This is an improvement of 161 frames (2.68 seconds) over the previous submission.

Game objectives

  • Emulator used: SNES9x 1.43 v17
  • Aims for fastest time
  • Uses warps and backtracks to save time
  • Abuses programming errors

Comments

This run features an improvement of the previous one by optimizing several levels. It doesn't have any new glitch, so a level-by-level comparison will be enough. A better lag management and score plan is present in the run.

Stage by stage comments

Misty Isle 1 (0)

Getting 3 corner boosts puts us 1 frame ahead, which is lost in the ending scene.

Misty Isle 2 (3)

Instead of using 6/5 while grabbing the shell, we use the hopping glitch. After releasing the shell, we lose speed, but it's regained in the downhill. This saves 3 frames afterall.

Misty Isle 3 (17)

As proposed by Dawn, we save 11 frames by getting the Goomba instead of the shell, in order to hit the Fireflower block. 3 frames of fadeout lag are saved.

Misty Isle 3 -> Misty Isle 4 (16)

1 frame is lost for some unknown reason.

Misty Isle 4 (20)

Delaying level entrance by 1 frame is better for Mario-sprite interaction. This allows us to get the shell earlier and additional corner boosts saves overall 1 frame. With the cape, we make more corner clips to save another frame. 1 frame of fadeout lag is also saved.

Misty Isle 4 -> Misty Castle (21)

1 frame is gained for some unknown reason.

Misty Castle (22)

1 frame is lost in the level entrance for some unknown reason. During the flight, 1 frame is saved due to a better rise cancel. In the next room, we avoid 1 frame of lag and in the boss we save another frame.

Salty Isle (24)

2 frames are saved due to additional corner clips.

Desert World 1 (27)

1 frame is saved due to additional corner clips and 2 because of a better fadeout lag.

Desert World 2b (32)

In the first room, we save 2 frames by simply getting the key and falling faster. In the other, a better corner clip and approach to the key saves 3 frames.

Desert Pyramid (44)

A ceiling boost plus a corner clip saves 1 frame. Optimizing the P-switch grab saves 5 frames and a better corner clip sequence in this room saves 2 frames. In the next one, several minor optimizations shaved off 3 frames, but we had to slow down 1 frame at the end to manipulate Reznor's oscillation. In the battle, we save 2 frames by hitting Reznor faster.

#2 Desert Castle (55)

We activate the P-switch later, as proposed by Dawn, saving 7 frames, and optimized the pipe entrance, which saves 2 more frames.

Desert Secret (57)

While entering the level, 1 frame was gained for some unknown reason. Another frame was shaved off because of small optimizations.

Desert Secret -> Desert Star World (54)

3 frames are lost because of the frame rule on the Star Road.

Desert Star World (73)

The stun glitch was executed in the left side, saving us 29 frames.

Water Star World (73)

No change.

Crystal Star World (76)

We can surprisingly duplicate the block 5 frames faster, but due to some frame rule related to Yoshi Wings, we have to wait 3 frames to activate it. In the next room, 1 frame was saved by simply falling faster.

Sky Star World (88)

Following Mister's improvement, we don't use 6/5 until the Lakitu cloud, but hop with 51 speed and regain P-meter later. Going upwards is good to avoid lag. This new route shaved off 11 frames and 1 frame of fadeout lag was saved at the end.

Ice Star World (91)

A better score reduced 3 frames of fadeout lag.

Pipe Star World (111)

Several small optimizations in the pipe entrances and while passing through the ceiling with the keys saved 11 frames. A better score during the laggy part saved 5 more. The jump at the end of the level was 2 frames faster and we saved 2 frames of fadeout lag.

Bowser's Star World (118)

A better corner clip and turn at the beginning saved shaved off 7 frames!

Backdoor Star World (144)

Again, for some unknown reason, 1 frame was lost upon entering the level, but it was recovered because we entered the pipe faster. As we didn't wait for the Spike Pillars to go up and had a better score, 23 frames were saved! We also saved 2 frames of fadeout lag.

