New "Wolverine" Multi-processor DSP

FYI: Sound Design Technologies is now marketing its new Digital Signal Processor design, which is a 90 nm CMOS, single die, multiprocessor design, with five micro-DSPs, multiple special hardware units, and write-once PROM memory. The first incarnation, known as the SA3400, is a two-die package that carries both the DSP and a 1Mbit EEPROM die. SDT claims “Ultra Low Power” for this DSP, which might result in greater battery life for the HOH. See Overview.

The clock speed is 2MHz to 16MHz, which SDT says can deliver 50 “equivalant” MIPS. I interpret this to mean that each of the five micro-DSPs are capable of 10 MIPs. See Data Sheet.

O.K. So for dummies like me why don’t you explain in very simple layman’s language what you just said/wrote.

Thanks.

Shi-Ku (Who doesn’t understand all that technical stuff.) Chishiki

ShiKu.Chishiki@Gmail.com

OK. IMO, this DSP is better, faster, and will ultimately be cheaper than earlier models. It should be able to process sound faster so that (for example) hearing aids would leave fewer artifacts, have improved adaptive directional mic performance or offer more complete frequency correction. There may be sufficient power and memory storage to implement new, better sound processing algorithms. It may also be suitable for entirely different applications (other than HAs), which could boost production volume and bring the cost down quickly. HA batteries should also last longer.

Stupid question: how does this compare with the best that’s out there now? David R

Thank you for clearing that up.:frowning: And since this is made by “Wolverine” I assume I can wear this on my feet?:wink:

Good question! Ask Phonak to publish the technical specs on their DSPs and maybe we could begin to find an answer. Unfortunately, the biggest players in the hearing aid market are very secretive about the details of their products. They are perfectly happy to sell them for outrageous prices to desperate people who think they are getting something better.

Many of us on this forum have called for head to head comparisons of hearing aids but they have been slow to emerge. Consumer Reports did publish a report in July, 2009 - the first timid step toward getting comparative information. Specific HA details are here but you have to have a subscription to read them.

IMO, the next best way to estimate hearing aid quality is to look at the capabilities of their components. Most of us do that automatically, when buying computers, HDTVs, etc.

Well, the most important component in today’s digital HAs is the DSP - the computer that processes sound and adjusts it to compensate for the wearer’s specific hearing loss.
The DSP, with a microphone and a tiny speaker, fundamentally, constitutes a hearing aid. So, it seems to me that the DSP is the most important component to look at.

So, try to get a technical Data Sheet from Phonak! I’d love to start comparing them on the basis of their performance features - and - what they actually cost to produce.

I find it very difficult to make such a comparison, 1) all the big companies have a different approach on how to compensate for a hearing lost, meaning you don’t know how the different companies uses there raw power MIPS. 2) all companies have to some extent unique features, meaning you you can’t compare something to nothing.

If you only look at the price for the chips, today can buy electronics with a DSP for a few $ meaning that the raw cost of making a DSP it can’t by very high, but the cost of designing/develop the DSP are properly gigantic read millions of $. I can’t see why some are so focused on the price of making a HA, I believe that it’s more important does a HA make my life easier. Again a Ferrari is expensive but it’s just a car(I know some would disagree here) and compared to a cheap Kia that both have four wheels and a engine a few doors and if I owned any of them, the both would be able to transport me fra A to B, if i’m not to drive to fast the both would properly do it on the same time. So why can’t I buy a new Ferrari for a few thousands $ :).

Of course comparisons between hearing aids are difficult today.
The purpose of this thread is to add some facts to our tiny hearing aid information base. My firm opinion is that information is a good thing and will eventually improve everyone’s understanding of how they work (the mechanics), what features are effective in helping people and what features are just pure fluff.

Specifically, this new DSP has mulitple processors that operate simultaneously. These will be able to do several different operations at the same time (in parallel, rather than one-at-a-time, as older DSPs did). This makes it able to execute more complex and exotic sound processing algorithms faster.

Well, when somebody figures out how to recreate perfect hearing with this chipset it will be a big seller, until then it’s just more technology laying in wait.

My home stereo has 10M+ transistors in it no doubt. You will never be able to convince my audiophile friends it sounds better then their 2 dozen tubes. I am not convinced that the performance of the hardware is the limiting factor in having accurate hearing restored. The weak link in my mind is the characterization of your hearing for fit with an instrument by half a dozen points on an audiogram. We need a way to hook a network analyzer to our ears!

Faster CPUs alone does not help to eliminate the frequency/temporal distortion that is introduced when variable gain in time and frequency is applied to the sound. What is needed is smarter sound processing and a lot of development is yet to be done in that area.

I’m not sure that any hearing aid ever will be able to restore peoples lost hearing, it can help a person to a easier life(hear easier). This is what the big companies HAs are doing today. Another problem today is that some people prefer one brand some another, here are we taking about persons with audiograms that look the same.

I can agree with you that a threshold measurement is properly not right way to do a characterization, but it’s a easy way that to a large extent works. But as long we don’t know how we hear things and how this information are processed in the brain, then could it be the best way to do it.

