PDOS atomic contribution for SOC don't match expectation & comparison to calc w/o SOC

I am calculating the Bands and PDOs of 2D WS_2 with a negatively charged carbon substitution on one of the Sulfur atoms.
I first calculate "atomic relaxation" without SOC and get following PDOS

As expected, the in-gap state resulting from the impurity have atomic contributions from C,S,and W.
However, when calculating the same structure with SOC ("structure as" is, taking the resulting struc from 1), I would still expect the same contributions, but the In-gap states show no contribution from C in the PDOS anymore.

Varrying PDOS smearing and the energy grid spacing or the supercell size seems not to have any influence on this result.
Example calulation for the SOC calculations are:
Anouk Droux
2742 (for 771 supercell) , 2326 (for 661 supercell), 3189 (less degauiss smearing, smaller E-grid)

Thank you for your help or ideas