Top or bottom? image quality analysis

[karlotta]

As someone interested in graphics technology my answer would be : the bottom image looks best.

As someone who enjoys playing video games my answer would be : who cares, you don't walk around in an FPS staring at the floor

This "lie" has been around since Q3 came out and the Radeon DDR, or even with 3Dfx, Both cards had have much better IQ than any nVidia card at the time but where just "to slow" in game play/benchmarks. What a crock! of course you look at the IQ when playing, notice the mipmap lines, all the blurr, banding along with the awesome rendering. How could you not?! I guess if you dont have the differnt cards in you comp to compare. But IQ is more importain than fps, having to settle for less IQ to have high fps is all wrong. That is what resolutions are for.
Bottom pic is better

 

[BRiT]

OMG.4 people voted for the top picture?!?

 

[Diplo]

What a crock! of course you look at the IQ when playing, notice the mipmap lines, all the blurr, banding along with the awesome rendering. How could you not?! I guess if you dont have the differnt cards in you comp to compare. But IQ is more importain than fps, having to settle for less IQ to have high fps is all wrong. That is what resolutions are for.

Well, firstly I never said IQ wasn't important, I just made a slightly facetious remark about not spending my time staring at the floor when playing an FPS game. Perhaps most people do, but I certainly don't.

Second, saying IQ is more important than FPS is totally subjective, and depends on how and why you play video games. Go and read any of the 'pro gaming' websites and you'll see people who will deliberately sacrifice all eye-candy for a higher frame rate. My clan used to play in Division 1 of the Barrysworld and Jolt Unreal Tournament league, and you would have been laughed out of the IRC channel if you ever said anything like, "having to settle for less IQ to have high fps is all wrong".

When you are playing extremely fast-paced online games against good human opponents you really don't have any chance to take in the scenery, let alone notice a tiny bit of banding. Infact, you want the background textures to be as low as possible in order that your opponent(s) stand out more. You don't want lens flares, coronas and screen flashes distracting you.

On the other hand, when you play a game like Far Cry then you do have time to take in the scenery, and most people (myself included) are willing to except a lower frame rate for better quality graphics. It really depends.

I'm just hoping there's a meaningful point to this thread and it's not, as I suspect, someone feeling insecure about their video card because a faster one has come out...

 

[Cyberon]

Well, if i bought a 500 dollar VGA-card i sure as hell wants crisp and clear image-quality. Aldo i never will, since the price of these cards drops very hard an fast in notime. That's why i still use a 8500 Pro, bought for 120 dollars and just recently i am looking for a 9800 Pro. The best bang for the buck, that's what i am lookin for and afcource a card that will hold for 2 years. My G400 did, the 8500 did and the 9800 Pro will take me there too

 

[AlphaWolf]

I'm just hoping there's a meaningful point to this thread and it's not, as I suspect, someone feeling insecure about their video card because a faster one has come out...

I am certain this thread was quite simply a reaction to the ps3.0 farcry fud that was going around where people were led to believe or assumed that nvidia was comparing the quality of ps2 to ps3 when in fact they were comparing ps1.1 to ps3.

I suppose you could make the point, is it really faster? Perhaps they should be turning down the quality of the other card to match their output?

ps. for the 4 people that chose the top image, I didn't know b3d was available in braille.

 

[thatdude90210]

I am certain this thread was quite simply a reaction to the ps3.0 farcry fud that was going around where people were led to believe or assumed that nvidia was comparing the quality of ps2 to ps3 when in fact they were comparing ps1.1 to ps3.

Are you sure those supposed ps1.1 shots had ANY pixel shader at all? Someone posted some GF4 shots, and even those looked better than the now-downgraded-to-ps1.1 shots.

 

[Diplo]

I am certain this thread was quite simply a reaction to the ps3.0 farcry fud that was going around where people were led to believe or assumed that nvidia was comparing the quality of ps2 to ps3 when in fact they were comparing ps1.1 to ps3.

Oh, I agree, the so-called PS3.0 Far Cry pictures that were posted looked to me like fakes. I've just completed the game, as it happens, and I was the one who posted the screenshots taken on my GF4 using PS1.3 There is no way the first screenshot was using PS1.0 on the water; infact, the water looked like it was photoshopped on. The compartitive PS3.0 shot doesn't even contain any water!

