ComboVid.com

Posted: **Mon Mar 05, 2007 12:10 am**

This came up on another board, in some random topic about super-basic game engine fundamentals ...

6) When two characters walk toward one another at point blank range, what determines the outcome? Everyone assumes that the character with the fastest walking speed will push the other character back. This notion always seemed kinda suspicious to me. There are some instances where two characters walk towards one another and neither moves forward, yet frame data says that one character is supposed to have faster walking speed.

In ST, if Guile and Rog walk towards one another, Guile pushes Rog back. If Ryu and Guile walk towards one another, neither moves forward. Strangely enough, if Ryu and Rog walk towards one another, neither moves forward. That's fishy because if Guile pushes Rog back, then you'd expect Ryu to do the same. The weird thing is that the frame data lists Rog's walking speed as 56, Guile's as 54, and Ryu's as 48.

Then there are some instances (Vega vs Cammy and Honda vs Blanka) where one character pushes the other even though their speeds differ by only two points. In fact, Honda pushes Sagat back and ties with Ryu, even though he's supposed to be slower than both of them. It's just not an accurate gauge of relative character speed. This arbitrary shit might also affect one character walking forward while the other stands still.

7) When Blanka and Vega walk away from one another, starting from the exact middle of the screen, the stage scrolls slightly in Vega's direction since he's got faster walk speed. However, when they get to the edges of the screen, the stage stops scrolling even if they continue walking backward. Why doesn't the stage keep scrolling towards the faster character?

Posted: **Mon Mar 05, 2007 6:47 am**

About your #7, Maj, I did a Old character vs New character walking speed test in ST awhile ago and found out that if you start in the middle of the screen then walk back, the game scrolls the screen towards 1 side of the screen and it's always that side. I thought I proved that old characters moved faster because the screen scrolled 1-2 pixels but when I changed who was 1p and 2p and tried again, the same side scrolled 1-2 pixels. So you might want to switch Blanka and Vega's sides to try to eliminate that weird ST background scrolling effect.

BTW, the old characters and new characters have the same walking speed in ST. Proved after 2 hours of walking back and forth with each new and old character. I did however found out that some old characters have different amount of jump frames than the new characters. I remember Ryu, Ken, and Sagat for sure to have this weird property about them. I basically mapped the 4 directions for 1p and 2p to the same buttons so that they would mirror each other. Then I just held straight jump for about 10-20 reps, let them stand, then diagonal jump for 10-20 reps. One of them lands faster than the other. I wonder why they made it like that for some of the characters and not all of them. It's not even the grappler characters that have this property but shotos. NKI?

Posted: **Mon Mar 05, 2007 10:07 pm**

That's interesting, but that's not what happens in the Vega vs Blanka case. Vega's walk speed is so much faster than Blanka's, that the screen scrolls way more than 1-2 pixels in Vega's direction. It's probably like 30-50 pixels.

Some of those test results are really bizarre. I wonder if those old characters have less airborne frames during their jump or less ground-based startup/recovery frames. Cuz if it's the latter, then it would make them harder to throw and it would make it easier to do Shoto air hurricane kicks using the tigerknee motion.

Posted: **Tue Mar 06, 2007 3:22 am**

It is also wierd that you can not push a grounded character with a character in flymode but you can push a character in flymode with a grounded character.

Posted: **Tue Mar 06, 2007 5:11 am**

Mike Z wrote:Random note: CPU Zangief in XSF can cancel this throw directly into standing attacks, but the CPU can do that with everyone (launch->s.Jab as a combo). XSF CPU Zangief can also do chains on the ground (c.Short->s.Strong->c.Fierce FTW).

My favorite that I've seen is still Gambit doing Crouch Roundhouse chained into Crouch Roundhouse chained into... walking!!! Didn't even need another move. Roundhouse connected and he was instantly just walkin' along, chilling.

- James

Posted: **Wed Mar 07, 2007 8:23 pm**

The CPU has a lot of oddities in the Alpha series as well. Most notably the Guy AI being able to combo his FF chain in strange ways (such as two fierce elbows or two roundhouses) or whiff his entire FF chain out in the open. "Rock Lee" on SRK has noted that the CPU Sodom is able to combo two roundhouses out of nowhere as well.

I'm sure there's plenty of other examples.

