From: Rossmanith, Peter Dr. (ZKM) <PETER.ROSSMANITH@MSM.BASF-AG.DE>
  To  : 'rasmb@bbri.harvard.edu' <rasmb@bbri.harvard.edu>
  Date: Mon, 1 Jun 1998 12:29:43 +0200

wierd fringe displacement

Hi RASMBers,

I was just thinking about an idea of altering the dcdt analysis for
interference data. Walter brought up something that sounded quite
reasonable: If you take the fringe displacement at the upper plateau,
this is supposed to be a value that decreases because of dilution - in
the simplest case. Due to the procedure of Fourier analysis this value
may jump because of black areas at the meniscus. But these jumps should
be of integers! (Is this correct?) So if you subtract suitable integers
from the n+1st file you should end up with the "correct fringe"
displacement relative to the n-th file. Especially the lower plateau
should be identical. This would have the advantage that the
determination of the "0"-position in dcdt would be redundant, as it is
for the analysis of absorption data.
So far about theory! I tried to do this with some of my data and got
something that is against theory: The shift from one file to the next
was less than 1 fringe but way too much to give the "correct value". I
have no idea about the reason! Does anyone have a suggestion? Every hint
is appreciated!


Thanks

Peter Rossmanith



Peter.rossmanith@msm.basf-ag.de