[daip] SPLAT funniness

[Eric Greisen]

Joseph Lazio writes:

 > I'm running 31DEC00.  I've been running the MNJ, and my version is
 > current upto late November or early December (whenever access was shut 
 > down in preparation for the rollover).

     BTW - the MNJ was shut down for only a couple of days.  It can be
rerun so long as you now call it NEW.  31DEC01 may also be installed.

 > 
 > I have a two-IF, P-band, single-source u-v data file.  It is the
 > output of FLGIT.  If I run SPLAT on it---with BCHAN=1, ECHAN=0, BIF=1,
 > and EIF=2---the visibility data in the second IF are clearly wrong.
 > They are measured in kiloJanskeys, but the data in the first IF look
 > reasonable (look like a resolved source with a flux density ~ 100
 > Jy).  If I SPLAT off the individual IFs, they look reasonable.  If I
 > set BCHAN=0, BIF=0, and EIF=0, the visibility data in both IFs look
 > reasonable.  If I set BIF=1, the visibility data are screwed up
 > again.  
 > 

     Can you make the data set available - on an anonymous ftp
location at your site or (2nd choice) on out ftp.aoc.nrao.edu incoming
site?  And can you show all the inputs - are you averaging in time,
channel?  or why are you running splat?

Eric