I apologize if I sounded harsh in my previous posts. It's not my intent to upset anyone.
SectionX, if I get some time in the next few weeks, maybe I'll write my own commentary for GuanoBowl's run. I hope to make it text-based, and actually explain things so that people who aren't familiar with OoT can understand what's going on.
No, it wouldn't be acceptable at all. Normal walking is approx. 1.5X slower than backwalking, while continued forward rolling is about 1.2X slower than backwalking (not to mention that the rolling yell gets _really_ annoying). And I'm not pulling these numbers out of my @$$--these were actually timed by dedicated OoT speedrunners. So doing anything except backwalking would tack on much more than ten minutes over the course of a 2.5-hour run.
This is a speedrun, not a tutorial video. If you have trouble following, you should (1) refresh your memory, and/or (2) use the ROUTE I provided on http://ocarinaoftime.net .
That's because I haven't updated it yet. When I do, it will be put in a separate category.
Can you please post this on the TAS corrections page in my sig?
By the way, I'll probably be able to host a low to mid-quality .avi of this run on the dedicated server I'm now sharing with my friend. We have 1000 GB of data transfer per month, and I can allocate at least 400 gigs to this. That would allow for ~2000 downloads if the file is ~200 MB, and ~1000 if the file is ~400 MB. This should be plenty for TASvideos users until the run gets posted to archive.org or something. Let me know if interested.
I'll start researching this game in a week or two when I'll have more time. I'll begin with playing the game once more to refresh my memory, and then I'll look at what the great folks over at SDA have been coming up with and hopefully try to plan out an efficient TAS route.
I think the number of "flaws" in this run may be overestimated. Take a look at the link in my sig...almost nothing, right? =P
In all seriousness, here are the ones that come to mind most easily:
* Deku Tree B1 shortcut: Saves exactly one minute.
* Deku Tree savewarp in slingshot room: Not implementable given technical limitations, but saves five seconds.
* Rupee route: About 4 seconds, the most noticeable error being climbing the fence en route to the Kokiri Sword when it's possible to just backflip over.
* Dampe's race: TSA has figured out a way to get a few seconds under 49", but he has yet to show us how. And he won't until he gets there in his in-progress segmented run, or so he says.
* Super-sliding: Super-slides could be triggered off Peahats and Stalchildren for all those treks across Hyrule Field. Saves about 30 seconds to a minute total.
* Forest Temple: Soft block-pushing and ground jumps. A few seconds.
* Gauntlets/Scale/Bottle skip with bomb bag: This could save anywhere from two to six or seven(?) minutes, but it's really hard to say because of all the different factors involved, such as manipulating the day/night cycle with the new route and other potential issues that may come up.
* Then, there are the technical limitations, such as the inability to savewarp and the subscreen delay. Those add to around 2-3 minutes, I think.
So in total I'd say that at most 15 minutes could be shaved off given our current knowledge, without counting Bottle Adventure. Since BA is version-specific, runs using the glitch should be put into a separate category.
There are two conflicting opinions here:
First, there are those who would like to see an improved run as soon as possible. I can understand that people are excited, but they fail to consider that new tricks are still being discovered and, for the time that it'd take to rerun this game, it's very possible that what is thought to be the most efficient route right now may be obsoleted later.
Then, there are those who want to wait until all the discoveries have pretty much settled down before attempting to redo this. However, what is a good waiting period? It's hard to say, but why deny the general audience a chance to see this thing, by not publishing it? Especially one that will undoubtedly attract a lot of attention given the game's popularity?
I know that the lone dissenting opinion won't make a difference, but it's the reasoning behind his vote that I was puzzled by.
And why is that a reason to vote "no?"
The Deku Tree shortcut (which saves only 1', not 4') wasn't discovered until he was already on the Shadow Temple. It just doesn't make any sense to hold that against him. Apart from newly-discovered tricks, all possible improvements are only worth seconds each.
A run of this length took more than three months to produce, and you're saying that because of a few minutes' worth of possible optimizations (most of which were discovered during the run's production), it shouldn't be published.
Are you serious? I personally like MM a lot more than OoT--it's just that OoT is so much more well-known because people saw it as revolutionary in the 2D->3D transition. MM is a very fun and somewhat different experience with its time limit, all the different forms you can take on, the sidequests, etc. Of course, there is still a lot a room for sequence breaking and glitching, and like for OoT, the discoveries are still pouring in. The game runs on the OoT engine, so you'll be backwalking, super-sliding, and doing all that good stuff. You should definitely give MM a try and think about doing a run if you enjoy it.
I'm not a big fan of 100% runs in general, though. There are way too many things to keep track of, especially in this game, not to mention that there is no difference in the ending sequence. If you want to do such a run, I'd fully support it, but to be honest it doesn't really matter to me whether a full-completion run is done at this point or not.
You have to watch the movie from beginning to end in order to really appreciate it. As AKA said, the pre-Deku Tree stuff was noticeably unoptimized, whereas a chunk of the child portion was indeed done using slowdown instead of frame advance. Simply compare the time spent playing ocarina songs to tell the difference. Dodongo's Cavern is really where it starts looking very impressive, and by the end...well, you just have to see it to believe it.
