We've seen an "X, 100%" TAS on our site. Now here's a run with the opposite goal, beating Mega Man X4 without magic shoes, magic gloves, etc., but only with the magic of TAS.

Game Objectives

  • Completing the game with no hearts, E/W/Ex-Tanks, or any piece of armor.
  • Movie stops after skipping the final animation. Since all the previous animations are skipped, I think I might as well skip the final animation, and get to the ending credits faster.

Low% and 100%, How Different?

This low% TAS should be about 1 minute shorter than sparky's 100% TAS. Movie time seems to show this run being 20 seconds longer than sparky's 100% TAS, but mostly because of the emulator difference. BizHawk has a really long intro, and produces way more lag frames. It's hard for me to compare and calculate it accurately, because lag differences happen everywhere.
In general, regular stages are usually faster, because no need to spend time collecting items, making regular gameplay more fluid, but boss fights usually take longer.
Arm Parts make the most difference, because it allows charged special weapon shots, as well as four Charge Shots at once, making them very useful against bosses like Sigma first form, Cyber Peacock, General, and Split Mushroom. Body Parts reduces damage, not only the amount but also time, and Nova Strike does more damage on almost all bosses, especially Storm Owl. Foot Parts are also useful on bosses like Web Spider and Gunner Sigma.
Boss order is different from sparky's TAS. But at this time, I'm not sure what the perfect boss order for 100% would be. In comparison, the route for low% TAS is much more straightforward.
The difference could be larger, because I'd predict that Nova Strike would be used more often for future 100% TAS.
I think it would be great if we could have "X, 100%", "X, low%", and "Zero, any%" as categories on TASVideos for this game.

How I Started This Project

At first, I was looking for improvements over sparky's 100% TAS, and I found some. But the main issue for me in the 100% TAS is route (or boss order) choice. Some routes may end up really close, so I have to test many boss fights in various ways.
But then I thought, while having to test those boss fights anyway, I might do a low% TAS first. During the whole month of making this TAS, I've learned a lot, and hopefully, all the improvements I've found can be used in the future 100% TAS.

New Tricks

Faster Charge

By pressing square and triangle cleverly, 1 frame can be saved for every 2 continuous charged shots, and 2 frames can be saved for every 3 continuous charged shots.
It's used a lot in this run, but it only works when no special weapon is selected, so when you see I changed to no weapon mode, usually it's because of this.

Soul Body of Death

Mainly used for Storm Owl miniboss (Generaid Core), Storm Owl fights, and breaking boxes during the final boss fight.
Normally, Soul Body does 1 damage every 9 frames, but if you combine it with direction change, it can do 5 damage every 9 frames.

Frost Tower Double Damage on Jet Stingray

Normally Frost Tower does 4 damage on Jet Stingray. Because the frozen animation takes a long time, hardly any speedrunner or TASer actually used it.
But when Frost Tower and Jet Stingray move reletively, every 1/9 chance, double damage occurs. This saves lots of time in this run, and it would save time in 100% TAS as well.

Lighting Web Pause Cancel

Used to skip Split Mushroom miniboss (Tentoroid).
Pausing is the only known way of canceling Lightning Web without touching it, except using Nova Strike.
With this technique, you could climb high anywhere. I've tested it almost everywhere, but unfortunately, the only time save I've found is this miniboss skip, because most ceilings in this game are sealed off.

Faster Wall Climbing with Rising Fire

Surprisingly, Rising Fire can sometimes save time during regular gameplay. It's used once in Slash Beast stage 2 where there's no train, and once in Web Spider stage 1, to make X climb on the wall faster.

Stage Comments

Volcano (Magma Dragoon)

Stage 1 contains several improvements over sparky's TAS, including climbing the last slope without taking damage.

Snow Base (Frost Walrus)

Stage 1 miniboss (Eyezard) is very hard to deal with without using charged Rising Fire attack. Unfortunately, the miniboss can't be killed in one cycle, because it will close after being hit too many times by regular shots.
Stage 2 is 50 frames faster than sparky's TAS, due to a different strategy of using charged shots to break the ice blocks.

