Hello polycount community, curious if someone of you has the technical background knowledge to answer this.
I'm currently working on a rather big personal project of a futuristic sniper rifle with a lot of moving parts. Workflow is that I will render both the highpoly in some raytracer, probably Redshift, and create a pbr textured game model, both versions animated. So I need a game ready rig anyway.
The thing is the highpoly will have a shedload of polys, probably around 15-20 million, and I wonder whats better for performance.
I could "rig" the highpoly within maya just with groups with a proper hierarchy and the right pivots, so no skin clusters necessary (all parts are completely mechanical, nothing has to deform in any organic way). But since I need a joint based game rig anyway with the same setup as the group-"rig" I could only create the game rig and skin the highpoly to it like the lowpoly.
My thoughts are now that I believe that the highpoly animation would perform faster without skin clusters from the joints. That it's faster to compute the translation and rotation of groups and their assigned objects than appling the translation and rotation of a joint to its skinned mesh vertecies, right?
Just a technical thought out of curiosity, or is it bad practice to create a non-joint based rig for cinematic animations anyway?
Thanks for the read! Have a nice day everyone! - Martin
Replies
Having a proper rig means you can reuse it on your game model as well.
Basically, make a proper one