search

ratt-ru / meqtrees

A library for implementing radio astronomical Measurement Equations
http://meqtrees.net
10 stars 2 forks source link

EMSS beams produce 0 at center #837

Open gijzelaerr opened 10 years ago

gijzelaerr commented 10 years ago
at 2011-11-19 15:53:39 Oleg Smirnov reported:

EMSS beams produce 0 at center

gijzelaerr commented 10 years ago

Original comment thread migrated from bugzilla

at 2011-11-19 15:53:39 Oleg Smirnov replied:

With the current MeerKAT sims, putting a single source dead center seems to produce null visibilities (or perhaps NANs? At any rate, the resulting image is empty). Perhaps this is due to something else besides EMSS beams, but in any case ought to be fixed.

at 2012-02-16 22:27:21 Tony Willis replied:

This problem is also there with using e.g. a gridded sky. I explicitly set the flux of the field centre source to 1 Jy but nothing appears, whereas the rest of the sources in the grid are present.

at 2012-02-27 21:51:27 Tony Willis replied:

Actually the gridded sky remark only applies to JPL simulations where the beams are already supplied in Ludwig III coordinates so no beam rotation is needed. When I do a meerkat sim with a gridded sky, the central source is seen.

at 2012-02-27 23:30:31 Oleg Smirnov replied:

I have a vague suspicion this may be due to a problem with their beams at 0. The reader code currently averages the theta=0 values across all values of phis -- there are, after all, 360 different values for the theta=0 point in the file, and we need to do something sensible with them all. Will you please check what their numbers are doing at theta=0? It could be a quirk of their theta-phi to Ludwig3 conversion causing the E vector to rotate at the pole, and thus averaging out to 0.

at 2012-02-28 13:21:28 Tony Willis replied:

I don't think the JPL beams are the problem. I updated emss2fits.py with an option --neg_rot_xy (short for negate rotation of XY coordinates) and when I run the JPL beams through that script the fits images appear perfect with full intensity at the field centre etc. Since emss2fits.py makes use of the EMSSVoltageBeam.py script etc the way rest of your emss system does, I would assume that at least the beams are being initially digested OK. BTW I just checked in the updated emss2fits.py script.