[Sidefx-houdini-list] [HOM] Subframe evaluation of PythonSOPs

Andras Ikladi andras.ikladi at gmail.com
Sun Feb 3 01:46:10 EST 2008


I evaluated the case further:

If you put a time dependent node (like what Gene suggested, a pointsop with
a minimal amount of time dependent jitter on it), that doesn't help either.
When the export runs, I can tell based on the cooking time and some printed
logs, that it only gets evaluated on the integer frame, and not at shutter
close. Also, I checked point order/count consistency on subframes and it
seems to be have as expected (not surprising as this sop doesn't
delete/create new geometry at all).
I compared the output in the IFD and looks like when writing out deformation
geometry, the motion block is exactly the same on the reference and the
deformed states of the geometry.

At the end I came to the conclusion that this might be a SOHO bug, so
submitted a report.

cheers,
Andras

On Feb 2, 2008 7:22 PM, Andrew D Lyons <tstexture at gmail.com> wrote:

> Thanks for publishing the results of all this research Andras. This
> doesn't sound like ideal behavior, and hopefully it gets some
> attention.
>
> Cheers
>
>
>



More information about the Sidefx-houdini-list mailing list