The HLP file for CALIB claims that SMODEL(4)=2 specifies a optically thin sphere for the model. I believe it should be SMODEL(4)=3 (assuming its the same as VISDFT). michael b