[daip] 2 Gb bandpass

[Eric Greisen]

bob at physics.ucdavis.edu wrote:
>> bob at physics.ucdavis.edu wrote:
>>> hello. once again I am trying to keep up with the new data.
>>> I got my first 2 Gb bandpass data ( 6 cm). is there a
>>> recommended procedure for reducing this data?
>>> does one do a single image over the entire bandpass or
>>> 10 images or 20 images?
>>>
>>> any advice would be welcome.
>>>
>>> bob becker
>> See Appendix E if the CookBook for EVLA advice incl wide bands.
>> I would image each spectral window separately and you may even have to
>> calibrate them that way although that probably is not needed.  If you
>> have anextended source so that RFLAG must run with UVRANGE set then you
>> will have to separate the spectral windows ("IFs").  You can put the
>> IF-separate images back together with SPIXR whoch solves for spectral
>> index and total from the 16 separate ones.
>>
>> Eric Greisen
>>
> thank you. does each IF need a separate selfcal solution?
> 
> bob

Possibly - it would depend on how much spectral index structure there 
is.  Note that with these large data sets, tasks may run very much 
faster if you run UVCOP to separate the IFs into separate UV files 
before the imaging and self-cal.

Eric