I don’t think what you’re saying is 100% accurate. If we can cancel at any point in an attack, then why hasn’t been tech regarding late cancels? I’m talking reallllllyyy late canceling. Imagine me doing st.mp at max range, and instead of canceling while my hand in pressed up against the opponent’s face, I input the cancel when I’m retracting my arm… doesn’t that sound a bit odd? I mean, have you ever tried to cancel cr.LK into fireball, but not inputting the :p: when he retracts his foot rather than doing it when making contact? No fireball comes out.
My understanding of what you said is that when you whiff a move, any normal move, recovery frames don’t exist if it doesn’t touch anyone. I reached that conclusion because you said the OS should happen during the active frames of the previous move (yes I read that second bit and I’ll get to that)… which shouldn’t make a lot of sense because if you inputted a special move that should’ve canceled the normal move’s recovery frames. Every move has recovery frames and during these frames, one cannot do anything. Obvious example is st.HP. If, for the sake of an easy example, st.HP was special cancelable, can you imagine being able to cancel it when he’s retracting his arm? Probably not. During the super long time he spends pulling his arm back, you can’t do shit. Going by what you’re saying, if we did an OS involving st.hp (it’s still special cancelable for the sake of making an example), and we were to catch a backdash that was able to dodge st.HP, and we inputted EX Tatsu, we wouldn’t see him retract his arm and EX tatsu would just come out and catch the motherfucker.
That doesn’t make a lot of sense to me. Don’t forget recovery frames exist even if you whiffed a move, so if we don’t disregard that, what we end up doing is st.HP, wait for his arm to come back and then EX tatsu comes out. “How’s that an OS?” you may ask. “That’s equivalent to linking…” you might think. Well that’s what I believe is happening when you do an OS. Obviously you wouldn’t pick something slow like st.HP (if it was special cancelable), but I used st.HP as an example because those recovery frames are significantly easier to notice. We can see and feel the huge amount of time he spends pulling his arm back. So if we were to go apply this way of understanding back to cl.MP, the OS must’ve come out after its recovery, despite how hard it may be to notice. It couldn’t have been during the active frames because that would’ve been a special cancel. At the same time, if you were stand away from your opponent, and do cl.mp followed by EX Tatsu, you should get the full animation of cl.MP, and EX Tatsu would come out right after. That’s the OS maneuver in essence. However, if you were to record that with a dummy and the cl.mp in fact hit the other person, EX Tatsu wouldn’t have come out (like we saw in the video). Why is that? We got EX Tatsu to come out precisely after cl.mp ends and recorded it onto the dummy, no human error can possibly be factored in. So what’s going on? It’s the hit-freeze. Analyzing this mathematically, the whole duration of cl.mp’s animation is 17 frames, and EX tatsu is timed to come out right after, so on “frame 17” of this whole scenario. If it should come out on the 17th frame each time we did cl.mp (standing away from the opponent), then how come it doesn’t work when he hit them? That’s where the hit/block freeze comes in.
If a move doesn’t come out, it must mean we were in a state where we cannot input anything. Agreeable statement?
If so, then that must mean at frame 17 when EX tatsu gets inputted, we were still in recovery frames (the only frames in the case of cl.mp) where can’t input anything else (i.e. link or special cancel). But if it worked when cl.mp whiffed, then how do you explain it now when it does? I believe one way to see it is to give those hit/block freeze snippets “frames” of its own in the time line of animation frames. It doesn’t make sense to have it after the recovery frames and obviously it doesn’t happen during startup, therefore I located the “F” frames in the code diagram between active and recovery, where the attacking limb makes contact and when he pulls it back. By adding additional frames, we can see that at frame 17, Gouken is still in recovery frames of cl.MP, therefore EX tatsu can’t initiate when cl.mp makes contact with another body.
Note I totally didn’t mean to shoot down your post, climaxter. Over the course of an hour, I think I figured out the whole thing down to technical bits while typing out an explanation that may have been too wordy. But I went wordy to ensure there was no misinterpretation.