Re: [Openeeg-list] FM EEG signal recommendations

[Oren T. <ore...@hi...>]

On Mon, May 12, 2003 at 07:16:08PM +0200, Andreas Robinson wrote:
> If we assume the EEG amplitude is 50 - 100 uV, and the bitstep is 0.5uV, we
> get 7-8 bits for representing the EEG itself. The rest is for handling
> offsets.

There is no "bitstep" on the result of FM decoding. The equivalent 
measurement would be 42-48db SNR.

Any db of dynamic range on the FM system beyond that is headroom for 
things like DC offsets. Of course, the signal must be attenuated to 
actually make use of this headroom.

> PS.
> 
> When you guys (Joe, Oren ...) are talking about dB, are you using the
> 20*log10 measure, or the 10*log10 measure? I remember Joe referring to 70dB
> as 24 bits a while back => 10*log10. I prefer 20*log10, so for me 70dB is
> roughly 12 bits.

Always 20*log10. 6db per bit of voltage resolution (6.0206 to be precise...)

   Oren