Backdoor Star World -> Bowser's Back Door (156)

We lost 3 frames because of the Star Road rotating time, again... We are barely affected by the Bowser's Statue Fireball frame rule, so we gained 16 frames over PangaeaPanga's TAS.

Bowser's Back Door (161)

The 5 frames of improvement come from Bowser's fight, because we hit the first Mechakoopa sooner and could end the input a bit earlier.

Acknowledgments:

We would like to thank the following people:
  • PangaeaPanga: for his previous TAS, work in progress and some help
  • Dawn: for his 120 exits WIP that showed us a lot of ideas.
  • Mister: for his improvement on Sky Star Road.
  • Everyone who watches and rates the movie!

Nach: Accepting as improvement to existing run.


TASVideoAgent
They/Them
Moderator
Joined: 8/3/2004
Posts: 15667
Location: 127.0.0.1
Amaraticando
It/Its
Editor, Player (162)
Joined: 1/10/2012
Posts: 673
Location: Brazil
This is an automatically posted message for discussing submission: #3819: Amaraticando & Dnnzao's SNES Super Demo World - The Legend Continues in 17:55.96
I don't understand. Playing the movie on the emulator, the last frame of input says 17:57.72, not 17:55.96, as the site is claiming.
Masterjun
He/Him
Site Developer, Expert player (2057)
Joined: 10/12/2010
Posts: 1185
Location: Germany
Great improvements, easy Yes vote
Warning: Might glitch to credits I will finish this ACE soon as possible (or will I?)
Emulator Coder, Skilled player (1144)
Joined: 5/1/2010
Posts: 1217
Amaraticando wrote:
I don't understand. Playing the movie on the emulator, the last frame of input says 17:57.72, not 17:55.96, as the site is claiming.
The site framerates are emulator-independent, for SNES NTSC, the framerate is 39375000/655171 fps (which is about 60.1fps). Whereas Snes9x uses 60fps as the framerate.
Active player (451)
Joined: 3/21/2011
Posts: 127
Location: Virginia (United States)
Excellent job, definitely worth a yes vote.
YouTube Channel - Twitter Current projects: Sutte Hakkun, Hyper VI, RTDL, own hacking projects
Post subject: Super demo world: YouTube + lossless video file
Joined: 3/18/2006
Posts: 971
Location: Great Britain
Amaraticando
It/Its
Editor, Player (162)
Joined: 1/10/2012
Posts: 673
Location: Brazil
Ilari wrote:
Amaraticando wrote:
I don't understand. Playing the movie on the emulator, the last frame of input says 17:57.72, not 17:55.96, as the site is claiming.
The site framerates are emulator-independent, for SNES NTSC, the framerate is 39375000/655171 fps (which is about 60.1fps). Whereas Snes9x uses 60fps as the framerate.
Did it happen because I selected 'USA 1.1'? For that TAS is 161 frames faster than Panga's and it means around 2,68 secs. If I can change this unfair advantage over the previous run, I'd like to do so. === By the way, thanks for the encode and for the positive feedback. I've just edited the submission and put the level-by-level comments.
Active player (264)
Joined: 4/15/2010
Posts: 198
Location: England
Great optimization and entertaining throughout. Yes vote.
Retired smw-96, smw any%
Expert player (2639)
Joined: 12/23/2007
Posts: 831
Awsome run, nice optimization. Easy yes vote.
Recent projects: SMB warpless TAS (2018), SMB warpless walkathon (2019), SMB something never done before (2019), Extra Mario Bros. (best ending) (2020).
Emulator Coder, Skilled player (1144)
Joined: 5/1/2010
Posts: 1217
Amaraticando wrote:
Ilari wrote:
The site framerates are emulator-independent, for SNES NTSC, the framerate is 39375000/655171 fps (which is about 60.1fps).
Did it happen because I selected 'USA 1.1'? For that TAS is 161 frames faster than Panga's and it means around 2,68 secs.
The frame counts in submissions reflect that, and should be consulted to determine which is faster for frame-based emulators (at least every console emulator).
Amaraticando wrote:
If I can change this unfair advantage over the previous run, I'd like to do so.
The frame count and time is solely determined by the sent movie zip file contents. It is not an unfair advantage. And similar things have been happening for a long time with emulator changes (e.g. FCEU vs. FCEUX).
