Skip to content
HearingTracker
Log in

Oticon Speech Rescue

This topic is closed.

38 replies8.4k viewsLast reply

Jump
Link to this topic

Was reading their white paper on this. Apparently if it’s activated, one has to turn off other features because of lack of processing power. I don’t know if this is still true of the OPNs or not as I believe this referenced hearing aids that were older. Volusiano, have you read up on this?

7 people in this conversation19 min readJump to latest

I didn’t know this white paper exists until you mentioned it, MDB. So thanks for bringing it up and of course I downloaded it and read it right away. :slight_smile: It’s a pretty informative paper, I enjoyed reading it.

With respect to their answer on question 24 about some of the features like Transient and Spatial Noise Management (Dynamo only), Music Widening, Power Bass and Voice Priority needing to be disabled because they cannot run simultaneously with Speech Rescue, I must say that I never really heard of these features mentioned on the OPN. Also, this white paper was published in 2015 when the OPN wasn’t introduced yet. The fact that the OPN is on the Velox platform which has considerably more processing power compared to Oticon’s previous platform, I’d venture to guess that conflict of resources is probably no longer an issue on the OPN. But that’s just a guess and when I get the new firmware update hopefully some time in June and have my audi activate the Speech Rescue feature for my OPN, I’ll for sure ask her whether it’d come with any trade-off at the expense of any other features having to be disabled or not.

I really can’t wait to try this out! And for sure I’ll report back.

One thing I didn’t understand in the beginning was why in the transposition method, the high frequency source signals have to be cut out and transposed to lower bands. It seems like they (not Oticon because I assume the transposition method is not employed by Oticon) could have just left the HF source signals alone and transpose a copy of it down to the lower bands. But then I realize that maybe this has to do with resource limitation of both memory and processing power. It’d be less resource intensive to simply cut it out, transpose and be done with it, instead of having to make a copy and then transpose the copy, all the while still having to maintain and process the HF source signals in parallel as well on top of it.

I wonder if Oticon has a patent on this composition method or not. If yes (and presumably so), and Speech Rescue turns out more successful than the other 2 methods, compression and transposition (as implied in the white paper), Oticon might have a golden nugget in their hands especially with folks with severe and profound HF loss who can benefit a lot from .

Another interesting thing I picked up from the white paper was the fact that Oticon thinks there’d still be value to preserve the HF sounds just the same and not cut it out because there may still be valuable cues the hearing impaired person can pick up for music and non-speech sounds in the HF range. I never thought about it but I guess it’s a valid point because speech sounds in the HF range may not be loud enough for the amplification to help make audible, especially in noisy environments. But music sounds in the HF range may be loud enough for amplification to be more effective in making those music sounds heard compared to speech sounds. So there’s some validity in wanting to preserve the HF range for music and not cut it out just for speech frequency lowering at the expense of music.

Always good to remember. I mainly shared because I found it interesting which features could be activated was hardware limited.

@Neville said: (unavailable)

Also, remember that white papers are marketting, not research.

In reply to #5

@Neville said: (unavailable)

Also, remember that white papers are marketting, not research.

Not to Volusiano they’re not. He soaks up white papers like a drug addict soaks up heroin. Oticon can do no wrong in his eyes.

@alesha said:

Not to Volusiano they’re not. He soaks up white papers like a drug addict soaks up heroin. Oticon can do no wrong in his eyes.

Instead of doing negative personal bashing like this without any intelligent discussion, I’d rather invite you to read the white paper and discuss the merits and/or demerits of the 3 different approaches and explain to me why you think the white paper has the demerits that you don’t agree with. Then I’d gladly engage with you in a cordial and civil discussion about it. Otherwise it’s not worth engaging with you any further beyond this response.

