🚀 go-pugleaf

Here we go again, boys and girls.  :(

For brevity's sake <g>:

1) scores from differing test machines and setups should not be compared
2) I don't think multi-sampling scores should be compared to super-sampling
scores because multi-sampling is basically an edge AA solution whereas
super-sampling addresses both edge and texture aliasing.  Not exactly an
apples to apples comparison because the image quality will not be the same.
Any gamer who's spent hours with Porsche Unleashed on a GTS, V5, or Radeon
and brings home a shiny, new GF3 and fires up 2x2 AA might be more than a
little puzzled where all the texture shimmering is suddenly coming from.
The fix for this?  Enable anisotropic texture filtering (or perhaps Quincunx
mode in the AA options, if it turns out to be a reasonably effective aid in
alleviating texture aliasing), which will impact performance on the GF3.
Kind of makes Tnaw's comparisons null and void, eh?  Unfortunately, Tnaw has
twice now demonstrated that he doesn't understand the differences between
multi- and super-sampling.

John


"tnaw" <tnaw@MailAndNews.com> wrote in message
news:3AA98990@MailAndNews.com...
> GeForce3 can be preordered at a price of about $550 now.
> Is the GeForce3 worth the price? Under antialiasing modes,
> GeForce3 gets ahead Voodoo5 6000 by +18% to +41%.  Is it
> a good deal to get a GeForce3 that is not only faster
> than Voodoo5 6000 but also having extra features such as
> hardware T&L and DirectX 8.0 compliance with a lower price?
> And for your entertainment, there are also some Voodoo5 6000
> card pictures and a 8xFSAA screenshot grasped from the net.
>
> http://www.geocities.com/tnaw_xtennis/NV20-8.htm
>