Steve
EX CADC: 2F (5)
s.MP: On Block +5 (+6)
cl.MP: On Block +3 (+4)
c.HP: On Hit -2 (-1)
c.HP: On Block -8 (-7)
Foot Stomp Right Hook: On Hit hard knockdown (knockdown)
British Edge: On Hit +9 (+5)
Shoulder Rush: On Hit -3 (-2)
Shoulder Rush: On Block -9 (-8)
Uppercut: On Block -12 (-11)
Spin Low Hook: On Block -10 (-9)
MP Sonic Fang: Start Up 17F (14)
EX Sonic Fang: Start Up 14F (15)
EX Hellfire: Start Up 10F (11)
Super Art: Start up 4F (3)
Xiaoyu
EX Shooting Star: Start Up 29F (30)
EX Hakkesho: Start Up 25F (22 according to 1.08 Patch notes)
Wave Crest: Start Up 7F (8)
EX Wave Crest: Start Up 7F (8)
Turn of Fortune: On Block -8 (-7)
Super Art: Start Up 9F (8)
1 to go!!!
Sorry if this was discussed, but the last thing on the first document is the first frame the character can cancel their charge move into a dash, correct? Man with a 4f cancel maybe Bryan has some cadc shenanigans…maybe.
Yes.
Aaaaaaaaaaaaaaaaaaaaaaaand we’re done!
Yoshimitsu
cl.MP: On Block +2 (+3)
c.LK: Start Up 5 (4)
Oni Killer: Damage 130 (120)
Added missing b+LK (air) > MK target combo
Poison Wind Steel Fist: On Hit -2 (+2)
Poison Wind Silver Fist: On Hit -30 (-38)
Poison Wind Silver Fist: On Hit -43 (-51)
f+MK > MP: On Block -11 (-10)
Flea to LK Poison Breath: On Hit 0 (+1 according to 1.08)
Flea to LK Poison Breath: On Block -22 (-20)
Flea to MK Poison Breath: On Hit +1 (+2 according to 1.08)
Flea to MK Poison Breath: On Block -22 (-26)
Flea to HK Poison Breath: On Block -21 (-4) [wtf - how can this happen?]
Flea to EX Poison Breath: Start Up 15F (16)
Flea to EX Poison Breath: On Block -22 (-6)
MP Suicide: Immune against “counter” moves (instead of strike invincible)
EX Poison Breath: Start Up 16F (17)
EX Gehosen: Start Up 6F (7)
Super Art: Start Up 7F (6)
You did a good job.
Many thanks!
Or are we? 
Fantastic work gentlemen. 
Found a notable thing for Cody:
c.MK cannot hit opponents during jumping or jumping attacks.
Also I discovered something else interesting that affects all characters. Counterhit frames are not universal. Unlike in SF4 where it’s always a counterhit if you hit the opponent out of the move before the move becomes active, in SFxT a BUNCH of a moves will grant a counterhit if you hit the opponent at any point during the ACTIVE frames of the move as well.
What is even WORSE though is that isn’t a universal trait for all moves. SOME moves are counterhittable during their active frames, some moves are not.
Rufus for example, his cr.MP is 7F startup 3F active 15F recovery. If this where SF4 he’d be counterhittable for 7F. However in this case he is counterhittable for 10F
However Rufus’ far s.HP is 9F startup 4F active 17F recovery. So you’d think “Ok well then he is counter hittable for 13F, right?” NOPE! Thanks to Capcpom Rufus is only counterhittable for the standard 9F in this case. It seems to be completely arbitrary which moves have counter hit state till the recovery starts and which moves have counter hit stats until the active frames start.
From what I have been observing, crouching normals seem to have CH state until the end of active frames, rather than just during start-up. I didnt test every single crouching normal tho.
Maybe because most anti air normals are crouching?
So c.MK “whiffs on airborne”, but you can juggle with it after launcher, right?
I’m not sure why they did it tbh. Maybe they wanted to discourage ppl from doing so many crouch jabs? (But then it encouraged stand jabs lol)
Alright, I just finished testing all of Cody’s standing normals, crouching normals, air normals, command normals, knife attacks, throws, cross cancel, raw launcher…everything, and the only discrepancy I could find was with his diagonal jumping LP (4 frames in testing vs 5 frames in the frame data). I did see that each of Cody’s animations that don’t have attacks (e.g. Fake Bad Stone) have an extra frame compared to what the guide says, but I’m assuming this is because you’re counting the frames from the button press to the first pre-jump frame and aren’t counting the frame before the button press when the animation actually starts (e.g. Cody’s knife pick-up. I start counting frames from when the animation transitions to knife pick-up, a frame before the button press appears. Then I count until the first pre-jump frame, which is on frame 29, so Cody’s knife pick-up has a total of 28 frames).
Other than that, everything seems to be in order with Cody’s frame data. Though I should mention that I found that Cody’s bad spray appears to have 25 frame start-up. There’s a weird animation error with the start-up of bad spray where it looks like Cody is stuck in the floor; if you count frames starting one frame after that part of the animation, that’s how I got the start-up.
If you want I can also check the frame advantage of Cody’s moves to make sure that there aren’t any discrepancies there either.
j.LP is 4F. Also fixed the start up for the fake / pickup stuff.
Will test Cody’s c.MK on pc version. You can use it during combos tho.
Sometimes animation doesn’t vastly change at the 1st start-up frame, so simply looking at animation might not be the most accurate way. Abel’s CoD is an example (his 1st start-up frame simply looks like what he was doing before the qcf+P input, so if he did cl.st.HP xx qcf+P the 1st start-up frame animation looks just like his cl.st.HP animation)
I can’t hit Juri’s f.MK, so I think you’re right. I don’t know if this is intentional tho. It also dodges many airborne attack.
So he doesn’t have an expression change either? Because Cody has a couple of moves where the animation itself is basically the same as the animation before it (I think believe knife throw was like this), but even though the animation itself looked almost the same, there is a clear expression change on his face to indicate that he’s started the move.
Either way, in that case it’d still be easy to check the start-up since you can count frames according to the button press rather than the animation. Though that may become a problem when trying to check the start-up of super charge moves since it’s easier to get eliminate the charging frames if you cancel into it from a normal move.
I don’t see expression change, tho I might have missed it since it would be so subtle.
Checking most special moves would be fine without checking the animation itself. You just have to know the hitstop/blockstop then count from the frame in which the character himself moves again regardless of animation, provided you use the cancel method.
What street11 means is when you cancel a normal into a special move you sometimes have moves that use the normals 2nd active / 1st recovery frame as their oiwn first animation frame and then have still a distinct cut to their own 2nd animation frame. You just have to count as soon as the dummy leaves the hitfreeze. You need to use cancels to measure charge moves as you’ll get additionally charge frames when you do it raw (due to button press / release)
Ah, alright, so if that’s the case then there may not be an expression change either. Thanks for clarifying that.
And yeah, that’s how I double checked the start-up of Cody’s ruffian kicks; cr.hp xx ruffian kick. Otherwise I kept getting start-up increases of 3-5 frames which obviously isn’t right.