[daip] Re: Help with AIPS (please!)

[Eric Greisen]

Andrew Fletcher writes:

 > I am using AIPS for the first time and cannot get the task IMAGR to work
 > with missing spacing flux in ZEROSP. I have two sets of 3cm VLA
 > observations of the galaxy M51 and have calibrated and SPLIT the UV files
 > separately and then DBCONed them into one file. There are two IFs.
 > 
 > I expect the zero spacing flux to be about 1 Jy/beam but no matter what
 > I enter as parameters in ZEROSP there is no change in the cleaned
 > image. For example I have tried ZEROSP(1) = 0.5, 1, 10, 1000, 1000000 with
 > weight ZEROSP(5) = 5, 50 for NITER = 100, 1000, 10000 all to no avail. I
 > think one of the parameters in IMAGR may be switching off ZEROSP or making
 > it ineffective but can't see which one for the life of me. I have read the
 > Cookbook and explain file but am not much wiser.
 > 
 > The parameters of IMAGR for a typical effort are below. Any advice you can
 > give me would be greatly appreciated.
 > 

The zero spacing flux is an over-sold commodity.  When you add a
sample at u,v = 0,0, it changes the DC-level of the image by a little
and that is all.  In your case you have 1024x1024 pixels, with
probably a 20 pixel beam area, so to have the flux integrate to 1 Jy
over your field you would only get about 20/(1024x1024) change in the
DC level.  Actually it might be worse than that since I do not know if
the integral needs to be 1 over your field or over all space.

Run IMAGR with no ZEROSP and with ZEROSP set but with NITER=0.  Then
run COMB (opcode 'sum'; aparm=1,-1) to compare the 2 images.  You will
find an essentially constant (except at the edges) difference but it
will be very small.

What one really needs to fill in things is all the very short but not
quite zero spacings.

Eric Greisen