Official Ask Anyone Frame Things Thread

[Kadano]

What are all of the factors that come into play for moonwalking?

I’m still way too bad to find offsets and read out memory adresses that point to the data being read during the initial dash, so I can’t give useful information.
All I know is that the farthest moonwalks are done by dashing forward for only one frame and then keeping pressed backwards until the second last frame of the dash. On the very last frame of the dash, dashing forward again carries over the most momentum to the next moonwalk iteration. Also, the moonwalk distance is determined not only by the angle, but also – and moreso – simply by the x value of your control stick. A moonwalk with [220,128] on all backwards inputs produced slightly less distance than one with [245, 128]. Thus, it’s possible to gain a minimal increase in moonwalk distance by 1. making sure your stickbox has a small dead zone and 2. sanding out the octagonal shape on the GCC shell. (For those who actually feel like applying this: I also recommend adding additional bumps at the (mod45±17)° angles for easier perfect wavedashes, spacies recoveries etc. Also don’t take away too much or moving around the stick will make him slide off of the stickbox post, which is a really off-putting sensation)

This is pretty funky stuff, so maybe it’s best to view this post as nothing more than a bump for Strong Bad, Magus or someone else elaborating on why moonwalking works like it does.

 

[Bones0]

Nice bump, Kadano. You should have waited 2 more days so it would have been a full year. lol

Also, I'm not sure we should condone modding controller faces in that manner tbh. I could definitely see a TO taking the position of not allowing that, and frankly, I wouldn't blame them. Adding stuff that makes certain inputs easier is comparable to turbo functions considering how important precision is in Melee.

 

[DRGN]

Interesting. Thanks for the note. What are you using for unit of distance?

 

[Kadano]

I simply made savestates with Dolphin at the same frames and compared them. At 480p, the difference was about 2px. I know that internal units are much more accurate, but I don't know how to access them.

@Bones: I see your point, but I don't think this is an unfair improvement. It's certainly arguable.

 

[DRGN]

Do you argue whether it offers gameplay improvement?

Even if it's only very slight, it's fundamentally a bit unfair for mods in competitive play. Even if the option is open to everyone, there are plenty of cases where people will not have the mod. Some people aren't comfortable with opening up their controller; some may use rare controllers or a controller that they really like that they don't want to tamper with or damage. Or some people might be having to borrow a controller to play. They may even just simply not have knowledge of the mods. If they give gameplay advantage, their use would be imposing on the non-users, at best.

However, having said that, in small tournament settings where everyone knows about it but doesn't care, or if the mods are just being used outside of tournament for fun, I think that's totally fine. I definitely agree about their exclusion in larger events though.

Personally I've thought about making a controller just for fun that has like perfect wavedash buttons (maybe waveshine, that follows the angle you hold the stick at) or other cool stuff. Maybe a button that super-wavedashes with Samus, every time. lol. Could be done with a programmable chip, since we can get the input data signals from controllers.

 

[TKD]

Can you provide a complete breakdown of the frame data for Fox's wavedash?
What's the IASA for landing from a wavedash? How long before a character starts moving after inputing an airdodge?

Also, For Fox's JC shine? (I'm guessing it's 9~10 frames total including 3 frames jump start-up, 1 frame shine start-up and 6 frames before Fox can jump cancel)

I want to analyze the data to maybe develop a new broken-*** Fox tech.

 

[DrinkingFood]

Can you provide a complete breakdown of the frame data for Fox's wavedash?
What's the IASA for landing from a wavedash? How long before a character starts moving after inputing an airdodge?

Also, For Fox's JC shine? (I'm guessing it's 9~10 frames total including 3 frames jump start-up, 1 frame shine start-up and 6 frames before Fox can jump cancel)

I want to analyze the data to maybe develop a new broken-*** Fox tech.

wavedash is 3 frames jumpstart, airborne frame 4 where you input airdodge, hits ground frame, 5 can act as early as frame 15
jc shine hits frame 1, jump can start at frame 4, he's airborne frame 7
Put them together, shine comes out frame 1, jump can start as early as frame 4, he's airborne on frame 7 where you input airdodge, frame 8 you hit the ground, frame 18 is first free frame to input w/e

 

[standardtoaster]

You actually move on the same frame you're airborne if you air dodge. This would make it 13 frames for perfect wavedash and 16 frames for perfect shine wd

 

[Bones0]

Yeah, if you are WDing properly you don't spend a single frame airborne.

 

[TKD]

Oh, it's just that, you can actually waveshine Marth with Fox. It goes like this:

shine > wavedash > dash > jc shine

JC shine hits on frame 4, btw; not 1. 1-frame jc shine would be broken and probably a lot easier to input.

So you hit marth with a shine. He has 35 hitstun frames.
Fox spends 17 frames on waveshining (lag before being able to jump (3) + jump startup (3) + wavedash (11)). Marth has 18 hitstun frames remaining.
If this is correct, and the Fox does it perfectly, after taking away the 4 frames it takes to JC shine, Fox has 14 frames available to dash (maybe around 10 with human error).

