[daip] AIPS question regarding CVEL

[egreisen]

On 2020-10-30 10:44, Rebeca Soria via Daip wrote:
> Hi,
> 
> I'm calibrating 43GHz spectral line VLBA data with AIPS DEC20 (and
> tried also DEC19) of
> 
> our recent project BS285A. When I run CVEL, everything seems normal but
> 
> when applying the BP table, I plot the data and something goes wrong 
> with
> 
> the bandpass. I'm sending you an example in the two files, you can see 
> that
> 
> the original UVDATA (bs285A-SINCVEL.eps) looks fine applying the 
> bandpass.
> 
> Once I run CVEL, the bandpasses are not well applied 
> (bs285A-cvel+bp.eps).
> 
> The same problem appears if in CVEL I don't apply the BP table, and 
> then with
> 
> POSSM I plot it, so I think it is something related to the CVEL task.
> 
> I've also checked the SU table (maybe for something wrong that I may've 
> done
> 
> within SETJY), but I think the frequencies are set ok (yellow in
> bs285A-SU-table.pdf).
> 
> I'll be very grateful if you could help me solve this mistery!
> 

CVEL works by doing an FFT on the spectrum.  If there are features in 
the spectrum that are very narrow and the shift is not an integer number 
of channels (which is the usual case), then this sort of ringing occurs. 
  It would have been helpful to look at the phase as well which should 
also show this ringing.  Consider APARM(9)=2 in CVEL.

Eric Greisen