[gameprogrammer] Re: Optimizing opengl rendering

On Wednesday 06 April 2005 09.27, jorgefm@xxxxxxxxx wrote:
[...]
> I don't really have a slowdown because i'm rendering with an average
> of 
> 43 fps.

Is that intentionally restricted, or why 43 fps?


> The quads average per frame is 220. I was thinking in 
> optimizing because now i have a working platform and i'm curious
> about how much improve i can get using display lists, vertex arrays,
> etc.

I might be misunderstanding what you're doing, but I have a feeling 
you're wasting your time and potentially complicating your code for 
no real reason. Find a minimum spec system and see if there's any 
need for optimization at all.

If you want to push the minimum spec down as far as possible, do some 
profiling on hardware that isn't too far from what you're optimizing 
for, and deal with the actual hot-spots.

Of course, if you just feel like optimizing for fun, you define your 
own rules and do whatever you have to. :-)


> In other email I've read a quote about texture size from David. I've
> used your glSDL idea of texture fake size. In the Intel extreme 2
> opengl driver i have a maximum texture size of 1024, but i'm using
> a maximum texture size of 256. With values over 256 I get a very big
> slowdown. Then to use textures bigger than 256x256 i have to tile
> it.

Yeah, this seems to be pretty universal... You get better texture 
space utilization with smaller tiles, at least if the "real" surfaces 
have non-power-of-2 dimensions, but I don't see how that would 
explain the rather substantial differences seen on most cards.


//David Olofson - Programmer, Composer, Open Source Advocate

.- Audiality -----------------------------------------------.
|  Free/Open Source audio engine for games and multimedia.  |
| MIDI, modular synthesis, real time effects, scripting,... |
`-----------------------------------> http://audiality.org -'
   --- http://olofson.net --- http://www.reologica.se ---


---------------------
To unsubscribe go to http://gameprogrammer.com/mailinglist.html

Other related posts: