On Sun, 19 Feb 2012 17:53:37 +0000
Fons Adriaensen <fons(a)linuxaudio.org> wrote:
On Sun, Feb 19, 2012 at 04:11:55PM +0100, Renato
wrote:
I'd like to understand how the sensors work,
but his site doesn't
seem to go much in depth on this... it says the sensors register
the sounds of the undercutaneous muscles. But are those even in the
audio frequency realm? And how would you capture them, with a piezo
contact mic?
There was a paper on this at LAC 2011, see the proceedings.
<http://lac.linuxaudio.org/2011/?page=download>
again, thanks Fons. For the lazy, the actual sensor device is described
at page 160 (170 of the pdf) of the proceedings pdf [1]. It's basically
an electret condenser microphone, kept in place by a silicon case that
should also filter out external noise. So he's (knowingly) leaving out
part of the spectrum, since as he describes earlier in the article
muscle sounds are in the 4-45hz range and those mics go down to only
about 20.
I must say I'm quite surprised to learn muscles make loud enough
sounds to be used this way... and this project of his looks freaking
cool IMO.