>You can either set the MAX L IN WF in in1 (limiting it to 2 will give large >discont at sphere). That is what I have done first. But there are non-zero values for |<f|psi>|² in the qtl file (and then non-zero f-DOS). Why? > Or set the linearization energy of the unwanted l-states to >some large value (20 Ry or so) Regards Laurent