Phonak Marvel: DSL 5 Adult to Adaptive Phonak Digital

I work in ENT at a major academic medical center. I willingly let the audiologists use me as their guinea pig when I bring up an idea. It’s great!

Over the past week or so we wanted to try out DSL 5 Adult vs Adaptive Phonak Digital Contrast vs Adaptive Phonak Digital. All with REM. We called Phonak: Adaptive Phonak Digital is based closely on NAL-NL2, and it’s not inappropriate to use the NAL-NL2 for REM purposes.

Right now the clear winner is Adaptive Phonak Digital with REM tuned to NAL-NL2.

Soft voices are clearer and louder out of the box whereas DSL needed G50 to be cranked . SoundRecover2 “works better” — sh, ch sounds are clearer, more distinct and less harsh compared to DSL. (Same SoundRecover2 settings).

Adaptive Phonak Digital Contrast wasn’t for me. The words sometimes get muddy - not clear. Which makes sense given slower compression and slower attack times. I noticed it and I don’t notice it in Adaptive Phonak Digital.

I don’t plan to try out NAL-NL2 alone. With REM on Adaptive Phonak Digital, I trust the engineering and fine tuning that Phonak has done - a best of both worlds so to speak.

Additionally, I think a LOT less fine tuning will be needed with Adaptive Phonak Digital as compared to DSL. I’m shocked how good it sounds with just REM and nothing more.

What’s your experience? I’ll post more as I experience more.

4 Likes

I just switched about a month ago from Rexton SmartFit adjusted using NAL-NL2 in the REM step, to DSL v5 which can be adjusted directly in REM. I would agree that there may be a slight loss in the soft speech sounds due to the significantly less compression with DSL. However, overall I find DSL to be more natural and less harsh. I need to go back for a bit of a tune up in my left ear, as I am still getting a touch of feedback at times. But, overall I like it. What I didn’t like about the SmartFit/NL2 is that in a restaurant I would hear what people were saying 3 or 4 tables away. I call it the bionic ear effect. I don’t need to hear all that other stuff. I just want to hear what people are saying at my table.

A few weeks ago I tried a similar test due to all the talk on the forum about fitting algorithms.

I too found Phonak Digital Adaptive to be best but it is the algorithm I fine tuned over months of adjustments for best speech recognition. My hearing is pretty bad so I am only interested in speech recognition.

Adaptive Phonak Digital Contrast tended to slur words or be muddy as you mentioned.

DSL made things louder but less clear in the middle and upper frequencies.

My take on this experiment was any of the algorithms could probably be tuned to have similar speech recognition. To just switch from algorithm to algorithm from a properly tuned algorithm will not give best results.

1 Like

That is very interesting. My aids came set up as NAL-NL2. When I was doing my first programming session, I tried the different fitting formulas and thought DSL-5 adult was best. I am going to try the Phonak Digital for a while to see if it is better. I seem to remember trying it with my Venture aids and not liking it, but it might be just the thing with Marvels, which I am finding are a much different thing.

I’m intrigued by the substantial differences among fitting formulas. Here are four gain profiles from four formulas for my Marvel M90s, all from the same audiogram – the one shown by clicking on my “B” avatar:




NAL-NL2 generates high compression ratios:
.

Aside from the differences in the prescriptions, I don’t know why the fitting formula would make a difference in the feedback threshold – the purple line. In this regard, compare the first two – APD and DSL.

I am surprised you are getting such low feedback potential. Everything I run in Target shows all kinds of feedback limitation at high frequencies. What fittings are you using? Is there some kind of feedback reduction setting that has to be made that is not turned on by default?

Yes, the prescription formula used makes a big difference. I always wonder when I see comments that this hearing aid sounds better than that one, but no consideration seems to be given as to what formula each uses and whether or not it has been REM adjusted. And in addition to the differences in the three gain curves an compression there are differences in the kneepoints, and the attack and release times of the compression.

I have, as of yesterday, custom molds in each ear with, currently, 2 mm. vents. As you know, to reduce feedback, close 'em up!

Hi, I currently wear Resound and want to change to Phonak. Your technical discussions have lost me. Can you point me to the web pages that would educate me so I can better understand your technical discussions and lead me to an educated decision so I can work with my Audi. Thanks much, Larry

Audiology Online is a great source of information. It only costs if you want continuing education credit.
Most of us learn about this stuff for fun. You don’t need to be an expert to work with your audiologist. Just describe what problems you’re having and they should be able to work with you.

2 Likes

This is what I get using your audiogram, DSL v5 adult, and a totally occluded cShell mold. It hits the feedback limit at about 3 kHz and up. Phonak Target Version 6.1.7.47719.

Your feedback is the anticipated feedback, not the real feedback. Difference is the dotted line versus the normal line.

The point is that Brec is getting a different anticipated feedback point.

No, his is the true, measured feedback curve rather than the anticipated one which is usually more conservative.

3 Likes

I was assuming based on the posted graphs showing “not connected” as the status, that they were a simulation.

The feedback then would of been ran on a previous go and now when it says ‘not connected’, it would of just been looking at the software.

1 Like

Interesting take on Adaptive Phonak Digital vs NAL-NL2. I believe another provider never uses Adaptive Phonak and uses NAL-NL2 instead. I could be misremembering.
Also curious on your usage of SR2. (What kind of settings?) You’re loss isn’t the typical one I think of when thinking of frequency lowering, but I have read it’s useful even in flat losses as it’s harder to get gain and audibility to the higher frequencies.

At the time I took the screenshot, the HAs were not connected. But a “Feedback and real ear test” had been done in a previous session and results are saved with the fitting. As @Zebras noted, a solid line feedback threshold is the result of the test, and a dotted line is an estimate based on the specified instrument coupling.

Still, it seems puzzling that the estimate would be lower given that you specified “Occluded” for molds venting.

Perhaps you meant about 6 kHz and up?

Yes, I meant to say 5 kHz and up. It seems strange that the Target software seems to be so pessimistic when it comes to estimating feedback. The Connexx software on the other hand, seems to be overly optimistic (in my opinion, based on my own experience).

1 Like

Yours is an experiment of one sample. In this paper the researchers tested 60 different hearing aid wearers, comparing NAL-NL2 with DSL v5.0a and got mixed results. Apparently they did not try Adaptive Phonak Digital.

I’ve learned a lot from this post as I had never heard of any of these terms. I have no idea what my audiologist used for my 30-day trial last month.

That is a helpful paper, but is tough to follow exactly what they were doing and concluding. My superficial analysis is that the NAL-NL2 and DSL v5 formulas are different. They have some pros and cons, but the value is in the ear of the beholder. They seem to behave a bit differently depending on the type of loss you have, but in general the DSL has significantly less compression in the soft sound range, and somewhat less in the loud.

What I find interesting is that hearing aid users don’t blink about trying 2, 3, or more different brands of hearing aids, but do not even consider simply switching the prescription formula. In each brand there are typically three basic choices; NL2, DSL, and a manufacturer proprietary version. Seems to me that it may be more expedient to try the different formulas before trying different brands, especially if you already own the aid.