Submission Text Full Submission Page

Introduction

One day, I came across a new Sonic 3 & Knuckles TAS by BenInSweden, Chrezm, eandis, & kaan55. Its main feature was that it corrupted the lock-on technology, thus separating the games, essentially turning it into a Sonic & Knuckles run. Since I usually get the good ending of Sonic games when I play them, and how I like Sonic & Knuckles standalone more, I decided to use inputs from the Sonic 3 & Knuckles TAS for a new 100% Sonic & Knuckles TAS as a concept of how a TAS that corrupts the lock-on technology would look like if it had gotten the good ending.
Notice: This run copies inputs from https://tasvideos.org/6890M. See https://tasvideos.org/6890M for more details on the tricks being used.

Movie objectives

  • Primarily aims for RTA time
  • Takes damage to save time
  • Intentionally takes deaths to save time
  • Uses glitches to save time

Zone explanations

Mushroom Hill Zone Act 1, Mushroom Hill Zone Act 2, The Doomsday Zone and the Special Stages are done by me, so I'll explain these.

Mushroom Hill Zone Act 1

As you may have already known, the MHZ1 spawn point is different in Sonic & Knuckles. However, that doesn't actually change much in overall routing of the level and after the first giant ring, there are no major changes to account for the different spawnpoint.

Mushroom Hill Zone Act 2

No zips have been used in MHZ2 and I instead got the rest of the Chaos Emeralds. Since there isn't the additional giant ring in MHZ1, I didn't take damage near the cuckoo-like enemy guarding Special Stage 6. Instead, I took damage near the wasp-like enemy near the pullies and thus entering a giant ring that most 100% runs do not enter.

The Doomsday Zone

I didn't follow any routes seen in other TASes and made my own route, avoiding vertical movement. Within the final hit, as no more inputs can be provided to save time, no time was wasted by hitting the asteroids.

Special Stages

Within the first 3000 frames, you can turn on yellow spheres, thus changing the direction of the jump. The jumps that yellow spheres give you don't change depending on what direction you hit them. You can also do the same thing with red spheres but I didn't turn on any red spheres in this run.

Potential Improvements

  • As Sonic 3 hasn't been locked on in this TAS, there's less RAM allocated. Because of this, in 0:18 in MHZ2, I wasn't able to do a slope jump to get jump to Special Stage 4, thus losing 3 seconds. And in 0:34, also in MHZ2, I wasn't able to use the Flame Shield boost on the mushroom next to the speed up monitor to go left, thus losing around 2-4 seconds.
  • In DDZ, I avoid getting hit from the yellow dots and especially the rockets. If intentionally taking damage could potentially result in saving time, then it can also be improved.


