Professional Rendering

Started by
1 comment, last by Magallo 23 years, 10 months ago
Thinking about a real professional 3d game I was asking myself how the hell programmers can render in a devent frame rate so huge levels. I mean, think about a Quake or Half-Life or even Tomb Raider level. They are huge!!! They are made by millions of vertices and triangles. How can they render them smootly ? It must be a particular way to do it. They cam''t use glVertex3f() milions of times per every rendering loop. So, what is the solution ? Another question.... Does glDrawElements give a very good boost performance instead of using glVertex3f or not ? Now I''m using the glVertex3f method called ''n'' times where ''n'' is the number of my vertices. Should I pass to use glDrawElements or not ?
Advertisement
:-)
They just do not render what you cannot see
(BSP tree)
(you can find me on IRC : #opengl on undernet)
Current hardware cannot cope with the amount of data needed for a single quake 3 level ( not hardware available to average gamer at least ). So it is best to render only what is visible ...

Look for bsp tree, octree and portal rendering tutorials. These methods allows to draw only what is visible ( or could eventually be visible ).

I still dont know enough about opengl to answer your second question, but I wish you good luck ...

------------------"Between the time when the oceans drank Atlantis and the rise of the sons of Arius there was an age undreamed of..."

This topic is closed to new replies.

Advertisement