The Infinio AI spheric chip claims to greatly improve speech understanding in noisy environments. And by all accounts and reports on these forums, it seems to be helping many people.
For some reason, I did not have that experience. In fact, I question if the AI feature was actually working on the pair that I trialed. My audiologist did what they could to dial them in, but it resulted in all sounds being muffled and no human voices coming through audibly. The opposite of what the marketing folks at Phonak portray and users have reported experiencing as well.
I did notice slight improvements with some features, but no wow factor when it came to Spheric Speech in Loud Noise.
Overall, I only found a marginal improvement in the Infinio over the Lumity. But again, it could be that I received a faulty pair where the AI chip is not working.
I was going to ask if you wear open domes or at least bass domes with vents because your low frequency hearing is still fairly OK. With open domes or with the vents, they let the noise in and this would negate a lot of the noise blocking that the can do for you to achieve that wow effect.
But since you said that the sounds are muffled and no human voices were coming through audibly, I guess it must have been something else wrong at play here.
But since you said that the sounds are muffled and no human voices were coming through audibly, I guess it must have been something else wrong at play here.
I bet on underfitting medium and high frequencies (which correcting requires also tighter acoustic coupling).
Maybe upward spread of masking by low frequencies? (there is dr Cliff video about that).
@platypus said they were a DIYer. I’m still learning Target, but a lower fitting tab in Target is "Audibility Fine Tuning, in which you can play various speech sounds and adjust gain in suggested ways to make that sound audible, e.g., the F or S sound. IIRC, there are more global “dumb” fixes under the Automatic Fine Tuning tab that platypus could have tried. (click to enlarge)
Relative to @platypus’s complaint that Spheric noise reduction wouldn’t turn on, if one creates either a linked or a copied manual copy of Spheric Speech in Loud Noise, then in the MyPhonak app, you can turn Spheric processing on at any time. Invoking the program manually doesn’t depend on the noise environment. And you can see the effect by playing back the various Target media samples of speech in low environments through your computer speakers.
At home, though, I haven’t been able to manufacture any noise environment that gets Spheric processing to turn on automatically in AutoSense. It would be great if someone could come up with an easy-to-emulate-in-your-home test scenario so one doesn’t have to go to a noisy bar or restaurant to test out various program modifications to the Spheric Speech in Loud Noise program.
I’ve nearing the end of a 3-week trial of the new Sphere I90s, having used Marvel M90s for the last 5 years and having trialled Lumity L90s a few weeks ago. I too have not been impressed with the I90s - the noise reduction system has not been evident even though I visited several not-so-noisy restaurants. But what is absolutely surprising to me is the sound quality. It is undoubtedly inferior to the Lumity L90s: my wife goes to bed before me, so I therefore reduce the TV volume then from around 20 to 10. With my Marvels and with the Spheres, I crank the volume up to its maximum and turn the TV subtitles on. But with the Lumitys, I increased the volume by only 2 points instead of 6, and felt no need for the subtitles. Somehow the Lumitys are able to increase the speech volumes without distortion, leaving the TV background pretty well unaltered. I shall be purchasing a pair of Lumitys in the near future.
@jim_lewis , there is a lot of sound samples in Phonak Target also. You need to download them separately in the update section of Target software (see link below).
With domes or earmolds? I don’t see your tonal , but acoustic coupling can make a HUGE difference, along with making old hearing aids with proper coupling, um, “better” than the newest with bad coupling…
Sorry, but this sounds like a woeful fit. Did you have real ear measurement?
I definitely agree.
(The shorter bar = the better hearing in noise)
There is a graph that shows that OLD, analog, 1-channel fitted with are better in noise than the best-on-the-market, ~10 yrs newer digital multichannel HAs fitted with first-fit (without REM).
@jim_lewis , there is a lot of sound samples in Phonak Target also. You need to download them separately in the update section of Target software (see link below
Thanks. I know that and have done that. But I was looking for sounds I could provide a link to and everyone could access easily. I might run into copyright issues if I posted a Phonak MP3 file here (not even sure the Discourse website software can handle that file type).
