Posting this here because I need a place to post it and it doesn’t merit its own thread.
Azwel’s CE in sword mode is i13 and Geralt’s RE K is -13 on block.
Azwel’s sword CE can’t punish Geralt’s 1AA, Grøh’s 66K, or 2B’s BB, all of which are -12 on block. It can punish Grøh’s 44K, which is -14. It can also punish Geralt’s RE K on block, which I thought was -12, because it can’t be punished by Yoshimitsu’s b:A, Sophitia’s 236B, or Azwel’s 6B.
It’s possible that all 3 of those moves are only i14 at close or mid range, and that they’re all at least a frame slower at tip, preventing them from punishing RE K. To confirm that this wasn’t the case, I also tested to see what clashes with what after Geralt’s RE K on block. It turns out that nothing with an even impact frame clashes if you buffer the inputs properly, indicating that RE K is not an even number on block. I did 10 clash tests, setting 2P Geralt to do a recording in which he gets RE K blocked then does the move on the right in the parentheses, while 1P Geralt buffers the move on the left:
At least -11: i24 beats i14 (4K vs. BB)
At least -13: i24 beats i12 (4K vs. 2A)
At least -13: i30 beats i18 (66BB vs. 6B)
Less than -14: i26 loses to i12 (1A, 1B vs. 2A)
Less than -14: i30 loses to i16 (66BB vs. 3A, 6A)
Less than -14: i36 loses to i22 (A+B vs. 4B, 4A)
Less than -16: i32 loses to i16 (6[K] vs. 3A)
Scuffle also agrees that RE K is -13 on block (for what that’s worth).
The only explanation for this behavior is that RE K is -13 on block. This isn't all that weird. It’s possible that RE K is “actually” -14 on block, but because of how RE works, it only hits at tip range, where it’s 1F safer. (There are lots of moves that are 1F off their normal values at tip range; Groh’s BB is i15 at tip but i14 elsewhere.) What is weird is that Azwel’s sword CE punishes it but not -12 moves.
I tried some additional tests to verify if Azwel’s CE was i13, but couldn’t find a definitive answer. It’s invincible from F12 onward:
F11: Loses to i17 (Inferno 6K) at -6 (blocked AA)
F12: Beats i18 (Inferno 2A) at -6 (blocked AA)
… which means there’s no reason to expect it to clash like normal moves do. I tried testing to see if it clashed with other CEs, but this didn’t go anywhere — Grøh’s i16 CE beats Azwel’s allegedly i14 sword CE when Grøh is at -2 (Grøh lands B then does CE).
So: Azwel’s sword CE must be i13. If anyone can prove this wrong that’d be great, because I hate anomalies. But it seems to be the case.
Some video evidence (I didn’t record the clash tests because it’d take forever to edit them together):
Azwel CE can’t punish 1AA, can punish RE K; 6B can’t punish RE K.
Stab can’t punish RE K.
b:A can’t punish RE K.
60FPS footage of Azwel’s CE in sword mode, with captions showing the start. You can see when the game enters and exists the CE cutscene by looking for duplicated frames — there’s a stitch in the animations where you see the same frame twice. If you look for these and count up the frames you’ll see that it freezes on F11, then resumes on F12 and hits on F13. This method may not be reliable (the capture software might not be synced to the game’s framerate).
Watch "here" on Streamable.
streamable.com