RMSF and FDF Faraday depths differ by very tiny amount. #178

I have two different reports of slight differences in the Faraday depth spacing between the FDF and RMSF.

From Alex Kiehn:

Because of the different length error during my rmclean3d run, I checked and found out:

CDELT3 from RMSF: 12.0458984375
CDELT3 from FDF: 12.045898

It seems like rmsynth3d uses a different accuracy for CDELT3 in the FDF and the RMSF files, which can cause different reconstructed channel shapes during the rmclean3d run. In my case 820 and 821 in the Faraday depth.
Inserting the CDELT3 from RMSF to the CDELT3 of FDF in the header solved my problem for the rmclean3d run.

From Matthew Maguire:

The phi fdf and
phi rmsf Faraday depth grids are confirmed identical in sampling
across all sources, though their nominal spacings differ very
slightly (∼5.9016 vs. ∼5.9034 rad m−2. See Section 3).