VC is the fastest version of MM to speedrun on, but the sensitivity on the control stick is really bad compared to on N64 (aiming and ESS are generally considered way harder on VC than on N64). VC saves about 1 minute and 30 seconds in an any% run (obviously not exactly that since lag is variable, but that's the closest estimate we have for right now).
For those curious about the 1:30 figure:
-37 seconds from loading times (0.3 seconds per loading zone on average with 125 loading zones in the current any% route)
-23 seconds from cutscene lag (darkeye timed this a while back)
-About 10 seconds from faster text in first cycle (this is nowhere near exact since slight differences in mashing speed could make up to 1-2 seconds difference here)
-11 seconds from faster unpauses (N64 unpauses are always exactly 19 input frames longer than VC ones). This is assuming you use a lot more pauses than N64 for lullaby skip ISG or don't buffer it on either (VC would generally want to advance 1 frame at a time since the unpause is so fast that the game's video output can't keep up with it)
-Around 12 seconds from explosion lag (this is a horrible estimate; explosion lag is extremely variable and the number of explosives used can vary. I assumed about 1/10 second explosion lag based on timing a few and taking the average and assumed 120 explosives used which is the max you would ever use in the current route. Obviously 12 seconds is probably a huge overestimation.)
-The rest of the lag is negligable, since faster playing can easily cancel out the tiny bit of lag there is (deku palace can be done just as fast on N64 as on VC despite all the lag there for example), and VC actually lags MORE in some areas such as Zora Cape.
I timed all of this VERY extensively in Sony Vegas over the past few months (with the exception of the explosive lag obviously), so it should be pretty accurate. If someone knowledgable like darkeye wants to look this over, go ahead; it's always possible I missed something.
Last edited 03 Dec 2011, 01:27 by ING-X.