Jump to content
  • Advertisement

Archived

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

ade-the-heat

BSP trees - worth knowing inside out ?

This topic is 5227 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, I''ve read about BSP trees- there''s also a tutorial about using them in www.gametutorials.com with a quake level. Is it worth writing my own one from scratch - what I mean is that on www.gameinstitute.com there is a whole course for $125 about them. Is this money well spent in order to get up to speed on something that at a high level seems reasomnably understandable to me anyhow. cheers

Share this post


Link to post
Share on other sites
Advertisement
hi,

I'd say its worth it from a 'learning' experience, although in terms of game production they are rapidly losing favor these days. Doesn't mean you can't or shouldn't use them though.

As to whether its worth paying out for the gameinstitue course thats a different matter. I saw much of the course tutorials before it got taken offline for use in it and felt it was proberbly the most well written and indepth article on the subject you'll likely to see. Obviously I can't comment on the actual course as i've not seen it, bu I would presume it expanded on the original tutorial. However there are plenty of other BSP sources online (google), just depends if you can get an implementation working from them.

There are two main uses of BSP trees (at least for me), one is to get an accurate draw order (avoiding things like the painters algorithm issues), the second is for simpler collision/line of site detection. Its biggest problem is long computational times and limited usefulness for destructable/deformable geometry.

On a personal note, I usually find the journey of learning some new method and its implementation often more useful than the method itself. Each new thing you learn becomes a building block for something else.


[edited by - noisecrime on June 3, 2004 7:17:54 AM]

Share this post


Link to post
Share on other sites
Thanks,
You said they''re rapidly losing favour - what''s replacing them then ? Octrees ? (which again are simply recursive ways of splitting up the world).

Share this post


Link to post
Share on other sites
Well Octree''s are pretty useful, I''ve used them myself for polygon soup levels. However I hear very good thing about ABT''s, but haven''t had the time to investigate them yet.

I think you''ll find a few threads about ABT here (or in graphics programming and theory forum).

Starting point
http://www.gamedev.net/community/forums/topic.asp?topic_id=163240

Share this post


Link to post
Share on other sites
I''ve implemented both, the octtree is just a smarter way to go these days. The back-front order of polys that the format features is somewhat useless, now that we have the depth buffer.

When in doubt, learn both anyways.

Share this post


Link to post
Share on other sites

  • Advertisement
×

Important Information

By using GameDev.net, you agree to our community Guidelines, Terms of Use, and Privacy Policy.

We are the game development community.

Whether you are an indie, hobbyist, AAA developer, or just trying to learn, GameDev.net is the place for you to learn, share, and connect with the games industry. Learn more About Us or sign up!

Sign me up!