Also, EX Senpusha is indeed airborne from frame 10 (the 1st active frame). I do not research airborne info in most cases I’m afraid. That data was a leftover from the Brady guide. Thank you for info.
I tested it out awhile ago. I never brought it up until now…not sure why I didn’t…Anyway I can help out…I can…or I will. Also (again :D) If dud’s counter whiffs he is apparently fully invincible till he can move again (meaning he can block in time for whatever thing tries to punish the whiff). I tested that out yesterday.
Yea, if he successfully counters an attack and whiffs he is invincible all the way. I did not add the info in because I think it’s normal for the action to be invincible.
Lei’s cr.HP taking 150% damage during start-up isn’t (so it scales up to 187.5% since it is 150% x 125% for counter-hit. If he gets countered by Ryu’s HP SRK, he takes 300 damage!), so I added that info in!
Some info: ground bounce is a special state in which you retain your juggle count and goes into air knockdown state after getting hit, but it is considered standing. So if you have a move that whiffs on airborne opponent and hits grounded opponent, on ground bounce it still hits, but if you have a move that whiffs on standing and hits airborne, it whiffs. There actually is a separate table for hitting the opponent after ground bounce.
That makes true. One of the few reasons to combo into suicide…but I’m still unsafe afterwards…tis a shame…
This is not really a frame data stuff, but I have a hypothesis about Super scaling.
One thing I’m sure right now is that for each hit of Super, the damage before scaling slowly increases until it reaches 200% of the original damage.
Lei’s Super does 150, 50, 50, 100, and this eventually gets to 300, 100, 100, 200 before damage scale (of course, this is by the time scaling hits 10%, so I wouldn’t recommend it anyways)
The first hit’s damage does not start increasing until 80% damage scale, while the later hits would start scaling from the second 100% damage scale.
I’m almost certain that to achieve current damage, you multiply the unscaled Super damage by a certain percentage, round that to the nearest integer, apply the scaling, then again round it to the nearest integer. I’ve seen lots of cases with 69%, 59% scaling, etc., but you have to apply the regular 70%, 60% scaling to make the below values work:
details=Spoiler 1st:none
(100%) 2nd:0.67%
(80%) 3rd:4%
(70%) 4th:8.67%
(60%) 5th:16%
(50%) 6th:25.33%
(40%) 7th:36.33%
(30%) 8th:48.67%
(20%) 9th:64%
(10%) 10th:81.67%
(2nd 10%) 11th:100%[/details]
**Note: I’m not 100% sure on the above values. The values only apply for 2nd~ hits, below is what I assume for the 1st hit of SA:
First hit
details=Spoiler 1st:none
(2nd 100%) 2nd:none
(80%) 3rd:0.67%
(70%) 4th:4%
(60%) 5th:8.67%
(50%) 6th:16%
(40%) 7th:25.33%
(30%) 8th:36.33%
(20%) 9th:48.67%
(10%) 10th:64%
(2nd 10%) 11th:81.67%
(3rd 10%) 12th:100%[/details]
Using the above values, I can make a calculation:
[details=Spoiler]Hugo’s last hit does 150 damage unscaled. Let’s assume he hit the Super at the 4th hit therefore having 70% scaling:
- You add 8.67% to 150, so it will be 150*1.0867 = 163.005
- You round that to the nearest integer, so it becomes 163
- You apply damage scaling, so 163*0.7 = 114.1 (Like I said above, you simply apply 70% even if the training mode screen says it’s 69%)
- You round that to the nearest integer, so it becomes 114[/details]
I assume that everything (attacks, dashes, jumps etc) gets rounded down always?
When it comes to damage scaling, it usually rounds to the nearest integer, so 7.50~7.99 will get rounded up to 8, while 8.01~8.49 will get rounded down to 8.
Not entirely sure what you mean about dashes and jumps though. They’re just integers.
I say this because things like speed gems alter movement speed. So someone with a 19f dash gets like 17.1. That’s with a 10% speed gem on. So Would those round down as well?
Also…I’m pretty sure Juri’s ex senpusha I-frames didn’t change…it should still be 14 not 13.
Like how if Chun slaps on a 15% speed gem. Her dash goes to 12.75…so would you round up? Or down?
Oh I see.
It is impossible to simply calculate the speed gem based on the frames shown in the doc, cuz afaik it affects the multipliers within the game files that results in the frame data.
Also, I believe Juri is invinc from 1-13, because during my test, the 2nd hit traded.
yes because there is a 1f gap between the 1st and 2nd hits. the active frames go 3(1)4. It starts up on frame 10 so that 10 I-frames right there. she’s completely invinvible for the 1st 4 active frame (it ends right on the 1f gap) so the 5th active frame is now only projectile invincible.
Well wouldn’t it make sense to just use percentages when using the dashes and whatnot? like a 20 frame dash with an 20% speed gem would shave off 20% off 100 (the 100 being 20) so that gives you 18. I think of it as dmg scaling more or less but with speed gems. You’re scaling back the dash frames but in a positive way.
