Torizo skipped on NTSC! (Yes, you read that right.)

2007-05-16 07:09:30 pm

Hello. Today is the good day.

The stars aligned well and all the pixel and subpixel manipulation gave me the result I wanted to achieve for about a few months already:

- SMV (needs Snes9x 1.51, though works on any emulator if done right);
- AVI (needs FFDShow or VLC or what have you);
- YouTube (I recommend NOT watching it there unless you have no other choice, due to low frame rate concealing some movements).

Enjoy... this may well be the last big thing for this game, now that everything else is done by Kejardon, Saturn and all the others. :)

Needless to say, it was so hard to execute even with all the tools available that you can forget about doing it on a console. It will stay as one of the TAS-only tricks and can be used to make a complete no-boss/miniboss run.

***UPDATE***

Faster version:

- SMV;
- AVI;
- YouTube.

This is a copypaste straight from my upcoming low% WIP. The amount of time saved on this skip is approximately 20 seconds.

Update: Links fixed.

Thread title:

Hello. Today is the good day.

The stars aligned well and all the pixel and subpixel manipulation gave me the result I wanted to achieve for about a few months already:

- [url=http://dehacked.2y.net/microstorage.php/info/5574/torizo_skip_ntsc.smv][b]SMV[/b][/url] (needs [url=http://tasvideos.org/forum/viewtopic.php?t=5461]Snes9x 1.51[/url], though works on any emulator if done right);
- [url=http://moozooh.diinoweb.com/files/SM/torizo_skip_ntsc.avi][b]AVI[/b][/url] (needs FFDShow or VLC or what have you);
- [url=http://www.youtube.com/watch?v=4McBbP1bV-A][b]YouTube[/b][/url] (I recommend NOT watching it there unless you have no other choice, due to low frame rate concealing some movements).

Enjoy... this may well be the last big thing for this game, now that everything else is done by Kejardon, Saturn and all the others. :)

Needless to say, it was so hard to execute even with all the tools available that you can forget about doing it on a console. It will stay as one of the TAS-only tricks and can be used to make a complete no-boss/miniboss run.

[b]***UPDATE***[/b]

Faster version:

- [url=http://dehacked.2y.net/microstorage.php/info/5641/faster%20torizo%20skip.smv][b]SMV[/b][/url];
- [url=http://moozooh.diinoweb.com/files/SM/faster_torizo_skip.avi][b]AVI[/b][/url];
- [url=http://www.youtube.com/watch?v=4McBbP1bV-A][b]YouTube[/b][/url].

This is a copypaste straight from my upcoming low% WIP. The amount of time saved on this skip is approximately 20 seconds.

Update: Links fixed.

nate:

is in the group Administrator.

is in the group deutschsprachiger Moderator.

registered on 2003-09-15 06:16:34 pm.

Gender: male

Location: lubbock, tx, usa

view all posts by nate

nate

2007-05-16 07:24:59 pm

君だけの

光になると

very well done.

Spoo:

registered on 2006-06-03 12:06:23 am.

view all posts by Spoo

Spoo

2007-05-16 08:39:37 pm

Cpadolf:

registered on 2007-02-14 03:32:46 am.

Gender: male

Location: Sweden

2007-05-17 05:35:37 am

Hurr Durr

Great find, how big of a margin is it till the door makes you stuck? 1 frame 2, it looks very close.

moozooh:

registered on 2006-01-11 12:51:46 pm.

Gender: male

Location: Moscow

2007-05-17 06:33:05 am

Quote from Cpadolf:

Great find, how big of a margin is it till the door makes you stuck? 1 frame 2, it looks very close.

Actually, less.

The door timer, before it starts the countdown, oscillates between "1" and "2" each frame. Depending on what the value is by the time of triggering the bombs, you'll get either 41 or 42 frames to escape once you regain control. You need 42, it won't work with 41.

Then, you should take the bombs by abusing the walljump check, meaning you have to press away from them at the maximum possible distance (180.40xxx in my case). Picking up the bombs will put you at 180.65535.

Another important note is that you should land immediately after you take the bombs (which is why I jump from the highest point available to build up the downward speed -- I'm not sure it's possible without full downward acceleration). Ideally, you should have only one frame of air movement before you land.

Upon landing, start running with arm-pumping immediately, and jump at the last possible frame. If you get stuck in the door at X=28.0, everything is alright. If more, something is wrong.

