orbital visualization molden/molekel differences

[e.berardo]

Dear all,
I am trying to visualize the molecular orbitals of my systems. After running a dscf  I called the script tm2molden and visualized the molden.input both with molekel and molden.
The two visualizers don't agree on the shape of the HO-LU orbitals also when I use the same contour value. Molekel (which makes nice pictures) localizes in the LUMO some weird d-shaped orbitals on the oxygen atoms, while molden just shows p-shaped lobes (and that is what I would expect).
Has anybody ever experienced something like that?
I would use molden, but it is really hard to get nice quality pictures.

Thanks in advance,
Enrico

[e.berardo]

Hi all again,
I managed to use jmol, and the HO-LU orbitals are in perfect agreement with molden. So, I would assume that there is something wrong with molekel and its interpretation of the molden.input created through tm2molden in turbomole.

Enrico