search

EoRImaging / FHD

Fast Holographic Deconvolution
BSD 2-Clause "Simplified" License
20 stars 10 forks source link

Bug transferring polarized calibration #267

Closed rlbyrne closed 3 years ago

rlbyrne commented 3 years ago

Transferring calibration does not return the same polarized image results.

Here is the dirty Stokes V image for a calibration run: 1131454296_uniform_Dirty_V

Next, I took transferred the calibration from that run and re-imaged. Here is the new dirty Stokes V image: 1131454296_uniform_Dirty_V (1)

The difference is subtle, but the two are certainly not identical, as they should be. The transferred calibration run has more power in Stokes V and the appearance of large-scale structure.

We suspect this is because the calibration transfer is not transferring the crosspol phase.

rlbyrne commented 3 years ago

I reran with the fix in branch fix_transfer_cal_crosspol_phase (simply commenting out line 46 of vis_calibrate). The dirty Stokes V now matches the calibration run. We need to revisit PR #268 to implement this correction in a better way.