Well I did some testing last night to figure out what the best setup is. My testing was fairly informal though I tried my best to keep things constant while I changed settings. System Specs:
AMD64 3700+ @ 2.9GHz
2x1 GB ram
7800GTX 256mb @ 500/1300 driver tweaked 91.47
1680x1050 resolution
BF2 settings are a mix of High/med with 2xAA
Basically I had the in-game texture filtering set to High (which is supposed to be equal to 4xAF) and I would tab out of the game and change the Nvidia CP settings and then tab back in to see what happens. I would pick a spot on a map where I was getting less then max FPS (BF2 max is 100fps) and aim at a very specific spot. I would then try this with various settings to see the FPS. The whole time I had Fraps running to display the FPS. If for some reason someone was in the area or there was a lot of change happening in the area I did not use the numbers as the FPS would be to inconsistant.
Here is what I found:
Nvidia CP 4xAF_______BF2 High texture filtering
65--------------------73
66--------------------72
58--------------------62
So as you can see there was an advantage in each instance in using the BF2 settings v.s. the control panel settings. This really surprised me. While playing and watching the FPS you would see a constant difference of 4-8fps difference all the time. Certain areas it made no difference i.e. looking at a wall. I'm not sure if anyone can confirm that the BF2 settings are in fact 4xAF but I was hard pressed to notice any difference in image quality between the two.
Just for more info I did the same testing but used 8xAF in the Nvidia CP vs. the in-game BF2 settings on high texture filtering. Nothing else was changed and I used the same method to test.
Nvidia 8xAF_________BF2 High texture filtering
78-------------------85
76-------------------85
35-------------------45
33-------------------41
As you can see, and as expected, the 8xAF setting runs slower. The last two readings are low becuase I went into a room and let off two smoke nades, this really drops the FPS. However what I was surprised to see was that the fps hit was less then 10fps in every situation. The game ran just as smooth because anything below 30fps would be noticeable lag but 30fps and above is smooth. But the image quality difference in game from running 8xAF vs. 4xAF is huge!. The textures are sharper, extend farther and the game no longer looks like everything is covered in mud
The performance penalty is very acceptable, much less then I expected.
Running 8xAF the game looks great, which is no surprise as any game will usually look better at 8xAF vs. 4xAF. However people with decent systems who are not aware of this setting are missing out because the image quality comes at a very modest cost to FPS. Sorry to say that if you have a lower end or mid range system I haven't found a performance boost for you with the 4xAF vs. BF2 settings and you won't be able to benefit from the 8xAF
Well I hope that this made sence
