Sonic Chaos Any% TAS - 11:02.45
- Emulator used: BizHawk 2.2 (NTSC)
- Bios used: SMS BIOS 1.3 (USA, Europe)
- Display Type: NTSC
- Fastest any% run possible with Sonic as character
This TAS shows the fastest speedrun of this game currently existing and beats the non-TAS world record by more than 30 seconds (Sept. 2017). This is the first any% TAS of the game. It utilizes all known time saving glitches known to me (Sleeping Egg Zone glitch and 3x Special Stage glitch). The inputs are mostly frame perfect. The game is completed with Sonic as character in just a bit over 11 minutes (11:02.45) starting from the character select until the final boss jumps on the platform and starts smiling (speedrun.com rules for the game).
The boss glitch in SEZ3 and the special stage glitches in SEZ2, APZ1 and APZ2 are noteworthy. Also the MGZ3 boss is hit will being in hit stun. I described the glitches in more details (partially in videos) here: https://www.speedrun.com/Sonic_Chaos/forum This is the first TAS, I have done. It took me between 3 to 5 hours. It is noteworthy, that the game's visuals have a lag of roughly 3-4 frames and when running with 60fps (NTSC). The game is plaqued with lags, which are eating inputs and appear pseudo-randomly. This made it impossible for me to redo any stages, since "cutting out input frames" would make frame lags in the later game appear differently and therefore corrupt the rest of the recording. The strategies used for the run are mostly known. The inputs are for the most part frame perfect. The TAS is not perfect, in that the time can still be optimized marginally. I strongly believe, that a sub 11 minute time is possible, but (if no other glitches surface) only by a few seconds.
fsvgm777: Replaced movie file with one that removes about 8 seconds of blank input at the end.
Noxxa: As has been discussed in the submission topic, the quality of this run is a bit spotty. Comparing to the published best ending TAS (in levels that have identical and comparable goals), this movie has a few improved strategies and saves time in some levels, but in other levels it unnecessarily loses time. Even if it's a different category, a new submission is expected not to lose time to existing runs in comparable areas. For that reason, I'm rejecting this run.
It is most certainly a decent run for a first try, but it just needs a second pass, with that extra bit of polish.