Well if it's possible to derive almost a quarter of a minute from the first ~8 minutes of the gameplay, and then there are about 15 minutes of pure non-stop running, I think it is utterly possible.
One frame here, one frame there — they all sum up pretty well, you know. There are enough microscopic mistakes (on a frame level; maybe not even mistakes but just unwise moves) in Terima's run to squeeze another couple of seconds while cleaning them up.
Well yeah, except for those places where you can't save your momentum to fly through the next room because of its geometry, but still can carry your charge to another area. What am I thinking about is a possible advantage of combining short charge with a diagonal superjump, with its ability to "slide" through some edges and vertical walls.
And again, if it really doesn't save more than a few frames, we still have at least two benefits: the use of an entertaining trick and time saving.
Agree with you on Phantoon, but still, doesn't it just help you to actually land your hits faster in any boss/miniboss battle? This is another serious frame shaveoff, as I see it.
Oh, didn't know that. So there is absolutely no possibility to kill them faster than that?
And what about the other "steel door" monster sets (like those of Brinstar in a small Energy tank room)?
Ah, I see. I never bothered to use them on the bosses back on my SNES, anyway. :D
Time will show. ;)
BTW, thanks for the info Graveworm.
Just a question: was the "23" number picked up accidentally or intentionally? It's my favourite number, and it surrounds me almost everywhere, even in my birth date (23 of October, 1985; 1+9+8+5=23).
As for my nickname, it came after an interjection "mooh" (which I used back in the 2002), later transformed into "Mo0zOoH" (this wacky style of typing actually reflects the intended pronounciation, and thus makes it pretty unique; I still use it at Hydrogenaudio forums, Discogs.com; my world top-20 Crimsonland appearance also came with this nick), and later "moozooh", which is a simplified version that came up somewhere around the mid-2005. It's clean of that wackiness which noone really understood, but still fairly unique. Besides, it's not case sensitive, which makes it perfectly suitable for my current gmail and livejournal accounts.
That's all, I think.
Very impressive stuff, it was so pleasant to watch that I went somehow disappointed when you suddenly stopped. %)
P.S. BTW, I'm very glad you're working on it again, Megafrost. It seems Fusion will eventually get beaten soon. :D
Do they really give a damn about everything in the game? Doubt that…
Actually, most of the time that kind of runs are done by the fans and for the fans of the game, or else they're pretty pointless.
Well, if MF was 13.5 seconds ahead of Terima at super missile/mockball room, and Saturn was 0.3 seconds ahead of MF at Ceres, I expect three to five minutes overall improvement, based on the:
• faster moving while arm pumping constantly (a sure minute only for that);
• frame perfect jumping (especially walljumping and climbing small heights);
• frame perfect falling (also using the advantages of G-force to fall faster through the exploding blocks and vertical doors);
• creative and frame perfect mockballing (ideally, it should be used every time Samus is obliged to travel a distance in ball form, just like in the end of Saturn's second video);
• more creative use of superjumps and short charge boosts when possible (and it looks pretty cool, also);
• using doppler effect on every boss possible to shoot faster;
• exploiting the possibility of hitting an enemy (I'm talking mainly about Ridley, if it's possible at all) twice with a single charged plasma shot;
• manipulating unavoidable enemies (such as ninja pirates) such a way which allows you to kill them simultaneously or without any serious slowdowns;
• just more accurate playing (eg. setting a power bomb off just before a boss becomes vulnerable to inflict additional damage without wasting any time, manipulating luck to get all the powerups you need along your way, etc.).
I think even a 5 minute improvement to Terima's run is possible this way.
But this is the same as any%, the difference is that Murder Beam glitch looks so ugly no one really likes it. And it doesn't really save much time when done in a TAS, either.
Anyway, Terima's current any% is a bit outdated: with all these new tricks and glitches uncovered it can (and should) be redone much quicklier.
Well you've obviously missed my point. I was talking about "death simple" physics of the game ("physics" means such things as G-force, moving speed, momentum, air control, etc. — NOT the programming errors!), which affect every type and direction of moving. Rockman is clearly glitchier than SM, but not a single bit trickier (as almost all the speed tricks of Rockman are in fact another glitches). Yes Rockman is already hoyled to death by Bisqwit (and AFAIR, he mentioned the possibility), but as the game itself, it doesn't require such careful route planning and attention to small yet important physic aspects, IMO.
From what I learned reading Megaman tricks guide, all the main time saving glitches are based on a similar principles (like looped screen exploit, sprite collision exploit and such), and I see no problem using them on a console (I admit I haven't test them for myself, however).
Viewed objectively? :D You're joking, right?
I am not absolute enough for that, sorry. There won't be any objective viewing for a man, ever. You must cope with that, sorry.
And subjectively, I've found everything I needed to vote for a star for this particular run.
What's the fourth one, actually? We're talking about RBO, 100% and any% ones, aren't we? However I, too, think that giving a star to more than two different types of run for the same game is wrong; one run should be as fast as possible (any%), and the other should be really fancy (like a RBO run, for example). 100% runs are made mostly for the fans, as they don't usually attract newcomers (and they are much longer than any%, as well).
Well if you want to discuss this matter with me further, we could continue via PM. That will be ok.
Well that's exactly the point people invent those silly restriction: to make the run look more fresh, more complex, more innovative and just look not like the others.
Edit: See what I mean? ;)
Moreover, there are some well known rule sets for making penalty runs of different RPGs, like the "Iron man" rule: PC must never return to starting town (or up a level depending on the game); and all the others, again depending on the game itself.