Back to this new DSP, I fully convinced that big hearing companies at least those how are making there own DSPs have today multi core processes, but without a specification on there platforms it’s difficult say for sure. But how to the make use of these cores I can’t figure out, the way I see it to many things will have influence on each other. e.g noise reduction together with an adaptive microphone direction system, these systems have to work together one-at-a-time. a noise has to be detected before it can be reduced.

we are months away from 2nd gen wireless instruments… I would expect soon,
more mid price instruments with basic wireless capabilities…

I would predict that in about 2yrs or so we would see basic instruments with
wireless - streaming capabilities

IMO, as soon as hearing aids begin to fully address hearing loss, word comprehension, ambient noise, feedback, etc., we can begin to address “fluffier” features such as these.

BTW, you never did post specific studies that demonstrate and/or quantify the hearing benefits of wireless.

To my knowledge no such studies exist :slight_smile: What I mean is that I haven’t read any studies that to a convincing degree proves the wireless features are beneficial for the average hear impaired person. At present only Siemens and Oticon make use of communication between the hearing aids (transfer of settings), I see sync of volume and program only as a simplification of user control, as such can’t be counted as advantageous in sound processing. What can be found is that companies that currently offers wireless features have heavy biased information on that test users fund wireless features didn’t make thing worse, then not wireless HAs, and to a degree the test users preferred new HA with wireless technology. Considered that for some companies, there new wireless HAs are only a upgraded previous generation HAs with and addition radio/wireless features, and doesn’t offer new audiologically features, is it fantastic that some preferred these new HAs (I know that this statement will tread on somebody’s toes, and as such are very difficult to prove).

For me the introduction of wireless features is an indication that the top companies HAs have hit a limit on audiologically features (read the HA’s are to powerful, the have simply to much processing power) and to justified the making of new HAs the had to invent something new. This doesn’t mean that I find wireless HAs useless but compared to other users of this forum (this is my personal impression) I don’t find these HA to be the 7 wonder for hear impaired persons. For advanced users (incl all that read forums like this) the can offer an easier integration with a modern (IT) life, but to my knowledge most HA are senior, and for some of those just changing the battery is a challenge.

To xbulder. I’m convinced that we will see new wireless HA at EUHA (I guess this is what you mean with within a few months) I’m also convinced that in less then 10 years time hear impaired persons are going to demand wireless features, what I’m not so such of is that we are going to see a next generation wireless technology soon. Both Phonak and Oticon pronounced that developing wireless technology have cost a lot of money and time (the last part is at leased something Oticon declared) To just throw the current technology out, doesn’t make sense from a business point of view. It money out of the window. I don’t believe that the current technology is exhausted. It have it limiteds but currently it’s difficult to see how to improve it with the current restrictions (power and size).

I agree 100%. Wireless gives marketing something to work with. It’s benefit to the end user is not particularly relavent in this aspect, however they need to show that it really is worth spending $<large sum> for the latest and greatest instruments.

Plenty of MIPs are available on todays hardware, what is lacking is the knowledge on how to use them to solve a particular persons hearing defficiencies.

I respectfully disagree. It seems you have never fitted Oticon Syncros. Epoqs are not
syncros with radio or wireless capabilites. Syncro or Syncro 2 are perhaps one of the worst HI that Oticon had. XW performs substantially better.

I have also heard informal comments about the improve sound quality of the exelia over
the Savia. But peharps Admin could clarify this since I know he does fit tons of phonaks.

Prediction EUHA 2009 - Phonak- entry level instrument with wireless cap.
Oticon and siemens - Mid price instruments with wireless cap, SUMO dm Upgrade

AAA2010 new wireless instruments…

I don’t understand this anti-wireless focus here.

Technology moves on: wireless has the possibility of improving hearing aid operation … NOT having wireless removes that option.

If you have ever used or fitted a Phonak aid with the latest ear-to-ear audio streaming zoom mode then you will KNOW that wireless is useful!

So far today I have read here that wireless is a sign of a failing industry, remote controls are a scam to cheat customers, processors are so powerful they have no more algorithms to run because we have run out of processing ideas … all nonsense …

I agree with EnglishDispenser. As somebody who has been hard of hearing since birth, having progressed from some very good analogue aids, that indiscriminately amplified every sound, to the digital technology today, the idea that some of the top line hearing aid companies are making cheap DSP components is risible.

Having lived through countless number of situations over 2 decades, where I could not hear at all, for several hours at a time, in social or work functions, pubs, beside traffic, to being able to hear in virtually all situations to at least a 75% degree of comprehension, it is a testament to the aids I am wearing, produced by industry professionals. Believe me, if I wanted to be cynical, I could be, but not about this. Sure, the cost of these aids I know is painful for some, but there is no denying that the technology is moving in the right direction, and does not consist of generic components produced for a dime.

Can you please comment on the ear to ear audio streaming? I have heard this long time ago as a future direction. I would agree with you, that wireless do provide tangible benefits.


hearnow might be right in that I have not read formal studies about efficiency of Wireless techn. Perhaps they are there and Im not aware of those.

but for those, who had fitted many many patients, we know this seems to be the right
direction. Right guys? what do the disp. prof. feels about this?

We’re not anti-wireless, just anti hype!

If the manufacturer is spending $millions to develop a technology, why would’nt they do a double blind study that shows how it increases performance? We skeptics assume that this information would be available if it really did improve performance.

There are those of us who don’t want another instrument brochure with a picture of an elderly couple and a bird in the background written by marketing having little to no performance information. I would love for the fitters and audies to start a thread pointing to double blind studies on different instruments/algorithms/hardware platforms.