To me personally, with this being the first and only white paper I’ve read about so far, what they say seems to make a lot of sense to me. I would think that my electrical engineering background over the last 30 years would allow me to form my own educated and independent opinion about any white paper I read without having a bias. I guess based on what you say, you already formed your own opinion and bias that the Oticon white paper is no good simply because Volusiano soaked it up like a drug addict. Very astute opinion you made there, I guess…:slight_smile:

In reply to #4

@Neville said: (unavailable)

Limited literature on this, but so far compression is the winner. e.g. The effects of frequency lowering on speech perception in noise with adult hearing-aid users - PubMed

This literature is indeed very limited. Without the details of the full paper, it’s hard to form any opinion on your own what to think of the paper. Do you have a link or the full content of the paper you can post up here, Neville?

In reply to #5

@Neville said: (unavailable)

Also, remember that white papers are marketting, not research.

To me, white paper is neither marketing nor research. I see white paper as something in between. Marketing materials usually make a lot of claims without much rationales or detailed explanation to back it up. White paper at least provides a lot of details about the rationale and logic behind it to at least to back up what they’re saying to help you form your own opinion on whether you agree with what they say or not.

In reply to @Volusiano

If you look down just past Keywords, there’s a link for free PMC article that lets you see the whole article. My take is that it found that transposition tended to result in distortion. Oticon’s might be better because of faster processing (less temporal delay) and because it’s transposing into a narrower frequency range. I must admit that compression makes more intuitive sense to me, but I’ve learned from reviewing lots of medical research that intuition is often wrong. It’s a shame that there isn’t more cooperation between hearing aid companies to determine best approaches. It is nice to see some independently funded research like this article, but it is very limited. I must admit that if I were looking for the best approach to deal with that Phonak’s Sound Recover 2 sounds like the most promising.

@Volusiano said:

This literature is indeed very limited. Without the details of the full paper, it’s hard to form any opinion on your own what to think of the paper. Do you have a link or the full content of the paper you can post up here, Neville?

In reply to @Volusiano

@Volusiano said:

To me, white paper is neither marketing nor research. I see white paper as something in between. Marketing materials usually make a lot of claims without much rationales or detailed explanation to back it up. White paper at least provides a lot of details about the rationale and logic behind it to at least to back up what they’re saying to help you form your own opinion on whether you agree with what they say or not.

For someone who claims he has 30 years of engineering background, that’s a stunningly naive take on white papers. The purpose of a white paper is to present a one-sided opinion or position that’s favorable to the content creator. Companies use white papers to target and influence specific audiences. In other words, marketing.

To me, it seems you’re not aware of your own biases and how they appear to other readers. Due to your previous shilling for Oticon Opn, it’s very difficult to take you seriously whenever you make negative comments about other hearing aid brands or technologies.

In reply to @MDB

@MDB said:

If you look down just past Keywords, there’s a link for free PMC article that lets you see the whole article. My take is that it found that transposition tended to result in distortion. Oticon’s might be better because of faster processing (less temporal delay) and because it’s transposing into a narrower frequency range. I must admit that compression makes more intuitive sense to me, but I’ve learned from reviewing lots of medical research that intuition is often wrong. It’s a shame that there isn’t more cooperation between hearing aid companies to determine best approaches. It is nice to see some independently funded research like this article, but it is very limited. I must admit that if I were looking for the best approach to deal with frequency lowering that Phonak’s Sound Recover 2 sounds like the most promising.

Thanks MDB for pointing out the link for the full article. I’ve been able to get a chance to read the full article now thanks to that tip.

I was interested in understanding how NFC (nonlinear freq compression) vs LFT (linear frequency transposition) vs FT (frequency translation) are defined, as compared to the Oticon, just to make sure that we’re talking about apple vs apple here. To me, NFC and LFT in the independent paper Neville referred to is pretty close to the compression and transposition that the Oticon paper describes. However, I think the FT from the HHS paper is not really the same as the composition method described by the Oticon paper. That’s because FT is simply LFT with the HF source signal retained. The composition method by Oticon goes one step further and breaks up the transposed HF source signal into 2 or 3 smaller chunks before it superimposes these 2 or 3 smaller chunks on top of a lower band. The significance of this is that it now can present information from a wider input (HF) region into a narrower (LF) output region. This is of significance because it helps maximize the capturing of a wide HF source region where important speech cues are, all the while still able to minimize the distortion at the LF range due to the introduction of the HF transposed sounds because the superimposition is done on a much narrower band as compared to the HHS’s FT approach.

