I've built a poly mesh for a human head/face, and the intent is to convert it to a subdivision surface. With regard to deformers for animation, I have a few questions:
1) Is it best (or easier) to bind the deformers to the original poly mesh and let the 'history' of the subdivision surface work for you?
2) Or, is it best to bind to the poly proxy mesh? This method would be similar to the above method, except I don't believe you need to retain the 'history' for this method. However, I can only seem to display the poly proxy mesh as a wireframe view, which makes it difficult to 'see' the topology. With the first method, I can select vertices in shaded mode, which is much easier.
3) Lastly, does one typically just bind to the subdivision control points? But they seem to be very difficult to mentally follow with regard to the actual surface.
I prefer method #1, but it doesn't really seem to be what others are doing, so perhaps I'm missing some techniques with regard to the interface which would make method #2 or method #3 easier.
Or, perhaps I have it all wrong and there's a better way?