Xenoblade Chronicles: Difference between revisions

{{testing/entry|revision=3.0-204|OS=Windows 7 x64|CPU=Intel Core e6600 @ 3.15GHz|GPU=ATI Radeon HD 5850|result=PAL patched to 30 FPS both tried fully; 9+ hours in the game. So, with 3GHz I generally had no overhead and every big situation could lead to crackling despite HLE Hack or LLE DSP, LLE DSP fared generally a lot more well, when I overclocked to 3.15GHz I had an overhead of about 20% which is nice, this could also because I had far lower memory timings 700MHz @3.15 with 5/5/5/13 (could lower even more, but tends to be more unstable, T2, could imagine T1 gives another speedup but is often very unstable with picky programs) with 3GHz I had 888MHz at 6/6/7/20 and with 2.4GHz I either use 800 or 1066MHz at respective settings. IdleSkipping ON, accurate Vbeam emulation ON (this one helped a lot) also OpenGL gives the best results for me with the least fuck ups and most accurate results, LLE is locked to core, also I assume that with a dual core clocked less than 3GHz you should disable Sound all together if your rig doesn't give you good results, this will also result in a big speedup, for me e6600 @ 3GHz wasn't playable with stable best looking settings and memory management but I still have to try out very low memory timings on less FSB, it could be viable because an increase of 150MHz from 3000 to 3150 shouldn't let me jump from 76/84/96-98% to an even 98-100% (114/126/140 without framelimit/throttle) in big locations or with dozens of monsters. Also patching the PAL version to 30FPS doesn't seem to have a negative FPS impact whatsoever. Also don't cache display lists as it's sometimes needed, EFB copies are needed for caves and other locations (you can disable it safely and only have minor glitches, like no head but hair etcpp)|tester=SunnySalvadore}}