-
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.
7900GT or X1900 GT?
Page 4 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
Keysplayr
Elite Member
LOL, you're a sad individual. Actually, I said, and meant, this: "Oh I think I know all that there needs to be known about "Ulfhednar". Now go over to your fish tank, you know, the one with all the red-herrings in it. The fish food is to the right of the tank in a cylindrical container marked "Fallicies". You can't miss it. I heard it's relaxing to watch swimming fish and very therapeutic."
I had some questions, after reading the FS article, keysplayr2003, but I think CaiNaM and josh6079 have at least partially awnsered them. I'm reminded of the times in the past that I preferred to turn the sharpness control up on the tv a little too far. At that time I enjoyed the picture more with the subjective sharpness despite the increase in video "noise." Nowadays, I like smooth transitions and the like more. After a quick read of the article, it struck me that possibly ATI was AA'ing some of the smaller branches out of existence. In other words, the ATI card was starting to render the whole tree but the AA was killing some of the smaller branches because they were made up of too few pixels and were masked by the overlayed AA. Because if the all the parts of the "tree" the cards were rendering was supposed to be the same distance away from the player, wouldn't it either entirely be inside the draw distance or entirely outside of it? I'm afraid I may be showing my technical ignorance...
Have a :beer: on me everyone, let's all mellow out a little
Have a :beer: on me everyone, let's all mellow out a little
Cookie Monster
Diamond Member
8xS = 4xMSAA + 2xSSAA. (unless im wrong 😛)
TRAA is just another AA mode, to apply anti aliasing to alpha textures. This mode can be used with any other AA modes and the AA modes you use it with depends on quality of the TRAA. (So 2xTR MSAA would look worse than 4x TR MSAA)
From what i can gather, NV's super sampling AA applies to not only jaggies, but texture as well. The sumper sampled texture obviously means, more detail can be visible. Not to mention ATi cant to SSAA in OpenGL games according to BFG10K.
Just like hmorphone, i to wonder why ATi cards get rid of smaller branches in trees for example. Different drawing distance or something else? While NV cards show more detail of the branches. Anything thin in ATi screenshots are either more thinned out/blurred out when applying AAA. One of Nitromullet's HL2 screenshot showed this.
it's not opinion. whether they "back it up" or not is a factual statement, not a subjective one.
that's like saying "my opinion is the world is flat". and i suppose while you could indeed still label it as "your opinion", it does nothing to change the fact it's still wrong.
True, but if we are looking only at image quality we do not need a fps number to guide us. Like I said, in order to compare just image qualities the performance doesn't matter. If we are going to do a relative comparison to the image quality/performance, then an fps number would be more significant.
Keysplayr
Elite Member
I am really not understanding this whole "back it up" thing. The review is there. It shows the shots in magnified detail. That is enough for me. Obviously it is no where near enough for you. So where does that leave us? Your opinion, and mine. Pretty much done.
Question for you though. Any idiot these days can plainly see the world is not flat. We have photos from space to show this. FS provided photos that absolutely without question show detail "shortages" in ATI shots. So what do I need to back up? The article that I did not write?
but you're taking a literal stance on the title of an article which is completely irrelevant.
since we aren't purchasing high end video cards to display static images, IQ without taking performance into consideration is irrelevant. again, these are gaming cards, and regardless of how pretty the images they display are, the are worthless if they won't run your game.
the article doesn't back up your statement. how is that hard to understand?
if your statement was that nvidia had better FSAA, the article would support that, but that wasn't what you stated.
want examples?
"Generally they?re both very competitive with each other, but there are two slight differences, both along the telephone cables running above the battlefield. Specifically, near the hanging lamp, the ATI card does a poorer job of hiding the steps in the cable, while the NVIDIA card produces a nearly-smooth image. The top wire, after passing the far telephone pole, is also clearer on the 7900GTX as it passes from the pole through a brief section of sky before disappearing into the roof of the far house. On the other hand, ATI clearly does a better job with the same cable when it passes over to the house on the left as it curves upwards. Here we see more jaggies with NVIDIA than ATI."
"When we turn on Transparency AA and Adaptive AA for both cards, we see the same results as above, but with changes to the fence on the left. It loses some brightness with both methods, though the Adaptive AA on the ATI upcoming card seems to handle things somewhat better, and continues to draw the fence almost to the end, unlike the GeForce."
"The best image quality, bar none, is delivered by ATI in its 6xAA 16xAF mode with Adaptive AA turned on."
there are more examples, but i'm sure these drive my point home -- any "idiot" should see that, right?
so what you're doing is drawing conclusions based only on specific parts of the article which support you "opinion". while there are certianly areas where nvidia's AA renders slightly more details, there are areas in which ati does. it's a bit of a 'give and take' depending on the area being rendered, and the differences are slight. now, if you want to say overall you'd give nv the edge there, fine, but that hardly makes overall quality 'just as good' or better for that matter - especially considering FS didn't even cover any of the ati specific IQ advantages (HQAF, HR+AA), so not only are you tyring to draw finite conclusions based on selective portions of the article, but on an article that isn't even complete."