P.S. Here’s where I mentioned using Target media in the post you quoted:
At home, though, I haven’t been able to manufacture any noise environment that gets Spheric processing to turn on automatically in AutoSense. It would be great if someone could come up with an easy-to-emulate-in-your-home test scenario so one doesn’t have to go to a noisy bar or restaurant to test out various program modifications to the Spheric Speech in Loud Noise program.
I was able to trigger the spheric mode by putting YouTube on my TV and searching for a video “Loud restaurant noise” and turning up the volume. Then I just needed to start talk talking and it triggered it for me.
Some answers and comments to people’s helpful suggestions above.
I use power domes. No venting.
AutoSense 6.0 did invoke “Spheric Speech in Loud Noise” when I entered a noisy environment - as expected. I also had a manual version of the spheric program which I could invoke at any time. it was very clear that I was in the Spheric program when testing.
The Lumity aids, the ones I have had for two years, also have a speech in loud noise program. I also have a manual selectable mode for that program. It is this mode that I am comparing between the Lumity and the Spheres. I could not detect an amazing difference between Lumity and spheres in this mode.
Many, if not most, who have the spheres have had a very good experience with them including me if I was coming from Marvels or Paradise. If someone is due for new aids I would highly recommend Phonak’s Sphere I90s.
Lastly, I continue to emphasize that the best outcome for hearing loss treatment, no matter which modern hearing aids are chosen, will depend on the competency of the audiologist one works with. I’m grateful to the ones that drop in to this forum now and then and give us their advice.
My experience is that hearing loss is a very personal issue. That is each persons loss is unique, and a combination of their personal loss of hertz and decibel tones, and the type of hearing loss; conductive, sensorineural, or mixed. It seems uses algorithms based on its analysis of what surrounding sound it is capturing to make adjustments. But, is AI compensating for the personal hearing tones and loss of each individual? That is, does or can AI personalize adjustments based on the individuals hearing loss(es) and needs? Or, is it making adjustments based on the most probable adjustment most people with “x” hearing losses need? From my perspective AI program adjustments do not discriminate in a personalized way, which is why I still have to select from “clarity, speech, party, TV, music” and then continue to fine tune that for the environment and directional focus. Honestly the more AI the more challenged and deaf I feel.
The ai is taking a soundscape and reducing the noise around the signal to being the experience closer to that of being in a good listening environment. My understanding is that it is generic. It would work the same way for you and for me. There are some interesting implications that might wait for another topic.
You still want to be able to control how much noise you get in different situations.
If that is what happens then the fact that it is “generic” is the problem. You refer to soundscape, isn’t that a nice way to say ambient or surrounding sound? For example if someone has a mid range loss, and its mostly conductive, then amplification of all the soundscape overwhelms the ability to find the sound needed. Really am so tired of adjusting, and adjusting, every time the AI does its thing.
I have no actual experience of using -based noise reduction in a hearing aid which is the kind of ‘AI’ I meant. What I would hope is that sounds that aren’t nearby human voices lower in volume and sort of melt into each other, making the human voices clearer. That’s partly based on what some people have said in here.
The Phonak Infinio Sphere is the only hearing aid we know for sure does AI-based noise reduction. Is that what you’re refering to?
For example if someone has a mid range loss, and its mostly conductive, then AI amplification of all the soundscape overwhelms the ability to find the sound needed.
I’m not sure what you mean by “conductive?” Do you mean air-conductive or bone-conductive? 90% of hearing loss is sensorineural, it’s air-conductive, and aided by hearing aids. Bone-conductive or mixed-conductive hearing loss is a decided minority of losses. The Sphere is just a typical frequency-dependent sound amplification device for sensorineural loss.
So, I’d say it’s this way. If you already know that a standard “air-conductive” hearing aid helps you in a quiet environment, the Sphere, just like other good hearing aids, will help you. If hearing aids don’t do much for you and your is terrible in a quiet environment, the Sphere isn’t going to be a miracle cure. If you have bone-conductive or mixed-conductive loss, you should learn this from a good hearing examination and be directed to another type of hearing aid in the first place.