And one other thing I’d like to point out is that I disagree with Neville’s conclusion that compression is the winner over LFT and FT according to the HHS paper. What I conclude from the HHS paper is the compression didn’t help much, while LFT and FT made it worse due to distortion introduced due to superimposition of the LF bands. “Making no difference” to me is hardly “winning” in my opinion.

The distortion the HHS paper refers to is really not “distortion” per se in the strict sense of what I think distortion is. The HHS paper used the word distortion simply to describe the “confusion” and “masking of useful information” caused by the supersimposition of LFT and/or FT. That is why I think Oticon goes through the trouble to take the next crucial step to break up the wide band of HF source into 2 or 3 sections and superimpose them on a much narrower LF band, expressly for the purpose of minimizing this “distortion”/“confusion”/“masking”. So you cannot generalize and say that the FT approach the Starkey 3 Series il10 employs is the same as the composition algorithm that Oticon employs for Speech Rescue, because it’s not.

Hopefully if there are to be more independent papers like this HHS paper in the future, beside the categorization of NFC and LFT and FT, there’ll be a fourth categorization called Composition or something of the sort to more accurately reflect the algorithm Oticon employs in Speech Rescue. After all, the HHS paper did not use an Oticon HA with Speech Rescue as a test subject for their paper, they used a Starkey HA with a different algorithm compared to that used in Speech Rescue.

Another thing that irked me with the HHS paper is why did they do it only on 10 subjects? That’s such a small sample to make much definitive conclusion. But I can understand that it may be limited by resource constraints. I do hope for more independent studies like this to be done in the future so that we can all benefit from learning their findings.

In reply to @sgh

@sgh said:

For someone who claims he has 30 years of engineering background, that’s a stunningly naive take on white papers. The purpose of a white paper is to present a one-sided opinion or position that’s favorable to the content creator. Companies use white papers to target and influence specific audiences. In other words, marketing.

To me, it seems you’re not aware of your own biases and how they appear to other readers. Due to your previous shilling for Oticon Opn, it’s very difficult to take you seriously whenever you make negative comments about other hearing aid brands or technologies.

While I don’t really want to get down this rat hole further, I think you’re putting words in my mouth. I never said that white papers are unbiased. I challenge you to quote me on where I said that. I only said white paper provides a lot more details to back up a company’s claim. Whether one still believes in it or not after one reads it is up to themselves. If you choose to dismiss reading a 16-page white paper right off the bat because it’s still marketing junk to you then it’s your prerogative. If I choose to read up a 16-page and form my own opinion on it then it’s my own prerogative as well. At least I don’t attack you for forming your own opinion. Why are you attacking me?

And remember, too, that I’m not the one who instigated this thread. I was invited by MDB, the OP, who called me out by name, to read this white paper (which I didn’t even know exists) and to offer my opinion on it. And that’s what I simply did. I didn’t start this thread and never said “HEY EVERYONE, check out this phenomenal white paper by Oticon on Speech Rescue!!!”

While you’re accusing me of not knowing my own biases, maybe the same for you and Alesha who seem to have a pre-disposed bias against me prior to this thread already because my previous posts/threads spoke favorably of the OPN. But since when is it a crime to speak favorably of a brand especially if I’m wearing that brand and speak from my personal experience as well as understanding about that brand?

Finally, I again invite you, if you’re truly interested about discussing Speech Rescue and , which is really the topic of this thread, then let’s engage in that conversation and limit ourselves to that discussion only. This personal bashing is really not relevant to the topic raised by the OP. You’re welcome to express your own opinions about frequency lowering and Speech Rescue. Expressing your opinions about me is not really worth discussing here.

Have any of you read 20q Frequency Lowering Ten Years Later–New Technology Innovation? It’s by Joshua Alexander and dated 9/12/16. You have to look for it in audiologyonline.com4. It’s a long survey of six different companys’ efforts at . It’s a scholarly discussion, not a company blurb. And might add something to this discussion.