However, what I would object to is people blaming this on Nvidia fud. I watched the video of the Far Cry presentation and no-where did they show those shots. What they did show they said could be done in PS2.0. What is scary is that the Nvidia fanboys will believe that PS3.0 is the holy grail (I doubt it, in this card's life time) whilst the ATI fanboys will do whatever they can to try and discredit Nvidia (since they can't win on speed, it has to be IQ or power consumption or slots taken up).

My other point would be the relatively worse IQ seen on CURRENT builds of Far Cry on the NV40 should be taken with a grain of salt, since the paths were aimed at the NV35 card, which did struggle with PS2.0 performance. Far Cry has Nvidia path for shading which is lower quality than the ATI to boost performance. However, this path is also being applied to the NV40 too, which doesn't need it, as it can cope fine. Concluding from this that the NV40 suffers from bad IQ is just wishful thinking, I'm afraid.

 

[Pete]

However, what I would object to is people blaming this on Nvidia fud. I watched the video of the Far Cry presentation and no-where did they show those shots.

The question us, who other than nV and Crytek would have access to a PS3.0 FC build and the motive to provide comparison screenshots? Plus, it's been mentioned that the same sets of screens were presented previously by nV with the first shots labelled "PS2.0." I'm viewing this situation with nV's 2003 shenanigans in mind, so I'm definitely not giving nV PR the benefit of the doubt. (Their engineers get full credit for an excellent GPU, though.)

My other point would be the relatively worse IQ seen on CURRENT builds of Far Cry on the NV40 should be taken with a grain of salt, since the paths were aimed at the NV35 card

Agreed. Given NV40s speed and capabilities, I'm assuming any bad DX9 IQ is a result of defaulting to NV3x paths, which should be corrected soon enough.

 

[AlphaWolf]

Agreed. Given NV40s speed and capabilities, I'm assuming any bad DX9 IQ is a result of defaulting to NV3x paths, which should be corrected soon enough.

I agree with that assessment also, but will the fixes come with a performance hit?

I think it's quite likely that they can be fixed without, but it just bothers me the way the whole thing looks.

Look at our wonderful nv40 ps3.0 screens.

Here is what they would look like without. (insert crappy shot here)

The fact is that clearly r300 has the best looking option for playing farcry right now and it looks a lot better than what Nvidia would have you believe.

 

[jimmyjames123]

NVDA is not claiming that PS 3.0 looks way better than PS 2.0. I think people were just initially confused because we were seeing NV3x style pictures compared to PS 3.0 pictures. As far as I know, NVDA cleared up that situation for the website that hosted the pics.

Also, there is no good reason to believe that the NV40 will be slower in Far Cry once the PS issues are fixed. I believe that there have been some tests done showing that the NV40 may even perform better with PS 2.0 than with PS 1.1. And obviously, in all the synethetic PS 2.0 benchmarks, the NV40 flies right through them. It also does quite well with Tomb Raider. Far Cry should be no exception, because PS 3.0 is supposed to be a more efficient way to render PS 2.0. We'll just have to wait and see.

 

[Diplo]

The question us, who other than nV and Crytek would have access to a PS3.0 FC build and the motive to provide comparison screenshots? Plus, it's been mentioned that the same sets of screens were presented previously by nV with the first shots labelled "PS2.0."

Well, you certainly can't blame people for thinking the worst of Nvidia - they've brought that on themselves with their previous dirty tricks.

However, this time, I'm wondering.... For a start, as far as I can tell, these screenshots originated on http://www.pcper.com/ (and other sites copied). Have you ever heard of that website before? Is it known for being reliable? Then you have to look at the screenshots - they don't look convincing to me. Let's remind ourselves :

The first obvious thing about these supposed before and after shots is that the second one is competely missing the water from the first! Infact, it seems pretty obvious to me the water in the first-shot has been photoshoped in. It's totally out of context in the scene, and doesn't look anything like the water does in 'Far Cry' (I've finished the game, remember!). Even with PS1.x, on low settings, the water looks better than that - (check this shot I took on my GF4 for comparison).

The second thing is the lighting is wrong on both pictures. The roof is totally pitch-black, yet the wall is brightly lit. As I say, I've played Far Cry and even on a PS1.x card the lighting is far, far better than that crudeness.