Above all else, I really admire the author's perseverance, as others have mentioned. The route went through at least four revisions mainly due to problems with reaching Hyrule Castle during the daytime (as those of you who've played the game know, the drawbridge is raised after sunset and you'd have to wait in Hyrule Field until sunrise for it to lower, which wastes about a minute in real time), and each time, GuanoBowl had to redo significant portions of his run. Also, the desynch issue caused a lot of problems, and there were dozens of times when he had to restart just because the movie went out-of-sync after a certain frame and was essentially worthless from that point on. He came close to quitting once but he pulled through--and this is now one of the greatest speedruns considering its length and overall technical quality.
So, it's been a pleasure helping out, and with any luck this run will be published soon so I can get a full AVI and start working on a list of corrections that could be made for a follow-up run. When that time comes, it'd be very much appreciated if any others knowledgeable with the game could pitch in as well.
That was a spectacular finish. Every single battle was, as you said, perfect. I didn't think it was possible to crouch-stab both Iron Knuckles at once on the fourth floor of Ganon's Tower, but somehow you managed it. That was awesome.
I'd really, really like to say that you got a perfect 10 on the last segment, but since TSA beat you by one second on the escape when entering the fourth floor (you had 2'32" left on the timer, while he had 2'33"), I'd have to give you a 9.9. =P
But in any case, that was very, very well-done. It's been quite a pleasure walking you through much of this run, and I appreciate the mention in your submission text. Would you be up for doing a Majora's Mask run?
The Last SegmentInventory Check
* Light Arrows - Used to stun Ganondorf
* Bombs - Keep 'em there to save time
* Deku Nuts - Used in phase 1 of Ganon battle
Rainbow Bridge
You'll trigger it the usual way (by walking up to the pit). Unless someone figures out a way to do the Walk of Doom and super-slide in under 40 seconds altogether, save yourself the trouble for now.
Skip the Trials
Video by Smallmuck: http://youtube.com/watch?v=523XQCOoJxE
You push the Armos in the Spirit chamber into a position such that it forms an acute angle with the wall, and then you jumpslash right through. Pretty straightforward. Be sure to use "soft-pushing" on the Armos.Ascending Ganon's Tower
General advice:
* Use your Deku Nuts to stun enemies if necessary, but be sure to have at least 10 remaining for Ganon.
* Of course, the attack power of your initial jumpslash can be preserved throughout the entire tower.
* Use rolls, sidehops, and backwalking (if you can manage it) on the spiral staircases.
1F - Entry chamber - Keese
2F - Dinolfos x2 (12 HP each)
3F - Stalfos x2 (10 HP each) - Boss Key
4F - Iron Knuckle x2 (30 HP each)
5F - Battle chamber - Lower level - Just head to the other door
6F - Ganondorf Battle (see below)
Ganondorf (40 HP)
Video by me: http://www.youtube.com/watch?v=9zYbaVb31co
This is definitely the optimal strategy for this battle. Don't Hookshot the central pillar--it's faster to simply hover. You'll have to do a bit of luck manipulation so that Ganondorf doesn't do another time-wasting slam attack once you're on the pillar. Then, fire the Light Arrow and be sure to hit him on the downslash. You'll need to do nine crouching stabs instead of four--I used the Biggoron's Sword.
Escape
If I recall correctly, the record is 1'28" remaining on the timer. 1'30" should be doable. Here are TSA's splits in his 4h 57m run:
6F - Start: 3'
5F - Enter: 2'45" (near-optimal); Exit: 2'39" (near-optimal)
4F - Enter: 2'33" (near-optimal); Exit: 2'28" (near-optimal)
3F - Enter: 2'22" (near-optimal); Exit: 1'50" (very non-optimal)
2F - Enter: 1'45" (near-optimal); Exit: 1'38" (slightly non-optimal)
1F - Enter: 1'28" (near-optimal); Exit: 1'15" (non-optimal)
By "near-optimal," I mean no hits by falling rocks, and no apparent delays.
Ganon (10+10 hits)
Video by AKA: http://www.youtube.com/watch?v=zgT1bpr_HOg
Phase 1: The most efficient known strategy involves locking onto his tail and firing one Deku Nut after another. Just try to squeeze in as many Nuts as you possibly can...and once Ganon turns around, roll under his legs and repeat. 10 hits will allow you to reclaim your sword.
Phase 2: In this phase, only the Master Sword can damage him. (I've tested this too many times using other inventory items.) So, re-equip your sword and do what AKA does--roll under his legs and try to keep him down as long as possible with quick spins. Ten more hits and you'll be able to deal the final blow.
The Final Blow & Trailing Dialogue
There isn't much to say about the climactic blow--just stand next to Ganon and press B on the first frame you possibly can. You will need to forward through all the dialogue boxes that follow, and do so as quickly as possible since they count in your time. The final dialogue box is shown below:
Once you've provided the last input, stop recording as soon as you can, preferably on the next frame. Any extraneous frames will inflate your final time, and you don't want that.
Thanks, Nitsuja.
Play the movie all the way from the very beginning and make sure it reaches the final "THE END" screen (even though there is no recorded input for the last seven or so minutes).
Sorry, but I'm not quite sure you understood.
Here is a vid. This is phase 1, not phase 2. Your positioning and angle needs to be perfect so that the stream will hit the other witch dead center right after switching targets. It's only about 50 frames faster than your manual aiming, so it's your choice if you want to redo it.
I asked him to look into that Light Arrows glitch last night. He was surprised that no one had bothered even though Kazooie came up with it weeks ago.
So Guano, are you going to consider redoing that Twinrova battle? (See this post)