[daip] IFs and Channels

[Eric Greisen]

Shea D. Brown writes:
 > Hello,
 >     I am reducing a multisource Westerbork data set with AIPS 31DEC06, and
 > I have a question.  My data has 8 IFs with 32 Channels per IF.
 > 
 > 1: Is there any way to relabel the data so that AIPS thinks I have 256 IFs
 >    with one channel each?
 > 
 > 2: If I can do this, will tasks such as SETJY and LPCAL that work per IF do
 >    the right thing with so many IFs?
 > 
 > My problem is that I need to calibrate polarization leakage terms, and
 > with westerbork they vary across a 10 MHz wide IF by about 2-3%.  Any
 > suggestions would be much appreciated.  Thank you for your time.
 > 
 > Shea Brown
 > University of Minnesota
 > Department of Astronomy

I am afraid that this will not work.  AIPS can handle up to 90 IFs but
many tables etc would have to be reformatted in order just to change
to 16 or 32 IFs and we discourage any header changes since that can
affect not only the tables but even how the data get read from disk.

Clearly this is a long-term problem to which AIPS does not have a
solution.  It is worse because the WSRT is linear polarization and so
these leakage terms affect the solution for the gains of the Parallel
hands.  AIPS' CALIB, BPASS, and PCAL would need to be combined with a
new polarization cal table to do this problem right.  Even with
circular polarizations, some leakage versus frequency will be needed
on the EVLA.  Unfortunately, we have not been able to get help with
this problem (we tried hard a couple of times) and have not absolutely
had to solve it previously.  I remind you that aips++/casa has been
supposed to replace us "any day now" for 14-16 years, which has slowed
significant developments.

Eric Greisen