Even those highly-compressed jpg's can't hide the fact that ATi's 4x mode looks slightly better than nV's, though their 2x modes look similar. I can't rreally tell, though.
Did the review actually say the Det 53's improved AA quality, or are you inferring that from HH's analysis of those two screengrabs? No offense, but that seems to be a rather shaky inference. Both screenshots don't show a lot of high-contrast edges, either, which is where the difference in AA quality will be most pronounced (and most easily detected).
Did the review actually say the Det 53's improved AA quality, or are you inferring that from HH's analysis of those two screengrabs? No offense, but that seems to be a rather shaky inference. Both screenshots don't show a lot of high-contrast edges, either, which is where the difference in AA quality will be most pronounced (and most easily detected).
Thats funny, because I have heard this numerous times, just about for every release Nvidia has. According to some people nvidia's IQ has gone up every driver release as far back as I can remember, strange. Maybe it's like taking a placebo pill and thinking you feel better
Do you have any proof to bsack that up? The only company i've seen (appearing to) detect screenshots is ATi, as explained in my posts above...
The English version of 3DC's AA comparo is up. They declare 2x equal and 4x and above a decisive win for ATi. They didn't really cover texture aliasing, though. nV's mixed modes, while much slower than ATi's straight MS, should offer nicer texture quality by reducing texture aliasing (pixel shimmer). We're getting close to the next gen, but three months is a long time to wait to see what AA and AF modes they'll offer. I hope both companies put some effort into IQ beyond the expected improvements in programmability and precision.
G (Greg?), does that happen in both OGL and D3D?
G (Greg?), does that happen in both OGL and D3D?
G (Greg?), does that happen in both OGL and D3D?[/quote]
Yes, my first name is Greg
According to Edge the AA issue with the game occurs in both OGL & D3D:
Yes, my first name is Greg
According to Edge the AA issue with the game occurs in both OGL & D3D:
Read the nvnews forum thread, that is where the evidence is and also discussion on Gabes comments.
Basically, Gabe has done plenty of accusing but not much proving. I'd suggest he needs to put up or shut up.
BFG10K
Lifer
- Aug 14, 2000
- 22,709
- 3,003
- 126
Uh, did you even read that thread?
On the ATi card Fraps 2.0's output matched the on-screen image so obviously any problem was a short-coming of version 1.x of the program. You'll also note that nVidia's screen and capture images still didn't match even after he used version 2.0.
On the ATi card Fraps 2.0's output matched the on-screen image so obviously any problem was a short-coming of version 1.x of the program. You'll also note that nVidia's screen and capture images still didn't match even after he used version 2.0.
The burden of proof is not on Gabe but ironically the thread you linked to is presenting more proof in his favour. If he used version 2.0 of Fraps on both vendors then he would have all the evidence he ever needed.
Excuse me???
FRAPS is showing exactly what the user saw on screen - no AA. Yet if you do a printscreen screen capture you will see AA.
EDIT: to make this crystal clear:
Ati: onscreen = no AA, printscreen = AA, FRAPS = no AA
nVidia onscreeen = no AA, printscreen = no AA, FRAPS = no AA
The printscreen function works through the device drivers.
EDIT2:
It is Gabe's responsibility to prove his claims. Just like the legal system, the accuser has to prove his claims - Innocent until proven guilty.
FRAPS is showing exactly what the user saw on screen - no AA. Yet if you do a printscreen screen capture you will see AA.
EDIT: to make this crystal clear:
Ati: onscreen = no AA, printscreen = AA, FRAPS = no AA
nVidia onscreeen = no AA, printscreen = no AA, FRAPS = no AA
The printscreen function works through the device drivers.
EDIT2:
It is Gabe's responsibility to prove his claims. Just like the legal system, the accuser has to prove his claims - Innocent until proven guilty.
BFG10K
Lifer
- Aug 14, 2000
- 22,709
- 3,003
- 126
Fraps 2.0 is showing what's on-screen; Fraps 1.x isn't. Also Print Screen itself hasn't been an accurate tool to take 3D screenshots for quite some time due to reasons such as post-filtering and external gamma correction.
For what you're implying to be true, ATi would have to quickly enable super-sampling just for one screenshot to get AA on transparencies. And since we know that ATi's drivers don't support SS, I find something like that highly unlikely if not impossible.
And nVidia does the same except even Fraps 2.0 is showing AA images where there are none. But on the ATi card Fraps 2.0 does indeed show the correct image.
