Compression problems with Phonak Audéo M90 hearing aids

I am a senior with severe hearing loss/nerve/cochlea & tinnitus. I recently got new hearing aids, 2 phonak audio m90s from VA. they have been fitted several times with REM. I still have real problems trying to hear speech with them. They are worse than previous HAs I have worn.
I have a technical question. After several unsuccessful attempts to fit HAs , I am wondering if the problem is compression. audiologist says I have 29% hearing capability left, & it is difficult to fit the aids.
My thinking is that the compression/attack/release capability of there M90 phonaks might not be good enough in my case, & speech distortions result.

  1. does compression with a 29% capacity to start with & tinnitus produce distorted speech?
  2. is it possible that varying attack/release times at different frequencies might better solve distortion perception?
    Does phonak have ability to vary compression speeds for different frequencies?
    Are there other HA brands that handle this compression speed better?

going back jan 25 to audiologist for another shot at it. dont know what best course of action.

I think um bongo answered some of your questions on another threa. If by 29% “hearing capability”, you mean the word recognition score in your better ear, yeah, you’re going to have problems. If you’re talking something like an AI score, one might expect better results. Posting your audiogram and word recognition scores would give people a bit more info.


this subject is related to a question I just posted.
my question deals with the attack/release times that the compression executes.
I am getting very poor speech recognition with a new pair of phonak audio m90s.
I am wondering if its possible to manipulate the compression speeds to eliminate the speech distortions. Can manipulating the speeds improve speech perception? (would it matter)
can the speeds vary depending on frequency location, or do HAs just do fast or slow? (either/or)
My bottom line is , why are my new phonak worse than old ones?
My 2nd bottom line question is , how come I can hear talk radio speech with a sony headset/am/fm radio ok, but not HAs???

Setting faster attack and longer release times may help with better preservation of the speech signal. Some people find the Sonova defaults too aggressive, especially in respect of the interconsonant noise management.

Also having a mess around with the threshold knee points may help, people with more severe losses tend to do better with more linear settings than their prescription would otherwise suggest. Bunching the response to ‘fit’ apparent recruitment loses some of the natural loudness growth which people find helpful in discerning parts of speech. (Unless you’re a fan of Harvey Dillon who reckons all speech should be boosted to a similar level.) Oticon does this differently by preserving short term ‘Windows’ of linear gain within the overall gain curve.

So you can have a play with both the TKs and the noise management within the Sonova product, accepting slightly less overall ‘sharpness/processing’ might be more comfortable in the longer term.


after trying a bunch of settings, last time around, the audi disconnected compression & set everything back to level/analog type amplitude settings.
it was 4.30 pm/ time to go, (DESPERATION???)so, I have been exposed to level settings, & they are bad. /so speech perception with or w/o compression setup does not work for me.
*so how am I able to listen to am/fm talk radio with sony headset ok, but hearing aids bad.?? inquiring minds would like to know?

You might want to post your audiogram.

1 Like

I go to audiologist jan 25, will get copy & post my audiogram.

question. if problems in speech comprehention with compression due to sqeezing sound wave size, would it make sense to get a HA like phonak paradise 90 with the bigger cpu & memory? would bigger cpu /memory handle compression wave sizing better than phonak audio M90 that I have currently?

No, that’s not how it works.

1 Like

As he says. It wouldn’t help.

dont know too much about hearing aid engineering, but with a computer operating system, a cpu chip that can handle a larger word ie 64bit v 32bit , & more memory size so there is less , I/O activity can operate faster. wouldnt that result in less need to shrink wave or distortions?
so, does any of this effect speech quality?

Not really. The bit rate you need isn’t that large. The system demand comes from clever signal processing. It’s not particularly about speech quality per se.

definition of clever signal processing? elegance of program? or, fitting the instructions to the hearing loss condition?

anyone out there who can define what what clever signal processing means?

having terrible time with my phonak m90s, about to see audiologist nxt week. so would appreciate direction . AUDIOLOGIST HAS TRIED SEVERAL TIMES & HAS NOT GOT IT RIGHT YET? speech lousy. compression/attack/release ???
hardware, programming, my lousy hearing?
will post audiogram when I get it.

Eg The way the processor takes digital signal from the several microphones and uses these to cancel noise and steer focus, the handling of loud noises and sudden transients, the frequency shaping and compression, wind noise reduction and so on. Nothing to do with fitting software or elegance of the programming.

Try asking the hearing care professional to change the proprietary algorithm to DSL 5 or NAL-NL2

so is it all software? are you saying that the signal shaping & delivery output (fitting the listener with more/better perceptible speech) is the finessing of the algorithms?
Is the audiologist limited to what the the program is only set up to do?
do some HA brands algorithms offer more flexibility in solving speech perception output than others?
appreciate your insight. thanks.

The audiologist’s main job is to set the HAs up to compensate for your specific loss. This is much more complex that just a matter of getting the frequency response right. Ears are not linear in the first place and ears with hearing loss have the added problem that many sounds will be too loud or quiet. So setting the aids up in a way that works for the individual’s loss and in the sort of environment they live their lives in is complex.

This is not the same at all as the examples of clever digital processing I gave earlier. That’s all down to the different HA software designs. It’s highly competitive and different manufacturers have their own research programs to inform the way their HAs work.

Another challenge for the designers is that all this clever processing and amplification in the HA, along with streaming and wireless control has to be made to work on a tiny battery. If you take a 312 or 13 zinc/air cell, you are talking about just over 1 volt and a maximum current drain of a mA or two. That’s not a lot to play with!