Well of course, this may be so. I find Super Metroid a game that is really able to produce such a hype mainly because of the flexibility and variability I mentioned in my previous post. If you remember Michael Flatley's almost-legendary incomplete 100% run, you'll notice exactly the same hype. ;D
Why? Because there were tons of new tricks never seen before. Because he was actually able (and proved it, as well) to push the limits so far no one could really imagine. A clear reason, provided with WIPs admired by everyone.
The funny thing is at that moment everyone (including Michael himself) thought his run was frame-perfect in every aspect, and now, if you look here, you'll notice Saturn managed to surpass even that result. So, this hype has a fairly solid base. :D Therefore I have no doubts concerning the overall quality of this run (and not just the parts of it showed in the video).
Actually, nobody said it will definitely get a star, either. ;) We just express our opinion, based not only on the material already seen (the demos), but on the whole idea of the run either.
Being in fact one of the most nonlinear action games on the SNES, Super Metroid can be beaten using hundreds of combinations of different routes, tricks and glitches. What do we see here is a run that uses all the best speed tactics to date (mad props go to Michael Flatley for arm-pump running) with a rather shocking penalty (running suitless; nothing could be worse actually) and the craziest sequence break possible in this game at all. As far as I remember, starred videos are meant to show unexperienced visitors what can be achieveable when using emulators, with these emulator-specific tricks being clearly visible and entertaining as well. The only fact that can still prevent this run from getting a star is possibe sluggishness in the Maridia (without a gravity suit, it could turn out really slow) that will stretch the timer pretty well.
And what makes you think we didn't think about that? %)
I think I said enough on my reasoning, and come to think about it:
Besides, I never really liked Megaman (especially on 8-bit era consoles), because of its linearity and, well, "arcadeness", and, as I see it, most of the time saving tricks in its TASes can be performed on a console (based on game physics, which is death simple compared to that of SM; I'm not blaming Megaman for this, but it is really so) — the main difference is that TAS is a speedrun without mistakes (I could be wrong though, but not very much). If you're familiar with SM physics and actual gameplay at least as I am, you do know that most of the SM time saving glitches are pretty hard to perform on a console, and some of them are pretty well impossible to do without frame advance (just because they need to be frame-perfect to work). The time difference between normal SM speedrun even without mistakes on a top of human possibilities (!) and a really good TAS is measured by tens of minutes, not even tens of seconds! Hardly an arguable point, isn't it? ;)
If it still doesn't convince you I'm talking about getting a star as conciously as I could (however, the final result matters most, I agree with you on it), then I don't know what else to say. For me, this run already deserves a star once complete, without a doubt. If it doesn't for you, I'm fine with that. No reason to argue about.
What about JXQ's upcoming 100% run and the next any% run (I don't know who will be the next one ready to do sub-20 min run; Saturn started to do it but switched to his current project and I don't know if he'll return to any%): well, if they will prove themselves worthy, then why not? Our point is to provide entertainment and all that stuff. ;)
Well if that helps, bumping against Giant Sidehoppers or spikes and doing superjumps (especially the latter, if done right) are by far the most suitable (at least in that area) ways to consume a moderate amount of energy per second, AFAIR. Moreover, horizontal superjumps may prove themselves useful to take you to some points faster while still doing their job (like in the huge Brinstar pink room and the long horisontal room just to the left of it).
Well it's not that fantastic since Red Scarlet and other famous speedrunners have already performed this trick exactly at that point in the game back on the SNES, simply because there is no faster way of grabbing that super missile pack. :) If you're interested, have a look at the SDA, they've got it all.
BTW Highness, I second your thought about a star. If all shall be well, I fear, the star is unavoidable, since this run will serve a pretty good example on how masterfully and creatively can the emulator features be used (which is a point of giving a star, AFAIR). :) That demos alone already deserve a star, hehe.
However, it certainly will get slow in Maridia, if at all possible without waiting. Saturn, have you finally managed to get past Mt. Doom one of the faster ways while testing (if tested at all)?
AFAIR, there were two ways to speed up on the straight way: high jump + roll + twist or short jump + twist (which, I think, is more suitable with frame advance). I played this game pretty well back then on my SNES, but I've forgot all the tracks since those times� A perfect TAS for this game will be extremely interesting, but hard� I'm afraid, just TOO hard to accomplish (the game's pretty long as well).
Edit: On a second thought, I may have overestimated the difficulty of this run. If anyone is feeling he could do it, I'll try to help.
Well I think you should consider watching movies on the emulator itself, since the emulator movie files have as low filesize as you can possibly get.
DeHackEd has a good point: the precision. I watched some Super Metroid runs from SDA recently, and they were encoded at 29.97 fps. Every time Samus rode up the elevator, she disappeared. Some shots disappeared, too. It was ugly.
100% is already taken by JXQ. I think he should consult with Michael Flatley or Graveworm or someone else for an optimal route. Too bad I don't know this game good enough to invent an optimal route myself.
Just had a look at it, and liked it.
It's a pity the game becomes terribly repetitive when TAS'd, though… You just fly through the level without any noticeable action, same with bosses.
Well, yeah. I've beat it a couple of times back then on my SNES. My friends were quite surprised, heh. They couldn't get past the 30—35 levels. :)
It's also one of that games where passwords are totally useless. You can beat it only if you build up your strength from the very beginning.
Well it becomes even hilarious sometimes, but renders the game unplayable… That's why I try to hunt unlocalised versions down every time I can.
If you're familiar with the Engrish, then know: it's many many times worse with Russian. %)
I've got a nice selection of such goofs from Fallout 2, too bad none of you know Russian good enough to really enjoy those mistakes.