05 May 2006, 12:40 AM
No it's not a stupid idea, and no, it's not possible out of the box.
It's not particularly hard to write as a tool (I know, I have it in my toolbox :p ), but the solution isn't as trivial as you might expect when you work with weights.
A trivial but expensive solution is having two meshes deformed by the same deformers but with different weightmaps, and biasing a third between the two (or more), which is a simple vectors averaging, but it can be fairly dataheavy as you can imagine.
The alternative solution, of working weights-side, while not being anything phenomenally hard, isn't as simple as it may sound. The example you're thinking of is probably just a few bones and all points are already weighted to those bones and only need shuffling, but blending between very different sets of deformers can turn out unpredictable very soon.
I know it took me a while to get something that would average satisfingly for more then just an arm bend, and it's not particularly snappy.
Thanks Raffaele. Your reply was vey helpful. It's good to know that I'm not crazy for having the idea in the first place.
05 May 2006, 01:13 AM
This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.