SI
SI
discoversearch

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor.  We ask that you disable ad blocking while on Silicon Investor in the best interests of our community.  If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.
Technology Stocks : Advanced Micro Devices - Moderated (AMD) -- Ignore unavailable to you. Want to Upgrade?


To: golfbum who wrote (243227)11/3/2007 2:57:50 PM
From: graphicsguruRead Replies (2) | Respond to of 275872
 
after thinking about it a bit i realized that 2d and video will still benefit from having a "legacy" video system similar to today's integrated graphics.

Yes, I think that if we end up with an interactive ray tracing API, it will
be in addition to the current standards and will not replace them.

It's hard to do better than scan conversion for the first bounce -- it has
spatial coherence in its favor, the math is simpler, it pipelines well,
and there've been a whole lot of legacy effort in hardware and software
to optimize it. For secondary rays (from an object to a light or another
object instead of from an object to the eye/camera), however, it would
not surprise me if ray tracing starts to become popular.

With scan converted shadows, there are annoying shadow map
resolution and offset issues that rear their ugly heads more often than you'd
prefer Of course, ray tracing can also bump into precision issues
I've seen ray traced images of very thin surfaces where the
rays were erroneously calculated to hit the backsurface rather than the
front. Just another reason to keep the dynamic range of distances in your
CG geometry under control. If you have sub-1 mil thick surfaces on a model
of a whole city, you're probably asking for trouble . . . especially if you
ever see them anything close to edge-on. But most of the time, DP is
good enough.