• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

If you overclock, then Nvidia's the company for you.

Page 5 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
Originally posted by: BFG10K
umm.. there was a 16x. it is BROKEN at that res.
I find it very hard to believe that 8x works fine but 16x is broken. In my experience AF tends to be an "all-or-nothing" excercise.

In any case, your comparisons are invalid as you're only running the ATi card at half image quality.

lol.. you're such a pompous... sort ... at times.

you should actually read before you speak. aa/af works at resolutions 1024x768 and below. whether the game, the driver, or a combination of both, it appears to be broken above that.

as far as whatever the quality of the images are, bitch at ati; i am simply sharing my observation, nothing more, nothing less.

 
Originally posted by: Pete
Thanks for the screenshots, Cainam. Yep, the two cards look similar standing still. ATi's 16xAF extends a little further on the floor, and I can detect *slightly* better AA on a few lines, but otherwise nothing major at all. Of course, nV's AF is supposed to be better in motion, so ATi's extended clarity may be less preferable to nV's ostensibly lesser texture shimmer.

i think that's a fair observation. on the second shot, where you can see the the diff of 16x af on the floor tile, if you look at the textures, the nv has a bit more detail, and actually, i like the color of the "night sky" a bit better.. but overall, i really don't think the quality of both are far better than we had a generation or two ago. i think nvidia has recognized the importance of IQ, and has worked hard on addressing that.

as i said earlier, i think ati has a sligher better aa scheme, and nv has a bit more detail in textures, but again, it only shows up in certain situation, and IQ is subjective, so someone else may certainly hold a different opinon. performance wise tho, i think at least so far, it's pretty well been proven ati's aa/af methods excel.

I'd be interested in your opinion in performance differences between the two as another "real-world" and apparently unbiased source, but I won't ask you to take time to do so, because I won't be buying one of these cards soon. It'd just be an exercise in statistics for me. 🙂 Those two CPUs seem pretty evenly matched, though I don't know if DAoC prefers one architecture over another as some games do.

well, if it means anything, daoc has an nvidia "the way it's meant to be played" logo on the splash screen, but i think most developers work closely with both nv and ati to put forth the best product they can, regardless of the video card brand. i mean, it wouldn't make sense from a marketing standpoint if you're game only played well on one platform or the other; you'd be alienating a good chunk of your market.

Thanks again for your efforts. I'm sure others have found your results interesting, too. I'll leave it up to them to bother you for performance results. 😉

heh.. no problem. i find it all interesting. i may run it later, but i think had ati's drivers worked correctly, ati would run a little better with hi aa/af and hi resolutions, but i wouldn't say drastically so. as far as cpu's, there's no doubt the p4 outperforms the amd, but it's age is starting to show, and were it a current generation athlon 64 the tables would be turned, tho again i think it would be a marginal advantage for the a64, not an earth shattering one.

BFG, I think that's what he meant--neither AF nor AA work on his ATi card above 10x7, as demonstrated by his first set of pics. His second set shows a difference between ATi's 16x and nV's 8x.

exaclty.

peace 🙂
 
I hate texture aliasing. It is so annoying. I catch it now and then in some games at certain angles. Say a hill is to my right and I'm kneeling and the textures are close. If ATI has more texture aliasing, that would suck if I had one, but it's probably not much more anyway. With the Voodoo5 if you set the LOD all the way, even if you set it a little to the negative end, there would be textures aliasing. Setting it all the way, would worsen the aliasing so much that the texture would be distorted. And guess what fixes this issue. AA. Their AA would apply a blur to everything and clear up the Texture aliasing.
 
Right, I see. I must've missed the settings you used in your end shots.

But keep two things in mind about screenshots:

Only FRAPS 2.0 produces correct images on ATi cards but not even FRAPS 2.0 produces correct nVidia shots. There was a thread at nvNews and it was observed that even FRAPS 2.0 was adding additional IQ to the nVidia screenshots when it was not present on the screen. Standard screenshots (i.e used by the game) do not work at all with either vendor since they can't account for post-rendering.

AF comparisons are best done on moving scenes, not still screenshots.
 
Originally posted by: BFG10K
Right, I see. I must've missed the settings you used in your end shots.

But keep two things in mind about screenshots:

Only FRAPS 2.0 produces correct images on ATi cards but not even FRAPS 2.0 produces correct nVidia shots. There was a thread at nvNews and it was observed that even FRAPS 2.0 was adding additional IQ to the nVidia screenshots when it was not present on the screen. Standard screenshots (i.e used by the game) do not work at all with either vendor since they can't account for post-rendering.

AF comparisons are best done on moving scenes, not still screenshots.

i'm not familiar w/ the farps thing... either way, at least in daoc, af/aa does not work in 1280x and up, but does at 1024x and below.

i do agree what you see on the monitor is what's important, but in all honestly, there's no discenrbable difference in IQ that i can see when observing both in action. the only time really i can tell is in stills, where i can compare side by side.

the fraps thing sounds interesting tho.. if you happen to have some links i'd appreciate it 🙂


thanks!
 
Fraps.com

I cant believe someone from this forum doesnt know about Fraps.. Its been around forever. It is used in a lot of reviews ( HardOCP uses it in every review), and can capture video from virtually any game. Its free, but if you pay $20 you can use extra features.

nVidia doesnt like it, wonder why that is. They released a statement months ago saying it wasnt accurate.
 
Originally posted by: Ackmed
Fraps.com

I cant believe someone from this forum doesnt know about Fraps.. Its been around forever. It is used in a lot of reviews ( HardOCP uses it in every review), and can capture video from virtually any game. Its free, but if you pay $20 you can use extra features.

nVidia doesnt like it, wonder why that is. They released a statement months ago saying it wasnt accurate.

Many do know it... They're just too bored to mention it.. 😀

 
Originally posted by: Ackmed
I was speaking to the poster above me who said they didnt.

i know what fraps is. what i didn't know was what was quoted:

Originally posted by: BFG10K

Only FRAPS 2.0 produces correct images on ATi cards but not even FRAPS 2.0 produces correct nVidia shots. There was a thread at nvNews and it was observed that even FRAPS 2.0 was adding additional IQ to the nVidia screenshots when it was not present on the screen. Standard screenshots (i.e used by the game) do not work at all with either vendor since they can't account for post-rendering.

which is why i asked bfg for links....
 
i'm not familiar w/ the farps thing... either way,
You mean you were using standard PrintScreen/in-game screenshots? Dude, those haven't worked properly since the Voodoo2 days. 😉

nVidia doesnt like it, wonder why that is. They released a statement months ago saying it wasnt accurate.
That doesn't surprise me one bit. Anything that exposes possible cheats from nVidia is automatically ruled as inaccurate or incorrect by them.

It's highly likely that FRAPS 2.0 is in fact correctly picking up their fudged screenshots, though it's difficult to prove as the only evidence we have is sketchy at best.
 
Originally posted by: BFG10K
You mean you were using standard PrintScreen/in-game screenshots? Dude, those haven't worked properly since the Voodoo2 days. 😉

why is that? is there some resources/links you can point me to so i can understand this better?
 
Standard screenshots only capture the contents of the frame buffer. The problem is that post-filtering methods - many versions of AA, 3dfx's 22 bit colour up-sampling and so on - modify the image after it has left the frame buffer but before it hits the screen.

That means that standard screenshots will often be missing information that is present when you look at the same image on the screen. The significance of utilities like FRAPS is that they are capable of picking up the additional information that standard screenshots cannot pick up.
 
Back
Top