medical noise generator

[Hirsch2k]

Hello @all

First of all, please excuse my bad english 
Ok, I think I have to explain this a little bit. My mother is suffering from a heavy tinnitus which is caused by a genetic defect. It seems that the only cure for her is listening to a certain type of noise with headphones which kind of acts like an "anti-noise" to the noise caused by the tinnitus.

I've ordered a Clip+ for her which I rockboxed and which she appreciates very much, because she can listen to sounds from freesound.org (waterfall etc.) and can adjust it with rockbox graphical EQ. This already helps her very much. There are companys which sell some kind of in-ear noise generators for about 1000 (!!) dollars which not everyone can afford. I think it would be a blessing for those who are suffering from tinnitus if they had a device like the clip+ with rockbox which can generate noise like pink or white noise and which they can adjust.

I don't know if this could be an option for a plugin for rockbox, I thought about writing one myself but my programming "skills" are a little knowledge of C and thats it   :'( I think it could help a lot of people that are suffering from tinnitus.

Oh and thank you so much for your great firmware. I don't know why manufactures can't write such a great firmware as their original firmware...

[nls]

What is it that you would want to adjust about the noise?
Generating white noise is trivial and i think i saw a relatively simple algo for pink noise somewhere but this could easily be done on a computer and recorded to a file that you could play on the DAP so what benefit would a plugin have over recorded pregenerated noise?

[Hirsch2k]

That was what I thought, but one problem is that the tinnitus noise is constantly changing. I'm not suffering from it, so I can generate noises but I don't know if its exactly the right noise because I'm not hearing it. Sometimes a kind of "beep" seems better, it seems to change from day to day, even from one hour to another. Unfortunately my mother is a little older and she can't handle a computer not to mention a noise generator. And when she is in holidays, she don't even have me or a computer 
I also think of the elderly who often suffer from tinnitus aren't computer users or even have a computer. I think they could benefit from it.

[NiHaoMike]

How about expand this idea and make it a function generator? It can be very useful for testing audio equipment. It would basically be implementing a DDS in software.

[[Saint]]

I also think of the elderly who often suffer from tinnitus aren't computer users or even have a computer. I think they could benefit from it.

Not to rag on the idea, but this is completely true in most cases.
Most elderly people I have met/had much to do with struggle with a VCR remote, let alone installing a third party firmware on a DAP (which they probably don't own/see the point in) and maintaining/using it after that.

NiHaoMike's idea of expanding it actually holds some merit, and could be used by more people than just (the probable minor percentage of the population, and even lesser percentage of RB users) people suffering from tinnitus.


[St.]

[Hirsch2k]

I also find NiHaoMike´s idea of an function generator for audio equipment testing very nice. Some pink/white noise genrator and sinus sweeps and phase testing sounds i.e. This could then be used for a range of purposes.

[funman]

Do you know rockbox has a PureData plugin?

I don't get exactly what is it since it only runs on iriver now but I think the main point is audio creation

[nls]

Hmm that reminds me, i have a plugin that preglow made a long while ago somewhere that does sinus sweeps and i think i added a whitenoise generator to it. If i find some time and motivation i might try to dig it up.

[csavery]

If you want to see some kind of medical noise generator then a programmer is going to need some reference material that describes what parameters need adjusting by the user. I would guess that a medical journal article or two on how and what works for helping tinnitus sufferers would be very helpful. Otherwise I would guess a library of suitable recorded sounds to choose from is the simplest way to use Rockbox for this. You could produce such a library using an audio editor that has generator tools.

[Ezraghast]

I have tinnitus in one ear and would be interested to see this attempted. Might be an idea to have separate and adjustable sounds sent to each channel. My own noise is a 'low hum' or hiss which varies in intensity and sometimes 'pulsates' so it seems like i can hear my blood flow. Perhaps best compared to the sound of a distant car or plane.

[nls]

I read up a little on generating pink noise etc, but you guys still haven't said *what* it is that you would want to adjust.
Running a loop that generates noise statically would be the same as playing a short pregenerated file in a loop (ignoring that the same samples repeat after a short while instead of after the RNG's cycle time, not that i think anyone can hear the difference), this could of course be integrated into a plugin but i fail to see why.

[Chronon]

If the perceptual character of the noise fluctuates then it seems that the best that can be done is to mask it by burying it in a higher level of ambient noise.  Cancellation would require you to have a waveform of the perceived noise so that you can adjust amplitude and phase*.

It seems to me like mixing pre-generated clips (Brownian noise, 1/f noise, white noise, etc.) would give a bit of ability to tune the perceived properties of the noise.

*: Cancellation usually relies on linear superposition of physical waves.  This cannot physically happen here because one of the sounds does not correspond to a physical wave.  I don't know enough about human perception of sound to comment about whether a certain spectrum of physical sound could cause perceptual cancellation or not.

[Hirsch2k]

The idea of an tone/noise generator for equipment testing is nice too 

[Ezraghast]

Cancellation usually relies on linear superposition of physical waves

I would have thought cancellation relies on OPPosition of the waves so the sum is zero?

If the perceptual character of the noise fluctuates then it seems that the best that can be done is to mask it by burying it in a higher level of ambient noise.  Cancellation would require you to have a waveform of the perceived noise so that you can adjust amplitude and phase*.

It seems to me like mixing pre-generated clips (Brownian noise, 1/f noise, white noise, etc.) would give a bit of ability to tune the perceived properties of the noise.

I think adjustable/mixable brown/white noise clips would be great! You have hit it, spot on, when you say that "the best that can be done is to mask it by burying it in a higher level of ambient noise."

I don't know enough about human perception of sound to comment about whether a certain spectrum of physical sound could cause perceptual cancellation or not.

When a tree falls in a forest....!

[saratoga]

Cancellation usually relies on linear superposition of physical waves

I would have thought cancellation relies on OPPosition of the waves so the sum is zero?

Superposition just means combining two waves.