Posted: **Thu Mar 08, 2007 10:02 pm**

BB Hood wrote:It is also wierd that you can not push a grounded character with a character in flymode but you can push a character in flymode with a grounded character.

Can two flymode characters push each other?

jchensor wrote:My favorite that I've seen is still Gambit doing Crouch Roundhouse chained into Crouch Roundhouse chained into... walking!!! Didn't even need another move. Roundhouse connected and he was instantly just walkin' along, chilling.

Haha that sounds way too stylish for no good reason whatsoever. I remember in that one Evo DVD trailer showing SooMighty's legendary match against Justin Wong, there was one part where Magneto kills one character then dashes forward once. Then the video highlights his motionless standing animation while waiting for the next victim to step up. That was helllllluva stylish.

Xenozip. wrote:The CPU has a lot of oddities in the Alpha series as well. Most notably the Guy AI being able to combo his FF chain in strange ways (such as two fierce elbows or two roundhouses) or whiff his entire FF chain out in the open. "Rock Lee" on SRK has noted that the CPU Sodom is able to combo two roundhouses out of nowhere as well.

That sounds pretty dope actually. I'd love to see that double roundhouse.

I think at one point NKI was considering putting together a video of all the absurd bullshit that the CPU gets away with. But it would take way too much time (and patience) for one person to do alone.

Posted: **Wed Mar 28, 2007 2:51 am**

8) How much frame advance does superfreeze generate in 3S? For reference, here's the answer for CvS2:

Buktooth's CvS2 Systems FAQ wrote:------------------------
SUPER TIME FREEZE EFFECT
------------------------

Ever wonder why you can't block Sagat's level 3 hotfoot super if you weren't already blocking low when the super flash happened? Popular belief says it's because blocking and crouching aren't instant; they have start up times that are too slow to stop Sagat's hotfoot in time. That's only partly true. All supers have a small amount of time after the super flash where all inputs by the opponent are invalid. This explains why you can never jump out of Yamazaki's level 2 grab super on reaction, even though the grab has 4 frames of start up and it only takes 2 frames to jump.

added 7/15/03 - In case anybody was wondering, Sagat's hotfoot super comes out in 4 frames after the super flash, at any level.

The following chart shows how many frames each level of super "stops time" for.

level 1 - 3 frames
level 2 - 5 frames
level 3 - 7 frames

Actually the Yamazaki thing is an inaccurate example because CvS2 jumps are invincible to throws from the very beginning. Even though you don't become airborne until the 3rd frame (or later, depending on the character/groove), you're invulnerable to grabs even during the ground-based startup of jumps. That's why you can jump out of Yamazaki's lvl1 grab super after superfreeze but not the lvl2 or lvl3 versions. In fact, it seems he clarified this way back in 2004 but it's still all confusing in his guide. So yeah, just focus on the numbers at the bottom.

Posted: **Wed Mar 28, 2007 4:00 am**

Is that considered frame advance, where your animation is still frozen in time and only the character doing the super gets animated? If you were in the middle of a move during super freeze, would you still be animated normally during the frames you can't input?

I did a test few months ago to see the nature of the input freeze after the super flash. Here's how I set it up. Dummy Guile does a move to signal for the super, then Sonichurricane.

I've noticed that if you weren't blocking before the super flash you couldn't block, but if you were on like the last frame of an attack you did during super flash, you could block the super after the flash. This was months ago so I'm not 100% confident in saying this, but I'll probably go home and test this again today.

Also, isn't the leeway for RCs something like 3 frames? I've also done a down+roll right as the super flash happened (you can see the early animation of the roll during the flash) then inputed db,b+kick after the flash and got an RC hurricance kick.

This is unrelated, but I've found out the other day that the select button (taunt) in the ps2 cvs2 is actually a macro of some sort which i think is start+short. I mashed the select button with Chunli and got lk hundred legs. I haven't tested whether it's really start+lk or thought of a way to test if the other button is really the start button, but I think capcom might've done this deliberately to keep things in sync with the DC version. It makes sense to keep the taunt command from the DC version since the PS2 version and the DC version were cross-platform online-capable games. So keeping the command for taunt to be start+lk even on the PS2 version makes programming the online communication between the two systems easy (communicating every input from every frame).