In reply to @AlanM

Have definitely read it as well as anything else that comes up on the topic from audiology online. I don’t know if it’s any more scholarly than the Oticon piece. It’s basically explaining how the different versions of work and how to use them. It doesn’t make any claims that any version is better. Even after reading the audiology online piece, I was still fuzzy about how frequency transposition worked and what was different about the Oticon version. The Oticon white paper helped give me a better understanding of how it worked and created a plausible story of how it worked. Again, the main reason I posted was that I found it interesting that as we ask the hearing aid CPU to do more and more that we can run into limits and have to make choices. I mentioned Volusiano because I knew s/he had a notable high frequency loss and wanted to try out the feature when it was available. I think we could avoid the bashing. I don’t sense any deceit in Volusianos, just enthusiasm for his/her “toy,” Perhaps similar to my enthusiasm for KS7 which I think has annoyed some.

@AlanM said:

Have any of you read 20q Frequency Lowering Ten Years Later–New Technology Innovation? It’s by Joshua Alexander and dated 9/12/16. You have to look for it in audiologyonline.com. It’s a long survey of six different companys’ efforts at frequency lowering. It’s a scholarly discussion, not a company blurb. And might add something to this discussion.

Google Frequency Lowering Ten Years Later–New Technology Innovations. Search within for the article by Joshua Alexander. It is extremely difficult to find but I did find it this way a few minutes ago.

Edited

If you Google “ hearing aids”, it’s the second hit. Several other interesting articles show up too.

In reply to @MDB

I am interested in this discussion and have done a little research in the area because my audiologist has just ordered a pair of Phonak Audeo B (presumably the 90s–they always get the high end) mainly for the . My voice recognition went down noticeably to 49% within two years but the frequency response did not change noticeably. And I have been hearing much more distortion around speech in many situations. I am replacing a pair of Oticon Alta 2 pros. If the Phonaks make me unhappy I will try something else. I hope eventually some type of frequency lowering will be effective. I have been wearing hearing aids for 30 years and this is a new adventure.

R [50 50 55 55 60 80 90](tel:50 50 55 55 60 80 90) 90
L [40 40 45 50 55 60 85](tel:40 40 45 50 55 60 85) 85

In reply to @AlanM

Thanks. I should have thought about googling it (like I did with the Oticon paper), but it didn’t occur to me at first only because there’s the mention of audiologyonline.com so I thought maybe I’d have to login to find it and google may turn up empty. But I was wrong and thanks for your tip to just google it, I found it.

Boy, the Alexander one was a great paper to read, too. I enjoyed it even more than the Oticon paper because it gave a broader perspective of the different aspects of by all the major players in the industry. Also, I didn’t quite get what they were saying about the cochlear distance in the Oticon paper (they presume you know what it is), but the Alexander paper explains it in a little bit more details. I must admit, though, that I followed the compression trail of discussion up to a point then it was lost on me, both in understanding and in interest, because it started to get way too complicated for my attention span. But one thing for sure is that I didn’t realize that there was so much to it to compression than the Oticon paper simplified it to be.

I’m just excited that I’ll get to try out the Speech Rescue when the firmware update becomes available in June. If I could, I would want to try out all different frequency lowering forms and see which one I like the best. But that’s not going to happen so I’ll just settle for what I can get and will report back.

Like MDB correctly sized me up, I’m just overly enthusiastic about my new “toy”, so if this came across as annoying to anybody because they might have misinterpreted my being overly vocal about the OPN as me bashing other brands and technologies, my apology because that was not my intention. I just think that the OPN is different enough to be easily misunderstood so I just want to share my understanding and opinion of it. But maybe I have a tendency to overshare and become over opinionated. But I can tell you that if Speech Rescue turns out to be not helpful to me, I’ll report back my honest opinion of it without bias. If anything, my high expectation on things seems to have a history and tendency to let me down, including the OPN when I first started using it.

Related discussions