So i place a file cache after my white water simulation... but forgot to hide the collision object in the dopnet (thought it was just a display / viz thingy). 21 hours of white water sim for only 5 millions particles and caching over 150 frames... to find out it also cached the collision object into it. Grrrr.... what's up with that?

Cheers,

A.

It caches what you send it. How were you fetching from the dop network? The dop IO is usually setup to only get specific objects/fields.

SOP level DOPnet is outputting objects and data based on it's Object Merge tab
but default * which is all objects

its usually not a good idea to blindly be sending stuff to cache without even checking few frames, as even if it was pulling just objects you want you still may want to delete tons of attributes before caching to save disk space etc...

tamte
SOP level DOPnet is outputting objects and data based on it's Object Merge tab
but default * which is all objects

its usually not a good idea to blindly be sending stuff to cache without even checking few frames, as even if it was pulling just objects you want you still may want to delete tons of attributes before caching to save disk space etc...

Yeah, i realize it's obviously my mistake to have put the file cache right after the dopnet instead of after the dop IO import node, silly me, dunno why i skipped that step. I'm a bit distracted these days. I used the new ocean config nodes and the dop IO isn't something that seems to come along with it, had to do it manually... i dunno.

Cheers for teh reminder to both of you.

A.

Quick somewhat unrelated question here. But i'm using 19.5 new ocean config nodes, also the white water default nodes, and it looked like it was using openCL by default and i'm getting the feeling that in my case scenario it's just incridibly slow. old dual xeon here with 128GB and 4 x 1080s... should i turn that opcenCL off? Any recommendations?

Cheers,

A.

Adriano
But i'm using 19.5 new ocean config nodes, also the white water default nodes, and it looked like it was using openCL by default and i'm getting the feeling that in my case scenario it's just incridibly slow. old dual xeon here with 128GB and 4 x 1080s... should i turn that opcenCL off? Any recommendations?

sorry, I'm not familiar with the newer ocean tools or whitewater in 19.5. (last time I used it was probably in Houdini 16)