toms hardware puts up article on ati's filtering game

[hysperion]

you can read the article here.......
http://www20.tomshardware.com/graphic/20040603/index.html

 

[rbV5]

ATi deserves credit for the fact that the image quality of the cards is not visibly compromised by this filtering; at least no example has yet been seen of this.

 

[Megatomic]

Originally posted by: rbV5

ATi deserves credit for the fact that the image quality of the cards is not visibly compromised by this filtering; at least no example has yet been seen of this.

Way to find the silver lining behind the black cloud.

 

[McArra]

Look who and Tom... I don't trust that guy.

 

[kylebisme]

i want to compare this to Tom's article on nvidia's filtering game. what? tom's didn't have an article on nvidia's filtering game?

 

[imported_obsidian]

Wow, the same conclusion about NVIDIA's brilinear "cheating." Look at the ATI folks pounce on it when it is in their favor.

 

[imported_obsidian]

Originally posted by: TheSnowman
i want to compare this to Tom's article on nvidia's filtering game. what? tom's didn't have an article on nvidia's filtering game?

Maybe the same reason Anand doesn't have an ATI filtering review?

 

[rbV5]

Originally posted by: Megatomic

Originally posted by: rbV5

ATi deserves credit for the fact that the image quality of the cards is not visibly compromised by this filtering; at least no example has yet been seen of this.

Way to find the silver lining behind the black cloud.

That is the whole point isn't it? From what I've seen and read, looks like it improves performance without a negative effect on IQ. Its been blind "tested" in the field by owners of 9600/xt cards and has been looked at very closely over the past few weeks. We'll see more when we start seeing IQ scrutiny from the reviewers. I'm watching how it plays out, but I've yet seen this as anything but a positive for the cards that are supported with the exception of user forcing true trilinear in the drivers.

 

[Megatomic]

Yeah, I have to admit that I've not seen any IQ problems in my games since I got my 9600XT. Painkiller looked and played great and so does UT2K4. Maybe I need to get CoD to test with?

 

[LTC8K6]

I think it's neat. ATI has some talented folks working for them.

No doubt NV is feverishly writing similar stuff for their drivers.

 

[T9D]

That's low of ATI. They outright lied and hid it. Then they told reviewers to make sure they turned on trilinear on the nvidia card even though the whole time ATI didn't even do full trilinear. The tests weren't done fairly. They need to redo the tests, one with and one without the cheats. Seems without their optomizations ATI cards take quite a large hit.

 

[DAPUNISHER]

Originally posted by: tk109
That's low of ATI. They outright lied and hid it. Then they told reviewers to make sure they turned on trilinear on the nvidia card even though the whole time ATI didn't even do full trilinear. The tests weren't done fairly. They need to redo the tests, one with and one without the cheats. Seems without their optomizations ATI cards take quite a large hit.

That's the problem and all the IQ remains excellent appologies will not wish it away. While I herald the advancement, the tactics employed are reprehensible and ATi derserves a good flaming for it just as much as nV deserved the ones they've gotten for their tricks, period.

 

[Genx87]

ATI sure has a longer list of optimizations.

 

[Ackmed]

Originally posted by: DAPUNISHER

Originally posted by: tk109
That's low of ATI. They outright lied and hid it. Then they told reviewers to make sure they turned on trilinear on the nvidia card even though the whole time ATI didn't even do full trilinear. The tests weren't done fairly. They need to redo the tests, one with and one without the cheats. Seems without their optomizations ATI cards take quite a large hit.

That's the problem and all the IQ remains excellent appologies will not wish it away. While I herald the advancement, the tactics employed are reprehensible and ATi derserves a good flaming for it just as much as nV deserved the ones they've gotten for their tricks, period.

I agree what they did was shady, and they deserve some bad press. The difference is, NV degraded quality, while ATi hasnt.

 

[DAPUNISHER]

Originally posted by: Ackmed

Originally posted by: DAPUNISHER

Originally posted by: tk109
That's low of ATI. They outright lied and hid it. Then they told reviewers to make sure they turned on trilinear on the nvidia card even though the whole time ATI didn't even do full trilinear. The tests weren't done fairly. They need to redo the tests, one with and one without the cheats. Seems without their optomizations ATI cards take quite a large hit.

That's the problem and all the IQ remains excellent appologies will not wish it away. While I herald the advancement, the tactics employed are reprehensible and ATi derserves a good flaming for it just as much as nV deserved the ones they've gotten for their tricks, period.

I agree what they did was shady, and they deserve some bad press. The difference is, NV degraded quality, while ATi hasnt.

No arguments there brudda

 

[DrMindbender]

I think the law of unintended consequences is at work here. Nvidia got in trouble for mixing filtering modes. Ati took the idea and actually succeeded at making it work without IQ loss. Trouble is they might not have thought it was a good idea to market it, given the public might have a bad impression of a mixed mode just by the sound of the words. They should have bit the bullet and not lied about it.

 

[DrMindbender]

Another funny thing is that even with a completely fair comparison benchmark, ATi would still have been ahead in some of the benches, just not as much.

 

[Lonyo]

Is this the THG that may have been bought out bu some unknown media company?

 

[LTC8K6]

I recall one review that did the tests again fairly. IIRC, it didn't nelp NV much.

 

[LTC8K6]

In our X800 test NVIDIA's trilinear optimization was not disabled, so the comparable values continue to be valid and comparable.

So we can use Tom's benchmarks then? Okay.

 

[mamisano]

Why compare Nvidia's Full Trilinear to ATI's Full Trilinear when ATI has an algorithm that does not decrease IQ and increases speed. Is it ATI's fault that they have succeeded where Nvidia failed with Briliniar filtering?

 

[LTC8K6]

I wonder if this is where the speed in NV's new driver is coming from.

Nah, that would be too funny.

 

[Megatomic]

Apples to apples mamisano... Not apples to volkswagons.

 

[fsstrike]

I personally do not consider what ATI did bad at all. They made the cards faster, and no image quality has been lost.

If it looks trillinear, feels like trillinear, and has good FPS, whats the problem?

I can forgive ATI for lying, in fact, I never cared. ATI has done a good job.

 

[GeneralGrievous]

Apples to apples mamisano... Not apples to volkswagons.

If the image is the same, it is apples to apples.