> Artifacts would be noticable if, say, your were rendering an image of a 
> lizard with the Grand Canyon visible in the background.

Man, you'd need some kind of supercomputer to model the entire grand canyon 
in enough detail to do a close-up of a lizard on it...

Use some kind of level of detail scheme (ie further away parts are modelled 
at lower resolution) - then you'll find that 16 bits of vertical resolution 
is plenty for the scene you mention.

Post a reply to this message