Posted: **Sat Mar 31, 2007 6:46 am**

I call it frame advance but there really isn't a widespread term for it. If you're in the middle of a movie during that frozen period, you're stuck in that move for however many frames it lasts.

I'm pretty sure that Sonic Hurricane is unblockable after superfreeze unless you were already blocking. It's very reliable in that regard. However, there are a lot of lvl2 supers that tend to be unblockable post-superfreeze if you were standing but they are blockable post-superfreeze if you were walking forward. I'm sure walking forward isn't the only action that this applies to. Honestly i'm not sure if it's worth spending time to look into. Superfreeze is just one of those rare things that carries ±1 frame discrepancy on a random basis. Capcom games tend to be very consistent, but there's bound to be some unpreditable bullshit.

Anyway i asked eKiN about this 3S superfreeze stuff and he had a lot of interesting info.

IM conversation with eKiN wrote:Maj: 3S Ken c.MP link into SA3 is a one-frame link?
eKiN: no way
Maj: according to this, c.MP is only +4
Maj: http://web.archive.org/web/200411092322 ... m/ken.html
Maj: and SA3 is 3 startup
eKiN: karathrow data is wrong
Maj: dang, really? why?
eKiN: DC data
Maj: oh, ouch
eKiN: even on DC, i dont think it was right
Maj: so where do you get arcade data?
eKiN: http://sf33rd.web.fc2.com/
eKiN: w/ translation tool
Maj: what the hell is this? http://ensabahnur.free.fr/Baston/index.php
eKiN: i think thats relisted kara throw data
Maj: oh, awesome
Maj: i love it when bad info gets perpetuated foreva
Maj: i'm gonna email this guy letting him know
Maj: got a "karathrow data is wrong" thread i can reference?
eKiN: maybe im wrong
eKiN: on japanese site it seems to be the same
eKiN: +4 for c.MP vs standing
eKiN: but its pretty easy so i dunno
Maj: the good thing is i have no clue whatsoever how supers work in 3S
eKiN: umm let me see
eKiN: "I buy brady games book so i'm going to correct startup values of Karathrow data which are wrong (or which are for DC and not for PS2/Arcade)..."
eKiN: guess he might have changed it already
Maj: ok, so explain 3S supers to me plz
Maj: i mean, if you've got nothing else going on right now, hahaha
eKiN: what about them?
Maj: i find it hard to believe that c.MP link into SA3 is a one frame link
eKiN: me also
eKiN: but according to people like paullee
eKiN: kara throw is one frame
Maj: everyone and their grandma does that link
Maj: arthritis can not stop that link
eKiN: lol
eKiN: so if kara throw is one frame execution wise, which i also dont believe
eKiN: well then i dont know what conclusion to make about 3S
eKiN: okay, its a 2 frame link
eKiN: Ken's far s.HP vs crouch is +3 and it links to super
Maj: ok so what do you think about supers? is 3 frame startup wrong?
eKiN: 2 frame start up, hits on 3rd frame, imo
Maj: then why does it say 3?
eKiN: =[
eKiN: i dont know, but you can test it with far s.HP -> super
Maj: dude shinryuken says 1 frame startup
Maj: there's no way that's 0 frame
Maj: do we know anyone who would know?
Maj: complete this analogy
Maj: bas is to CvS2 as ____ is to 3S
eKiN: you can try ask jinrai
eKiN: he knows a lot of the system stuff
Maj: : (
Maj: i want someone who doesn't speak my language
Maj: so i can't understand what he says so i can just assume whatever the hell i want to assume and tell everyone i got it from a japanese expert
Maj: like, "alex 360 combos if you kara cancel his low short -> low forward chain and if you can't pull it off then you are simply not japanese enough and if you ever come to FFA we're going to ignore you cuz you don't know what the ume shoryu is all about"
eKiN: lol
eKiN: ok, from SRK
eKiN: the article got it wrong. Ken's SAIII has a 2 frames startup (that's the correct data), and that means that it hits on the 3rd frame. Chun's SA2 has a 3 frame startup. So at least -3 is needed for shippu punishing, and -4 for chun (and take into account that Chun takes 2 extra frame of blockstun when crouching, so she can punish crouching moves that put the opponent at -6 ormore on other characters.)
This is the standard notation for 3s frame data: startup frames are all the frames before a move can hit.
So if a move has n startup frames, it starts hitting on the (n+1)th frame.
eKiN: also
eKiN: as Leva pointed me out yesterday on MSN, there are still some weird things about the game (or either a 100% correct framelist does NOT exist, even if the japanese one - game restaurant - is definitely the most accurate one.

