Archived

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

Bsp always faster?

This topic is 6358 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

Hi Is Bsp-tree based rendering always faster than z-buffer-based rendering on the new graphic cards? I think the z-buffer is hardware accelerated (im not sure at all) and it will probably faster than usin BSP-trees (if u dont draw stuff that u cant see)

Share this post


Link to post
Share on other sites
If you use BSP''s, you''re still going to use the z-Buffer, what the BSP tree does is cuts out Geometry tht''s not in the view volume, (i.e., that can''t be seen). With out the BSP, the Z buffer has to do more work, so BSP''s will almost always be faster, unless it''s a situation where all geometry is potentially visible at all times (say something like a 3D tetris or whatever where the geometry never goes behind the camera).

Dare To Think Outside The Box
_____________________________
|____________________________|
http://www.inversestudios.com

Share this post


Link to post
Share on other sites
I''ve been wondering about this, too.

Since we''re on the topic of BSP trees, I have another question about them: how much work does it take to implement them? I want to write a demo or two to try out the process, but I''m sort of new to 3D (been working with OpenGL for a couple of weeks), and I don''t want to start a project I won''t be able to finish (yet).

Share this post


Link to post
Share on other sites
see nate millers page (search) hes got a nice program that uses them. if only all the ppl on the web including myselfcould write so cleanly as him

Share this post


Link to post
Share on other sites
Guest Anonymous Poster
(im kinda crossposting on this topic ... sorry ... got another thread under directx/opengl api) ...
hmm
the problem is ... i dont know "how much" i should use the bsp-splitting-routine ...

at the moment i really use EVERY polygon i have as split plane --> i get huge data (alot more than i would have without bsp cause of polygon splitting) ... so its slower for me ...

Share this post


Link to post
Share on other sites
Guest Anonymous Poster
just split until each leaf has an acceptable number of polys (ie. set a maximum poly per leaf value). if you''re using a z-buffer it''s not gonna matter although the smaller the leaf size the less the zbuffer has to do.. obviously because more is being clipped

Share this post


Link to post
Share on other sites