Marine Base (Jet Stingray)

The strategy for Jet Stingray fight is to use 2 fully charged shots at first, and then several double damage Frost Tower, and another 2 fully charged shots, and finish it with one Frost Tower.
Screen Y position affects where Jet Stingray shows up above, so screen scrolling is used to make Jet Stingray appear faster, saving time than only doing it at the top of the screen.

Military Train (Slash Beast)

There are frame rules at the end of stage 1 and the middle of stage 2 when the train stops, so I have several frames to spare for entertainment.
I love this stage because it has the most enemies, and I got many lives through luck manipulation (although still quite rare).

Memorial Hall (Colonel)

3 charged shots are used along with Frost Tower. Colonel sometimes loses invincibility during change of moves, sometimes for only 1 frame. So every opportunity needs to be used to maximize the damage of Frost Tower.

Jungle (Web Spider)

Web Spider quick kill without X's foot parts is really tough, because every Twin Slasher shot counts, with almost no room for error.
Web Spider has a maximum 64 frame window above the screen with no invincibility time, starting after hitting it when it is just about to go above the screen.
The whole setup also requires heavy RNG manipulation, because the RNG that makes the Spider drops to a certain vertical position, without web attack, and then reappears on the same required spot is quite rare.

Bio Laboratory (Split Mushroom)

In stage 1, the vertical room takes quite some effort. New strategies are used, ending up 37 frames faster than sparky's TAS, despite not having charged Rising Fire shots.
Stage 2 has a long vertical auto-scroller room, so I tried my best to entertain.
I've tested multiple strategies for Split Mushroom, and the fastest solution uses the period when Split Mushroom is in the sky and maximizes the damage of Lightning Web.

Cyber Space (Cyber Peacock)

Both stage 1 and 2 contains small improvements over sparky's 100% TAS.
Soul Body is used 3 times in Cyber Peacock fight, because although Soul Body does greater damage, using Soul Body would cause Cyber Peacock to disappear longer than usual, so charged shots are the main weapon against Cyber Peacock.
Taking damage twice is to cancel Soul Body, so 1 frame can be saved from faster charge each time.

Air Force (Storm Owl)

This is the first TAS that finishes stage 1 without using the Eagle Ride Armor. It's 69 frames faster than sparky's 100% TAS. Although I like the Eagle Ride Armor, not using it can also be an interesting strategy for TAS only.
Stage 2 has awesome usage of Soul Body against miniboss (Generaid Core) and Storm Owl. Storm Owl loses invincibility for about 64 frames after getting hit by Aiming Laser and a bullet shot afterward. But it flies away very fast, so I can't kill it in one cycle, so I took damage to cancel Soul Body, and add another Aiming Laser and a half-charged shot to finish the fight.

Space Port (Colonel)

For the second Colonel fight, I used a different strategy than the first one, because this time Colonel makes his moves faster. For the new strategy, I manipulated the time and place firing Frost Tower to make Colonel skip the blocking animation, which saves time.

Final Weapon (Double, General)

Lightning Web is cleverly used throughout stage 1 and 2. Sometimes it's faster to destroy the giant machine, and sometimes it's faster to skip it.
I'm lucky to witness a graphic glitch of the last enemy killed before meeting General.