Every framelist points ken's cMK frame advantage the same as Ryu and Gouki (that is, -3 on block, -2 on hit, -1 on crouching hit)
but testing evidence points differently. How? It's simple: reversal shippu (2 frame startup, hits on 3rd) can actually punish a CONNECTED Ken cMK (on another standing Ken). standing hit Ryu/Gouki's cMK cannot. Also, blocked Ken's cMK (upclose) can be punished by reversal Gouki SA1, that by evidence has, then, a 3-frame startup (hits on the 4th). Can't happen vs Ryu/Gouki.
So, at this point, one would guess that Ken's cMK is actually -4 on block and -3 on hit, while Ryu/Gouki confirm their -3/-2 (still no evidence about crouching hit)
Now, with dudley things get even messier. Corscrew Blow startup is listed as 1, hitting on the 2nd hit. That would mean that it can punish Ryu/Gouki standing hit cMk too, but guess what? it doesn't.
At this point one would guess that it's startup is actually 2 (hitting on 3rd frame) but that, doesn't make sense for his links to work, since stuff like sMK and f+HK on crouchers give a +2 advantage (if THAT data is correct).
Things get even more messier after that: Dudley's SA2 punishes ken's standing hit cMK, and that would put it at a 2 frames startup, hitting on 3rd (which is what the game restaurant frame data say), but it can't link from stMK, or f+HK like SAIII.
At this point I concluded that Dudley's SA3 is actually a 2nd frame hitter, but that Dudley takes 1 frame of extra blockstun/hitstun when picking SAIII, since those link work but vs shotos it can punish the same stuff as shippu (AND NOT MORE THAN THAT!!!)
But then, reversal corckscrew punishes blocked Chun's cMK while shippu can't. I don't really get this. Pushback should not be a factor. Also, Kikoshou seems to be even faster than corckscrew (though they are listed to have the same speed) sinceit punishes even a croucing hitting Ken's cMk (but again, not Ryu/Gouki's).

eKiN: After all of that, the only thing I concluded is that Ken's cMK is -4/-3/-2 while Ryu/Gouki's are -3/-2/-1

In the end, either it's the PS2 conversion that's busted, or there are many unknown factors that alter those should be "set in stone" frame data.
i'm really getting a headache out of this
eKiN: interesting? lol
Maj: um, wow
Maj: someone needs to shine the James Chen signal on 3S
eKiN: did you know for example
eKiN: chun gets -2 when she blocks low, added to whatever move
eKiN: and ryu gets -1
Maj: so chun recovers faster?
eKiN: slower
Maj: oh, then +2
eKiN: added to disadvantage
eKiN: yea
Maj: ok, that makes no sense to me but luckily you play more 3S than i do so i can blame you with the traditional "it's your game!" comment
Maj: vintage shgl
eKiN: lol
Maj: by the way, your lakers sucked today
Maj: ok, so seriously
Maj: the real answer to how 3S supers work is that nobody knows
eKiN: basically its a bit hard to tell wtf is going on
Maj: but i have a followup question
Maj: you know how in CvS2, a lot of lvl2 supers and almost all lvl3 supers are unblockable if you weren't blocking by superfreeze?
eKiN: yep
Maj: does that happen in 3s?
eKiN: nope
Maj: so there's no frame advance type thing, where the guy doing the super gets some frames while the other guy is frozen?
eKiN: i dont think so
Maj: oh, also, does the screen freeze on first frame of supers in 3S?
eKiN: seems that way
eKiN: because chun's SA2 can be done wakeup through fireball
eKiN: but her super has no invincibility
eKiN: it just moves forward during screen freeze
Maj: haha nice
Maj: ok, that helps a lot
Maj: and it kinda makes sense that supers would be so fast if there's no frame advance
Maj: i mean, almost all CvS2 supers have 4 frames startup before superfreeze and 4 frames after
Maj: so you need a lot of frame advance to make up for that
eKiN: oh i see
eKiN: yea
Maj: what happens with hugo 720?
Maj: can you jump out after superfreeze?
eKiN: nope
Maj: it says 1 frame startup, which might be the first superfreeze frame, which would mean 0 frames after
Maj: what about alex?
eKiN: i think you can jump as long as you werent crouching before
eKiN: something like that, not 100%
Maj: sounds right
Maj: says 3 frame startup
Maj: ok, i think i am done wasting your time
Maj: you can go back to signing autographs now
eKiN: lol im happy to talk about 3S anytime
eKiN: haha
eKiN: ok peace

