[Bf-funboard] ideas for the new animation system
hidarikiki at gmx.net
Tue Nov 30 15:07:36 CET 2004
one thing that blender still lacks a bit is the connection and flexibility
of it's animation tools. The interaction can be improved.
Maybe most of this is already planned I dont have much time right now due to
personal problems, so please don't judge it too harshly.
Just some ideas for that:
Lattices and the new curve deformation could maybe work with vertex
groups. The problem right now is that one can only use a single lattice or
curve to deform a mesh. The curve deform would be great to do ,for example,
a squid, especially since the length of the mesh is preserved (good job).
But how, if the squid has to have a single mesh for each arm?
Maybe deformations could be passed along (maybe a pass button or a bev obj
equivalent) to keep blenders policy of a single parent, which I find
On the other hand, it may be possible to put several lattices into one
object and use the vertexgroup approch. The same goes for curves.
Would it be possible to simplify the rigging of curve deformation the way,
armatures are handled right now? Would that make a pose mode for curves and
A huge obstacle for combining animation methods is the fact that blender
when entering editmode always jumps into the default position of a mesh,
although the mesh is already rigged to something and deformed. That makes
modelling complex muscle behavior or good facial animation with RVK quite
hard. It wouldn't be that hard if entering editmode of an Alt-D copy of the
mesh wouldn't affect the rigged twin.
Another idea are shared verts. If you use the mirror modelling combined
with subsurfs, you experience a seam, which only disappears if you join the
meshes and hit rem doubles. If blender had the ability to compare two vertex
groups and internally join the partner verts (defined by a perimeter like
with rem doubles), it could treat two meshes like one and deform them
correctly without unwanted seams. This may also be a solution to connect
several lattice- or curve deformations (squid problem).
Of course the presentation of vertex groups would then have to be more
accessible. Maybe the outliner can provide that.
On the whole, I hope that the new animation system will concentrate on
simplicity. If something can be achieved by a combination of simpler methods
(depends on the amount of pain to achieve it), a highlevel tool is often
unnecessary. I doubt that a large amount of similar constraints for specific
purposes is very useful (no offense meant).
Sorry for not doing a full though through proposal.
More information about the Bf-funboard