GeForceFX Follow-up:
Well, yesterday was certainly interesting as we were questioning our own IQ screenshots in our GFFX 5800 Ultra preview. We are still working on getting all the facts together, but we do have some initial thoughts on this issue.
First off and most important to many of you, is that our performance numbers are dead on, and show to be correct for every AA and AF setting shown utilized in our benchmarks (whew).
Second, it seems that our High Quality Image Quality comparison is correct as well as our IQ comparisons done at and above the 4XAA level.
Where we are sure they are wrong is solely on our 2XAA/QCAA screenshots and comparisons. Being that this will be a level that almost anyone purchasing the card would be using it is very important.
Going back yesterday and last night and doing in-game comparisons with the ATI 9700 Pro, it looks to us that NVIDIA's AA is still not up to par with the 9700 Pro's level of quality across the board, but certainly not as lacking as we pointed out at lower levels of AA.
Overall, we are still standing by our conclusions we drew from our initial results. The GeForceFX 5800 Ultra is loud, hot, expensive, and does not deliver the in-game image quality that the ATI 9700 Pro does. There is no doubt about the validity of that statement in my mind. This is not to say that GFFX is a bad product as it is marginally faster in many applications than the ATI flagship, but we all know that it all stopped being only about frames per second some time ago.
So to put it succinctly, NVIDIA gets a bit of credit for doing much better 2XAA than at first thought, but it surely does not change our opinion of the product. Now if we can just get NVIDIA to tell us exactly what is going on with their AA and why some forms of it seem to be done differently than others.
We are currently working on getting proper 2XAA images placed into our GFFX preview along with a reworking of the evaluation of those images.