Posted: **Sat Mar 31, 2007 7:47 am**

Yay, see I used to try and understand 3s and then I found out about all the random bs. My belief is there is no correct frame data for this for one reason or another. My other belief is that the way capcom did the hitboxes, in edition to all the animations involved with people just standing around or couching makes things hit funny so you don't get the exact data.

Oh and towards the dudley thing, maybe it doesn't hit at it's full extended point right away?

Which actually leads me to think maybe there are two sets of frame data for certain moves in this game. In GGXX AC Slayer has two sets of frame data for certain moves, like the move actually can hit you at two different points even though it's still only a one hit move. I just tried to look for the specifics of that someone posted but I couldn't find it on dustloop again and it's late, but basically say with slayers standing punch, the move hits at two different points, one near his shoulder and then at the very extension, you get better frame advantage if you hit with the latter portion as because you don't have to wait for the move to fully extend. Of course this doesn't explain all the wonky 3s frame stuff, but it explains most of it.

You can always jump out of alex's hyperbomb as long as you weren't doing something else. You can combo into the hyper bomb if you use the aegis relector for some help though i doubt you can do it off of crouching short..

Posted: **Sat Mar 31, 2007 7:53 am**

Kind of like how Cammy's standing roundhouse(highkick) will hit with the knee before the toe. Even though it's just one move and will never hit twice.

She spins, the knee extends and can hit, then she extends her foot and that can hit too. If the knee whiffs the foot can hit, but if the knee hits the foot will not hit.

Posted: **Sat Mar 31, 2007 8:13 am**

That part about Alex was just me joking around. I dislike the term "Ume Shoryu" because we have a much better term for it already: "Psychic DP" is what it's been called for ages. Also, no disrespect for Daigo whatsoever, but there have been plenty of local players doing that shit since World Warrior days. I hella respect the scene in Japan and all they've done for SF, but i also hate it when the contributions of non-Japanese players get overlooked by overzealous supporters of a country 5000 miles away.

Again, i have no problems whatsoever with giving props to Japanese players. That community has a mind-boggling combination of talent and dedication. It's simply a pleasure to watch any match videos or combo videos that come out of that country.

However, i do have a problem with local history being forgotten or belittled in the process.

Posted: **Sat Mar 31, 2007 9:21 am**

I think fullmetalross is on to something about the picky hit boxes in 3rd Strike. The hit boxes are waaaaay more complicated than just squares in that game for every animation frame, so I suspect that Frame Data is just a lot harder to calculate. I mean, when Chun's Super can hit Makoto a few times and then Makoto can suddenly block randomly, you know the game is just really specific. So I think it's almost impossible to count frame data in that game unless you can see the hit boxes appear or something.

Plus, even though I have almost next to no knowledge on Third Strike, I've always believed that some Normal Moves can be canceled by Supers even before they finish, not necessarily being Buffered. I mean, Makoto's Low Forward is a good example. With SAI, you can link Crouch Forward into Super, right? I swear that Super comes out before Makoto's Crouch Forward finishes. If you hold Up, Makoto jumps way after her Crouch Forward ends. But when you "link" it with the Super, it looks like you canceled it much earlier than her official end of the move.

Also, there has to be some Super Freeze where the character gets to move a few frames before the opponent can move. Otherwise, you could Parry ever Super after the Super Freeze, right? Provided, that is, you weren't already holding forward when the Super activated.

Third Strike definitely has the oddest rule set out of any Capcom Fighter. I don't think there will ever be a day when the rules of that game can be definitely written down.