Keysplayr
Elite Member
Ah. All I see is someone seeking an intellectual argument over something extremely simple. Image quality is a key performance factor. You raise IQ, there is a good chance fps goes down. But that is not what the article was about was it? Seems it was just about IQ. But some bells went off as I was reading it and viewing the pics. Said to myself, "Self, this could very well explain nvidia's larger performance hit as IQ settings are raised." And this is perfectly logical. It makes sense. It fits.
Hypo Situ: I'm building a house from scratch and another guy next door to me is building the same house (we share blueprints) and we start exactly the same time.
We use the same materials. And I use cletes between every floor joist, rafter and stud. The guy next door only uses a clete every other floor joist, rafter and stud. Turns out that there are a lot of cletes missing from my neighbors house, but when all the plywood, sheetrock and siding and roofing are in place, nobody ever knows the difference and he gets done a lot faster than I do by cutting corners and omitting material. Until you tear down the sheetrock, (like firing squad did with magnified pictures) you really don't suspect there are missing cletes.
TanisHalfElven
Diamond Member
ok so this threwad has gone down the drain. please ppl the OP asked fro some advise not a was over whose got better IQ (which most ppl who posted here don't have much)
OP tjhe x1900gt is going for 195 at newegg
http://www.newegg.com/Product/Product.asp?Item=N82E16814102022
so i would recommend that.
the features like HQ_AF HDR+AA, software voltage control much ofset the sl;ight performance loss.
now if you'll excuse me i have fear multiplayer to play on my x800gt (hey it look pretty to me)
OP tjhe x1900gt is going for 195 at newegg
http://www.newegg.com/Product/Product.asp?Item=N82E16814102022
so i would recommend that.
the features like HQ_AF HDR+AA, software voltage control much ofset the sl;ight performance loss.
now if you'll excuse me i have fear multiplayer to play on my x800gt (hey it look pretty to me)
Keysplayr
Elite Member
No, what I'm saying is what I have said. Not how you decide to interpret it. What the hell is your arrogant deal this evening anyways? Who the heck do you think you are to try to tell me what a review means to me, especially when I'm outright telling you what it means to me.
I'm gonna try one last time here, Cainam.
NV renders all, or at least much more detail than ATI in the FS review as per photos. Things seem to disappear or not show up on ATI hardware in the FS article and various people here on the forums have commented and/or complained about this in there own experiences in the past.
To me, this would indicate that ATI in fact has a "lower" IQ than NV hardware. What kind of quality is it when things are not rendered properly, or at all? Not very good right?
This DOES!!!!!!!!!!!!!!!!!!!!! have everything to do with FPS hits when IQ settings increase.
If ATI renders less, of course it's going to take less of a hit when IQ setting go up. COMMON SENSE!!.
Can I make myself ANY clearer???? Hello in there!!!!
except in other situations, as i've stated examples above....
again, i've quoted examples from the article you're referring to stating nv does the same thing in different areas...
again, you keep ignoring the fact that there are areas where nv renders less/worse than ati - quoting the exact article you're referring to... so, in your words...
Can I make myself ANY clearer???? Hello in there!!!!
except in other situations, as i've stated examples above....
again, i've quoted examples from the article you're referring to stating nv does the same thing in different areas...
again, you keep ignoring the fact that there are areas where nv renders less/worse than ati - quoting the exact article you're referring to... so, in your words...
Can I make myself ANY clearer???? Hello in there!!!!
The image quality is not affected by the performance of a card. As long as the card can manage one frame, that is enough to measure image quality.
If one would attempt to measure strictly the performance aspects of a card, the image quality has an enormous effect on that and is therefore relevant then. That is why performance measureing (benchmarking) is graphed with different image quality settings.
Cookie Monster
Diamond Member
Link
More screenshots comparing AA settings for SLi/CF. I thought i would throw this into the discussion for more fuel.
Interestingly, compare the ATi/NV IQ in the HL2 screenshot. The tree branches disappear and theres no grass visible in the ATi one, compared to one in the NV shot. Game bug?
Just food for thought.
More screenshots comparing AA settings for SLi/CF. I thought i would throw this into the discussion for more fuel.
Interestingly, compare the ATi/NV IQ in the HL2 screenshot. The tree branches disappear and theres no grass visible in the ATi one, compared to one in the NV shot. Game bug?
Just food for thought.
lol... umm.. ok, whatever...
yes, performance is affected by image quality.
your contention image quality is not affected by performance is.. well, plain dumb.
that's like stating screen shots tell the entire story of image quality in a game -- that's just not true. fluidity is certainly part of image quality in any game -- they all attempt to trick our eyes (since the human eye doesn't see things frame by frame, which is why if you wave your hand in front of your face quickly your hand will 'blur') and simulate motion, and 1 fps would be a complete and utter failure.
your argument would work for comparing photoshop images, but not games. again, we're not comparing static images, and therefore the rendering rate is completely relevant.
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: 25K
-
-
-