What on earth are you talking about? From the thread:
I already posted what follows once, but it somehow got eaten:
Read the following post of edges carefully: Then look at the image. That is a comparison of FRAPS 2.0 and printscreen. FRAPS 1.x has nothing to do with anything. The visual evidence is based off of FRAPS 2.0 and printscreen.
Read the following post of edges carefully: Then look at the image. That is a comparison of FRAPS 2.0 and printscreen. FRAPS 1.x has nothing to do with anything. The visual evidence is based off of FRAPS 2.0 and printscreen.
Pure SSAA is not a display panel selectable AA option in any current nVidia detonator driver. It has to be forced on with rivatuner or similar, and yes, it will display AA because SSAA is the only AA method that works properly with alpha textures.
BFG10K
Lifer
- Aug 14, 2000
- 22,709
- 3,003
- 126
Actually it has everything to do with it. On the ATi card when using Fraps 1.x he got differing images between the on-screen and captured version but with Fraps 2.0 he did not.
Yes and we've already established that Print Screen produces incorrect images on ATi cards while Fraps 2.0 does not. In fact it's been established in general that Print Screen isn't a good way to take screenshots even as far back as the Voodoo 3 days when it failed to pick up the post-filter.
Irrelevant. The issue here is that on nVidia cards it showed no AA on-screen but in the screen capture (Fraps 2.0) it did. That means that even Fraps 2.0 cannot match the on-screen output of nVidia cards.
BFG10K
Lifer
- Aug 14, 2000
- 22,709
- 3,003
- 126
I'll repeat myself for the third time: Fraps 2.0 captures screens correctly on ATi cards.
If you have a point to make then I'd like to hear it. Because you just seem to be repeating what I've been saying right from the start.
Also I'd like to hear what your response is to nVidia's screenshots where not even Fraps 2.0 could produce the correct results.
If you have a point to make then I'd like to hear it. Because you just seem to be repeating what I've been saying right from the start.
Also I'd like to hear what your response is to nVidia's screenshots where not even Fraps 2.0 could produce the correct results.
I have never once disagreed with that assertion. This is all about what happens when you do a printscreen capture, not a FRAPS capture. FRAPS just confirms what was actually onscreen (as would a photograph of the screen).
BFG10K
Lifer
- Aug 14, 2000
- 22,709
- 3,003
- 126
It's true that both vendors appear to show incorrect images with Print Screen but I've already pointed out many times that Print Screen is basically useless for taking 3D screenshots and has been for some time.
What's more interesting is the fact that even the Fraps versions of nVidia's screenshots are incorrect and show AA where there is none unlike ATi's, which are correct.
What's more interesting is the fact that even the Fraps versions of nVidia's screenshots are incorrect and show AA where there is none unlike ATi's, which are correct.
That depends on what you mean by incorrect.
nVidia showing less AA than is actually present is forgiveable - they are not show rendering detail that does not actually exist.
ATi showing AA where there is none present is blatantly misleading people.
nVidia showing less AA than is actually present is forgiveable - they are not show rendering detail that does not actually exist.
ATi showing AA where there is none present is blatantly misleading people.
His comments were made at the ATi shader day conference, and he does currently have a $6 million interest in making ATi look good vs nVidia.
BFG10K
Lifer
- Aug 14, 2000
- 22,709
- 3,003
- 126
The screenshot not matching the screen of course.
nVidia's on-screen image was not AA'd but all of their screenshots (including Fraps 2.0) were.
Actually that's exactly what they're doing - even with Fraps 2.0 - while ATi's screenshots are correct under said program.
Uh, but that's exactly what nVidia is doing and it's not even possible to get a correct screenshot out of them, unlike ATi where it is.
Gabe said nVidia were altering screenshots and that thread provides evidence to support this fact. Once again I fail to see the logic behind your comments.
TRENDING THREADS
-
Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)- Started by DisEnchantment
- Replies: 25K
-
Discussion Intel Meteor, Arrow, Lunar & Panther Lakes + WCL Discussion Threads
- Started by Tigerick
- Replies: 23K
-
Discussion Intel current and future Lakes & Rapids thread- Started by TheF34RChannel
- Replies: 23K
-
-
News NVIDIA and Intel to Develop AI Infrastructure and Personal Computing Products- Started by poke01
- Replies: 411
Facebook
Twitter