[daip] Re: FITLD and vis weights

[Eric Greisen]

David Wilner writes:

 > I have a uvfits file of IRAM PdBI data generated by the GILDAS 
 > software, read into AIPS using FITLD. Does FITLD scale or renormalize
 > the visibility weights? I cannot reconcile the weights with the 
 > values expected given the GILDAS documentation. (There seems to be a 
 > discrepancy of about a factor of 10.) Any thoughts?
 > 

I am assuming that GILDAS wrote a simple random-groups UV format.
This format is read literally by FITLD with no scaling.  I suppose it
is possible that GILDAS did something odd like use integer-valued FITS
- aips once had a scaling for weights separate from data in that case
but I thought we had abandoned it and certainly did not advertise it
at all widely.  If you sent me a PRTTP output of the data file
(prtlev=1) I might be able to say more.  Of course, if the weights
differ by a constant scale then that would normally not affect
anything.

Eric Greisen