TASVideoAgent
They/Them
Moderator
Location: 127.0.0.1
Joined: 8/3/2004
Posts: 18024
Location: 127.0.0.1
Player (183)
Joined: 12/17/2018
Posts: 46
Decent, but I think there are numerous improvements that can be made.
I also avoid finishing Special Stages moving backwards.
I'm not sure why this would be? as it's faster to finish moving backwards, there's also some optimisation that could be taken from joojamoi's variations, e.g: http://viewsync.net/watch?v=A9dTAh5u_Yk&t=1412&v=dEfWpSR0kf8&t=430&mode=solo https://viewsync.net/watch?v=A9dTAh5u_Yk&t=1494&v=dEfWpSR0kf8&t=503&mode=solo https://viewsync.net/watch?v=A9dTAh5u_Yk&t=1672&v=dEfWpSR0kf8&t=351.5&mode=solo https://viewsync.net/watch?v=A9dTAh5u_Yk&t=1957.5&v=dEfWpSR0kf8&t=28&mode=solo
As Sonic 3 hasn't been locked on in this TAS, there's less RAM allocated
This isn't how it works. The Genesis has the same amount of RAM regardless of whether S3 is locked on or not. But from a game loop standpoint, S&K and S3K are identical, there are some variations on what happens depending on what's in memory (i.e. Locked on flag), but that's it. You are likely seeing issues caused by subpixels. The run-up to the first big ring seems sub-optimal, as I reach it marginally faster in my RTA PB: Link to video Copying what the S3K TAS does would likely shave off some time. I think stacking the big ring collection in MHZ is likely slower than what it could be, but would need some testing, since you're forgoing skipping a lot of both stages to collect rings. My theory route would be 2 in MH2 (basically very quick to get), 2 in FB1 using AsIntended, and 3 in Sando2. There is the common issue with S3K, S3 and S&K where All Emeralds is combined with all glitches become more of a "scenic route", than making use of super/hyper sonic, which I believe is why many TASes in the past have gone for a No level wrap approach. It does also limit the choices of big rings as well - as a good and bad thing.
Player (183)
Joined: 12/17/2018
Posts: 46
This is not optimised, but just showing a different and faster route to the first MH1 big ring. User file #639199027881083040 Link to video Also, I noticed that you aren't doing the first input for the special stages before the lag frames - joojamoi documents it briefly here: #5639: joojamoi & Evil_3D's Genesis Sonic 3 & Knuckles "Knuckles, 100%" in 44:27.73
JankRusty
Any
Location: International
Joined: 4/24/2026
Posts: 25
Location: International
Thanks for the notes on the transition inputs and the Special Stage documentation! This run primarily uses level warps and overall glitches to beat the game instead of taking a scenic route for the Chaos Emeralds. I appreciate the technical tips on the RAM values and the special stage logic too. It'll be good to know for future projects!
DrD2k9
He/Him
Editor, Expert player (2582)
Location: US
Joined: 8/21/2016
Posts: 1190
Location: US
BenInSweden wrote:
This is not optimised, but just showing a different and faster route to the first MH1 big ring. User file #639199027881083040
This update desyncs after the first bonus stage. adjusting the initial movement following that stage resyncs until the end of Act 1, but then desyncs again. I'm guessing there will be a series of 1-2 frame adjustments needed throughout the run to fully sync.
Player (183)
Joined: 12/17/2018
Posts: 46
JankRusty wrote:
Thanks for the notes on the transition inputs and the Special Stage documentation! This run primarily uses level warps and overall glitches to beat the game instead of taking a scenic route for the Chaos Emeralds. I appreciate the technical tips on the RAM values and the special stage logic too. It'll be good to know for future projects!
For scenic routes, I meant that blue spheres in a glitched run doesn't really make much use of Super or Hyper. Hence why TASers have historically gone for not doing wraps etc (or other quests like the ring attack) to make it more interesting.
DrD2k9 wrote:
BenInSweden wrote:
This is not optimised, but just showing a different and faster route to the first MH1 big ring. User file #639199027881083040
This update desyncs after the first bonus stage. adjusting the initial movement following that stage resyncs until the end of Act 1, but then desyncs again. I'm guessing there will be a series of 1-2 frame adjustments needed throughout the run to fully sync.
Yeah, I wasn't intending for it to be a replacement file, as it was only really an example of what could be improved. I think it's up to JankRusty if they want to carry on updating it, as it would be unfair for me to do the changes that I would make to optimising it - as it would remove essentially all of JankRusty's inputs - MH1 and 2 would be completely different, I'd use joojamoi's blue spheres as they are more optimised, and I already have Doomsday from another TAS that I was working on that I'd use the inputs for.
JankRusty
Any
Location: International
Joined: 4/24/2026
Posts: 25
Location: International
