PCgamer81 :
Yeah, but he and I have gone at it before. He maintains that Metro 2033 is harder to run than Crysis. He maintains that it is harder to run on MY system, as well, seeing as how he "had a similar system in the past, and Metro 2033 was more demanding."
Tomorrow I plan to do a Metro 2033 vs Crysis benchmark video with a frame counter and put it to bed. Last time I did that, he complained that I chose a weak part of Metro 2033. So tomorrow I will go a little more in depth.
Last time you compared them with DoF off (PhysX too), as you said Crysis was overall harder. Last time you decided to say what was overall harder, then proceeded to play Crysis with everything maxed and x8 AA. I'm not sure how that meant Crysis was more demanding. That was Crysis was more demanding with a big asterisk.
Either you compare both maxed, which you conceded was more demanding on Metro 2033, or if you wanted to compare the mysterious "overall" setting, then at least compare Crysis with reasonable AA levels that people would actually use (x2 or x4).
Edit: I also might add, where and what you use for the test run makes a big difference too. I always used their built in benchmark utilities. DoF or not, Metro 2033 is more demanding, but I can see how that might be unfair now, as it clearly is more demanding than most if not all places in normal play, while Crysis' benchmarks are closer to average. (this was with two unlocked 6950's).
The reason I don't believe an "overall" type comparison should be with x8 AA or perhaps x4 AA, is because this overall comparison should be done with settings where you compromise on the settings that make little difference at high performance cost, like you did by turning off DoF.
Anyways, if you redo your comparisons, please do with everything maxed on both before you try a more reasonable "overall" comparison with x2 and/or x4 AA and DoF off. I don't really wish to continue our argument, it is all in the earlier part of this post.