From: Frank Murphy <fm@mr...> - 2005-08-09 23:51:12
I have been trying to coax Pymol to raytrace or output to pov
(make_pov.py) some very large surfaces and I am unable to. It works for
~33,000 atoms, but chokes on ~61,000. I would like to get up to >100,000.
Anybody have some good tips? I have tried adjusting hash_max, but I
think this my be a more basic problem.
Thanks in advance
Structural Studies Division
Cambridge CB2 2QH
LAB +44 1223 402295
MOB +44 7779 224206
FAX +44 1223 213556
From: Peter Adrian Meyer <pam52@co...> - 2005-08-10 21:17:37
> I have been trying to coax Pymol to raytrace or output to pov
> (make_pov.py) some very large surfaces and I am unable to. It works for
> ~33,000 atoms, but chokes on ~61,000. I would like to get up to >100,000.
> Anybody have some good tips? I have tried adjusting hash_max, but I
> think this my be a more basic problem.
Good tips, maybe not...but I'll give it a shot (~30,000 atoms is about as
much as I've had to use with pymol, and never tried povray export):
It could be that the conversion from pymol to povray is done in memory,
and that's exceeding the max list size that the python interpreter (or
physical memory) can handle. It might be possible to rework the script so
that the conversion is done in on smaller groups (resides/atoms maybe) at
the same time as file writing. This would depend on povray being able to
to clipping, otherwise the final rendering would come out wrong.
Of course, I could be wrong about the memory problem. When you say
chokes, does pymol slow down, raise an exception, or crash?
Hopefully somebody who's more farmiliar with povray or the specifics of
the pymol->povray conversion can comment on whether or not this is
actually likely to be the problem or not.
BMCB grad student