In case it's not clear, I've said in my explanation that my TAS is supposed to be a theoretical concept of if https://tasvideos.org/6890M had gotten the good ending. That's why I did MHZ2 myself instead of copying from another Sonic 3 & Knuckles Sonic 100% TAS, and why I copied the inputs from https://tasvideos.org/6890M instead of copying inputs from another Sonic 3 & Knuckles 100% TAS. As for Special Stages, well I got routing from https://tasvideos.org/5310M, which is another Sonic & Knuckles 100% TAS but it's as Knuckles instead, I've did this for all Special Stages except for Special Stage #3, I got the routing from https://tasvideos.org/5453M instead. However, I still did inputs myself. I also did try to do that MHZ1 improvement myself but I wasn't able to clip into the wall. I may consider deploying these new changes if my current TAS is unoptimized enough.
Player (183)
Joined: 12/17/2018
Posts: 46
JankRusty wrote:
In case it's not clear, I've said in my explanation that my TAS is supposed to be a theoretical concept of if https://tasvideos.org/6890M had gotten the good ending..
Yup, I understood that, just was giving context that historically Sonic TASers have opted to forgo certain glitch types to be able to add Super/Hyper usage into the run. I've redone MH1 from scratch, using a modified route and joojamoi's blue sphere inputs, looks like it should also be faster than [5453] Genesis Sonic 3 & Knuckles "Sonic, 100%" by ShiningProdigy9000, Takz15x & kaan55 in 49:47.213, slightly hard to tell with the different layout of the level. Compared to JankRusty's it currently gets the final hit on the boss on frame 12935 vs frame 13500. It's also fast enough for me to believe that getting these 3 big rings in MH1 is likely the faster route overall rather than getting them elsewhere - this route only loses 9 IGT seconds (probably ~11 RTA seconds - centiseconds is reset on each blue sphere exit) for 3 emeralds. It does leave a few options open, can either - delay the final hit on the boss, and use JankRusty's "glitchless" MH2 route. - re-reroute "glitchless" MH2 with the lightning shield. - use the early boss kill to use the glitched route from the TAS this is based off of, and get 2 big rings on the way, and the remaining 2 elsewhere - likely FB1. User file #639202709419498855 Link to video
JankRusty
Any
Location: International
Joined: 4/24/2026
Posts: 25
Location: International
Thanks for the nice improvements! I have put your new MHZ1 improvement into my TAS although I hit the MHZ1 boss slightly later so that I can get the Flame Shield for MHZ2. I've also optimized the MHZ2 boss. I have also improved the Special Stages, now using joojamoi's inputs. I have also not intended MHZ2 to be glitchless, but I was not intending any zips either so that I can get the other four Chaos Emeralds in MHZ2. https://tasvideos.org/UserFiles/Info/639203011295020938
JankRusty
Any
Location: International
Joined: 4/24/2026
Posts: 25
Location: International
Apologies for the inconveniences, but just now, I've seen that my "improved" TAS (the Userfile above) actually desyncs at LRZ2. Yes, an entire week later. The LRZ1 also seems to be desynced. I may still deploy BenInSweden's MHZ1 improvement later down the line, and I will inspect the LRZ1 zip and boss since that seems to be causing the issue. Even so, I also may need to adjust the SSZ boss RNG due to different CPU cycles.
JankRusty
Any
Location: International
Joined: 4/24/2026
Posts: 25
Location: International
My TAS hasn't really gotten any attention for a while now so might as well explain the problems of trying to implement BenInSweden's improvements and Special Stage improvements into my TAS. So, when I implemented BenInSweden's MHZ1 and joojamoi's Special Stages to my TAS in the userfile seen above, I hit the boss slightly later to avoid the MHZ2 glitch, and I then use the Lightning Shield to change where the signpost lands to then get the Flame Shield. After positioning Sonic correctly and deleting the leftover frames so that the rest of the TAS is unchanged, it desyncs at Special Stage 4. I have no idea why this happens. The Flame Shield Boost doesn't work for some reason even though I'm mid-jump. After changing the routing to Special Stage 4 so that I can enter the Special Stage, it then desyncs again at Special Stage 5. After changing the route to Special Stage 5 to a slower one, there aren't any desyncs in MHZ2, well at least from what I remember. But then there is the other desync that is in the userfile above in LRZ1. It syncs all the way until the boss fight where the boss fight plays out normally until the end where in the final hit, the hand is slightly in the left compared to what's intended, causing the sign post to fall down to a spot where a monitor is, meaning the LRZ2 zip fails. I don't know why this happens, and in RAM watch all of the X and Y positions and sub-pixel positions line up exactly so I don't really know what's going on with that. Even if I break the monitor so that the LRZ2 zip can be done successfully, it's a time loss because of the sign post bounce, not to mention that I wasn't able to successfully get the LRZ2 zip to work. I have also mentioned this before but the RNG for the Mecha Sonic mk.II boss may also most likely need to be changed again because of different CPU cycles. So yeah, just wanted to clear this up.

1789421972