[daip] LTESS question

[Eric Greisen]

Lawrence Rudnick writes:
 > I must be doing something silly, but when I run LTESS on two (processed) 
 > images from the NVSS  (after first having hgeom'd them to a larger 
 > template that I want to place them on, and having the external areas 
 > magick blanked), it normalizes them by dividing by 2 everywhere, and 
 > then adding the maps together (so that the overlap region is higher than 
 > the non-overlap regions).  In other words, it is not averaging only in 
 > the overlap region, as it should.
 > If I put in noise values, it simply sums in the overlap region, and 
 > makes the values extremely high in the non-overlap region.
 > suggestions?
 > (31DEC06 version).        thanks.  lr
 > 

I guess I should put LTESS on the check it out list.  It is one of the
Cornwell-based (VTESS et al) tasks that assumes that the output image
is going to be POINTS as the usual image from VTESS may be.

Note that LTESS and FLATN are not really useful for NVSS images.  Both
assume something about correcting for primary beam, but the NVSS
images have already been corrected (and multiple pointings merged
properly a la FLATN).  You probably simply want COMB with OPCODE
MEAN.

Eric Greisen