Boss Rush

Magma Dragoon: I used Lightning Web in the Magma Dragoon fight, and it saves 20 frames because it allows time for another fully charged shot.
Storm Owl: Unfortunately it's slower than the first Storm Owl fight, because I tried lots of plans but can't get the 4 bullet shots to hit Storm Owl in place during the Soul Body shot. The setup requires very precise bullet shots, otherwise, it won't cause extra damage. Maybe it's because of asymmetric left and right physic, or some unknown reasons, so I had to use a fully charged shot to finish Storm Owl.
Split Mushroom: A fully charged shot is used first instead of Lightning Web, because unlike the first Split Mushroom fight, this time it's higher above the ground, costing more time if hit with Lightning Web.
Jet Stingray: One extra double damage Frost Tower is used instead of more fully charged shots, because this time the ceiling is wide open, so X can hit Jet Stingray much faster than before, with no need for screen scrolling.
Cyber Peacock: 2 frames slower before, because X couldn't afford to get hurt anymore to save time after the Soul Body shots.

Sigma

Gunner Sigma: Soul Body is lined up perfectly after getting hurt, without any delay or time loss.
Earth Sigma: It has 2 attacking modes. Spike Attack mode is slightly faster, but without charged shots or Nova Strike, to end the fight in only 2 attack cycles, I manipulated luck to get Spike Attack first, and then Slime Body attack. During the Slime Body attack, I broke the blue "mini body" because it allows X to move more to the right after the fight ends.

About RNG Manipulation and Lag Reduction

This run has a great amount of RNG manipulation, often to reduce the lag during the "READY" animation. But I'm not for extreme lag reduction.
Some super rare RNG values can save about 8 frames per stage, they are almost impossible to calculate (some require planning way ahead), and very painful to manipulate; even if they are achievable, entertainment and other things need to be sacrificed.
In my view, I don't find lag reduction more important than entertainment, so for this movie, I think a decent amount of lag reduction is enough.
Suggested screenshot:
Special thanks to Bernka and sparky. From watching their TAS, I've learned a lot.
Special thanks to SCWeeb, currently the Mega Man X4 speedrun world record holder for many categories. His low% speedrun inspired and motivated me a lot.
If people like this TAS, I may consider starting a new "X, 100%" TAS project.

CasualPokePlayer: Replacing movie file with itself to apply cycle count time.

feos: Claiming for judging.
feos: Renaming the branch to meet Publisher Guidelines
feos: Great run, it also got good feedback, and aiming for lowest in-game completion is a traditional goal, so accepting as a new branch.

EZGames69: Processing...


