[daip] Change in cross-correlation visibility weights in (VLBA) UV files?

[Amy Mioduszewski]

Dear Dan,

I think this has something to do the with a bug fix in April, Eric
changed how the weights for Ipol are calculated from:

(Wr+Wl)/2

to:

4 * Wr * Wl / (Wr + Wl)  Wr-weight for RR
                         Wl-weight for LL

So that would double the the weights for the Ipol if Wr and Wl are
the same.  Does that clarify things?  You didn't say if you where
quoting I weights or not.  To check this you can run UVPRT setting
STOKES to 'RR' or 'LL' and the weights should be what you got before.

Cheers,

Amy