Posted: **Sat Mar 31, 2007 4:02 pm**

I thought that part about makoto was interesting so I did some testing and this is what I've found:
*note: all tests performed with 0 dropped frames, and using each field to reconstruct 60fps as described on my blog </plug>
**note2: first animation frame of c.mk referred to as frame 1

test 1:
c.mk hits standing ryu, I hold UP so Makoto jumps asap after move is completed
result:
the first frame of jump starts on frame 27

test 2:
c.mk hits standing ryu, I hold FORWARD so Makoto starts walking asap after move is completed
result:
the first frame of the walking animation starts on frame 27

test 3:
c.mk hits standing ryu, I palm mash all buttons to get any attack asap (I repeated this test several times to make sure I got it asap).
result:
quickest atk frame is on frame 27

test 4:
c.mk hits standing ryu, I buffer motions and piano buttons for SA1. (Repeated several times)
result:
earliest SA1 started on frame 27

In conclusion, I can't find any evidence that Makoto's C.MK is loose or weird with regards to canceling into jump, walk, attack or super.

I'm getting off-by-one problems with all normals (c.lp c.mk c.mp so I assumed it applies to everything). c.mk which is supposed to have 7 startup (game restaurant) has its first hit animation on frame 9, which would mean it actually has an 8 frame startup. You could explain this away by saying that though the impact animation as well as the other character's being-hit animation show up at frame 9, the impact has internally already taken place at frame 8. HOWEVER, I have absolutely no basis for this, but it seems likely since the 8th frame of c.mk is Makoto's fully extended leg overlappinog with Ryu. I know from testing also that CvS2 has its impact animations appear on the actual hitting frame, and I believe this is the case with all versions of SF2 also.

PS. I have 60 fps clips for each of these tests if anyone wants to see and count frames for themselves.

Posted: **Sat Mar 31, 2007 5:04 pm**

While it's true that her cr.mk isn't weird like that, Makoto has some other moves that are weird in a similar way. Specifically, her dash and her neutral throw. Dash -> hold forward to walk, she takes forever to actually start walking, as she goes through this "recovery animation" for her dash. But Dash -> hold up to jump, and she jumps after only 12ms.

Similarly, her Neutral throw has this long "recovery animation" after it. She's completely invincible during this - if you have Ken reversal shippu after she throws him, he'll pass right through her. But if she does anything, the "recovery animation" aborts and she gets hit.

Also, there has to be some Super Freeze where the character gets to move a few frames before the opponent can move. Otherwise, you could Parry ever Super after the Super Freeze, right? Provided, that is, you weren't already holding forward when the Super activated.

You can parry after the freeze, it's just really, really hard for most supers. Some supers (full screen abare, Q's critical combo) are slow enough that they can easily be parried after the freeze.

If you have a 10 frame window for parrying and a 2-frame start up on the super, then you have 8 frames before the freeze to input the parry, and 2 frames after the freeze. As a result, you'll have way better success if you input the parry before.

Posted: **Sat Mar 31, 2007 7:14 pm**

- Makoto dash recovery and neutral throw recovery are different. Dash recovery is cancellable by moves and jumping, neutral throw recovery is EXACTLY Tanden startup and is cancellable by anything. She also has a bit of Tanden after she connects with a Karakusa (try it!). In fact, it seems the only things that dash recovery isn't cancellable with are the things that would let you dash twice quickly (walk fwd/back, dash fwd/back, crouch), which is why it's quicker to do dash, cr.Jab, dash, or dash, F+RH(hold).

- karathrow is wrong, if it were right Makoto would be able to do s.Strong xx Abare.

- You can combo Alex's 360 after a Flash Chop or a F+Fierce (you don't need Aegis, heh.)

- 3s seems like it has to handle super freeze differently with regards to parrying. Look at Alpha Anthology (only for a little, I promise :^). You can parry anything after the flash simply by holding the parry direction during the flash. To your character, it looks like neutral (superflash) parry-dir, because they receive no input during the flash. Either this is not true of 3s (which it probably still is otherwise you could buffer your super during their flash) or they are handling it some other way. Someone chime in with CvS2/CFJ info?
Also, look at the Active Super Freeze thing in 3s - if it were the same as other games' freezes, there's no way they'd have that happen.

