nVidia is at it again!!! Raise FPS and lower IQ! AquaMark3 & Det.50 compared

[Candyman]

Seems like nVidia is up to their old tricks again. DH has just posted a preview article on AquaMark3 with some benchies. All seems good until the last page where they compared the IQ on the cards.

Look at the flash presentation....:eek:


Read it HERE (http://www.driverheaven.net/articles/aquamark3/index.htm)

[Sazar]

Now, im sure most of you have read Gabes recent comments regarding the detonator 51.75s, and Nvidia's offical response but I really do have to say, having seen this first hand it confirms to both myself and Veridian that the new detonators are far from a high quality IQ set. Alot of negative publicity is currently surrounding Nvidia, and here at driverheaven we like to remain as impartial and open minded as we possibly can, but after reading all the articles recently such as here coming from good sources and experiencing this ourselves first hand, I can no longer recommend an nvidia card to anyone. Ill be speaking with nvidia about this over the coming days and if I can make anything public I will.

well... it appears that the extra fps is gained by reducing IQ... lesser precision/swapped shaders... you name it... it is likely to have happened for the IQ to drop like that...

the flash thingy was not as helpful as their TARGA files though... I must say..

I am just wondering if nvidia expects people NOT to notice that there is a loss in IQ when they are doing their 'optimizations' ?

AFAIK they recently stated no optimizations will take place that reduce IQ... relative to what I wonder... coz this is getting ridiculous...

[digitalwanderer]

Well good for DH, looks like they've finally woke up and smelled the coffee. :cool:

Too little, too late, and too many nasty bridge-burnings for me to re-embrace 'em; but I'm glad they caught a clue. :)

[John Reynolds]

3DGPU has come to the same conclusion: http://www.3dgpu.com/modules/wfsection/article.php?articleid=74&page=0

The Detonator 51.75 drivers loses even more details over the 45.33 drivers. Especially note how the green on the foreground landrover is darker than the other two shots. Now, these drivers are beta, and not available to the public, so let's hope by the time they're released, the lowered image quality won't be in them. Because frankly, I'm just shocked how much better the Radeon 9800 Pro looks over the new Detonator 50 drivers on a GeForce FX 5600 Ultra, and any self-respecting gamer will choose the image quality on the Radeon 9800 Pro anyday.

[DMA]

Almost feel sorry for the guys on nV's "driver team" these days.
Bet they'll all quit if they have to do this heavy "optimizing" for the next generation of cards too ;)

"GIVE US SOME GOOD HARDWARE TO WORK WITH FFS!!!1111"


:afro:

[rth]

there's a flash animation there showing the different cards and drivers.

it keeps saying the pictures are changing but I don't see anything happening.

where should I be looking for the changes?

[Sazar]

Originally posted by rth
there's a flash animation there showing the different cards and drivers.

it keeps saying the pictures are changing but I don't see anything happening.

where should I be looking for the changes?

the links above... there are links for each card/driver set... it shows what happens quite clearly...

I believe there is something fubar with the flash animation too since I had the same experience... but the targa files are quite clear...

[G6-200]

Originally posted by Sazar
I am just wondering if nvidia expects people NOT to notice that there is a loss in IQ when they are doing their 'optimizations' ?


No doubt, nVidia simply HAS to know that every new "optimized" driver is going to be scrutinized under a microscope for image quality degradation.
It's also hard to believe that AquaMark is going to just let nVidia walk all over their game image quality wise and not object.

G6

[Sazar]

iq with 9800pro/cat 3.7

http://www.3dgpu.com/previews/images/r9800pro_iq.png

iq with 5600u/det 45.33

http://www.3dgpu.com/previews/images/5600u_4533.png

iq with 5600u/det 51.75

http://www.3dgpu.com/previews/images/5600u_5175.png

if the images are too large mods please delete tags and leave img links...

cheers :)

[Bopple]

Where is GlowStick?
:banghead:

[Sazar]

the differences in 3dpgu's images are quite interesting.. first off the 9800pro seems to be doing a lot more work... notice foreground and background IQ...

the 2nd image with the 45.33's also renders more stuff in it than the 51.75 though the difference is not as apparent as with DH's clips...

notice the vehicle in the background and its lights in the 51.75... I wonder whats going on there...

the background is also blurrier with the 51.75's than with the 45.33's...

[digitalwanderer]

Damn, I re-read it and I just want to say nice job to Zardon and Veridan3 on the piece.

Nice job (http://www.elitebastards.com/forum/viewtopic.php?t=2013). :cool:

[ChrisRay]

Originally posted by Sazar
iq with 9800pro/cat 3.7

http://www.3dgpu.com/previews/images/r9800pro_iq.png

iq with 5600u/det 45.33

http://www.3dgpu.com/previews/images/5600u_4533.png

iq with 5600u/det 51.75

http://www.3dgpu.com/previews/images/5600u_5175.png

if the images are too large mods please delete tags and leave img links...

cheers :)


I honestly dont see a huge difference here. Not one that would distract me during gameplay anyways. Anyway someone could flash these together? It certainly makes seeing the difference in this easier


*edit* ok I do see a slight difference in ground texture from the ATI card to Nvidia card, But not seeing a huge difference between the 2 det versions on that scene.

[digitalwanderer]

Zardon at DriverHeaven:
Now, im sure most of you have read Gabes recent comments regarding the detonator 51.75s, and Nvidia's offical response but I really do have to say, having seen this first hand it confirms to both myself and Veridian that the new detonators are far from a high quality IQ set. Alot of negative publicity is currently surrounding Nvidia, and here at driverheaven we like to remain as impartial and open minded as we possibly can, but after reading all the articles recently such as here coming from good sources and experiencing this ourselves first hand, I can no longer recommend an nvidia card to anyone. Ill be speaking with nvidia about this over the coming days and if I can make anything public I will.

