[daip] Re: RSTOR in AIPS

[Eric Greisen]

Teddy Cheung writes:
 > Hi Dr. Greisen,
 > 
 > I am hoping for some advice on proper implementation of RSTOR.
 > 
 > I have a series of vlbi images with different beams. I would like to
 > restore to same beam and add.
 > 
 > My inputs are:
 > 
 > IN2NAME: just each CLEAN image with the cc table attached.
 > 
 > The results vary with different INNAME.
 > 1. If I use INNAME=IN2NAME, the restored image looks to have the beam I
 > specified (circular, rather than elliptical from original file), but the
 > image stats seemed to have doubled (maximum, rms noise...from imstat).
 > 
 > I am tempted to just divide the RSTOR product by two.
 > 
 > 2. I tried to get around this by producing a "blank" image, basically
 > taking the file and subtracting it from itself with comb.
 > The RSTOR product, though the Max looks reasonable, and the rms about 30%
 > higher (i restored with larger beam), but then the Minimum is reduced to
 > some really small number (few e-9 Jy).
 > 
 > Do you have any advice how to do this? I searched google for AIPS RSTOR
 > and came up with some documents I wrote when I was an undergrad which
 > implemented (2) ... and it seemed to have worked then.

RSTOR does not subtract the initial set of CCs with its beam from the
image.  It is intended to be used with residual images or e.g.
images from VTESS that need some point components added to it.

Eric Greisen