[Wien] Interstitial DOS for SOC calculation

[Falke, Johannes]

I have a converged non-spin-polarized calculation using HF and SO. I calculated the QTL file using "x lapw2 -qtl -hf -so -p". I then calculate the DOS using "x tetra -hf -so -p".

However, this results in the interstitial DOS* being larger than the total DOS, which shouldn't be happening according to the documentation about "x tetra". The same thing appears to occur for a DOS calculation with just SO, no HF, so this may possibly be the relevant feature, as it did not occur for a calculation with both HF and SO disabled.

*via line "<n_at + 1> 1 inter" in case.int. The same j_atom seems to be used to get total up/dn DOS for SO calculations, although "meaningless" for non-spin-polarized calculations according to documentation of `x tetra`. The calculated interstitial DOS is also inconsistent with being the real interstitial DOS scaled by some factor, so I'm guessing it is making this "meaningless" projected spin DOS.

How do I get the correct interstitial DOS? 

Best wishes,
Johannes