Archived

This topic is now archived and is closed to further replies.

dot product3 bump-mapping question

This topic is 5148 days old which is more than the 365 day threshold we allow for new replies. Please post a new topic.

If you intended to correct an error in the post then please contact us.

Recommended Posts

If I am using dot3 bumpmapping for lighting, how many lights can I use in a single pass? I am using D3D, if that makes any difference. ___________________________________________________________ Proceeding on a brutal rampage is the obvious choice.

Share this post


Link to post
Share on other sites
It depends:

1) On the hardware you''re using. Different hardware has different numbers of texture blending unit.

2) On whether you''re using the fixed function pipeline or shaders.

3) On whether you''re just performing diffuse illumination or specular (or more) as well.

4) On what types of lights you''re using and how accurate you want things such as attenuation of your point lights to be.

5) On what you need for your base texture and whether there are any other demands (e.g. detail maps, gloss maps, environment maps etc).

6) On whether you need any alpha transparency and if you want to do anything special with it in relation to the other effects (Fresnel etc)

7) On whether you''re combining lights in some way so that nearby lights merge into a single light.


The better looking things you do, the more texture blending stages (and/or pixel shader instructions) you''ll need - which means the number of blending/texture units on your target hardware becomes the hard limit. There isn''t _one_ answer.

On the first generation of DOT3 capable cards (3DLabs Permedia3, nVidia GeForce256 and GeForce2) you can just about do ONE diffuse-only light blended with a base texture and maybe with an attenuation factor in a single pass. Oh and that won''t have an ambient term either...

On newer hardware I''d say 2-3 of the same. Once you have specular, that''ll go back down to 1 since you have fun like re-normalising the vectors every time and raising the result to the shininess/sharpness power.


DOT3 means you get tons of surface and lighting detail per-pixel - so many people have actually gone for LOWER than usual polygon counts, sacrificing smooth silhouette edges for that higher surface and lighing detail and multiple PASSES. Take another very close look at those popular Doom3 shots - in particular the silhoette edges and you''ll see what I mean (AFAIK a D3 scene has less polys per scene than Q3).

--
Simon O''Connor
3D Game Programmer &
Microsoft DirectX MVP

Share this post


Link to post
Share on other sites