There are numerous abilities that are directly related or interconnected but have notably different applications or ways to trigger them. Such things as animation lock, elevator hop, corner jump (or, say, passing through platforms/robots/zebetites) are all about the same thing, really, but the main point about them is that you'll likely not be able to predict the existence of each of those tricks logically even if knowing about the others. Besides, the differences in application justify the redundancy.
Waterball as an ability is merely a special case of exploiting the underwater physics to have the benefits of both the conserved horizontal speed and an exceptionally high bounce. I'd even say that the bounce itself is not as important as the ability to keep the speed while underwater so that you will keep it even if you hop onto the dry surface — you'll see what I'm talking about if you try to combine waterball with mock-springballing.
[EDIT]
Ok, here's a demo for you. Sometimes I feel sad that Super Metroid simply doesn't have spacious enough rooms to use things like this. :\
Kneeling doesn't really increase the speed, it only places Samus higher (=in the ceiling) initially… Seems like the jump power just isn't enough for that to work in plain SM. :\ I'll add the shinespark demo, though.
My run is currently trying to gather the 5 missiles I spent on Torizo's door back in as low time as possible. I'll release a new WIP when I will reach Super Missiles. The third and final public WIP will be released once I reach the Brinstar-Norfair elevator.
Also, go TOC bloat: 2006-12-31 vs. 2007-05-27.
Well, yes and no. :)
It isn't quite the same thing, although the pass-thru collision abuse with shinespark happens for that same reason I stated as well. I think doing it without a spark would be far more representative.
I'm not 100% sure there are thin platforms low enough from the ground to allow such thing to happen in Super Metroid (there is a lot of those in SMR, though), but freezing very vertically short enemies (slimes in Maridia?) should help.
Just so you know, I've made another major overhaul to the Super Metroid Tricks. The content is now much better grouped, there's a new tab with the relevant forum threads, a LOT of new demos*, explanations and such, and some new sections/subsections. What's most funny is the fact that, even though I wrote about 200 lines of new content during the last two weeks, I barely fulfilled the TODOs I wrote myself some time ago (see the "TODO" tab on the page). So there's much more yet to come.
* — Doesn't mean we don't need more. It's actually quite the opposite: the more text there is, the more demos we need to illustrate it. I encourage hero of the day, evilchen, P.JBoy, Tonski and others to record good examples for the things on that page that lack a [proper] demonstration. The guidelines to follow:
1) it should showcase the subject of the demonstration in its entirety;
2) it should look representative enough not to cause dumb questions.
Good examples: 1, 2, 3, 4, 5.
Post the demos here or PM them to me, and I will add them to the page.
Huh, I seem to be imagining things. When you jumped up the stairs in 1-2, it looked like it was faster then just walking. o_0
Also, for some reason I can't find the horizontal position RAM address to check that.
Because it has way too little entertainment value compared to the proper run. It's like taking everything that makes Super Metroid a cool game to both play and run, and throwing it out of the window.
Actually, my opinion is that a NBMB kejardon-run would be "the best of both worlds", since it would be faster than a normal any%, will have less items than a normal low%, will showcase the X-ray glitch and will skip all boss fights on top of that. How many items will it take is largely irrelevant as long as the amount is lower than a normal low%. Fastest completion should be the primary goal, imo.
NMB, not NBMB. Saturn's run skips all minibosses that were skippable by the time he started making it (Torizo skip wasn't known back then, and Botwoon isn't skippable even now).
Heh, you wish. :\
This is a "YOU CAN'T HAVE MORE THAN 2-3 RUNS OF THE SAME GAME HERE" kind of site, at least until Tub's idea is implemented in some way.
Possible categories, as I see them:
1) 100%;
2) any% — without X-ray glitch;
3) low% — without X-ray glitch (14%);
4) NBMB — with X-ray glitch (10%)
or
4) RBO/NMB — without X-ray glitch.
