All you need to do is have one model (for each LOD), the "base
model", with the texture. For the frames, all it needs are the
vertex positions. That's basically what's stored in the animation
file (plus the base models). I haven't released the version that
tweens the normals. The current version only tweens the vertices,
but normals would need to be tweened as well for the lighting to
be correct. I'll try to get a stable release soon.
You would apply the texture to just one model (for the base
model), then, you would batch convert all the other frames without
the textures. Set the base model to the textured model, and add
the frames (the models without the texture).
I see. Cool.
Perhaps you might consider having the tool handle skeletal
animation as well. I imagine it would save loads of memory.
That's much more difficult to do, plus it would need to support the
loading of skeletal data. With tweening animation, each frame uses
anywhere from 10-20k of extra memory. Usually you'll only need about
5-10 FPS for an animation to look like the original.
So really you'll only save about 20k for each frame of animation
(assuming 2500 vertices in the model). For a walk animation, you'd
likely need about 5-10 frames. That's basically about 100-200k of
memory. In total (assuming 10 animations), you're probably saving at
most a few megabytes per model which is not very significant. If you
copy the animation model by code within RB, it will not take up extra
space (except 1 separate vertex data to keep the copy separate from
the original).
_______________________________________________
Unsubscribe or switch delivery mode:
<http://www.realsoftware.com/support/listmanager/>
Search the archives of this list here:
<http://support.realsoftware.com/listarchives/lists.html>
|