What's different about DX8? (1 Viewer)

[Ryan Peter]

I read an article comparing the GCN to the X-Box, and it mentioned something about DX8 having a more open programming structure. I'm not a programmer (yet), so what exactly does this mean in layman terms? What makes the DX8 API better than the GCN API and more flexible for the future? Thanks.

 

[Jeff Kleist]

DX8 is designed to be more flexible because it has to deal with millions of possible hardware combinations. By comparison, NGC's API is very rigid because it's designed from the ground up for 1 set of hardware

 

[Gary King]

That's not really it, Jeff.

The big addition to DX8 (and why it is considered superior to technoglogy in the Gamecube) is that the programmer is given a significant amount of control over how graphics are drawn. Without going into the various maths used in graphics, this basically allows objects on the Xbox to be rendered with more realistic surfaces than the Gamecube - satin surfaces will look satin, brushed aluminum will look different than chrome, a jellyfish will be irridescent rather than a clear plastic bulb, fingers and toes can be animated without causing the joints to collapse, and the game can look as if it is being played through a physical camera (with any type of lens - fisheye, hemispherical, telephoto, wide angle - complete with aperture, f-stop, and focal length effects) rather than the generic computer graphics camera.

A number of these effects can be performed (or faked with reasonable-looking results) with expensive software tricks on the Gamecube; however, they are all fully accelerated on the Xbox (or on PCs using DX8-compliant hardware (GeForce 3, Radeon 8500)), so more developers will be likely to use them (and the size of the game's scope won't suffer as a result of them being used).

 

[Jeff Kleist]

I bow to the superior knowlege

 

[Graeme Clark]

ya, what Gary said

 

[Ryan Peter]

Thanks Gary, you rule!