Metroid Prime TAS

Edit history:

Baby Sheegoth: 2015-03-16 05:07:27 pm

2015-03-16 05:00:35 pm

Well, like I said though, they accept runs with unclean SRAM states. You are correct however, it mostly just matters what personnal wants to do :)

Hazel

Edit history:

Hazel: 2015-03-16 05:40:14 pm

2015-03-16 05:39:39 pm

rocks, locks, and invisible blocks

Quote from MilesSMB:

you backtrack through Chozo Ruins after World to get to Lifegiver

is this certainly faster than going through Tallon? makes sense to me that would be, but I haven't got around to timing late game stuff like this yet. (basically asking if I should waste my time actually timing the two, or if there's no way it could be slower.)

MilesSMB

2015-03-16 05:48:38 pm

In the current route, it's only 5-10 seconds slower than detouring for it on the way back to the ship after Magmoor. In the new route you'd have to add going to/from the ship to Tallon Canyon on top of that, so it should be faster... I think the Chozo backtrack has a lot of loads, though, so it may gain less time from optimization. I guess you should compare the two just in case.

T3

2015-03-16 06:02:53 pm

keep in mind that you can actually get world during your first pass through hote (good luck finding any reliable method though). you'd obviously only do this if it turns out going through tallon is faster, otherwise you'd be passing through hote on your way to lifegiver anyway.

Aruki

2015-03-16 06:10:05 pm

Quote from T3:

keep in mind that you can actually get world during your first pass through hote (good luck finding any reliable method though). you'd obviously only do this if it turns out going through tallon is faster, otherwise you'd be passing through hote on your way to lifegiver anyway.

so wait, how does this help you get World?

MilesSMB

2015-03-16 06:16:47 pm

If you're slow enough, it magically warps you from the cutscene into Elder Chamber.

T3

2015-03-16 06:27:45 pm

it's possible for the morph ball to stop rolling before it hits the cutscene trigger at the end there (though very, very rarely). this then allows you to use the statue while the trigger is still present in the aisle which turns the statue invisible and allows you to get world early.

MilesSMB

2015-03-16 06:28:35 pm

No, clearly it magically warps you.

MrSpEeDrUn

2015-03-16 07:08:06 pm

MP2 also a bit MP1 speedrunning

just ask some people on twitter to tweet your run and youll get a lot views too.

TheMG2

2015-03-16 07:09:06 pm

I'm pretty sure he won't need to ask anyone.

Hazel

Edit history:

Hazel: 2015-03-16 08:48:58 pm

Hazel: 2015-03-16 08:42:03 pm

2015-03-16 08:40:44 pm

rocks, locks, and invisible blocks

That early World looks/sounds crazy. I know it's still a while until I get there, so I might change my views, but I feel like this might be one of those route changes that future TASers can worry about. Unless someone does find a way to recreate that semi-consistently, being able to save a load state isn't going to actually make pulling it off any easier, not if I'm just going in blind and hoping it works.

Edit: also here's what I've got done so far (please critique)

I want to say it's around 5-6 hours work... And the final room was just a test.

Quote:

final room isn't optimized, I was just testing if jumping there is also faster than rolling (it is)

T3

2015-03-16 09:05:06 pm

soooooo... that was completely awesome. i'm probably going to end up watching that 20+ times before you upload the next part.

also no, you're definitely finishing this tas, i don't even care.

Baby Sheegoth

2015-03-16 09:45:28 pm

Quote from petrie911:

Sub-hour any% and 21% would both be pretty awesome to see.

hehehehehe

Also yeah, honestly this will never be perfectly optimal so I think we can just focus on optimal as possible :)?

JustinDM

2015-03-16 10:34:07 pm

I'm probably (most likely) way in over my head here, but what is the likelihood of a 21% TAS? Because that would be sick as fuck.

BioSpark

2015-03-16 11:12:06 pm

it's just a matter of there being someone who's willing to do it. not likely at this point

Daryoshi

2015-03-17 01:32:50 am

loving low%

Quote:

I'm probably (most likely) way in over my head here, but what is the likelihood of a 21% TAS? Because that would be sick as fuck.

Metroid177 and I are still finding various speedtricks for that route which is necessary to do for segmented runs. For 21% TASes, as of now, you'd have to do a shitload more testing even than for any%/ 100% which with 0:54 and 1:13 have come to a high richness of speedtricks that can be thrown in to make e.g. subhour in any% possible even if you screw up the final bosses, etc.
Besides, there's a slight chance you can use a TAS-only route and idk if that one would even be faster.

After all, though, I'd ultimately like to watch a 21% TAS the most but it will be in a year or so at the earliest that someone "can" start a 21% TAS. Probably it's 2-3 years when Metroid177 will have hopefully finished his 21% speedrun; that speedrun can then be taken as a basic fundament and out from that a TASer could explore further.

I want to say that my semi-TASed 1:31 in normal mode 21% was playing as a human and savestating several times per room and re-doing the room several times in order to find out which is perhaps the fastest way to cross a certain (part of the) room. That being said, in a run that is this long, I can see approaching sub-hour as totally possible for TAS and considering there're so many 21% (speed)tricks yet to be found, I'd say sub-hour should not be too far away from possible if not fully possible.