After you succeed in getting 28.0, just load a state at the moment of the final jump and press away from the door when the timer reads 6 or 7 or something like that. If everything is alright, you'll catch the trigger.

To put it shortly, there's no margin of error at all, and you need a subpixel-perfect positioning on top of that, too.

[quote="Cpadolf"]Great find, how big of a margin is it till the door makes you stuck? 1 frame 2, it looks very close.[/quote]
Actually, less.

The door timer, before it starts the countdown, oscillates between "1" and "2" each frame. Depending on what the value is by the time of triggering the bombs, you'll get either 41 or 42 frames to escape once you regain control. You need 42, it won't work with 41.

Then, you should take the bombs by abusing the walljump check, meaning you have to press away from them at the maximum possible distance (180.40xxx in my case). Picking up the bombs will put you at 180.65535.

Another important note is that you should land immediately after you take the bombs (which is why I jump from the highest point available to build up the downward speed -- I'm not sure it's possible without full downward acceleration). Ideally, you should have only one frame of air movement before you land.

Upon landing, start running with arm-pumping immediately, and jump at the last possible frame. If you get stuck in the door at X=28.0, everything is alright. If more, something is wrong.

After you succeed in getting 28.0, just load a state at the moment of the final jump and press away from the door when the timer reads 6 or 7 or something like that. If everything is alright, you'll catch the trigger.

To put it shortly, there's no margin of error at all, and you need a subpixel-perfect positioning on top of that, too.

Cpadolf:

registered on 2007-02-14 03:32:46 am.

Gender: male

Location: Sweden

2007-05-17 06:39:16 am

Hurr Durr

Sounds easy enogh... Rolling Eyes

moozooh:

registered on 2006-01-11 12:51:46 pm.

Gender: male

Location: Moscow

2007-05-17 06:47:55 am

You are free to try it yourself. :P

Cpadolf:

registered on 2007-02-14 03:32:46 am.

Gender: male

Location: Sweden

2007-05-17 06:58:54 am

Hurr Durr

Another quesiton, after you knew what to do, how long did it accually take before you got it right?

Tonski:

registered on 2007-02-15 01:16:13 pm.

Location: Finland

2007-05-17 07:05:38 am

Woah, no doubt that Hero of the day will now be getting 24 min end time to his any% improvement.

But this trick is just easy, I just did it on console. Not a topic-worthy.

Cpadolf:

registered on 2007-02-14 03:32:46 am.

Gender: male

Location: Sweden

2007-05-17 07:18:03 am

Hurr Durr

Quote:

Woah, no doubt that Hero of the day will now be getting 24 min end time to his any% improvement.

How high was his 00:25 anyway? was it far from or close to 24?
Is he going for lowest realtime this time through to or is he choosing lowest ingame?

Quote:

But this trick is just easy, I just did it on console. Not a topic-worthy.

Hahaha

moozooh:

registered on 2006-01-11 12:51:46 pm.

Gender: male

Location: Moscow

2007-05-17 07:42:47 am

Quote from Cpadolf:

Another quesiton, after you knew what to do, how long did it accually take before you got it right?

About 30 minutes of testing to determine separate values, about 5 minutes to align the stars. And a few hours worth of failed attempts previously, which apparently took a completely different approach (for which I feel ashamed, actually).

The first time I managed to get stuck at 28.0 was January, 21st of this year. That time I completely forgot about turning around at the door, and accidentally overwrote my attempt before I had a chance to try. Further tests have shown that by using the turnaround, the room transition trigger could be picked up at 28.6xxxx, which would mean triggering it before the door slams shut. Since then, I convinced Hero of the day and Saturn it's possible, and our efforts were put into replicating my X=28.0 attempt, which apparently was the key to escape.

I tried doing it about once per every month, to no avail, until yesterday, when Saturn reminded me about it after I made some other discovery at SMR. I decided I could give it one more shot, but this time, I first gathered the maximum values I could use to ensure the success. I came up with the coordinates that would get me as far from the bombs as possible upon landing.

Then I started recording, aligned my position as precise as I could, and rocked on. When I noticed that achieving all the values I should have achieved makes me land when and where I planned to, it was already a great joy. Then, when I realized I got stuck at 28.0, my hands were shaking. And then I loaded a state a few frames after I pressed the jump button, and turned around without bothering to check if it even was a perfect frame to turn around on, and -- everything stopped, the screen started to fade into black. I was happy. :D

Quote from Tonski:

Woah, no doubt that Hero of the day will now be getting 24 min end time to his any% improvement.

