> It seems to me that if there is no internationally accepted
definition
> of signal to noise ratio in powder diffraction, then let me suggest
> that
> this forum might be a way to define it once and for all. How about
> this?
> SNR = (P-B)/SQRT(B).
>
> Of course this would only give a value for a single peak. I suppose
> one could also define a global SNR as a sum over all
> Global SNR = sum over peaks (P-B)/SQRT(B).
>
> Does this make sense?
> Mike
A more overall value would be a normalized sum of the ratios
(Ip-Ib)/sqrt(Ib), where Ip and Ib are the profile intensities of the
pattern and the background for each datapoint.
Leonid
____________________________________________________________________________________
Looking for last minute shopping deals?
Find them fast with Yahoo! Search.
http://tools.search.yahoo.com/newsearch/category.php?category=shopping
