[daip] AIPS memory allocation / Re-defining phase-center {External}

[egreisen]

On 2023-03-06 03:22, Felix Pötzl via Daip wrote:

You could shift the position with UVFIX which I believe is the most
accurate way to do this.

Eric Greisen

> Hello,
> 
> I have had some trouble with AIPS dealing with datasets with small
> time and frequency spacing. I am working on some archival data set now
> with 0.1 s integration time and 62.5 kHz channel spacing, where
> sources are sometimes expected to be found several arcseconds from the
> correlation phase center.
> 
> To find the source far off-centre, I set large windows in delay and
> rate space while running FRING. The output that AIPS gave initially
> for too large windows and/or short solution intervals looks like this:
> "FRNSRC: DID NOT GET WHAT I NEED FOR MEMORY" and then proceeds to ask
> for shorter solution intervals and/or smaller windows for delay and
> rate. I have set the maximum allocatable memory with SETMAXAP, which
> now allows for larger windows to be tested.
> 
> However, I am still wondering if there is a more elegant solution to
> alleviate the memory issue and dealing with very large delay and rate
> windows. Is it possible in AIPS, post-correlation, to re-define the
> phase centre that is used in the fringe search? In this way, one could
> use potential prior information about the source position to use
> smaller windows, and thus make the fringe-fitting process faster and
> increase the SNR of solutions.
> 
> I would be very thankful for any feedback on this. Maybe I am missing
> something obvious.
> 
> Best regards,
> Felix Poetzl