I would only compare the Classic data and Remix data to get an idea of how different a Remix attack is. I don’t know why else anybody would want to compare them.
Rufus captured Cammy’s Thrust Kicks when they whiff, not on hit or block as hitstop would mess up the frame counts. I assume the Classic data is also whiffed as Rufus captured the exact same data.
Ganelon’s tests. An interesting read.
Well I mentioned it since you were using alot of akiba’s stuff. :chat:
Most of the Classic data is YBH.
A problem with T.Akiba’s data is that he seems to use both the arcade and Dreamcast versions.
On his site, some data such as stage speeds have to have been calculated using an arcade version (since I could find no difference in the DC version). In different stages (Spain for instance), there’s noticeably more frameskip for all moves.
Yet T.Akiba specifically mentions using the DC version on speed 0 to calculate some frame data (understandable since there’s no 0-frameskip speed in the arcade versions of ST). Because there’s no 0-frameskip speed in arcade ST, it’s impossible for moves to have the same number of frames (at least visible) from stage to stage in that version.
More frame silliness…
http://www.pedantic.org/~nate/sim/
http://www.pedantic.org/~nate/thawk/
(I know the alternative jab is missing)
This was a question posted by mikeidge. I’m reposting here cuz I’m really interested in a response to this, I really hope you don’t mind mikeidge.:pleased:
The animation is frame-doubled - i.e the game is running at half speed - while the opponent is reeling. In the case of Chun’s Jab fireball, it seems to last a full second at turbo 0 (what would normally take 1/2 second takes a full second) I think this is true for all of the projectiles… At higher turbo settings, the slowdown time is reduced by the turbo speed-up.
During the slowdown, the system will drop inputs. The SE stick’s turbo on roundhouse will sometimes fail to continue lightning legs through her slowdown, but does so reliably otherwise. The system probably ignores input during the ‘extra’ frames.
Charge times are determined by ‘game time’ calculations - so you have to charge up for longer to compensate for the slowdown.
Superflash pause doesn’t have any strange interaction with the slowdown. Before and after is at half speed - total slowed time appears to be unaffected.
Hit pause times are doubled by the slowdown.
In my experience, button inputs are harder, but directional inputs are not ignored. If I’m using Guile, and I want to combo Backfist after a Boom hits my opponent, holding forward and rapid firing Fierce works. Also, when I have a charge character, and a fireball hits me, but I was already charged down, holding up and rapid firing punch or kick will get my charge attack to come out. I don’t have to do b, f or d, u any slower than normal. I can say for sure that charges are definitely not dropped. But you are right, actual charging takes longer during slowdown.
As for non charge characters, I think the same thing applies, directions aren’t dropped, but button inputs have to be rapid fired. If a Hadoken hits my opponent, and I want to throw another Hadoken, I do the qcf at a slower speed and rapid fire punch while holding f, and that usually works.
Super flash is a weird phenomenon. I’ve seen videos of top players like Aniken able to Shoryuken after wakeup to escape/punish meaty Supers. But I haven’t found a way to consistently get an attack to come out that I was trying to do before the super flash. With charge characters, I’ve tried holding f or u during the pause, and rapid firing punch or kick, but this has worked sometimes but not other times. I think if the direction had already been input before the flash, that’s the only time it works, but I could be wrong. Same thing with non charge characters. IMO super flash definitely drops directional inputs.
Wouldn’t that make ken’s super unblockable as a meaty? Or is it all non back or non down back directions are dropped?
Maybe they’re not dropped. Maybe they’re frozen, and all other directional inputs during the super flash are ignored.
This is the sort of behavior that would be expected if the system ignores input on every other frame. (This is probably the side effect of a hack to prevent Guile and Deejay from using the slowdown effect to recover charge at real time rates.) If the edge detection only happens within individual frames, then that means that special move button inputs are much more likely to be lost and mash moves like lightning legs would be the most affected. I don’t really have the tools for testing this hypothesis.