- The Slayer thing about the move hitting differently: The whole point of that was that the move is active at a frame where it's not fully extended, and continues to be active through full extension. With Slayer, his 5p is active as soon as it appears as a little fist, but continues to be active through full extension. The difference is that if you hit with the initial frame, you have to go through more of the animation after the hit, so your frame advantage at the end is less. It's the opposite of a meaty move. They were trying to explain that and didn't do it very well.
They also ended up getting it confused with the fact that Slayer has extra hitstun on all of his moves so they stun for longer than they should for their Level.
This probably applies to Cammy hitting early as well.

Mike Z

Posted: **Sun Apr 01, 2007 12:21 am**

Mike Z wrote:- Makoto dash recovery and neutral throw recovery are different. Dash recovery is cancellable by moves and jumping, neutral throw recovery is EXACTLY Tanden startup and is cancellable by anything. She also has a bit of Tanden after she connects with a Karakusa (try it!). In fact, it seems the only things that dash recovery isn't cancellable with are the things that would let you dash twice quickly (walk fwd/back, dash fwd/back, crouch), which is why it's quicker to do dash, cr.Jab, dash, or dash, F+RH(hold).

- karathrow is wrong, if it were right Makoto would be able to do s.Strong xx Abare.

- You can combo Alex's 360 after a Flash Chop or a F+Fierce (you don't need Aegis, heh.)

- 3s seems like it has to handle super freeze differently with regards to parrying. Look at Alpha Anthology (only for a little, I promise :^). You can parry anything after the flash simply by holding the parry direction during the flash. To your character, it looks like neutral (superflash) parry-dir, because they receive no input during the flash. Either this is not true of 3s (which it probably still is otherwise you could buffer your super during their flash) or they are handling it some other way. Someone chime in with CvS2/CFJ info?
Also, look at the Active Super Freeze thing in 3s - if it were the same as other games' freezes, there's no way they'd have that happen.

- The Slayer thing about the move hitting differently: The whole point of that was that the move is active at a frame where it's not fully extended, and continues to be active through full extension. With Slayer, his 5p is active as soon as it appears as a little fist, but continues to be active through full extension. The difference is that if you hit with the initial frame, you have to go through more of the animation after the hit, so your frame advantage at the end is less. It's the opposite of a meaty move. They were trying to explain that and didn't do it very well.
They also ended up getting it confused with the fact that Slayer has extra hitstun on all of his moves so they stun for longer than they should for their Level.
This probably applies to Cammy hitting early as well.

Mike Z

Ohhh okay thanks for that slayer thing mike. And I forgot about the flash chop hehe. Oh well.

Posted: **Sun Apr 01, 2007 7:28 am**

Hey, thanks for testing that out, ZenFire. Guess it was just a perception thing. It always felt like it canceled earlier than Makoto recovered, but I guess not.

- James

Posted: **Sun Apr 01, 2007 8:15 am**

ZenFire wrote:In conclusion, I can't find any evidence that Makoto's C.MK is loose or weird with regards to canceling into jump, walk, attack or super.

Did you try testing how long it takes for her c.MK to recover into standing animation? I know it's kinda tough to distinguish but i'm just looking for an approximate number. It seems to me like there's a lot of extra animation at the end of 3S moves that can be canceled into anything you want, including walking and such. It's just there to make the animation look more fluid but instead it makes everything deceptive.

Mike Z wrote:- 3s seems like it has to handle super freeze differently with regards to parrying. Look at Alpha Anthology (only for a little, I promise :^). You can parry anything after the flash simply by holding the parry direction during the flash. To your character, it looks like neutral (superflash) parry-dir, because they receive no input during the flash. Either this is not true of 3s (which it probably still is otherwise you could buffer your super during their flash) or they are handling it some other way. Someone chime in with CvS2/CFJ info?

As far as i know, CvS2 is way less ghetto in that regard. There are some supers that you can parry after superfreeze, but frame advance ignores parry attempts too. Therefore, most lvl2 or lvl3 supers must be parried before superfreeze. At any rate, you can't just hold a parry direction during superfreeze and have it count automatically when gameplay resumes.

Honestly it's not particularly surprising that HSFA parries ended up kinda ghetto. They couldn't have had very much time to polish each of those new ISMs.

Posted: **Sun Apr 01, 2007 7:23 pm**

jchensor wrote:Hey, thanks for testing that out, ZenFire. Guess it was just a perception thing. It always felt like it canceled earlier than Makoto recovered, but I guess not.

