Hello all. Working on a game project with a cubic-3d look (similar in style to Minecraft or Voxatron).
My present workflow for models begins with creating an array of colored cubes. Runtime speed is important, so obviously I want to trim these down. To optimize, I load this geometry into modelling software and remove or merge non-corner points, and rebuild the outward faces. I then use the original model as a reference to construct a UV for the optimized model (renders with nearest-neighbor).
Obviously this is a naive approach and is creating an unnecessary workload. I have read that processing the original models with a marching-cubes algo would generate useful results, but unfortunately this is beyond me.
Can anyone suggest a better workflow from the art-tool side of things? Free software is preferable but I'd be willing to shell for an all-in-one solution...
Replies
Maybe this is useful: http://www.scriptspot.com/3ds-max/scripts/pix3led
Or this: http://www.scriptspot.com/3ds-max/scripts/voxelizer