Image quality vs Speed

[mdhome]

I was reading a message in one of Rage3d's forums about ATI increasing the speed of the Radeon at the expense of image quality in their drivers. This led me thinking. If Matrox were to drop the image quality of their products to *snigger* geforce levels, would we get GTS performance levels. Joking aside, what speed increase could we expect if something similar was done.


regards MD

[Greebe]

I doubt they have the slightest clue what they're talking about.

Got a link?

[Wulfman]

I got one, but it's a german site...

http://www.3dcenter.de/artikel/2001/10-24_a.php

you will get the meaning of the word "cheat" in a picture, right?

you have to move the mouse over some images to see the difference...

(and follow the "weiter" - button)

mfg
wulfman

[Gurm]

What ATI are doing, if I understand correctly, is disabling certain features, or putting them into "quick-n-dirty-hack" mode. The same way that nVidia for the longest time didn't REALLY do Trilinear filtering... but the driver said it was turned on, and it passed the tests for it - but looked like bilinear ('cuz it WAS).

- Gurm

[mdhome]

Think this is it.

http://www.rage3d.com/board/showthread.php?threadid=33586522

regards MD

[Guru]

Benchmarking Ethics and the ATI Radeon

http://firingsquad.gamers.com/hardware/radeonquack/default.asp

[paulcs]

Usually, I look at screen shots and think, "well, I notice a difference, but I wouldn't during game play." The differences in the Firing Squad screenshots, however, are HUGE!

This isn't the first time ATI has added really dubious optimizations. I recall a couple of years ago, they released a driver set and touted an increase in scores with a particular benchmark, but testing established they slowed the board down with everything else.

You've got to hand it to them. They've single-handedly established Quack 3 as the new standard measure for OpenGL gaming performance.

Paul

[Wombat]

Man, this really bothers me. ATi really tried to pull a fast one. It would have been one thing if they had just optimized the game, but they sacrificed image quality to do it, with no information or choice given to the consumer. Pretty fraudulent to me. And all just to increase benchmarks. This is just as bad as GDI manipulation. Worse, maybe, since GDI hacks don't reduce image quality.

I really was considering an 8500, but no ATi for me now.

[TdB]

TdB

they should atleast give the user the option to turn the optimization off, most players can´t feel the difference between 190 fps and 154 fps anyway.
my own gaming card (kyro2) also has some game specific optimizations, but i can choose the settings i want.

and even 154 fps is just crazy, if i get that kind of performance i begin to examine, if i can turn even more detail on, like fsaa or better filtering.

do we really need that kind of speed anyway??

if i turn v-sync off then i wont notice anything higher than 75 fps(im talking about constant framerate here), with v-sync enabled i noticed every single dropped frame, so a constant framerate equal to the monitorrefreshrate, is really all i need.

and i think 99.9% of all computer players feel the same way(im just guessing here).