[daip] DBCON error message I can't decipher

[Eric Greisen]

Cornelia C Lang writes:

 > I wonder if this is the case (i.e. the weights issue) because when
 > I combine the D + C data, it's as if the dbconned dataset doesn't
 > "know" about the C array data and still has a fairly low resolution.
 > 
 > How do I tell about the weights and what can I do about it -
 > recalibrate? or resplit?
 > 

You can just run PRTUV and look at the weights (be sure to check the
scaling of the display which is given near the top of page 1).  The
weights in the old system tended to be numbers like 1.0 (for 10 sec),
2.0 for 20 sec etc.  The new numbers are all over the map and try to
be 1/sigma^2 in 1/Jy^2 - so with a noise in the vis of e.g. 10 mJy the
weights would be 10000 or so.

The choice of weighting scheme unfortunately is made with FILLM.  One
does have to be careful with DOCAL if one has the old weight system -
one must NOT calibrate weights if they are the old form.  In 31DEC07,
DOCAL must be some magic number (100) to avoid applying cal to
weights.  DBCON does have a weight scalingfactor so that you can just
scale the weights of one data set to make them more like the other.

Note - if you have all your cal tables in the C array data - you could
re-FILLM and then copy all the flag, CL, etc tables and resplit
without having to do the flagging etc over again.

Eric Greisen