So, you have to ask yourself, would Nvidia be that stupid to release pictures that are so obviously not representative of how Far Cry looks using PS1.0 ? Knowing that anyone who had played the game would recognise this? If so, why not leak the screenshots to some well-known sites, for maximum publicity? Sorry, I just don't buy it... Those shots looks phoney to me.

 

[AlphaWolf]

Water isnt missing in the 2nd shot, its just transluscent.

 

[nobie]

Maybe the first shot was taken on a Voodoo Banshee 8) . But seriously I think most of the time these IQ comparisons get out of hand. In the original post, the 6800 shots definitely look worse, but only upon close inspection and I personally would never notice the difference while playing a game. It's kind of like the difference between 100 FPS and 101 FPS, in the end it doesn't really matter.

 

[hmmm]

Sorry, I just don't buy it... Those shots looks phoney to me.

Well, the site does claim to have gotten them from NVIDIA. The updates where they checked on the PS version in the "before" shot says "We've got an inquiry in with NVIDIA about this..." Then the second update says, "NVIDIA said it was PS 1.x not PS 2.0."

We also have seen the shots in NVIDIA marketing material elsewhere (one time the "before" shot was labled PS 2.0).

Clearly, the comparison is jacked, but it seems to me that NVIDIA is the most likely culprit for having done it, not some no-name web site. And as far as the site being no-name and not a popular site, it was pretty predictable they'd get around regardless (viral marketing and such). And we all take Anand and Kyle and Tom with a big grain of salt, but do you really think they would be foolish enough to fall for that bait-and-switch? I shouldn't think so.

My guess is NVIDIA pulled the wool over some young web admin's eyes because they couldn't get anyone more knowledgable to post the pictures. They knew they'd get plenty of circulation on fora anyway. Besides, if you gave it to Anand, it would be too obvious that NVIDIA was spoon-feeding the FUD. Perhaps they had hoped the PC Perspective web site wouldn't come back to them asking about the first image or that they would at least not say "We asked NVIDIA about their picture..." Who knows? I'm just saying it seems more likely, especially given the last 18 months, that NVIDIA is being just a little underhanded here. That's stupid because they needn't be this time. They actually have a good card to sell now.

 

[Jima13]

"My guess is NVIDIA pulled the wool over some young web admin's eyes because they couldn't get anyone more knowledgable to post the pictures"

I doubt that's the case, Ryans been running Amdmb.com and Pcper is the fairly new sister site

 

[thatdude90210]

I doubt that's the case, Ryans been running Amdmb.com and Pcper is the fairly new sister site

And he couldn't see that Quake 1 water in the picture?

 

[silence]

and doesn't look anything like the water does in 'Far Cry' (I've finished the game, remember!). Even with PS1.x, on low settings, the water looks better than that -.

i played for few minutes on friends notebook with GeForce 5200 Go...it looks like that. no reflections, nothing, just something blue and blurry.....

his settings were 1024x768, all LOW, no AA & AF.

i was shocked how bad does it look on his comp after completing the game on mine.

 

[Pete]

Well, you certainly can't blame people for thinking the worst of Nvidia - they've brought that on themselves with their previous dirty tricks.

So I'm blameless! Excellent.

However, this time, I'm wondering.... For a start, as far as I can tell, these screenshots originated on http://www.pcper.com/ (and other sites copied). Have you ever heard of that website before? Is it known for being reliable?

I think PCPer was started by a former HardOCPer. As it's new, though, I don't think anyone can speak of the site's reliability.

So, you have to ask yourself, would Nvidia be that stupid to release pictures that are so obviously not representative of how Far Cry looks using PS1.0 ?

My question (on which my suspicion rests) is why PCPer would have to ask nV for clarification? My inference is that nV provided the shots, and PCPer was double-checking with its source.

 

[Calavaro]

I think PCPer was started by a former HardOCPer. As it's new, though, I don't think anyone can speak of the site's reliability.

Sean is a former [H] reviewer. Mainboards if I recall correctly.

 

[epicstruggle]

OMG.4 people voted for the top picture?!?

I guess if you havent cleaned your bathroom (tiles) in a while, it will look like crap, hence the votes for the top pic.
epic