[daip] Aips calibration seems to have changed

[Eric Greisen]

Julie Hlavacek-Larrondo wrote:
> Hi Eric,
> 
> Many thanks for your quick reply and for forwarding the email to others.
> The idea that it could have been related to the flux model updates is
> interesting (although we haven't ran a MNJ, so they should not have
> changed..?), and it really looks like something similar has occurred.
> 
> 1. Unfortunately, it doesn't look like its related to the flag tables.
> I've written down in my log book most of the data that I've flagged, and
> the flag tables seems to have been properly applied. The way the current
> data look like seems to suggest that what used to be good data is now
> filled with bad data. It really looks like something in the calibration
> has changed.
> 
> 2. I only have backups of the BP tables, and when comparing the old ones
> to the new ones, nothing stands out. They old and new tables both look
> quite similar. I've attached a ps file that shows these tables for
> different dates (May 20 which is before the change and June 16-17 which
> is after the change). I've also looked at the current "new" CL tables,
> and they dont look particularly good (which is not surprising
> considering that there's a bunch of new bad data), but nothing really
> strange stands out.
> 
> Any other ideas/pointers are more than welcomed.

I am confused some and need some details.  The SN files rather than the
CL are useful to look at - the CL are interpolated from one or more SN 
files.  The flux model updates are just a change of the absolute scale
and so should make no difference to the cal quality.

What sources are your primary calibrators?  Did you use a model (image 
+CC) for them?  From your descriptions I do not see how the data are 
changed by anything other than flagging and possibly the BP tables.
Have the adverbs you are using in CALIB changed (e.g. ICHANSEL)?

Eric Greisen