Ah, the topic of FSAA, the focus of many a dinner conversation and many a fun-filled night. Full scene anti-aliasing, after all what could possibly be more enthralling a topic than the discussion of jaggies, crawlies and sparklers? Believe it or not after that first comparison we spent quite a bit of time going over this topic. After many colorful remarks, good and bad, we decided that some more testing was definitely needed. Many things went wrong on that ill-fated review. Basing the article on screenshots was only the beginning of the fall.
The initial tests, a few weeks ago, were performed on a single computer. Having to swap video cards out, and then take screen shots as best we could. As a result image comparisons between the games were a tad difficult. We never got to see both cards in action at the same time. Combine that with flawed images, and we definitely had a problem. By the amount of response we got from the first comparison there was a cause for concern. Inadequate testing systems were partly to blame for this little fiasco. Inadequate testing methods were greatly to blame.
In the best interests of all involved, we have decided to do a bit more testing. Just how much more will be quite surprising, in fact I'm surprised my eyes don't see the world in jaggies permanently. One thing is for certain; FSAA has scarred me for life. The mere mention of the term FSAA sends shivers through my body. I am forced to try to think of happy thoughts (… Garfield… Odie… and that adorable Pookie…). The FS crew had to coax me out from under a table with a can of Strawberry soda when James happened to open a box containing a GeForce2 in front of me. The mere sight of the video cards causes me pain now.
Why, you ask? After the first comparison we decided a second one was in order. As mentioned way back when, blind tests were conducted; our Pepsi challenge of sorts actually took place. Just as we were finishing up that review 3dfx sent us the shipping driver for the Voodoo5 5500. That meant that the tests had to be run again. Unfortunately, E3 came along; we were all gone for a week.
All I Wanted Was Another Pepsi!
Unable to conduct another Pepsi challenge with our fellow chums (as hard as it is to believe, they have their own work to do). Yours truly was to sit in front of the monitors and look for jaggies and all the other fun things. Just as this article was about to be posted my hopes and work were dashed, yet again. NVIDIA released Detonator driver version 5.22. I begged James to plunge a screwdriver through the base of my neck, as it would be far less painful than having to go through this ordeal again. He refused, and whipped me severely for being such a coward.
The following comparison was done using the 5.16 driver for the GeForce2 and version 1.0 of the shipping driver for the Voodoo 5500. We were going to include the 5.22, but 3dfx also sent us a new set of drivers a few days later, ver. 1.1. So instead of updating twice or having new drivers compared to old, we left it as old to old, our next update will contain all new drivers.
For those of you that read the article a few weeks ago, the next page is an explanation of FSAA. If you need a refresher, or this is your first time reading about the wonders of FSAA, by all means read the next page. If not, you might want to skip ahead and get straight to the meat of the article.