Suggestions needed on benchmarking the NV35

[Ostsol]

Here's an idea: don't use 8 bit .gif files for FSAA quality screenshots, like VR-Zone did. . .

http://www.vr-zone.com/reviews/NVIDIA/FX5900U/page9.htm

 

[DadUM]

Here's an idea: don't use 8 bit .gif files for FSAA quality screenshots, like VR-Zone did. . .

http://www.vr-zone.com/reviews/NVIDIA/FX5900U/page9.htm

Or change the jpeg compression levels between supposedly equal screen shots like [H] did.

(the 8x NV35 shots are 1/2 the size of the 6x ATI shots)

 

[MikeC]

Don't know if "perfect" landscape was used

The IL-2 results were based on High Picture Quality, which sets landscape detail to High. The Excellent Picture Quality option, which sets the landscape detail to Excellent, causes a pretty large hit in performance. I'll be happy to follow up with results from using Excellent Picture Quality.

 

[MikeC]

The second page has the IL-2 gameplay numbers based on FRAPS and some various other settings.

I'd fry my comp before trusting anything they put out.

My gameplay benchmark results are legitimate. I don't lie about benchmark results. Never have. Never will.

 

[MikeC]

Here's an idea: don't use 8 bit .gif files for FSAA quality screenshots, like VR-Zone did. . .

http://www.vr-zone.com/reviews/NVIDIA/FX5900U/page9.htm

You'll find an accurate representation of the output from FSAA Tester on the NV35 here:

http://www.nvnews.net/previews/geforce_fx_5900_ultra/index.shtml

The thumbnails and full-size images were saved in PNG format.

 

[radiATIon]

Hi,

Please use Temple 1.0.6 & Splinter Cell.