RE: [alma-sw-ssr] Paris meeting agenda

[Steve Scott <scott@ovro.caltech.edu>]

Brian Glendenning wrote:
> 
> >> 8) Correlator data processing
> >shouldn't this all be done in post-processing?  just save the lag
> 
> Sorry if I've missed follow-ups in my email backlog. Is there any comment on
> this suggestion of Bryan's? My assumption would be that we might want to
> throw channels away before archiving, which requires spectra rather than
> lags, but if I'm wrong and we can push the problem off to post-processing it
> would make our lives (somewhat) easier for control software.

I thought Bryan Butler partially answered his own question in his next line:

>>i guess the argument is in data rate though...

And also a general philosophy that the archive contains science data, which 
means spectra at the requested resolution of the proposal. Bryan's next 
paragraph proposed nice ways for the experienced PI to specify the lag 
apodization functions. Yes, the PI should be in control, and there should be 
sensible defaults for the inexperienced user that help avoid complexity and get 
good results.

BTW, time constraints did not permit discussion of this topic in Paris, so we 
will probably want to pursue it through email.

Steve