(I won't say it, honest. :angel2: )

[Bopple]

Originally posted by ChrisRay
I honestly dont see a huge difference here. Not one that would distract me during gameplay anyways. Anyway someone could flash these together? It certainly makes seeing the difference in this easier
No huge difference i agree. But what is the point of Highest Quality then?
Every company may adopt the same strategy. Just delicate cheatings.

[Sazar]

Originally posted by ChrisRay
I honestly dont see a huge difference here. Not one that would distract me during gameplay anyways. Anyway someone could flash these together? It certainly makes seeing the difference in this easier

well I am sure you notice the difference between the ati and nvidia cards rendering... just look @ the foreground...

further.. from the 4x series to the 5x series dets... notice the more blurry background and parts of the foreground as well...

there is something going on for sure.. its like the 3dmark03 sky issue people had with the det's not rendering properly...

DH has better pictures and both of the testers have stated clearly that the IQ difference is noticeable..

also to note... the IQ of the det 51.75 drivers may not be representative if what is claimed is true about nvidia's drivers detecting screen grabs and kicking up the quality on the shot compared to ingame viewing...

[Joe DeFuria]

Originally posted by digitalwanderer
(I won't say it, honest. :angel2: )

Actually, you can't say how big the difference is until you see it in motion. Things like high-lights due to dynamic range (which appears might be going on here...not too dissimilar to Half-Life2 high dynamic range effects), can have a drastic impact on quality in motion.

both 3dgpu and Driverheaven commented on the obvious difference in image quality.

[Hanners]

Oh no, we're not going to have the 'If you can't see it, then it's okay' argument again are we?

This is a benchmark!!! Lowering image quality for framerate gains in a benchmark is a big, big no-no.

Makes you wonder what happened to these, doesn't it:

http://www.hanners.nildram.co.uk/quality-test-procedure.gif

http://www.hanners.nildram.co.uk/automated-tests.gif

[Sazar]

Originally posted by Joe DeFuria
Actually, you can't say how big the difference is until you see it in motion. Things like high-lights due to dynamic range (which appears might be going on here...not too dissimilar to Half-Life2 high dynamic range effects), can have a drastic impact on quality in motion.

both 3dgpu and Driverheaven commented on the obvious difference in image quality.

I wish they show video captures with the differences...

maybe then it will be easier to see... though it would hammer their server I spose :(

[reever2]

Remember 3dmark? Darker sky, darker water, darker textures? They must be doing the exact same thing they did then with Aquamark

[particleman]

Originally posted by ChrisRay
I honestly dont see a huge difference here. Not one that would distract me during gameplay anyways. Anyway someone could flash these together? It certainly makes seeing the difference in this easier


*edit* ok I do see a slight difference in ground texture from the ATI card to Nvidia card, But not seeing a huge difference between the 2 det versions on that scene.

I couldn't see much of a difference initially either. But then I put it in a slide show, and saw the differences in lighting, and the small differences in the textures. The lighting is the biggest thing IMO, since that would probably be noticed even more during gameplay than static screenshots. It definitely looks like the GFFX is skimping on some lighting effects. The ppl at 3DGPU have seen it in motion and if they claim that they can see a big difference, that means a lot to me, since 3DGPU was/is a nVidia fansite, while Driverheaven has always had closer connections to ATi. I have no issue with nVidia trying to find the best combo of IQ and performance, but they need to give users the option to disable it if they don't like the IQ loss, and they definitely shouldn't be doing it in a benchmark. This goes against the very rules nVidia outlined as cheating a while ago.

[ChrisRay]

Originally posted by particleman
I couldn't see much of a difference initially either. But then I put it in a slide show, and saw the differences in lighting, and the small differences in the textures. The lighting is the biggest thing IMO, since that would probably be noticed even more during gameplay than static screenshots. It definitely looks like the GFFX is skimping on some lighting effects. The ppl at 3DGPU have seen it in motion and if they claim that they can see a big difference, that means a lot to me, since 3DGPU was/is a nVidia fansite, while Driverheaven has always had closer connections to ATi. I have no issue with nVidia trying to find the best combo of IQ and performance, but they need to give users the option to disable it if they don't like the IQ loss, and they definitely shouldn't be doing it in a benchmark. This goes against the very rules nVidia outlined as cheating a while ago.

Hey I agree. I'm not even saying. if you cant see it. It Doesnt make a difference. I'm just looking for the differences.

I really need to see it in motion. Perhaps my eyes are just not what they used to be I see very little difference between the Det versions and a rather drastic version between the Nvidia/ATI versions.

I wish I had a Geforce FX card to test it with then I could tell you.

[digitalwanderer]

Candyman, CANDYMAN!!!!

(The Dig runs across the forum and grabs up Candyman in a HUGE warm fuzzy hug, spinning him around repeatedly in his exuberance!)

By the gods and all that I hold holy it has been a while and it is DAMNED good to see ya friend! How is life treating you? :D

[Hanners]

Originally posted by ChrisRay
I wish I had a Geforce FX card to test it with then I could tell you.

I have the card, I just need the benchmark! :p

[Behemoth]

actually i see quite a huge difference from them, its like playing the game with the contrast/brightness/gamma turning half way down.... but at least its playable, still better than a slideshow experience. those not so popular/targeted games that do not get "optimized" worry me lots more :p