Oticon More3 (vs More1) Might Be Capable Enough For Some

Hearing Aids
#21

What I find interesting is that even if you get the More 1, Oticon set the Neural Noise Suppression default values in the MoreSound Intelligence (deep neural network stuff) for less than the max suppression it can handle. You can see in the first screenshot that the default max neural noise suppression is 0 dB for Easy Environment and 6 dB for Difficult environment. The default is denoted by the little “power” icon by the value. I had to manually set it to the max in this example.

So if your audi leaves these settings of your More 1 at the default values, you’re essentially setting your More 1 to perform only up to the max level of the More 3. Keep in mind that the max neural noise suppression is just the Max, and the More is smart enough to use only the right amount, so it’s not like you can tell it to overkill all the times (see second screenshot below in the Help section that confirms this). But it’s capping the More 1’s max neural noise suppression to the max level of the More 3 only by default. You’d have to override it to set it to the max More 1 level. That’s a very fishy thing to do in my opinion by Oticon. Maybe they’re trying to level the playing field after the user has spent the money for the premium version. It’s almost like cheating the premium paying users out of what they pay for unless they’re smart enough to recognize that and override it.

Even one audi commented to their patient that the max values are too aggressive. But they’re max values, they only get used up to those values only when necessary. So you can’t really be aggressive with max value in a smart system that knows which proper values to use in which situation in the first place.

image
image1604×915 123 KB

image
image754×214 11.8 KB

4 Likes
#22

I prefer it to give me all the sound that I can stand and to progress up from that as I need it. I am someone that loves to hear as much of natural sounds as possible, but to clamp down as needed so I can understand speech. I am someone that has a lot of issues understanding speech. But I am not in a noisy environment all the time, but even have issues with speech understanding in reasonably quiet environments.

#23

@cvkemp: Chuck … we’re going to have to call in a TechnoGuru - but maybe you’re hearing fewer sounds wearing your powerful Tier 1 instruments than I am with my cheapo Tier 3s., since yours have more adjustment bands and more capable noise suppression.

I dunno …

#24

No I believe it is my hearing loss in the 1500 to 4000 hz range that is my issue, not only with general hearing but also with my poor speech understanding. And also it may be my desire to hear everything that I possibly can, even the things you don’t know you aren’t hearing.

1 Like
#25

@cvkemp: I just downloaded the spec sheet for Opn S1/2/3, and that getup has a lot more noise suppression modules into it than my More3s.

That’s probably part of the reason they call them “Mores”: less sound suppression = More sounds to the brain.

I’m like you … the More sounds to the brain, the better. More has 50% More adjustment bands, too.

1 Like
#26

I really don’t not what is best for everyone! I just know what has worked best for me so far. And unlike so many on this forum I trust my Audiologist. He not only knows my hearing loss and needs. He knows the environments I deal with, and he loves the woods and hiking some of the same trails that I hike. And I have helped him with some of his patients’ technical issues with Bluetooth connections

#27

@cvkemp: You make an excellent point, Chuck. Just for the record: I am one of the people - like yourself - who trusts his audiologist implicitly.

After 6 years of “wandering in the wilderness”, having decent hardware that wouldn’t work for me, becoming despondent, and giving up hope/faith in the industry’s ability to help me … after all that, my audiologist has given me back a great measure of my quality of life.

How could I ever treat him with anything less than courtesy, gratitude, and respect?

And I agree with what you’ve said several times in the past: the audiologist you choose is at least as important as the make of hearing instrument.

MESSAGE RECEIVED/SPUDCOM OUT//sg

1 Like
#28

This is one area where VA spares no expense. They always have offered me top of the line hearing aids for over 40 years. The top of the lines aren’t working so well for me anymore, but it is not the fault of the equipment or the fitter/audi…my ears are just that screwed up…speech recognition stinks…
…

3 Likes
#29

This is a bit off topic but, @Volusiano, do you know what is the definition of “noise”?

I ask because, in particular, i struggle understanding my wife in the car over road noise at highway speed. Wondering if changing the Noise suppression would work and whether this would be considered “easy” or “difficult”.

Many thanks.

#30

Although “noise” is mentioned a lot in the Oticon whitepaper on MoreSound Intelligence (the deep neural network stuff), there’s no clear definition of what noise is. This is to be expected because especially in the open paradigm, a sound can be noise to one person and of interest to another person. Even to the same person, a sound can be noise at one time and be of interest at another time.