10 - 1st active frame
11 - 2nd active frame
12 - 3rd active frame
13 - the 1f gap <- she is invincible here
14 - 1st active frame of the 2nd hit <- she’s not invincible here
That’s why I assumed 1-13 invincible, since she is not invincible starting at the 14th frame.
Also, it doesn’t work that way I’m afraid. For example, Chun-Li’s forward dash with Pandora speed (20% boost), is actually 13f, not 12f (15*0.8=12, but it doesn’t work), and that’s because of how the frames are calculated, and how the speed gem is applied to that calculation. In order to get the exact frames you have to read the BAC files and calculate it yourself with speed gem applied in the multiplier value.
Well yeah that’s what I meant. during that 1f gap she’s still invincible. I count that as part of the active frames. Since she’s still invincible during the 1st hit which is 3 active frames and the gap. But after the gap is over and the start of the 2nd hit starts up then she is no longer invincible.
I see…so how do I calculate with speed gems then since I can’t read the BAC files myself I have no way of knowing. That’s what I was going by this whole time.
There is no other way to calculate the exact number other than reading the files (if I’m correct).
If you simply divide the frames by speed gems, then you will get some correct, while on some you’ll probably be like 1~3f off, which is not too bad imho.
The problem is that dash frames in the game aren’t calculated directly by frames, they’re calculated by animation speed. For example, here’s a look at the ono editing tool for Juri’s Back Dash in SF4:
Sorry if that may be a little hard to see, but essentially it has different speed multipliers that are applied to Juri’s backdash animation. Starting on position 10, the animation is changed to go at 1.75x speed. The speed multiplier is applied until another multiplier is applied in the table, which happens here on position 25, putting it back to 1x speed until the animation ends, thus ending the move (the Header section lists the move as having IASA frames on position 32, the backdash ends on position 35 otherwise).
So if I’ve got it right from what Street11 is saying, the speed gem increases are applied to the speed multipliers here, rather than being calculated for the total frames the move will last. So a 20% speed gem increase would presumably increase the default speed multiplier to 1.2x (as far as I know, a move script defaults the animation speed to 1x unless the table says otherwise), the 2nd multiplier to 2.1x, and the third multiplier to 1.2x again. These multipliers are then applied to all aspects of the script and the total frames will be based on when the animation ends after the multipliers are taken into account. I have no idea how rounding works here.
So let’s at least try some math here…Juri’s SF4 backdash has IASA frames on position 32, and it has 35 “frames” of animation if there is no speed multiplier. The 1.75x speed multiplier kicks in on position 10, so that’s 11 frames of animation (the count starts from position 0), and the 1.75x speed multiplier lasts until position 25, so the animation is sped up by 1.75x for 15 positions. So that’s 8.5 frames of animation rather than 15, and then the 1x multiplier kicks in on position 25 until the end of the animation. And the IASA frames are on position 32, and the speed multiplier is at 1x until that point, so that’s 7 frames. So…
11 + 8.5 + 7 = 26.5
I assume that the frames here are rounded up to 27 in-game, which matches what the SF4 frame data says in regards to the total frames on her backdash.
Now let’s take this for if she had a 20% speed gem:
11 positions until the 2nd multiplier, gets put down to 9.17 frames of animation (multiplier increased to 1.2x rather than 1x).
15 positions until the 3rd multiplier, gets put down to 7.14 frames of animation (multiplier increased to 2.1x rather than 1.75x).
7 positions until the IASA frame, gets put down to 5.83 frames of animation (multiplier increased to 1.2x rather than 1x).
So…
9.17 + 7.14 + 5.83 = 22.17
And also worth noting when I modded the backdash to include the above multipliers, it translated to 23 frames for her backdash down from 27.
If you just went with 27 frames and calculated from there you’d get 21.6 frames of animation.
So as Street11 said, calculating the dash frames from a speed gem by just using the raw frames won’t work all of the time due to this. The only way to know for sure is to either calculate from the game files or to record the dash and do a frame count (which is how I got the 23 frames when doing that quick mod SF4).
Street11 can probably clarify/correct most of this if I got anything wrong. And of course this is assuming that SF X TK uses similar scripts/tables as SF4 does (which I would assume is the case since they use the same engine).
Also street. Since Juri’s Sekku is LC…does that mean it’s only LC on hit?
TY for the thorough and clear explanation LordWilliams.
I changed Heihachi’s counter frames and added in the list of link-cancelable moves. Now I can say with confidence that the frame data is complete. Except BCs
It should be LCable period (both on hit and on block). I am still unable to explain the result behind Ken’s SA, but it should be -1 in other cases.
Oh is it because in your notes it says “just not against SA’s/CA’s” though I don’t think that’s true since Steve’s super can’t punish Juri’s Fuhajin store despite being -4 on block.
I’ll take a look into it. It might just be Ken’s SA. I’ll delete the SA note for now. TY for the info.
Anytime.