for a new specific project, I'll use DepthKit.
It produces per-pixel Depthmap (based on kinect) which is perfect.
That one will alter a mesh for point cloud style rendering.
BUT, I'll need to make a precalculation (outside of max) of that moving/altered mesh.
So it means that I would need to store the mesh vertices coordinates at each step (wow, huge) and export that as ...? OBJ ?
Ideal of course would be to NOT store that and to make all my calculations inside max (no precalculations) but not sur I can do that.
Anyway, the idea of storing that could be interesting.
I thought about story x,y,z as a rgb movie but not sure if that would be ok with all these colore compression & encoding stuff.