But maybe I can tell you how Oticon treats sounds as noise sources in their MoreSound Intelligence flow (the deep neural network stuff) as seen in the screenshot below.

In the Scan and Analyze module, they just look at the amount of sounds in a sound scene (every 1/500 seconds) to see how much sounds there is. Then based on the user input in Genie 2 on what the user think is Easy and Difficult, it’ll route the sound scene to either the top flow for Easy or the bottom flow for Difficult.

Noise is not a concern in the Easy Environment, so the Virtual Outer Ear with 3 different models is used to obtain spatial clarity without really attenuating any noise.

Noise is a concern in the Difficult Environment, so Oticon uses what they call a Spatial Balancer. It’s basically the same stuff they use in the OPN OpenSound Navigator to attenuate surrounding noise. It uses 2 types of mics, an omni (all around) mic, and a back-facing cardiod mic, and mix them up to attenuate noises from the sides and behind to favor speech. So in this case, noise is any sound from the sides and behind that is not speech, but in the presence of speech. If there’s no presence of speech (checked every 1/500 times per second), then the other sounds don’t get attenuated and you would hear them. In summary, this Spatial Balancer is basically a directionality system that uses a special type of beamformer to rebalance the soundscape in favor of speech, but only when speech is present. See the 2 figures in the second screenshot showing how the noise attenuation is done inside the Spatial Balancer.

Finally, after doing some “preliminary” noise attenuation via the Spatial Balancer, the soundscape goes into the deep neural network for processing. This AI has been trained to differentiate between speech and non-speech sounds, and trained to provide a balanced sound scene between all the sounds at that moment based on real life examples (12 millions of them) When you tweak the Neural Noise Suppression here in terms of dB, you’re effectively tweaking this balanced sound scene further, more in favor of speech and attenuating the other non-speech sounds (which is now considered as noise), but only if speech is present. If speech is not present, you get your normal balanced sound scene just the same. The higher max value you assigned in Genie 2 for the Neural Noise Suppression, the more leeway you’re giving the More to be as aggressive in the attenuation of the noise as needed, up to the max. But if there’s no speech, or only single speech, the attenuation of the noise should not be employed by the More the max. The More is supposed to attenuate less if less is needed, and more only if more is needed, up to the max set.

Having given you a long winded explanation, back to your original question of what can be done to understand your wife in the car better over road noise at highway speed? I think setting the Neural Noise Suppression value to max and include Moderate (or even Simple to begin with then ease back to Moderate later) as part of the Difficult Environment when your audi program Genie 2 for you can help.

But remember that this help from the More is only while your wife is speaking. As soon as she stops speaking, the road noise will come roaring back. So the other half of the equation (beside the help from the More) is your brain hearing acuity development over time, so that you can learn to tolerate the road noise when your wife is not speaking, and focus in as soon as your wife starts speaking. Some people get adapted to this quickly and thrive. Others just want the peace and quiet of no road noise even if their wife is not talking to them -> these people are probably much happier with the traditional paradigm instead of the open paradigm.

image
image1588×793 105 KB

image
image1376×721 69.5 KB

6 Likes
#31

Thank you this is great information. I am someone that wants it all. I need better speech understanding, and I also need better all around sound understanding for when I am out in the forest hiking, and need as much capacity as possible to pin point where the sounds are coming from.

1 Like
#32

Thanks for the thorough explanation @Volusiano . I will try this below. The road noise does not bother me per se… Not being able to easily have a conversation in the car is what i would like to improve so having more aggressive noise cancellation in the presence of speech in this scenario is probably what i need even if the road noise comes back later. I read in one of the white papers that the transition can happen as quickly as 6 seconds.

Many thanks and will give this a try.

1 Like
#33

In the MoreSound Intelligence whitepaper, they’re saying that the changes in the sound scene must take more than 2 seconds before the hearing will transition between the Easy and Difficult Environment (adapt the level of help, in their words).

But if you mean transitioning between applying the neural noise suppression when detecting speech, this whitepaper doesn’t really say how long, although I think it would be almost instantaneous because the Scan and Analyze module and the Spatial Balancer update themselves 500 times per seconds.

