[Bf-committers] Re: added index-of-refraction presets popup menu

[Ed Halley]

GSR wrote:
> BTW, now that you seem deep into ray code... make the IOR be able to
> simulate "camera is already in something, not always vacuum/~air" so
> submarine scenes with bubbles can be done with more correctness (from
> physics class I think that means first hit would be 1/ior but slider
> does not go below 1.0, which would be the other way to make such
> scenes). Thanks. :]

Invert the normals on your bubbles, put a plane of water material just 
barely in front of your camera (past the hither clip plane), and you 
might be lucky, without changing the code.  Windows on your submarine 
would probably cause problems, but getting rid of the glass and making 
an inward-facing water plane would approach correctness.

I see no reason that the IOR slider can't be changed to allow [1/3,3.0] 
instead of the current [1.0,3.0] range, though.  This would make some 
things easier.

The raytracer is "context free", which means it has no idea what is 
inside and what is outside.  Even if you constrained yourself to 
manifold meshes that never intersect (which is a good thing, 
geometrically speaking), the raytracer doesn't have a clue as to what 
IOR to enter when it escapes one mesh into another mesh.

-- 
[ e d @ h a l l e y . c c ]