• Join Us |
  • |
  • Sign in with:

V5 Benchmarks w/ 1.1 Drivers

by Maarten Goldstein, May 30, 2000 9:25am PDT
Related Topics – 3dfx

Over at The Pulpit, Reverend has put up Voodoo 5 Quake3 benchmarks using the 1.1 reference drivers. It seems there's a big FSAA speed difference with 16 or 32-bit desktop when using the "Normal" settings. The Shack Benchmarks were done with a 32-bit desktop.




Comments

7 Threads | 13 Comments




  • The most important thing there:

    The most important consideration is that of your desktop color depth. It totally influences the "Fastest" and "Normal" pre-configured settings (but not the "High Quality" one). Why? Because whenever you see the word "Default" where it applies to Color Depth and Texture Quality in Quake3's System/Graphics menu, it ("Default") means your desktop color depth. This is a very important thing to consider. To illustrate this, take a look at the three settings as depicted by the images below :

    Notice the words "Default" in the "Fastest" and "Normal" settings. What "Default" means in these two settings is that Quake3 will use your desktop color depth setting. The "High Quality" setting doesn't give rise to any confusion as it is specified what bitdepth is used for "Color Depth" and "Texture Quality" respectively.
    This is one of the reasons why I have never used Quake3's pre-configured settings in any of my reviews/benchmarks - I would specify exactly what bitdepth I'm using for each resolution for both "Color Depth" and "Texture Quality". Why? Simply because by doing so, I'm making it easy (and irrefutably so) exactly what conditions I'm testing under. This is, again I repeat, very important because you have so many websites reporting benchmarks and many of these websites, when using Quake3's preconfigured "Fastest", "Normal" and "High Quality" settings, almost never report what desktop color depth they're running these benchmarks under. The most important consideration for any of these three settings is the word "Default" - one website could be using a 32bit desktop and give you a set of benchmarks that actually means 32bit benchmarks (without them knowing it) while another website could be using a 16bit desktop and is actually giving you 16bit scores (again, without them knowing it), all the while when both are using the Quake3 pre-configured "Fastest", "Normal" and "High Quality" pre-configured settings. If both websites have similar systems, do you think you'll see a difference in scores? Of course you will!