Player (138)
Joined: 9/18/2007
Posts: 389
That's a pity... We SHOULD be consequent and reject this run for using a deprecated emulator, as this run is not listed in the allowed continuances (only the 120 exits run is listed). I'll vote yes regardlessly.
Warepire
He/Him
Editor
Joined: 3/2/2010
Posts: 2178
Location: A little to the left of nowhere (Sweden)
partyboy1a wrote:
That's a pity... We SHOULD be consequent and reject this run for using a deprecated emulator, as this run is not listed in the allowed continuances (only the 120 exits run is listed).
Hacks are a little different, due to how they were made, a lot of hacks do not obey the rules of how the hardware should be used by the game, some inaccurate emulators allow such things while accurate ones will not. I cannot speak for this hack specifically, but there is a big risk it does not work on SNEShawk or lsnes.
Player (138)
Joined: 9/18/2007
Posts: 389
So... is there someone out here who tried to play this one with SNEShawk? If it's playable, it'll be quite easy to create a movie which just completes the game. So... you would vote on rejecting this movie if this game can be beaten using SNEShawk or lsnes? (And if it cannot be played on a real console, we might invent a new "console" named "snes9x" to deal with it... just some crazy idea.) It might be a hard decision to reject a good movie just for emulator choice, but I guess that Super Mario World was one of the most important reasons to deprecate Snes9x, as the improvements became highly emulator dependant...
Active player (264)
Joined: 4/15/2010
Posts: 198
Location: England
partyboy1a wrote:
I guess that Super Mario World was one of the most important reasons to deprecate Snes9x, as the improvements became highly emulator dependant...
What? The only difference between emus in smw would be handling of lag. All glitches in smw (including arbitrary code execution bugs) work across all emulators, except that one antaasas did... There is no reason to reject any smw run based on emulator.
Retired smw-96, smw any%
Amaraticando
It/Its
Editor, Player (162)
Joined: 1/10/2012
Posts: 673
Location: Brazil
The main reason I didn't wrote the any% TAS on the list was because I had already written it long ago and I was really believing we'd finish it in 2012. A comparison video will be uploaded tonight... EDIT: http://www.youtube.com/watch?v=hwjaKKei0TE
Post subject: Movie published
TASVideoAgent
They/Them
Moderator
Joined: 8/3/2004
Posts: 15667
Location: 127.0.0.1
This movie has been published. The posts before this message apply to the submission, and posts after this message apply to the published movie. ---- [2266] SNES Super Demo World: The Legend Continues "warps" by Amaraticando & dnnzao in 17:55.96
Active player (315)
Joined: 2/28/2006
Posts: 2275
Location: Milky Way -> Earth -> Brazil
Mario flies through horribly designed levels for 15 minutes. I bet you people also enjoy the Paranormal Activity movies.
"Genuine self-esteem, however, consists not of causeless feelings, but of certain knowledge about yourself. It rests on the conviction that you — by your choices, effort and actions — have made yourself into the kind of person able to deal with reality. It is the conviction — based on the evidence of your own volitional functioning — that you are fundamentally able to succeed in life and, therefore, are deserving of that success." - Onkar Ghate
Bisqwit wrote:
Drama, too long, didn't read, lol.
Joined: 5/10/2014
Posts: 1
So, since it's a hack, you can TAS this using Snes9x 1.43 and it may be accepted?
Guga
He/Him
Joined: 1/17/2012
Posts: 838
Location: Chile
NoviceTASer wrote:
So, since it's a hack, you can TAS this using Snes9x 1.43 and it may be accepted?
It won't be accepted because Snes9X 1.43 is not accepted anymore. This was an exception since it was proved that this TAS (this specific movie file) was being worked before the deprication was done.
Joined: 4/25/2022
Posts: 3
Location: brazil
great work