Announcement

Collapse
No announcement yet.

Interesting FAA article

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Interesting FAA article

    16xFAA article (german)

    This article shows screenshot of some FSAA testprogramm, the conclusion is that the FAA has great quality, but there are specific things which didn't get anti-aliased many times though the edges are clearly noticable. That place is the middle of the triangle to be more specific:
    Parhelia 16xFAA:


    GF4 Ti4600 4xFSAA:


    These pictures are the best example, I am wondering how many times that happen in actual games and why the FAA doesn't anti-aliase these edges, though you can really see them.
    Specs:
    MSI 745 Ultra :: AMD Athlon XP 2000+ :: 1024 MB PC-266 DDR-RAM :: HIS Radeon 9700 (Catalyst 3.1) :: Creative Soundblaster Live! 1024 :: Pioneer DVD-106S :: Western Digital WD800BB :: IBM IC35L040AVVN07

  • #2
    Yes this is interesting since I too have noticed the jaggies present in a comparison pic of an entrance in the UT2003 demo that was shown on the Anandtech.com review of the Parhelia card HERE. The demo clearly illustrates the parhelia's 16xFAA mode enabled and it seems do most of the scene perfectly with an exception to the canopy in the scene for some reason? I wonder if this is the way FAA is going to be or if Matrox will be in the works of making a fix in their driver support for this in the near future?
    Last edited by APEXNETHOR; 10 July 2002, 18:44.
    Hardcore PC gamer with a sweet tooth for EXTREME eye candy!

    Comment


    • #3
      seems like the edge detection algorithm in 16xfaa doesn´t detects intersections.
      This sig is a shameless atempt to make my post look bigger.

      Comment


      • #4
        Hmm...that really sound like what is actually happening here. I originally thought it was a detection threshold, that needed to be adjusted. Now, if it is intersections that are causing the missed edges. I hope Parhelia's 16xFAA can be fixed to detect intersections. In other screenshots, I have seen missed edges that were not caused by an intersection. So this explanation could still be wrong.
        I should have bought an ATI.

        Comment


        • #5
          could this be solved if adaptive tessellation is used? Then you have an edge in middle of those triangles. Or am I missing something?

          Dc

          Comment


          • #6
            Why the hell edges of the polys should be detected?
            They are already defined into the scene geometry!!!
            Sat on a pile of deads, I enjoy my oysters.

            Comment

            Working...
            X