If you can remember which whitepaper you saw the mention of the 6 seconds from, I’d be interested to know to see what context they’re talking about there. Thanks.

image

1 Like
#34

The “Oticon MoreTM competitive benchmark” white paper states under the “Adapting rapidly to new sound scenes” section (p.7):

“For the competitor hearing aids, it takes up to more than 20 seconds for the contrast between speech and noise to become visually stable. In contrast, Oticon More reaches a steady state of opera- tion within approximately the first 6 seconds of being in this newly encountered sound scene.”

Not sure if I interpreted that in the right context.

Thanks

1 Like
#35

Oh, OK. Thanks for sharing this. Yeah, they’re talking about how fast the More transitions from one sound scene to the next in terms of having the noise reduction become stabilized as a metric to determine the successful completion of the transition.

So it is consistent with the 2 seconds criteria for the Scan and Analyze module to decide whether the change in the sound scene is persistently long enough to decide to make the switch from Easy to Difficult environment or not (or vice versa). Then about 4 more seconds for the new processing to take place for the new sound scene before the user makes effective use of the newly processed sound scene.

So this is equivalent to for example, you start merging into a noisy freeway from a quiet street until you can fully benefit from the new processing (like speech over noise suppression) 6 seconds later. This is not to be confused and to think that every time your wife starts speaking in the car, it takes 6 seconds after she starts speaking before you can hear her better. Once you’re on the noisy freeway, 6 seconds into it, for the rest of the trip on the noisy freeway, her speech should be brought to the foreground almost instantly every time she starts speaking (by way of the More rebalancing the sound scene to reduce the other non-speech sounds’ volume, up to however much you have your max neural noise suppression set to, as necessary).

2 Likes
#36

Volusiano - I have OPN1 RIC with 85dB receivers. It took a year of tuning to get the “noise” settings that worked for me in most situations such as noise in restaruants and road noise. Here are my settings:

image
image738×439 98.5 KB

1 Like
#37

@FredTX: These terms and settings aren’t familiar to me. The More3s seem to have a different vocabulary.

Thanks anyway for posting them … maybe @Volusiano can translate. I’m also not sure that I (or anyone else) could use your settings. I suspect that may be an individual thing.

#38

While it’s not the same, the parameters are fairly equivalent.

@FredTX basically has the Noise Reduction (OPN equivalent of More Neural Noise Suppression) for Simple Environment (OPN equivalent of Easy Environment) set to the max value possible of -3 dB (More 1’s max is -4dB). For the Complex Environment (OPN equivalent of More Difficult Environment), he has it set to the max possible of -9 dB (More 1’s max is -10 dB).

The OpenSound transition set to High is probably roughly equivalent to including at least Moderate and maybe even Simple as part of the Difficult Environment select in the More.

The rest of the stuff are pretty much 1 to 1 equivalents between the OPN and the More as far as I can tell.

The take away here is to set the Noise Reduction Max values to the highest possible that’s available in the tier you have, so as to not cap the max unnecessarily into the equivalent of a lower tier version. This way you let the hearing aid choose the most appropriate level of noise reduction for you without tying its hands behind its back by capping off the max at a lower value than it can be. That’s how I have my OPN 1 set up as well.

1 Like
#39

:upside_down_face:Okay … well, that’s very useful to know. Thanks to both @FredTX and @Volusiano for the clarification.

[The table Fred posted is useful to me because I reached my saturation point on these technical details a while back! :joy:].

Anything that I can take into my Thursday audiology appointment will be very useful to permit me to communicate unambiguously with my long suffering audi, whom I know to be doing everything in his power to adjust my HearingEngines to my advantage.

Usually, it’s my ambiguity/lack of technical savvy that keeps him from hitting the mark.

Thanks again.

[Thank you for your service, @FredTX.]

#40

Went to the audi today for this one adjustment.

  • “Moderate” and “Simple” were already part of the “Difficult” group.
  • In their version of the software, the default noise suppression for the Difficult Environment was 8dB (which was selected), not 6 as in the screenshot earlier in this thread. We made it 10dB as per Volusiano’s recommendation.
  • Sound Enhancer was already set to “Detail” (the default is “Balanced”).
  • Virtual Outer Ear is set to “Balanced” which appears as the default value (as opposed to “Aware” being the default in said screenshot).

Not sure if the differences pertain to their version of software or whether they are configured as a result of the patient’s audiogram.

Will report once i have a chance to go on the road with the wife :slight_smile:

Thanks!