[evlatests] WIDAR bronze dataset - C band dual pol.

[Vivek Dhawan]

This data is of lower quality than Rick's 'golden' dataset,
despite being vanilla, C band, RFI-free. Is it the correlator
or fumbles in post-processing? Can _you_ un-stump the chumps??

---------------------

July 29 UT 14:20-16:30, 11 antennas, 2 Stokes RR & LL only.
2 x 128 MHz subbands at 4808, 4936 MHz. 256 chan of 0.5 MHz per
subband; 1sec integrations.

3C48=0137+331, 0119+321, 0057+303 cycling every few minutes.

The Super-Short Summary:

__ NO dropouts. No symmetric artifacts.

__ LCP raw data has wider range of uncalibrated amplitudes across
     antennas, but steady in time and calibrates out OK, finally
     only 10-20% worse rms than RCP. (Ant 3,IF2,LCP notably bad)

__ Using 5 MHz channels, the rms of the central few channels is a
     factor 2 worse than theoretical=0.2mJy. Edge channels worsen
     gradually by another factor ~2. This is true of R & L.

__ V image cubes (5 MHz) are a bit better, compared to theoretical
     they are a factor 1.5 worse in middle, factor 2 at edge chans.

__ Averaging up to 105 MHz, the image rms should improve by a factor
     4.6. It improves by less than 2. The artifacts look more-or-less
     like the dirty beam.

---------------------

Datasets - If you want to grab a copy:

A) original: 5 GB
/home/casa-dev-16/mrupen/WIDAR0/3c48c.55041.596322349535.uvfits

B) parboiled: 1.6 GB    /home/aakash2/AIPS/DA01/FITS/WJUL29C.FITS

Flags + 1 delay, 1 bandpass both time-invariant on point-source
of 5.5Jy; then selfcal on 3c48 (~6s solint), split out with all
channels intact; calibration applied. Makes reasonable images,
about a factor 2-4 from theoretical. Plots available if req'd.

Averaging up in frequency does not give the expected improvement,
and at this point the artifacts are easier to see and look more-
or-less like the dirty beam. Is this a a signal of time-variable
errors not calibrated out? So I tried time-variable BP - see next
item - not much help.

C) hardboiled: 0.3GB  /home/aakash2/AIPS/DA01/FITS/WJUL29C-AV.FITS
data of B) averaged up to 5MHz channels, self-cal using central
5.5Jy src + 25mJy sidekick 2' away. BP every 3min, using same model.
Split with all cal, BP applied, central 21 chan (105 MHz) retained,
edges of 128 MHz dropped. Images are a bit improved compared to (B),
not quite a factor 2. Still a factor 1.8-3 worse than theory, even
for the central channels. Averaging in f does not improve as expected.

---------------

Details of Processing sequence:

QUACK, TVFLG, CLIP to remove off-src times. NO dropouts were present
of either the zero or low amplitude type, other than when antennas
were slewing, or some antennas were off source for ~30min at various
times, flagged those as well.

SETJY: force 3C48 to 5.50 Jy both subbands. (Other sources come out
        at the known fluxes, 1.5Jy for 0119+321; 0.6Jy for 0057+303)

FRING: 1 scan on 3C48 for delays, zero rate, apply w CLCAL to all.

BPASS: 1 bandpass for all time

CALIB: 6s solutions intervals on 3C48, apply to all sources.

SPLIT: averaged 10chan (5MHz), 21 chans out i.e. central 105 MHz.

IMAGR: Subbands and RR,LL were kept separate. Images were made with
        (a) single 5MHz channels. (b) averaged 105 MHz.
        So, for 3C48, 4 cubes of images were made with 21 chan each,
        plus 4 images with 105 MHz averaged. Also V image cubes.

RR amplitudes are more consistent than LL, before careful flagging
and calibration. After calibration the RR, LL differ little on a
baseline vs. time or baseline vs. frequency comparison. At this
stage the data look fine. No baseline is dramatically worse than
another.

--------------