Max Payne 2: DX8.1
Max Payne 2 is based off of the MaxFX engine, modified from the Max Payne engine. New items include the rag doll engine, larger textures, and more realistic modeling. I used Fraps to monitor (and record) the fps while I ran through the first level which has plenty of bullet time, and blurring effects.

Here are the maxed out settings I used for Max Payne 2. Of course for the higher anti-alias benchmark, and higher resolution benchmarks, I scaled them accordingly.

At both 2x and 4x fsaa at 1024x768 seems like the most optimal playing resolutions and settings. Most of the frame drops were during portions of in game movies, so the actually in game playing wasn't effected as much, however it does mess up the experience if the in game graphics are all choppy.
Here is an example of an in game blurring effect. This of course takes some more rendering power and slows down the card a smidge.

Need for Speed: Underground: DX9.0
This is a gorgeous DX9 game put out by EA games. It features tons of reflective surfaces, terrain conditions, and plenty of lighting effects. I decided to go ahead and simply tested a quick race, and used the course Olympic Square and gave the game a little run around. I tested each resolution at 2 laps, with a medium computer and no traffic.
Here were the settings I used:

Motion blur, and light trails were turned off as to maintain the anti-alias setting.

You'll notice that the highs and the averages stay pretty close to the same at both of the resolutions (this game doesn't allow 1600x1200). However the real thing to note is the minimum fps on both settings. Often the game was at the lower end, especially during crash scenes or anything involving particle effects which made driving even more difficult to say the least. At 1024x768 with 2xfsaa and all the settings boosted up, the game was readily playable, and never dipped below the 30fps deadzone mark except during high intensity crashes. This was definitely the more playable resolution.
Unfortunately this game had some of the same issues that were found in Farcry, such as objects suddenly becoming bright, and objects in the distance still completely in focus. Both of these items detracted from gameplay immensely.
As you can see, the car is only moved by a few feet (in the game of course) and the building brightness changed drastically. This was present in all large objects in the distance, and made the landscape constantly change, which certainly is distracting.
Serious Sam 2: DX8.1
This one is showing some age, but when you get into a battlefield with tons of enemies, it gets pretty intense.

With each increase in resolution the game play at the maximum resolution takes about a 20-30 fps hit, at average fps it drops around 15-20fps. I noticed that this game wasn't really affected by turning on anti-aliasing. There wasn't much of a performance difference, and there wasn't much of a visual quality differences either. Looking even at 1600x1200 and 30fps and below being the limit where things get really unplayable, Serious Sam 2 looks playable even at this resolution.
Splinter Cell: DX8.1

As you can see, I didn't bother to show 4x fsaa at 1024x768 due to the intensity of this game, and also because there is a known issue with fsaa or rather the lack there of. Even at 1024x768 this card had some trouble. Splinter Cell is definitely hard for this card to swallow. Most of the run-through seems quite smooth, with a notable slowdown near the end of the test in which an explosion occurs.
NEXT
|