The problem with X-ray glitch is that it can be used to create not only a full no boss/miniboss run, but also both shorten the item collection to 9% (5% less than a proper low%), and shorten the any% completion time to about 17-19 minutes (about 5-7 minutes less than a proper any%). However, it looks much less entertaining as it cuts pretty much everything that makes a SM TAS entertaining away, and substitutes it with such boring sequences as killing bosses without anything but perhaps Ice beam, horizontal bombjumping and many other dumb yet inevitable things. 10% NBMB is, by far, the only run with X-ray glitch that has a certain concept (doesn't fight ANY boss), is fast and moderately entertaining.
RBO, on the other hand, is another classic SM challenge that is heavily geared towards entertainment, but the only person who's working on such a run at the moment is Saturn — and he won't be finished until 2008 for sure.
I see now. I haven't updated the package in a while. I replaced the file, redownload it from the same address. The newer version should recognize more emulator versions. If it won't, use plain Snes9x 1.43v9, not the bugfixed version.
First of all, what emulator (name, version) do you use and what is its executable's name? Then, make sure you're using the right ROM. It should be identified as "Super Metroid (????)" by the tool (each time you open a ROM, the tool tells its internal name in the bottom row). I'm not sure what you could do wrong, actually.
Experiment in each particular case. Monitor your vertical speed addresses. Keeping the overall speed above 3 pixels/frame is a good idea, but the amount of jumps is also important. The key is to distribute the jumps across the distance so that the speed won't drop below a certain level after each jump. Also, keep in mind that normal jump's force is higher than that of a walljump, so the first jump should preferrably be 1-2 frames longer or so.
Actually, just watch what I'm doing in the old Tourian escape shaft, you'll get the idea.
Yeah, that can happen often. Seems like the contact with the ground is lost under certain conditions I haven't succeeded in determining… yet. Having your subpixel position shifted by a certain amount sometimes helps. Skipping a pump in the problematic zone also may help. Moving at a higher speed helps, either, since the game seems to be adjusting the maximum shift in vertical position according to the speed you're running with (otherwise you wouldn't be able to use Speed Booster on any slope).
You can adjust the subpixel position while running without stop if your acceleration hasn't reached 2.0 px/f yet. Watch what I do with my speed in my WIP at frame ranges 750—775 and 8225—8240.
No, they affect nothing until they haven't added up to a full pixel. Except that contact with the ground thing.
You need reset-modified Snes9x.
Ask judges. I'd go for it.
Each pixel is actually divisible further, though you obviously can't see it on the screen. By optimizing subpixels, you merely achieve more precise positioning, which in some cases can save a frame, or, as in the case with Torizo, more than 20 seconds.
A quick and sudden improvement will likely not be the case with a run that took 1.5 years to make by two very experienced TASers. Then again, it's an improvement to a recommended movie; it can't hurt if it's given a priority.
Well, it's not like you aren't allowed to do your own version of a certain run… It just might turn out to be inoptimal in case the other author has come up with something you haven't. It happened many times, including the Super Demo World 120-stars run, which has been attempted by Chef Stef (he had to quit it halfways since jimsfriend's WIP was faster at that moment).
Then again, you could try improving one of the existing runs of the games you like, even if you think they're "good enough".
moozooh casts the first yes vote on a Zelda submission! Something is definitely wrong here. :P
To tell you the truth, the previous LoZ run was pretty much the only non-glitched Zelda run I enjoyed watching. This does all the same, but better and faster. Manipulation is near-perfect everywhere, damage boosts — everything is top-notch. I think it was worth it to work 18 months on such an improvement.
Yes, exactly. :)
I see no point in making it look 100% authentic since most of the time they fail to do so anyway. Why not enhance the quality while we are at that? We are already doing it for certain consoles, why not for the other?
I have read the encoder guidelines a lot of times, and have yet to see something related to the game image authenticity there. Quite the contrary, I mostly see the concerns with image quality and pleasantness. If anything, that also helps discerning emulator recordings from unassisted play.