[alma-config] CLEAN and difference results for B arrays

[Mark Holdaway]

I am looking at the MARS images in B array.  The image quality
is largely a function of short spacings (or lack thereof), and
is not necesarily indicative of the general superiority of
one configuration over the other.

Bryan,

How would you view starting with an initial disk model
for Mars?  We know how big Mars is, its integrated flux can be optained
from single dish measurements, and the astarting assumption made of
a uniform disk of the brightness obtained from the single dish
measurement.  Details, such as 'how edge darkened or edge brightened',
are known in the measured visibilities, and the deconvolution will
adjust the starting model so that it agrees with such details.
(This gets us out of making a more complicated Mars model.)

It seems that the generation of an appropriate initial disk model may not
be straightforward, as you probably need to diddle the edge pixels
and all, and then convert it to clean components for AIPS.

In the C array, the zero spacing problem is not so bad, but still present.  
More damning is the ringing caused by the PSF's sidelobes interacting with
Mars' edge.  Again, a disk starting model would improve the situation
greatly.  I think it is at this stage (ie, getting the ringing down) that
knowing the details of the limb brightening or darkening in an initial
model would really help.

--------------------------------------------------------------------------

On another track altogether: some of the imaging that has poor results
can probably be helped significantly by including a mask indicating
where you can and cannot extract clean components for the model image.
(I haven't seen any use of masks in your scripts on the web.)

	-Mark