[Radiance-general] oconv -r and obj2mesh -r parameter?

[Greg Ward]

Hi Lars,

If I understand your question correctly, you do not need to worry  
about the difference between the minimum voxel size set by -r in  
obj2mesh and the minimum voxel set in oconv, as they are unrelated.   
Oconv incorporates the mesh created by obj2mesh, as well as instances  
created by other oconv runs, as transformed cubes in the scene, and  
does not subdivide them beyond that unless they intersect other  
instances or meshes or geometry.  If they intersect each other too  
much, oconv can take a long time to resolve them.

-Greg

> From: "Lars Grobe" <grobe at gmx.net>
> Date: May 16, 2005 5:51:39 AM PDT
>
> I have a question regarding the obj2mesh tool. So far I always used
> instances generated by oconv, and the -r setting was always the one  
> that I
> also used when assembling the whole scene (the same with -n). Now I  
> use
> obj2mesh, and a question comes to my mind. If I follow the rule -r >
> bounding cube / smallest face, I can use a much smaller -r for  
> obj2mesh then
> for the finally assembling oconv. However, does that make sense? Or  
> will I
> produce trouble with differing resolution settings for instances,  
> meshes and
> the complete scene?
>
> The reason for my question is that the model I am working on is  
> currently
> growing a lot, and I a getting trouble with really large oconv  
> processes
> when assembling the scene ( > 1 GB) or, when using the -n  
> parameter, slow
> renderings...
>
> TIA+CU Lars.