- James

Np, glad I could contribute.

Maj wrote:Did you try testing how long it takes for her c.MK to recover into standing animation? I know it's kinda tough to distinguish but i'm just looking for an approximate number. It seems to me like there's a lot of extra animation at the end of 3S moves that can be canceled into anything you want, including walking and such. It's just there to make the animation look more fluid but instead it makes everything deceptive.

I just did, and it's still frame 27 counting in the same way as the previous tests. It's not hard to distinguish in this case and probably any crouching moves, since the first frame after the move is completed is the exact same pose/animation-frame as the one you get for walking forward after a move (since she needs to get up before she can walk).

Btw, testing this made me appreciate the animators that made her crouching -> standing animation, the attention to details like this is why I'm in love with the art in 3S.

Posted: **Wed Apr 11, 2007 4:24 pm**

That's weird because for example, i think Shoto close s.HP has a few frames of extra disposable recovery animation that you can avoid if you do anything else. Maybe i'm just seeing things, but it always seemed that way to me. That move always seems more laggy than it really is.

Posted: **Wed Apr 11, 2007 4:55 pm**

well there definitely are moves like that in 3s, the best one I can think of off the top of my head that is easy to see is when Q parries. If you parry something with Q, he will go into this long animation where he parries and then wipes off the spot on his coat where you hit him. The wiping off the coat animation is all completely cancelable at any point. There are more examples I am sure just thats the easiest one I know of.

Posted: **Thu Jun 07, 2007 7:46 pm**

9) Does anybody own a PS2 promo version of Street Fighter Alpha Anthology? Is there anything special or different about it?

10) What versions of MvC2 were there for Dreamcast? Which version is best for doing (Guile) combos? Is there anything special combo-wise about the PS2 version?

Posted: **Sat Jun 09, 2007 12:56 am**

Maj wrote:9) Does anybody own a PS2 promo version of Street Fighter Alpha Anthology? Is there anything special or different about it?

10) What versions of MvC2 were there for Dreamcast? Which version is best for doing (Guile) combos? Is there anything special combo-wise about the PS2 version?

10) - PS2 Version sucks. <~ Period. Best version is for DC (JP or SUPER RARE USA version)

My suggestion is Japanese version, that is the one that has the Guile blade lock glitch. However you'll most likely need a GameShark for the DC. The USA version that has the glitch is ridiculously rare so best hope is Japanese version. The GS codes (provided by Joo) are Here
Just use a translator (NKI?) program, it's easy. So get the Japanese version and gameshark from someone else.

Posted: **Sat Jun 09, 2007 4:19 am**

On a related note, I have a copy of DC MvC2 that I am looking to get rid of...

Posted: **Sat Jun 09, 2007 9:26 am**

Magnetro: Thanx for the info. Actually i just subscribed to gamefly for the sole purpose of renting PS2 MvC2 and they didn't have it in stock, so they just skipped onto the next game in the queue.

Robyrt: Which version do you have? Is it in good condition? Is anything missing? How much do you want for it?

Speaking of buying used games, where do you guys get your replacement jewel cases? I looked around and most places only have clear slim cases or they have standard size "clear" cases but they are blurry plastic, not totally clear. This is way more of a hassle than it needs to be.

Posted: **Sat Jun 09, 2007 9:32 am**

Wow that sucks, cause in the PS2 version of MvC2 you can't do this

http://www.zachd.com/magnetro/jmvc2guile.avi

Posted: **Sat Jun 09, 2007 1:11 pm**

My copy is in good condition but it's missing everything but the CD, because I'm lazy and I don't use jewel cases anymore. I have no idea how much these things sell for, but I would rather see it go to a good home

I had the PS2 version, but sold that too after I realized that nobody in my area (metro DC) is actually interested in Marvel, and that PS2 DC is full of terribleness but still sells for $40 at Gamestop.

Posted: **Sat Jun 09, 2007 7:58 pm**

Oh ok. Actually i want a complete one, with booklet and insert and everything. It's got the dopest Guile cover picture ever, plus Psylocke is hanging out too. But i'm sure you could sell it on SRK. There's a forum for trading/selling.

ComboVid.com

Random Lazy Questions