Posted by Liviu Ungur on January 11, 2016 at 14:11:08:

In Reply to: Vast difference in g-tensors when calculating with and without symmetry posted by Asha Roberts on November 20, 2015 at 16:42:35:

For such calculations you have mentioned, please clarify the following issues:

1) What is the active space of the CASSCF in both cases?

2) How many roots did you mix by the spin-orbit coupling?

In case you use symmetry, the CI roots of different irreps are optimised in different CASSCF calculations. The resulting orbitals are also different. Such issues do not appear if the calculation is done without symmetry.

I suggest you use no symmetry, and mix all ligand field states by spin-orbit interaction in RASSI.