[daip] BLCAL Woes

[Rick Perley]

	Gentlemen:

	Something is weird with BLCAL.  

	I have a 3C273 database, and am seeking the ultimate dynamic
range (what else!)  I used BLCAL in two ways.  Both used the same 
data, and the same model.  

	a) I run BLCAL on a single-source database, then use SPLIT to
apply the corrections.  

	b) I convert the database to multi form with MULTI, run 
BLCAL, then SPLIT to apply the corrections.  (The reason for the multi
format is to permit a PCAL to be run).  

	The results are DIFFERENT!  Result (a) improves the map 
beautifully.  It's exactly what Dr. Rick wants.  Result (b) is *much* 
worse.  But it's worse in a very peculiar way -- not at all as if the 
corrections went in 'backwards'.  Mostly, the larger angular scale
visibilities seem to have been increased.  

	Is there any simple explanation?  A correctable error?  

	Dr. Rick