New Miyamoto interview (must read)!

[Private Hudson]

Originally posted by OB1
For one thing, there's the dynamic lighting that's never been done like this before. Instead of using a simple lightmap like most games use for lighting, the Doom 3 engine gets rid of that and is able to calculate the light value of every pixel in the screen at each frame. The dynamic lighting in conjunction with the clever usage of bump-mapping gives the illusion of much higher poly counts. Of course, you need a powerful GPU to handle that kind of dynamic lighting, but it's more efficient than cranking out super high poly environments and models. I could go into more detail but I think I might be getting carpal tunnel syndrom from typing all day.

Splinter Cell used per pixel calculations for it's lighting model, as well.

Not before Episode I was created.

Dude, just watch the Episode I docs. You'll get the picture.

Uh.. yah. Just not to the extent that was in Jar Jar.

And dammit, I just said I have watched Episode 1 docs.