Initial split = 10:06.00 Actual End = +15 Frames(!) Count-in Miscalculation due to emulator change = -17 Frames(!) = 10:05.93

Meaning I DID it! I tied my TAS on my first run! 10:05!

This is run on BizHawk (2.3) instead of whatever emulator I ran it on almost 6 years ago(!)

Since I TAS'd this using BizHawk TAStudio, I figured I'd try to run it on BizHawk instead of Kega (which I couldn't even come within 10 seconds of my old run on). That ended up saving 16 seconds off of the old run. The final second was from learning I could just buffer movement and item use inputs and from nailing the 3 Key menus and, 1 Matchbox menu in the run.

Very interesting to find a way lower in this run. Seems like this is the end though unless we want to start counting times in milliseconds...

[UPDATE]

Uhhhhhh........oh wow. Watching the TAS back...it's also 2 frames into a 10:05......... I legit TIED the TAS on my first attempt........ I can't even- Fucking legend.

What that means in reality is I (obviously) buffered everything correctly movement and item use wise. The real clincher was hitting the only 4 non-bufferable inputs (the 4 times you open the menu to use an item) frame-perfect. Nailed it. That's so immesurably dope that I didn't have to grind for this time at all. I was unsure I could hit a 10:05 non-TAS...but first run?! 4 frame-perfect menus FIRST RUN?! With only 1 real frame I could have missed?! HOW did I get this first run?!