People suggest the frame data is wrong all the time. The best way to prove it is correct(or wrong) is to test it against a bunch of other situations. Ken’s LP super for example, you could test against all other attacks that leave the opponent at -2 and see if you can hit them. Then test against all attacks that leave them at -1. There’s lots of these wierd situations for specific moves in SF4. Guile/Fuerte’s super is one of them. They can punish certain moves on block, they normally shouldn’t be able to punish according to the frame data. Have you ever wondered why Ryu can punish a point blank blocked low tiger shot from Sagat with his c.rh but not his c.forward? They both have the exact same startup and it looks like Ryu’s foot is in range for both attacks. Testing Ryu’s c.rh and c.forward against other situations show that the frame data is correct. SF4 is just really really wierd.
Out of interest, is there any difference between the PAL versions of the game and the NTSC versions, frame data wise, since one runs at 50Hz and the other 60Hz?
Perhaps it is some sort of hitbox glitch? As opposed to inaccurate frame data.
But I guess this isn’t the thread for speculation…
For close to ten years now, PAL SD televisions have been capable of running at 60Hz.
Thanks Richard Nguyen, this is really great information.
I want to add something that I discovered and am hoping for a discussion. I stumbled upon a thread somewhere that mentioned you can dash out of a focus attack by pressing :qcf::qcf: but I tried it and it doesn’t work so I went experimenting and found out something interesting.
To dash, you press :r::r: or :l::l:. You can’t dash by pressing :r::l::r:. However, from experimenting, I found out that you can dash by pressing :r::l::r: during a focus attack. From this I tried dashing out of a focus attack by pressing :db::qcf::db::qcf: and it worked.
I was testing this with Ken and Ryu and in both cases after the dash, all I needed was to press :3p: for their ultras to come out.
So that means you can FADC to ultra with any characters who have a :qcf::qcf: ultra by doing:
FA, :db::qcf::db::qcf:, :3p: or possibly FA, :qcf::db::qcf:, :3p:
Correct me if I’m wrong.
Hello, UltimateDavid. :wgrin: Very weird findings you got there. Just tested it out myself and it holds true. One of the mysteries of sf4 I guess.
I just tested it and Ken’s LP super is indeed 2 frames startup. Also, I previously read about your tests with Guile’s MK super against Boxer’s dash punch. Another weird thing.
Oh, and I think Ryu can’t punish the low tiger shot with the cr. mk because he gets pushed too far causing the first few active frames of the cr. mk to miss Sagat’s hitbox. I have tested this out before and it just looks and sounds like the cr. mk is making contact later than usual.
I probably do not have the correct technique down, as I have been messing up like 99% of the time doing your way. 
I do believe it’s true though with my few successful attempts.
The concept of the unusual dash inputs during the focus attack has also been brought up by Javits in the Boxer forum regarding a backdash to ultra to punish air tatsu’s. I guess the focus attack allows to “remember” the inputs for the dash longer. Another mystery. 
But sorry guys for not contributing to this thread recently. I wrote a section for kara’s like last week but it got deleted as my browser closed. :arazz: Shoulda wrote it in Word. Haha, well I will probably be writing that up again. Dang, I just remembered something else I was gonna do but I forgot…
Perhaps you should add the misconception about DP FADC. It’s actually -1 on block.
this is like all crazy to understand
I argue that the :df: :df: :p: shortcut works. It’s just not that good compared to a 3 move shortcut. My theory is that the :df: :df: registers as :r: :d: :r: :d:, with the second " :r: :d: " counting as a :df:. The timing needs to be faster than the 3-move version, or the normal dp for some reason though.
I guess what I’m saying is, if the shortcut doesn’t work how come it does?
It doesn’t work. Diagonals are unique inputs. They do not have horizontal or vertical components (in the same way that a horizontal has no vertical component and vice versa). The reason they appear to is because that is the way the game is coded. The motion for a DP is hardcoded as 3 inputs. There’s no way you can get around that.
If :df::df: appears to works for you, then you must have inputted another motion somewhere along the line. Try doing it Training Mode with input display on. You’ll see it’s impossible to get a DP from just two directional inputs.
yo guys… anybody can give me a hand on where I can find information about armor breaking??
I mean, why some moves that are not armor breakers, sometime breaks your armor (when focusing)???
By the way the info about the landing frames is true cause I have noticed that when I hit somebody’s jump ins on Blanka’s slide (df.HP)… if the opponent starts any move when airborne he cannot block, however if he jumps in without any move, he blocks it!!
^ When performed as a REVERSAL, some - if not all (?) - specials break armor
With Reversal, all specials except projectiles become Armour BReaking
I don’t think this applies to special moves. From what I can tell, if you attack in the air and it whiffs, the game allows you to buffer the motion and button press before you land (I’m not sure how many frames before you land though), and the special will execute on the second landing frame (that’s why you can get hit by sweeps). If however your aerial attack connects, then you can buffer the motion, but the game will ignore any button presses until the first landing frame. In either case, the special will execute on the second landing frame.
thank you guys… you helped me a lot…
anyway, I still hate this function! hahahah…
Not sure if anyone’s touched on this, but I’m just gonna verify. With Chun-li I cancel cr.mk into super like this:
:db::d::df:+:mk::db::r:+:mk:
However, if I used :df:+:mk: at the end, the super wouldn’t come out.
So when charging for a move there’s leniency on the charge input (could be :db:,:l:,:ub:) but the input to release the move must be a true horizontal. Specials will come out when the last input is a diagonal that is also registering a horizontal. Like for her ultra:
:l::r::l::r::df:+:lk::mk::hk:
Don’t know if that helps, just sayin.
The input leniency for the generic charge :l::r::l::r: super motions are lenient for the first three directions but not the last :r: (which is restricted to :r: or :uf:).
Input leniency for charge special moves depends on the move itself. So, Blanka’s horizontal ball is a strict :r: while Dictator’s scissor kicks is a lenient :r:.
Strange, I can’t pull out an ultra with :uf: at the end without there first being a :r: then :uf:…
So just from testing it, I guess it may be Chun-li specific where using an :uf: motion to finish an ultra or super won’t work.
This works :l::r::l::r::uf:
But I can’t get this :l::r::l::uf: or this :l::r::l::df: to work.
Also note that her kikouken has a lenient back charge motion but is strict to :r: when releasing, like Blanka. The characters that have a lenient :r: are lucky. I wonder what boxer is like.
[edit] Sorry, I should clarify that I’ve only tested charge moves with Chun-li, I’m interested to check everyone else out later though.
I think you can end a special on any motion as long as the button press follows a legal end direction within a certain number of frames (and of course you’re in a state to perform it). These frames are hardcoded onto each motion itself. So basically, it seems like the end motion for a horizontal charge move has to be :r:, but you can input anything afterwards as long as you press the button quickly enough. That’s why pre-charging works.
:l:, :r:, :l: + :p:
That’s what I said earlier, for the special to come out, your input can be :uf: or :df: at the end as long as it immediately follows a true horizontal :r:, but from testing, ending only in a diagonal (no horizontal before that last input) will not work.