I have been thinking that the problem may be related to corruption of
the first few data points. It should be possible to correct for this using
linear prediction. I haven't a clue how to do this (using UXNMR 970101.3).
Our documentation must be for an earlier version of uxnmr because some of
the parameters listed in the docs aren't there and the command to do the
operation doesn't exist either. Can anyone help out with the necessary
parameters and commands for uxnmr 970101.3?
Why should this suddenly become a problem? The only hardware difference
is a new (used actually) array processor that was installed at about the
same time as the baseline problems started showing up. Is there any other
reason that you can think of for this problem to occur?
Keith Brown
Department of Chemistry
University of Saskatchewan