worm on 3DMark03

ClyssaN said:
The software is way behind the hardware, so i think AA should be default, but hey, it's only a benchmark :)
Certainly not! If you think that FSAA should be on by default in the benchmark, then please tell my why it isn't on by default in the display properties?

We simply think it is a tad early to have it on by default. MAYBE(!) next time, but that's another story.. ;)
 
worm[Futuremark said:
]
ClyssaN said:
The software is way behind the hardware, so i think AA should be default, but hey, it's only a benchmark :)
Certainly not! If you think that FSAA should be on by default in the benchmark, then please tell my why it isn't on by default in the display properties?

We simply think it is a tad early to have it on by default. MAYBE(!) next time, but that's another story.. ;)

One reason not to have it on by default is that there can be different implementations with different advantages and associated costs. How would the benchmark be able to differentiate between these? A vendor could have som absolute crap that they implemented mostly for benchmarking purposes. This is already happening, the Xabre being the most obvious example.

Using AA as default would absolutely kill vendor to vendor comparisons. These are dodgy enough as it is.

You can always complement the standard run with different combinations of AA and AF switched on, and review sites already do this. Couple that with the new facilities for image quality comparisons, and you have a better scenario for card evaluation than just "using AA as default".

Entropy
 
BTW, first bench i saw using 3Dmark 2003 ;)
IMG0005900.gif

http://www.hardware.fr/articles/454/page5.html
 
worm[Futuremark said:
]
ClyssaN said:
The software is way behind the hardware, so i think AA should be default, but hey, it's only a benchmark :)
Certainly not! If you think that FSAA should be on by default in the benchmark, then please tell my why it isn't on by default in the display properties?

the default for the 9700pro is 4x :)
 
laGadU said:
the default for the 9700pro is 4x :)
Err.. But it is set to "application preference" which means that it is not "ON" (forced on that is) by default. At least not last time I checked! :rolleyes:
 
But it is set to "application preference" which means that it is not "ON" (forced on that is) by default

It's not off either..(application selects if its on or off).
Tests with no AA will most certainly be CPU/Platform bound and this option should be selected if the adapter that is detected is marketed as

1)High end product
2)A FSAA card

So a Ti4600 was marketed as a FSAA card, every review pushed that and comments like:
Some moaners are criticizing GeForce4 Ti for its lack of really new technologies. I would like to remind those people of the performance jump we are seeing here. Isn't that what really counts? GeForce4 Ti has so much power, it can run with anti aliasing enabled in virtually any game right now. Features alone don't win customers -- they ought to make sense, and the performance has to be right, too. GeForce4 Ti seems to be a clear winner here.

The 9700 was also marketed as a FSAA card, and is one of the main reasons people buy these cards.
FSAA shouldn't be on by default on all cards, just the $400 + premier cards, disappointing to say the least.
 
You can't have AA enabled for some cards, and not others. The right decision is to have AA disabled for the "standard score", and allow the option to enable AA and advanced filtering techniques.
 
Why..benchmarks are listed in the ORB with FSAA on or OFF...sorry doesn't fly

They way is it important to have FSAA on by default?

Does the ORB distinguish between SS AA and MS AA? Gamma corrected or not? Rotated vs. Ordered grid?
 
Who cares ?? People pay big money for FSAA, most people don't even know what MSAA is, as long as the Database logs FSAA was on then whoopee doo.
If we are waiting for ALL IHV's to get cards that run with AA on by default then we'll be waiting a very LONG time.
These high end cards are marketed that way, they should be benchmarked that way.
 
Of course... there are a lot of people like me: lazy. Set up Opera to reload every 15-30-60 seconds or whatever you want... :D
 
Thi was the very last second:

theendsm.jpg


Also notice the difference between the counter of IE6 and Opera... :oops:

Edit: smaller pic
 
Back
Top