[daip] Overlapping Fields in IMAGR CLEAN

[Eric Greisen]

landon at inaoep.mx writes:
 > I have been experimenting with the use of multiple fields in AIPS's IMAGR. I
 > have found that running IMAGR with 2 fields, field one with boxes on the areas
 > where emission is expected, and field two with a box over the central 25% of my
 > 512x512 pixel image, results in the best images (ie: most extended structure,
 > almost no artefacts, low noise). Both of my fields were set to size 0 (ie:
 > IMSIZE-10). As this is likely the procedure I will follow with all of our
 > cleaning of these observations, I need to understand better what AIPS is doing. 
 > 
 > >From what I understand from the AIPS manual, AIPS is simutaneously searching
 > both fields for components and then subtracting the componets at the end of each
 > major cycle. The clean box in field 2 full encompasses the clean boxes in field
 > one. Does this mean some areas are being "double cleaned"? The AIPS CookBook
 > also says "...never allow the search areas of one field to overlap with the
 > search areas of another field." However, this is exactly what I have done. Why
 > is it resulting in the best image I've been able to create thus far?
 > 

IMAGR in mode OVERLAP = 2 (or greater) avoids the double cleaning by
doing one field at a time.  MX and IMAGR with OVERLAP 0 or 1 will
double clean and it will take a very long time to converge.  The
CookBook may not address this subtle point properly.

Eric Greisen