TASVideoAgent
They/Them
Moderator
Joined: 8/3/2004
Posts: 15594
Location: 127.0.0.1
This topic is for the purpose of discussing #7529: HappyLee's PSX Mega Man X4 "X, no items" in 40:05.25
Expert player (2573)
Joined: 12/23/2007
Posts: 831
I have a question with the frame rate. For my last two PSX submissions, the time showed on the submission page are different from the time showed on BizHawk. On the submission page, it says Frame Rate "60.009601536245796", but in our Platform Framerates page, it says PSX 59.94005994005994, and I can't seem to find where "60.009601536245796" comes from.
Recent projects: SMB warpless TAS (2018), SMB warpless walkathon (2019), SMB something never done before (2019), Extra Mario Bros. (best ending) (2020).
Editor, Skilled player (1440)
Joined: 3/31/2010
Posts: 2109
Very cool TAS, very fast and entertaining. I don't think there was a single dull moment in the TAS (save for maybe the loading screens breaking the flow a bit, but there's nothing you can do about that), and the constant weapon usage really made it exciting. The Split Mushroom stage stood out. Yes vote. Great work!
Editor, Expert player (2078)
Joined: 6/15/2005
Posts: 3282
Looks good. I was wondering though. Is it faster to beat Earth Sigma last, instead of Gunner Sigma last like in the X 100% TAS?
Reviewer, Experienced player (920)
Joined: 11/18/2011
Posts: 315
Location: Morocco
Superb TAS! I enjoyed boss rush more (Split Mushroom dance was so funny lol!) Overall, very good and entertaining. I even noticed some improvements to 100% TAS in refights. Voting Yes of course! EDIT: Most improvements already mentioned by HappyLee so my comment was somewhat redundant haha
I still learn more about English. https://www.youtube.com/user/McBobX100
I wrote:
Working is the best way to achieve goals in speedruning. Hardworking is a pain.
Expert player (2573)
Joined: 12/23/2007
Posts: 831
FractalFusion wrote:
I was wondering though. Is it faster to beat Earth Sigma last, instead of Gunner Sigma last like in the X 100% TAS?
Good question. I didn't test that during this TAS, because in order to beat Gunner Sigma last, you need to take damage two times more (otherwise the Soul Body won't reach Sigma), and X simply doesn't have that much HP left. But I did some quick calculation. The final boss fight is 166 frames slower than sparky's 100% TAS. 60 frames are because of lack of charged shot and Nova Strike, and 73 frames are because having to take damage to fight Gunner Sigma. So I guess they're close, and maybe fighting Gunner Sigma last would be slightly faster for 100% TAS.
Recent projects: SMB warpless TAS (2018), SMB warpless walkathon (2019), SMB something never done before (2019), Extra Mario Bros. (best ending) (2020).
Emulator Coder, Judge, Experienced player (730)
Joined: 2/26/2020
Posts: 783
Location: California
HappyLee wrote:
I have a question with the frame rate. For my last two PSX submissions, the time showed on the submission page are different from the time showed on BizHawk. On the submission page, it says Frame Rate "60.009601536245796", but in our Platform Framerates page, it says PSX 59.94005994005994, and I can't seem to find where "60.009601536245796" comes from.
I'm not sure where the 60.009601536245796 comes from other than something weird in the db. 59.94005994005994 should be the "correct" framerate here. Although even then the PSX has a variable framerate in reality due to interlaced/non-interlaced having different framerates. Nymashock (which this movie uses) saves a cycle count, which using that shows the correct time to be 40:05.25 (81463062909 cycles at 33868800 cycles per second). Although this cycle count isn't parsed correctly yet, so the site doesn't consider it. A fix for that is pending though.
Expert player (2573)
Joined: 12/23/2007
Posts: 831
CasualPokePlayer wrote:
I'm not sure where the 60.009601536245796 comes from other than something weird in the db. 59.94005994005994 should be the "correct" framerate here.
Thanks for the explanation. I think 40:05.25 should be the correct movie time, and 59.94005994005994 should be the correct frame rate. I hope someone can fix it soon. Also, thanks to scrimpeh and McBobX for your warm comments. :)
Recent projects: SMB warpless TAS (2018), SMB warpless walkathon (2019), SMB something never done before (2019), Extra Mario Bros. (best ending) (2020).
Site Admin, Skilled player (1254)
Joined: 4/17/2010
Posts: 11479
Location: Lake Char­gogg­a­gogg­man­chaugg­a­gogg­chau­bun­a­gung­a­maugg
Would it be correct to call this "no items" and [1839] PSX Mega Man X4 "X, all items" by sparky in 39:45.00 "all items"? If percentage is not present in the game, we prefer to mention which entity constitutes completion. Also if the final item quantity is zero (RTA rules say "No amount of items can be collected for this category"), then low% wouldn't be accurate since it means "minimum but not 0". Same difference as between "all" versus "max/high".
