search

ratt-ru / bullseye

A GPU-based facet imager
GNU General Public License v2.0
1 stars 1 forks source link

Trying to generate an image with widefield distortions #13

Closed bennahugo closed 10 years ago

bennahugo commented 10 years ago

Have tried 2 and even 20 degrees, but to no avail.

I have ensured w-projection is off. I'm using the Turbo-sim TDL script, input a model fits file and ask it to write out into the DATA column of the Measurement Set. When I then image it there are no distortions.

I'm using simple point sources across the image (see the QA Wiki).

Where am I going wrong? Maybe we should just use a dataset with known distortions?

bennahugo commented 10 years ago

Okay so I've finally gotten hold of a reasonably sized EVLA observation. I've been having issues with the UCT network timing out on my connection to the NRAO server in the US and I finally gave up and downloaded the measurement set I needed at Rhodes :(. I'm planning on following this tutorial: http://casaguides.nrao.edu/index.php?title=EVLA_Wide-Band_Wide-Field_Imaging:_G55.7_3.4 as soon as I copied it across

o-smirnov commented 10 years ago

Ahh, but with a model FITS file you will not see distortions because the FITS forward predict does not take w-projection into account. So the predict is wrong, and the imaging is wrong in the opposite way, so the result comes out without distortions.

Forget the FITS file -- use the gridded_sky sky model, and ask for a grid of sources. Say 7x7 at 30 arcmin spacing -- then you'll immediately see distortions.

bennahugo commented 10 years ago

Thanks Oleg. Will try out now. Actually also just managed to get the dataset calibrated in casa... respect_for_astronomers++

bennahugo commented 10 years ago

Thanks that worked very well :)