Sure. Skipping Torizo + CWJ at the missile lake would be already enough for that, not to mention other possible improvements. Actually, I think it would be well possible to achieve 24 minutes even without the skip, since at least 2+ seconds could be shaved off by the end of the fight with Torizo just by greater precision.

Quote from Cpadolf:

How high was his 00:25 anyway? was it far from or close to 24?
Is he going for lowest realtime this time through to or is he choosing lowest ingame?

He got something around 00:25:30.xxx, IIRC. He is aiming for the lowest realtime but lowering the ingame minute count is also a good goal. I believe times as low as 22 minutes are possible now using heavily ingame-oriented routes (like the one Saturn is using for his any%).

[quote="Cpadolf"]Another quesiton, after you knew what to do, how long did it accually take before you got it right?[/quote]
About 30 minutes of testing to determine separate values, about 5 minutes to align the stars. And a few hours worth of failed attempts previously, which apparently took a completely different approach (for which I feel ashamed, actually).

The first time I managed to get stuck at 28.0 was January, 21st of this year. That time I completely forgot about turning around at the door, and accidentally overwrote my attempt before I had a chance to try. Further tests have shown that by using the turnaround, the room transition trigger could be picked up at 28.6xxxx, which would mean triggering it before the door slams shut. Since then, I convinced Hero of the day and Saturn it's possible, and our efforts were put into replicating my X=28.0 attempt, which apparently was the key to escape.

I tried doing it about once per every month, to no avail, until yesterday, when Saturn reminded me about it after I made some other discovery at SMR. I decided I could give it one more shot, but this time, I first gathered the maximum values I could use to ensure the success. I came up with the coordinates that would get me as far from the bombs as possible upon landing.

Then I started recording, aligned my position as precise as I could, and rocked on. When I noticed that achieving all the values I should have achieved makes me land when and where I planned to, it was already a great joy. Then, when I realized I got stuck at 28.0, my hands were shaking. And then I loaded a state a few frames after I pressed the jump button, and turned around without bothering to check if it even was a perfect frame to turn around on, and -- everything stopped, the screen started to fade into black. I was happy. :D

[quote="Tonski"]Woah, no doubt that Hero of the day will now be getting 24 min end time to his any% improvement.[/quote]
Sure. Skipping Torizo + CWJ at the missile lake would be already enough for that, not to mention other possible improvements. Actually, I think it would be well possible to achieve 24 minutes even without the skip, since at least 2+ seconds could be shaved off by the end of the fight with Torizo [i]just by greater precision[/i].

[quote="Cpadolf"]How high was his 00:25 anyway? was it far from or close to 24?
Is he going for lowest realtime this time through to or is he choosing lowest ingame?[/quote]
He got something around 00:25:30.xxx, IIRC. He is aiming for the lowest realtime but lowering the ingame minute count is also a good goal. I believe times as low as 22 minutes are possible now using heavily ingame-oriented routes (like the one Saturn is using for his any%).

Tonski:

registered on 2007-02-15 01:16:13 pm.

Location: Finland

2007-05-17 07:49:36 am

I thought Saturn hasn't even started to work on his any% original SM TAS? I know he's almost at wave in SM:R TAS and RBO is kinda stuck in Maridia or so but no any% yet.

I'd badly want to see a full run from Saturn, I just wish he completes his RBO or Super Metroid Redesign TAS quickly. :P That any% would also be very interesting if the clear time would actually get as low as 22 min.

Saturn me want full run! (and many other too i believe :) )

Cpadolf:

registered on 2007-02-14 03:32:46 am.

Gender: male

Location: Sweden

2007-05-17 08:11:29 am

Hurr Durr

Would'nt the fastest possible be the so called "kejardon run" or am i totaly wrong? Still that would not be nearly as fun to watch. If you could go sub 20 in that run there would be a reason though Rolling Eyes

moozooh:

registered on 2006-01-11 12:51:46 pm.

Gender: male

Location: Moscow

2007-05-17 08:18:55 am

Quote from Cpadolf:

Would'nt the fastest possible be the so called "kejardon run" or am i totaly wrong? Still that would not be nearly as fun to watch. If you could go sub 20 in that run there would be a reason though Rolling Eyes

Well, you are correct. A decently optimized TAS using the X-ray will surely end up being sub-20 minutes and will get less items than a proper (14 items) low%, but considering its poor entertainment value, no-one really wants it to obsolete one of the current runs published on TASVideos (and I'm working on a 14% run improvement myself). We are just running out of categories, as 3 is already much and 4 would be hardly possible with our harsh judges. :)

Tonski:

registered on 2007-02-15 01:16:13 pm.

