[daip] imagr

[Eric Greisen]

Shea D. Brown writes:
 > Eric,
 >    Thanks for the help.  I changed the CH WIDTH (column 3) in the FQ table
 > from -1.5625e5 to -3.125e5, and it fixed the problem.
 > 
 > I used SPLAT to apply the Hanning taper, and because we were using DEC 04
 > at the time, I had to use SPLIT to select every other channel and DBCON to
 > put it all back together again.  Not sure where the problem happened.
 > Does CH WIDTH in the FQ table mean the channel spacing, or the actual
 > channel width? If it means the later, it's strange that IMAGR uses that
 > number and not CDELT in the frequency axis of the header, which was
 > correct after SPLAT and SPLIT.
 > 
Great and thanks.

We used it as a channel spacing with total bandwidth = Nchan * spacing
although the total is never used.  But each IF could be different
(GMRT does this!) so we have to be careful.  The header can contain
only one number and so cannot handle this complexity.

Eric Greisen