Warning: When making decisions, I try to collect as much data as possible before actually deciding. I try to abstract away and see the principles behind real world events and people's opinions. I try to generalize them and turn into something clear and reusable. I hate depending on unpredictable and having to make lottery guesses. Any problem can be solved by systems thinking and acting.
Expert player (2573)
Joined: 12/23/2007
Posts: 831
feos wrote:
Would it be correct to call this "no items" and [1839] PSX Mega Man X4 "X, all items" by sparky in 39:45.00 "all items"? If percentage is not present in the game, we prefer to mention which entity constitutes completion. Also if the final item quantity is zero (RTA rules say "No amount of items can be collected for this category"), then low% wouldn't be accurate since it means "minimum but not 0". Same difference as between "all" versus "max/high".
I'm not really sure about the perfect name for this category, so it's up to the judges to decide. I called it low% mostly because RTA called it low%, and I didn't call it 0% because it still completes the game with all bosses defeated.
Recent projects: SMB warpless TAS (2018), SMB warpless walkathon (2019), SMB something never done before (2019), Extra Mario Bros. (best ending) (2020).
Site Admin, Skilled player (1254)
Joined: 4/17/2010
Posts: 11479
Location: Lake Char­gogg­a­gogg­man­chaugg­a­gogg­chau­bun­a­gung­a­maugg
HappyLee wrote:
I didn't call it 0% because it still completes the game with all bosses defeated.
Which bosses are optional?
Warning: When making decisions, I try to collect as much data as possible before actually deciding. I try to abstract away and see the principles behind real world events and people's opinions. I try to generalize them and turn into something clear and reusable. I hate depending on unpredictable and having to make lottery guesses. Any problem can be solved by systems thinking and acting.
Expert player (2573)
Joined: 12/23/2007
Posts: 831
feos wrote:
Which bosses are optional?
None. But "0%" may sound like that the game is not completed or mostly skipped. Maybe that's why very few speedruns are called "0%".
Recent projects: SMB warpless TAS (2018), SMB warpless walkathon (2019), SMB something never done before (2019), Extra Mario Bros. (best ending) (2020).
Site Admin, Skilled player (1254)
Joined: 4/17/2010
Posts: 11479
Location: Lake Char­gogg­a­gogg­man­chaugg­a­gogg­chau­bun­a­gung­a­maugg
HappyLee wrote:
None. But "0%" may sound like that the game is not completed or mostly skipped. Maybe that's why very few speedruns are called "0%".
I see. The thing with 0% and low% is that one is a subset of another. They both indicate "fewest possible", one of them simply being a case where known minimum is "none". Same relation is between "maximum" and "all": if it's impossible to collect all of something, it makes sense to aim for maximum. So since the items are countable and limited, and optional, you can collect all, some, or none. "Maximum", "minimum", "low", and "high" are simply redundant in this case.
Warning: When making decisions, I try to collect as much data as possible before actually deciding. I try to abstract away and see the principles behind real world events and people's opinions. I try to generalize them and turn into something clear and reusable. I hate depending on unpredictable and having to make lottery guesses. Any problem can be solved by systems thinking and acting.
Expert player (2573)
Joined: 12/23/2007
Posts: 831
Thanks for helping change the category name. "X, no items" seems accurate to me. A question for the judges: would an "X, Ultimate Armor (any%)" TAS be acceptable on TASVideos? I see that the rules are changed and in-game codes can be accepted. It would be very different than normal X runs, and faster than a Zero TAS.
Recent projects: SMB warpless TAS (2018), SMB warpless walkathon (2019), SMB something never done before (2019), Extra Mario Bros. (best ending) (2020).
Site Admin, Skilled player (1254)
Joined: 4/17/2010
Posts: 11479
Location: Lake Char­gogg­a­gogg­man­chaugg­a­gogg­chau­bun­a­gung­a­maugg
HappyLee wrote:
A question for the judges: would an "X, Ultimate Armor (any%)" TAS be acceptable on TASVideos? I see that the rules are changed and in-game codes can be accepted. It would be very different than normal X runs, and faster than a Zero TAS.
Ultimate Armor is just getting all upgrades right away? If the result is different from existing branches and entertaining, it may get good enough feedback to go to Moons, so yeah.
Warning: When making decisions, I try to collect as much data as possible before actually deciding. I try to abstract away and see the principles behind real world events and people's opinions. I try to generalize them and turn into something clear and reusable. I hate depending on unpredictable and having to make lottery guesses. Any problem can be solved by systems thinking and acting.
Post subject: Movie published
TASVideoAgent
They/Them
Moderator
Joined: 8/3/2004
Posts: 15594
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. ---- [4743] PSX Mega Man X4 "X, no items" by HappyLee in 40:05.25