Do AMD's Radeon HD 7000s Trade Image Quality For Performance?

[opethdisciple]

Interesting article.

 

[arc@css]

I seem to recall there was something similar mentioned when the AMD 6000 series were released, though I've not heard or seen anyone complain about the 6000 IQ at all.

 

[MichaelM]

TL;DR version

To be clear, the discrepancies we found on the Radeon HD 7000 cards aren’t obvious until you start poring over screen captures.

AMD responded, clarified the situation, and provided proof in a fixed pre-release driver, plus a commitment to include that resolution in the upcoming Catalyst 12.4 WHQL driver. The driver it sent us for testing suggests that the issue can be solved without sacrificing performance.

I’m a bit surprised that the original findings we published in the Radeon HD 7870/7850 launch article didn’t seem to generate a significant amount of concern from our readers or the community.

 

[PGdude]

Good read, nice to hear there wasn't any funny business going on.

 

[James J]

Good to see AMD immediately sorted this with a new driver release. If it waited or didn't have a good excuse, this would have been a PR disaster.

 

[Boomstick777]

The first comment was spot on..

"What a pointless article. Title reads "Do AMD's Radeon HD 7000s Trade Image Quality For Performance?".
Answer: No, because the new driver fixes the problems while keeping the performance."

Lol.

 

[drunkenmaster]

Still find it funny that we get articles like this showing the "massive" difference in a tiny part of the screen zoomed in so it errm, looks awful either way on AMD or Nvidia hardware. They excluded comparing to a 680gtx....... maybe image quality was sacrificed by them

Most importantly, I've seen guys like [H] randomly throw in all these insane comparisons of 74xmsaa + TRAA x 672 + TXAA x eleventy billion to come up with daft performance issues, and I don't think a single one of the main reviews put up a single image quality comparison on a completely new architecture...... something I've yet to see before. Anandtech for instance normally show their AF pictures and some zoomed in AA comparisons, yet with all the new AA modes together no comparisons. Same for [H], normally you get some image quality info, this time around nothing, but we just have to assume that while enabling 3 different types of AA, nothing gets disabled or optimised, and we have to take their word it looks better when a screen shot could have proved as much?

If AMD make a mistake, point it out, people can't fix things if they don't know they are broken, I have no issue. But it seems to me a complete lack of anyone even talking about 680gtx IQ amongst the review sites........ which for a new architecture is odd, for new AA modes is odd.

 

[Freddie1980]

Good read, nice to hear there wasn't any funny business going on.

It didn't stop the author from using a loaded flame baited headline though did it?

 

[Sin_Chase]

It didn't stop the author from using a loaded flame baited headline though did it?

Gets reads, traffic, £££.

Why would you NOT use a flamebait headline?

I would. At least it was somewhat diluted with a question like approach and '?'

 

[titaniumx3]

I think review sites should start to make a point of including image quality tests as standard in all their graphics card reviews, especially considering the numerous occasions image quality has come into question over the past few generations of graphics cards from both manufacturers.

I made a thread on this forum about the anisotropic filtering 'dead zones' observed in the AMD HD5000 series back when I got my 5870. This wasn't a minor problem either, because I could notice it in actual games fairly easily, yet many people either didn't care or accepted AMD's claims of perfect, angle independent anisotropic filtering. This problem therefore never got fixed untill the HD6000 series, which was pretty annoying considering that it was a driver issue which eventually got fixed in a driver update (I think it did anyway).

 

[Gerard]

I’m a bit surprised that the original findings we published in the Radeon HD 7870/7850 launch article didn’t seem to generate a significant amount of concern from our readers or the community.

Probably because barely anyone gives a monkeys anymore, an article like this seems to crop up for either camp every so many years, the drama has just died out since any differences you'd need a magnifying glass to spot.

 

[mmj_uk]

It's good to hear AMD aren't intentionally cheating but not so good that their QA evidently hasn't improved.

 

[KillBoY_UK]

unlike in the past this was a mistake/bug rather than done on purpose and has been fixed and so is a non issue as far as i care now, although shouldn't of been there to be found in the firstplace.

Sadly I'm sure this will be used from flamers and trolls for years to come yet.

 

[Purgatory]

unlike in the past this was a mistake/bug rather than done on purpose and has been fixed and so is a non issue as far as i care now, although shouldn't of been there to be found in the firstplace.

Sadly I'm sure this will be used from flamers and trolls for years to come yet.

Well it didn't make any difference to the performance, so the flamers are wasting their time, it was a mistake missed by AMD and they resolved it and the testing shows it was not done to make the performance look better than it really is and it was just a visual quality bug really. Atleast AMD put their hands up and said yep we missed that and fixed it.


Thanks to the op for the post was a nice read. It's nice to see people spot visual quality problems and hardware makers resolve them quickly. As killBoy_UK said it really is a non issue.. and has been fixed anyway now.

 

[Lightnix]

Could've been worse, it could've caused the card to overheat.

 

[Secret_Window]

I've always felt the image quality has been superior on AMD/ATI cards. This might be slightly off topic but nonetheless still slightly relevant. I've always believed the DVI connection to the cards has produced a sharper image than Nvidia's. By this I mean general use, like the desktop, videos and images. It doesn't seem to matter either which monitor I use, I always get better results on an AMD/ATI card. This time it made even more sense as the colour profile for my Samsung screen was better on an AMD card than an Nvidia one - but then this screen favours AMD tech more like 3D etc.

 

[old gamer]

Ive always preferd the image quality of ati, but in fairness to nvidia ive not realy seen much of there gpu's for a while

 

[Purgatory]

AMD is still better for image quality, I had an amd 5870 and moved to gtx 580 and I noticed the video quality was not as good as the AMD card out of the box... it's a trade off I guess.. but a little tweeking made the gtx 580 pretty good but I still feel AMD card was better for 2D image quality, for gaming you can't tell the difference and I think the GTX 580 may even match or beat the 5870 in the image quality for 3D but not even close on 2D. That was my main worry moving back to Nvidia but I started to loose my patience with the AMD drivers and wanted things to just work well and FSX likes Nvidia cards and works better on them.. and i'm a keen simmer have to say.. I don't regret it I love the 580 and it just works without me stressing with it.

 

[V F]

AMD is still better for image quality, I had an amd 5870 and moved to gtx 580 and I noticed the video quality was not as good as the AMD card out of the box... it's a trade off I guess.. but a little tweeking made the gtx 580 pretty good but I still feel AMD card was better for 2D image quality, for gaming you can't tell the difference and I think the GTX 580 may even match or beat the 5870 in the image quality for 3D but not even close on 2D.

Can't say I've ever really noticed any difference in 2D/movies. I have a 6800 Ultra in a Mac as well as a 8800GT in another and from what I've seen of the x1950 and 5870 in 2D I wouldn't be able to tell except the different OS's. All DVI and hardware calibrated I just can't see any difference.