Location: Finland

2007-05-17 08:23:44 am

As a matter of fact, the "Kejardon run" would in my opinion be very interesting to watch. Around 15 min completion or less? I hope that at least evilchen will do such a run since no one else is really working on it.

Cpadolf:

registered on 2007-02-14 03:32:46 am.

Gender: male

Location: Sweden

2007-05-17 08:33:40 am

Hurr Durr

Well i deffinetly agree that a low% 14 run would be more entertaining to watch, still that dos'nt mean that no one should do a proper Kejardon run. Seing an endtime below 20 would be magical... Still i rather se the 14% category survive more then i would like to se it be obsoleted by the Kejardon run.

moozooh:

registered on 2006-01-11 12:51:46 pm.

Gender: male

Location: Moscow

2007-05-17 08:40:16 am

Quote from Tonski:

As a matter of fact, the "Kejardon run" would in my opinion be very interesting to watch. Around 15 min completion or less? I hope that at least evilchen will do such a run since no one else is really working on it.

Well, the point is that it would skip all the highlights of any particular categoty of a normal SM TAS, including only the stuff all of them do anyway (which is nearly identical to them all up to the big pink Brinstar room), then doing a long sequence of X-ray climbs and bombjumps, then running through Tourian (as usual, except with messed up graphics), dying, and skipping Mother Brain -- which is by far the most (or the only?) interesting thing to watch. You can see it all in Kejardon's 13% no-boss demo, there isn't anything else to it, sadly. :\ All you could do with it is to skip a couple more redundant items and increase the precision.

It's hard to disagree that proper any% and low% bear far much more entertainment value to them due to showing more parts of the game done blazingly fast, using many tricks possible only with Speed Booster and alike, and other cool things.

I guess I'm finished.

[quote="Tonski"]As a matter of fact, the "Kejardon run" would in my opinion be very interesting to watch. Around 15 min completion or less? I hope that at least evilchen will do such a run since no one else is really working on it.[/quote]
Well, the point is that it would skip all the highlights of any particular categoty of a normal SM TAS, including only the stuff all of them do anyway (which is nearly identical to them all up to the big pink Brinstar room), then doing a [i]long[/i] sequence of X-ray climbs and bombjumps, then running through Tourian (as usual, except with messed up graphics), dying, and skipping Mother Brain -- which is by far the most (or the only?) interesting thing to watch. You can see it all in Kejardon's 13% no-boss demo, there isn't anything else to it, sadly. :\ All you could do with it is to skip a couple more redundant items and increase the precision.

It's hard to disagree that proper any% and low% bear far much more entertainment value to them due to showing more parts of the game done blazingly fast, using many tricks possible only with Speed Booster and alike, and other cool things.

I guess I'm finished.

Saturn:

registered on 2005-09-01 03:34:43 pm.

view all posts by Saturn

Saturn

2007-05-17 12:10:46 pm

All I can say, thank you very much for finally managing this unbelievable escape on NTSC, moozooh!

All the hard optimization-work was worth it, and I'm really happy it's finally done.

Regarding my any% TAS, I started working on it almost a year ago, but decided to start over, mainly because of a special shot in the Red Brinstar Shaft, that hero discovered a while ago, which would save more than one second.

Though now I decided to finish the SM-Redesign and the long awaited RBO TAS first, and put any% on hold until then. Reaching a 0:22 by aiming for ingame time only, is still a very very risky statement, and as it seems to me currently, still not possible despite the Torizo skip. But let's see and surprise ourselves if it might indeed work. :)

Tonski:

registered on 2007-02-15 01:16:13 pm.

Location: Finland

2007-05-17 12:22:02 pm

Quote from Saturn:

Though now I decided to finish the SM-Redesign and the long awaited RBO TAS first, and put any% on hold until then.

REVERSE BOSS ORDER RUN!! \:D/ =D> whoa (@c.bags)

Although I'm also interested in your other runs this new beats it all. I very much look forward to this one! grin new

But I just thought couldn't this run had been even faster with Torizo skipped? Although as this is more like entertainment run (while still being really fast) the Torizo fight just makes it more interesting. :P

Cpadolf:

registered on 2007-02-14 03:32:46 am.

Gender: male

Location: Sweden