Official "Speedrun" Timing

[Jason L. Davis]

I was wondering about this for a short while earlier today while I was on the site pondering a game to go for, Does the timing count consist of "The Moment You Press Start Until The Final Blow Of An Enemy" Or crossing a finish line of a race or anything equal to? Or must we wait until a credit or " The End" pop up?

[Kong]

It is emulated time - how many seconds the game thinks you played.

So it is from the second the game loads to the second you press F12 adjusted by your Frames Per Second.

Your Frames Per Second must remain within 95% of the real time.

[Jason L. Davis]

I understand that, but I was speaking as far as certain games themselves, like what is the most acceptable submission for a speedrun?

[ARedLetterDay]

My understanding it's whatever the actual gameplay is. For instance, Super Metroid's timing is the moment you land on Ceres Station at the beginning to the moment you escape from Zebes at the end of the game. For RE3, it's the moment you control Jill until you escape Raccoon City. So usually, anything right before an ending cutscene or credits.

[Jason L. Davis]

cool, thank you, just game variation

[TRB_MetroidTeam]

Nice question!
So, it seems it uses the general rule that "TG" uses, something like what JIG said above.

It would need to be manually edited to have the correct "time" for speedrun records purposed... too bad. Is there a way that RU can track it according to the rules? But if ChamPOW's idea is it, would be nice start using here on RU the exact time with that rule, for future comparisons in world records (yep, a pain for the referees here).

I recently played Adventure (A2600):
Adventure - Atari 2600 - Score Board
and my time is reported as: 00:00:32. I thought it would be the INP full lenght time, but after playbacking/watching, I remember it gets 33:xx, very odd.

I also know MESS reports this game/console as not 60 fps, but 59.9xxxxx fps... I did the conversions and yet it didn't match. Probably something about what Kong said... the "adjusted" and "thinks you played". I also think NTSC x PAL issue is not in effect for our emulations here right?

And about Adventure WR… TG uses whole second rule for "lower time best": (32"00 ~ 32"99 = 32"). TG record is 33" (also in ChamPOW)… my RU record is 32" for now, but if you use frame-by-frame on MESS, and count from "character moves" to "castle freeze" it can be yet lesser (31"?), despite my submission is very far from perfection.
And to finish, these "time" doubts can affect how the current RU contest rolls. IE: It will be tracked frame-by-frame to tie-breaks, or it would tie in the whole second "32" for example?

---------- Post added at 07:17 PM ---------- Previous post was at 06:44 PM ----------

Detailing Adventure frames: 1929 (end) - 22 (1st move) = 1907 frames.
A2600/Adventure = 59.922743 fps
Message in the ending of my playback: 33"86 = 1998 frames
Considering 60 fps: 1907 frames = 31"78
Considering 59.922743 fps: 1907 frames = 31"82

[Barra]

Personally I think speed runs are more hassle than they are worth, especially when getting down to the nitty-gritty of a single frame. Best left to original hardware for sure!