prior to maya 2011 everything worked fine, now I have the strong feeling I do a lot less work in the same time. I used to move the joints, set some keyframes and paint on deformed mesh, until I get to a result wich is good in even in the extremes.
So i have to use interactive weight normalization. Something is slowing me down, somehow the weights I paint are not the weights I want. It is so strange for me, I can’t even explain.
How do you guys manage to work with the new options?
well if you turn dual-quaternion off, set normalization mode to “interactive”, you should be in a pre-2011 mode, except that undo should be a bit more stable (don’t remember if they fixed that in 2011 or 2012).
it’s much easier to create an animation for problem poses to the area you’re painting. My technique is this: select all controls in a rig, set key on each 10 frames (0, 10, 20, …), then on every second keyframe (10, 30, 50) I add a problem pose, for example shoulders up, shoulders down, shoulders forward, shoulders backward. This way, while scrubbing timeline, i can easily go to some target pose and back to T pose. When moving on to other parts, I leave keyframe on zero, delete all the other keyframes, and create new ones.
I make sure normalization mode is set to interactive, to be honest that post thing is something I dont understand and every time I give it a shot, I return to old school mode(interactive), which works fine for me. And as uiron pointed out, creating a little animation to scrub through is really nice to show yourself and others the deformation quality.
normalization is a process where a sum of influence weights for a vertex is made to be 1.0. if you had one influence with 1.0 weight, and another one with 0.5, they’re reduced to 0.666 and 0.333.
With interactice normalization, like you’re used to, when you add some weight on one influence, weights are reduced on others. Post normalization keeps whatever you’ve painted for each influence, doing the normalization only at the time of deformation. Some general math applies here, like the fact that if you paint smooth weights for each influence (which is much easier on post-normalization rather than interactive), the resulting interpolation in normalized sum of those will also be smooth.
While this sounds good in theory, it’s still not very intuitive to use. Like a simple situation where you’d just want to paint a 1.0 weight for an influence, but instead you’ll have to go and remove weights for all other influences instead. I find my idea of skinning layers (shameless plug for ngSkinTools, heh) much more suited for situations where you’d want a fine control between separate groups of influences.
Thank you for the replies.
I knew about the middle-mouse button thing, I used it in the past, now I use a quick animation as others.
The problem was that the default is “post” normalisation, so no wonder I had wrong values.
Is there any situation when one would like to use that option?
I just watched the tutorial on ngSkinning tools. Looks great, I think I will try them in the near future.
just out of curiosity: is this handled with the maya skin cluster node, or it uses one of its own?
it uses custom node to store intermediate data for layers, but the deformation itself is handled by skin cluster.
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.