Btw awesome job, personnal! That "not-rolling-but-jumping" looks so sick!

[quote]I'm probably (most likely) way in over my head here, but what is the likelihood of a 21% TAS? Because that would be sick as fuck.[/quote]

Metroid177 and I are still finding various speedtricks for that route which is necessary to do for segmented runs. For 21% TASes, as of now, you'd have to do a shitload more testing even than for any%/ 100% which with 0:54 and 1:13 have come to a high richness of speedtricks that can be thrown in to make e.g. subhour in any% possible even if you screw up the final bosses, etc.
Besides, there's a slight chance you can use a TAS-only route and idk if that one would even be faster.

After all, though, I'd ultimately like to watch a 21% TAS the most but it will be in a year or so at the earliest that someone "can" start a 21% TAS. Probably it's 2-3 years when Metroid177 will have hopefully finished his 21% speedrun; that speedrun can then be taken as a basic fundament and out from that a TASer could explore further.

I want to say that my semi-TASed 1:31 in normal mode 21% was playing as a human and savestating several times per room and re-doing the room several times in order to find out which is perhaps the fastest way to cross a certain (part of the) room. That being said, in a run that is this long, I can see approaching sub-hour as totally possible for TAS and considering there're so many 21% (speed)tricks yet to be found, I'd say sub-hour should not be too far away from possible if not fully possible.

Btw awesome job, personnal! That "not-rolling-but-jumping" looks so sick!

Hazel

Edit history:

Hazel: 2015-03-17 02:34:41 am

Hazel: 2015-03-17 02:31:45 am

Hazel: 2015-03-17 02:31:28 am

Hazel: 2015-03-17 02:30:37 am

2015-03-17 02:30:26 am

rocks, locks, and invisible blocks

I don't know how to go about posting progress with this TAS. I want to post every room somewhere for people to help think of potential better strats, because unlike with SA2, I don't have a group of knowledgeable TASers to poke over Skype whenever I'm not sure about something.

Here's where I'm at now.

I'm a little confused currently, I need to double check a lot of things (which I'll do tomorrow) but based on another test I did that used different dashes and was 6 frames slower, that door should have opened sooner with this method. Unless I'm completely mistaken, I hit the loading zone 23 frames faster, not 6....
where the loading trigger is:

T3

2015-03-17 09:29:38 am

door loading is not 100% consistent down to the frame. i've noticed in very many instances where doors can open a handful of frames faster/slower depending on how you pass through the room/section before hitting the trigger. this is especially noticeable for automatic loading rooms where you can just time the amount of time it takes from the transition to the door opening. there are also some cases where pausing the game can make a door open faster like the second door during the pal escape. this doesn't really affect console runs because it's pretty negligible and probably just balances out by the end of the run anyway but for a tas this obviously matters. basically loading is a mess and it doesn't exactly help that you're using dolphin.

MrSpEeDrUn

2015-03-17 09:52:48 am

MP2 also a bit MP1 speedrunning

Sounds like fun. I bet you lose your motivation before ice beam. aiwebs_011

Skull64

2015-03-17 10:59:38 am

I approve of everything that has happened here

Hazel

Edit history:

Hazel: 2015-03-17 04:35:19 pm

2015-03-17 04:34:16 pm

rocks, locks, and invisible blocks

That's going to complicate things... I always figured the load difference on console was actually from imperfect consoles (in most cases). I know there are rooms like GTH where the way you go through them can really impact how long it takes for the door to open, but I always figured that was due to avoiding load triggers and/or weirdly placed triggers, not seemingly randomness.
The fact that a TAS recording can play back without desyncs at doors suggests that playing the room the exact same will have the exact same amount of load delay at the door, or else a TAS would be impossible to sync. That confuses me... because we've got two conflicting things going on here.

kirbymastah

2015-03-17 05:14:41 pm

so... time to test every possible input and see which loads faster? :^)

Baby Sheegoth

2015-03-17 07:03:59 pm

Quote from itsPersonnal:

That's going to complicate things... I always figured the load difference on console was actually from imperfect consoles (in most cases). I know there are rooms like GTH where the way you go through them can really impact how long it takes for the door to open, but I always figured that was due to avoiding load triggers and/or weirdly placed triggers, not seemingly randomness.
The fact that a TAS recording can play back without desyncs at doors suggests that playing the room the exact same will have the exact same amount of load delay at the door, or else a TAS would be impossible to sync. That confuses me... because we've got two conflicting things going on here.

Wow that's actually something I didn't even think about... it doesn't desynch? That is intense

Littlenino

2015-03-17 07:05:43 pm

Well, at least so far with the first ~4 rooms, it has yet to desync

Baby Sheegoth

2015-03-17 07:09:11 pm

Quote from Littlenino:

Well, at least so far with the first ~4 rooms, it has yet to desync

Well didn't Personnal say that he tried it with other rooms that didn't desynch? Maybe I'm dumb.