What I mean is that inputs made during the super flash - most likely shifted by the input lag - are ignored as if the game were paused during that time. So, for example, if you tap up on the stick in the 1/3 second that the super flash is on the screen, your character won’t jump. It would (once again) basically require programmable controllers to check whether the input of moves like fireballs can be split across the super flash. Directional controls are continuous, so if you hold the stick in some direction, the character will start jumping/crouching/walking immediately, so if you continue to hold the direction past the end of the super flash (as if for blocking or jumping) it’s still executed at the end of the flash. (Holding directions like that can be useful if you know you want to jump or move immediately after doing something. There are several examples where jumping immediately after a move recovers will lead to safe jump-ins so a player can just hold the stick up/forward to get a ‘perfect’ jump.)
Hmm… perhaps the testing doesn’t require programmable controllers…
The reason I mentioned ken’s super specifically, is because it hits on the first frame. Here’s a way that might test that.
Ken stands point blank, opponent holds down back. When the oppoent sees the super flash, they let go of down back (let it go to neutral) then go back to holding down back or back and see if they can block the super.
There’s also the problem that first frame moves like ken’s hp srk/super and blanka’s rolls are still 25% unblockable.
You mean 50% unblockable? That was fixed for Remix, tho it can still be turned on via dipswitches. If the opponent was already holding back before the super flash, all supers will be blocked IMO. If the opponent was not blocking, they will be hit by 1st frame hitting supers, regardless of whether they held back during the flash IMO. Tho Ken’s Super hits on the 2nd frame: 1?(17)?1?2?6?13?2?2?26?30?2 as per NKI’s site and YBH.
If you read the top of that page NKI says that the game “pauses for 1 frame before and after the super flash.” That means that when the pause ends, he immediately has hitting frames, meaning there is no technical startup.
No, its 25% without the dip switch. If you block ken’s hp srk 4 times in a row, one of those will still hit.
Its easy to see it if you fight CPU ken in training. When he does hk tatsu into hp srk, there’s roughly a 25% chance that the hp srk will blow through your block.
Turn the CPU on and just hold back against his hk tatsu, he will always let an hp srk fly.
This is actually annoying because against people who whore out hp srks online, you can actually get hit, by this occasional glitch. It is definitely reduced from 50%, but its not fixed.
CPU ken is the original flowchart :tup:
Ah ok. *That *I didn’t know. Thanks Coth.
For some reason, this rings a bell. Do you have a link to an article or a thread that proves this?
http://forums.shoryuken.com/showthread.php?t=168773&highlight=unblockable&page=14
Pretty much posts 344 to 350 was where we talked about it. Originally I thought it had to do with me crouching after blocking the hk tatsu. That isn’t it though, just go into training and fight the ken computer, block his tatsu standing and just hold back and sometimes the hp srk is unblockable.
This confirms that they did something to it, but it is still there, just to a lesser degree.
It takes less than 10 minutes to test against the computer. Hell you could tell the computer to all block and push them into the corner and do repeated hp srks too. I think I’ve done that before too.
I’m going to go read the thread, but I would very much like to see a PvP test of this (which should be easy) since the computer gets to cheat in other weird and wonderful ways. (Yesterday I got hit out of the Raging Demon by Shin Akuma…) If it’s always happening in the corner, it might be a tricky cross-up.
Oh yeah, totally forgot about that. = ( It would be nice to figure out exactly what is happening frame by frame.
That’s what I thought at first too, but it can be done in the middle of the screen by a human opponent as well.
It’s definitely not hitboxes registering on the reverse side for the unblockable FP uppercut. Rather I’m guessing that it’s the uppercut hitting before the opponent changes from walking backward to blocking.
I cought one on frame-by-frame video, and the victim doesn’t switch to the first blocking frame on the first frame of the DP. Has anyone checked if point-blank FP Reversal DPs are unblockable?
YAH! That’s what it was in ST, they had a 50% chance of being unblockable on the first frame. It was suppose to be fixed in HDR, but its more like 25% now.
Reversal? Just if it hits on the first frame. Reversal or not.