
And if you think about a system like Phonak’s Autosense - how do you determine which mode the hearing aid is in when you perform the test.
1 Like
In my 20 years of hearing aids not once has fitting rationales ever been discussed. I suspect they only use the software default.
Where I am pediatric standards are that we verify all subprograms individually and all ANR and dmic settings, if activated.
And this is what the proprietary rationale is built for. Minimizing returns across the large number of clinics that do not follow preferred practice.
2 Likes
Isn’t the “official” line from Phonak (and I think most of the big HA companies) that the fitting should be verified with REM set to NAL-NL2 values? It’s like they give two messages. If you want to follow best practices, you should do REM. If you just want to sell hearing aids, just use First Fit.
Nevelle told me many years ago that most self programmed hearing aids are under fit, not enough gain or volume.
My thoughts on this are the Insitu fittings are under fit because of the manufacturers proprietary algorithms being a little low concerning gain.
So fitting with RIM will consistently need a little more gain to satisfy the RIM targets. This can be uncomfortable for many until acclimated.
Please correct me if I am misunderstanding this.
2 Likes
That agrees with my worst ever button pusher over Thanksgiving with Resound One 9 trial. The speech performance was no where near what I needed.