[daip] OHGEOM and OGEO

[egreisen]

On 2021-01-11 12:37, Farhad Zadeh wrote:
> Hi Eric,
> 
> I have been comparing MeerKAT and VLA images and have been converting 
> from RA, DEC to l,b coordinates using the above two tasks. MeerKAT
> image is a mosaic and is huge, more than 8k by 8k. After doing lots of
> work, realized that astrometry is not good. I was curious if aips
> provides a better talk for conversion and rotation? What is the best
> way to do? Break the image into smaller size?
> 
> Also, I have made a comparison for a recent 6cm data. I uvsorted the
> data to galactic and then imaged. I also made  images in RA, DEC  and
> then rotated. Which procedure  is more correct astrometrically when
> using  4kx4k images of  6cm data taken in the A-array?

OHGEO and OGEOM are the best way to regrid images and are as accurate as 
possible
with the geometry.  I would not know about the "astrometry is not good" 
business.

AIPS should not allow you to UVSRT to l,b coordinates.  IMAGR does not 
require a
UVSRT and the u,v coordinates with the data are computed with respect to 
RA and Dec
only.  Imaging in RA/Dec and then converting to l,b is the only path 
that makes
any sense.  I would worry that really large images in the A array might 
have
wide-field issues that would suggest using multiple facets.  However, I 
would then
probably use FLATN and then CELGAL and OGEOM.

Eric