The Sphere processing can eat into normal voice sound. I have limited experience, but sometimes a processed voice can sound too soft or too loud. It can warble a bit. But masking noise is definitely removed. My working theory is the quality of the voice that comes out depends a lot on the quality of the voice that goes in.
My wife speaks perfect American English. Yet she tends to slur her words. My theory is that her childhood language was Cantonese, which features tonal inflections on continuous intonation, whereas the Anglo-Germanic side of English features more distinct syllabic enunciation. So, I think her voice survives processing less well than a voice I heard through the PA system in our grocery store that was a high-pitched, clear-toned female voice that came through the processing in distinctly enunciated syllables.
The other weird and most intriguing grocery store experience I had is that our Texas HEB store likes to play a lot of good ol’ rock n’ roll overhead. Usually, I hear the remnants of the music over the store cacophony and the lyrics get obliterated. With Sphere processing turned on, the music receded and I could hear the voices and understand the lyrics much better than in past experience.
If hearing aids work for you in a quiet environment, I think the nature of the soundscape itself and the quality of the speaker’s voice will be the biggest influences on what you hear for any hearing aid, even for the Spheres with their DNN processing running.
Your should have sample media sounds provided by Phonak in the Target fitting software. The HCP could play any one of several examples of speech in noise with a very low . In the HCP’s office, you could turn Spheric noise processing on via a manual program and see how well the speech is cleaned up for you. I’ve done that, and it works very well for most of the media samples that I’ve listened to.
The other weird and most intriguing grocery store experience I had is that our Texas HEB store likes to play a lot of good ol’ rock n’ roll overhead. Usually, I hear the remnants of the music over the store cacophony and the lyrics get obliterated. With Sphere processing turned on, the music receded and I could hear the voices and understand the lyrics much better than in past experience
Wow, I was wondering about lyrics with music, but forgot to ask about it.
If hearing aids don’t do much for you and your WRS is terrible in a quiet environment, the Sphere isn’t going to be a miracle cure
Indeed. If someone has a low in quiet - there is low probability to break the “ceiling” of e.g. 50% WRS (without ) by using hearing aids.
The only exception I’ve seen is @cvkemp writing about the effects of meticulous fitting of Oticon Intent 1. However, I am not aware of the quality of his previous (i.e. before Intent 1) proceedings of WRS in quiet tests.
My wife speaks perfect American English. Yet she tends to slur her words. My theory is that her childhood language was Cantonese, which features tonal inflections on continuous intonation, whereas the Anglo-Germanic side of English features more distinct syllabic enunciation.
I believe the answer to your considerations can be found here:
I only speak for my own experiences. And I will always believe that the right audiologist is the key over and above the hearing aids. But i must add for me having the best combination of audiologist and hearing aids has given me almost normal hearing as long as I am wearing my INTENT1 aids. I have my Real1 aids as my primary backup aids and I notice the difference. My speech understanding suffers and I don’t enjoy music, I cannot hear as well in meetings or lectures like I do with my INTENT1 aids. To give something of a comparison I wear 105db receivers in both sets of aids, i have custom acrylic ear molds for both sets of aids. Now the receivers are a newer version for my INTENT1 aids and the vents are much smaller for the INTENT1 aids. Vents do make a difference. My hearing hasn’t noticeably changed in the last 5 years.
My hearing normally changes every 5 years so I expected it to have changed at my last hearing exam in October but it hadn’t changed. My word recognition scores have actually improved due to intense therapy. I continue that therapy myself with the reading and listening of audiobooks. I read a long while listening to the book. If I miss understanding a word I replay that part of the book. I make sure to mix up male and female voices as much as possible. When watching TV I don’t use captions but I read lips. I can detect even the slightest delay between the lips and sounds. Sometimes I use the TV adapter but almost never while wearing my INTENT1 aids, but most of the time if I am wearing my Real1 aids.
You are asking why I wear still wear my Real1 aids, well I keep the batteries healthy. I never know when I may have to depend on them
Poland
NJ