I've been able to waveshine Marth at all times except when he DIs away. So away DI probably escapes it 100% (not enough time to reach it; I don't know if this changes if executed frame-perfectly). But you can react to the long slide and just reach with a JC grab. You can also reach Marth offstage with another shine if you send him off and he misses the fastfall to grab the ledge (probably resulting in a KO since another shine is pretty easy to land after grabbing the ledge, or a pretty easy follow-up).

Waveshine to dash to JC shine is an extremely technical feat, so this is a great technique to practice techskill with (you just leave a controller assigned to the dummy Marth so he's not controlled by the CPU).

It's also possible to pull this off during actual matches at times when you're confident or warmed up enough that you're pretty sure you can pull it off.

 

[Bones0]

Oh, it's just that, you can actually waveshine Marth with Fox. It goes like this:

shine > wavedash > dash > jc shine

JC shine hits on frame 4, btw; not 1. 1-frame jc shine would be broken and probably a lot easier to input.

So you hit marth with a shine. He has 35 hitstun frames.
Fox spends 17 frames on waveshining (lag before being able to jump (3) + jump startup (3) + wavedash (11)). Marth has 18 hitstun frames remaining.
If this is correct, and the Fox does it perfectly, after taking away the 4 frames it takes to JC shine, Fox has 14 frames available to dash (maybe around 10 with human error).

I've been able to waveshine Marth at all times except when he DIs away. So away DI probably escapes it 100% (not enough time to reach it; I don't know if this changes if executed frame-perfectly). But you can react to the long slide and just reach with a JC grab. You can also reach Marth offstage with another shine if you send him off and he misses the fastfall to grab the ledge (probably resulting in a KO since another shine is pretty easy to land after grabbing the ledge, or a pretty easy follow-up).

Waveshine to dash to JC shine is an extremely technical feat, so this is a great technique to practice techskill with (you just leave a controller assigned to the dummy Marth so he's not controlled by the CPU).

It's also possible to pull this off during actual matches at times when you're confident or warmed up enough that you're pretty sure you can pull it off.

Your frames are off by a bit. You only have 2 frames of lag after shine hits before you can jump. Shine lasts 3 frames with the first one hitting, so after both characters are out of hitlag, you only have 2 more frames of shine before you can begin jumpsquat. WD also only has 10 frames of landing lag, not 11.

Even disregarding the frame data, it's certainly possible to waveshine Marth. I've done it myself and been able to do things like waveshining actual players all the way across the stage. This really isn't surprising when you consider that it's possible to drillshine him too, and it takes more frames to get drill out after dashing and jumping than it does to shine.

 

[Kadano]

I always thought waveshine with walks instead of dashes was guaranteed in NTSC on Marth. While I would like to just throw this in and see what the consensus is, it would probably result in two parties stating contrary claims, so I guess I better cut this and go straight to TAS tests. Sigh …

Angles assume Fox is waveshining Marth from right to left. Unfavorable factor, favorable factor, bad outcome, good outcome.

Shine 1
1.1 • No SDI by Marth, 197° wavedash, walkfast afterwards: 8 frame window for shine 2.
1.2 • No SDI by Marth, 197° wavedash, dash→run afterwards: 5 frame window for shine 2. Frame 5 of this window will be a reverse hit.
1.3 • No SDI by Marth, 225° wavedash, dash→run afterwards: 5 frame window for shine 2.
1.4 • 1SDI+ASDI with 180°, 197° wavedash, dash→run afterwards: 5 frame window for shine 2.
1.5 • 1SDI+ASDI with 180°, 225° wavedash, dash→run afterwards : 2 frame window for shine 2.
1.6 • 1SDI+ASDI with 180°, 197° wavedash, walkfast afterwards: shine 2 whiffs.
1.7 • No SDI by Marth, 225° wavedash, walkfast afterwards: shine 2 whiffs.
1.8 • 1SDI+ASDI with 180°, 225° wavedash, walkfast afterwards: shine 2 whiffs.

At this point, we can already establish that perfect wavedash to walkfast is not reliable because Marth can SDI out. Wavedash to dash→run, on the other hand, is perfectly reliable. Even if the wavedash is as imperfect as 225°, it will still connect on a SDIing Marth.
You might wonder about the risk of overshooting against a Marth who SDIs in as a sort of counter against your wavedashrunshine. I didn’t include this option in the list because it hardly changes anything – even if you do a 197° wavedash and then dash, you have 3 frames to hit him forward and 2 frames to hit him backward with your shine 2, so all his SDI in will do is add another frame at the end that will hit reversed compared to 1.3.
The hard part about the wavedashrunshine is that you need to have a precise timing on the dash. If you dash 5 frames later than possible, your run will start to late and Marth can shield.

Shine 2
2.1 • No SDI by Marth, 197° wavedash, walkfast afterwards: 8 frame window for shine 2.
2.2 • No SDI by Marth, 197° wavedash, dash→run afterwards: 5 frame window for shine 2. All frames of this window will be reverse hits if shine 1 hit out of run. If shine 1 hit out of walk, the first 3 frames will not hit reverse.
2.3 • No SDI by Marth, 225° wavedash, dash→run afterwards: 5 frame window for shine 2.
2.4 • 1SDI+ASDI with 180°, 197° wavedash, dash→run afterwards: 5 frame window for shine 2.
2.5 • 1SDI+ASDI with 180°, 225° wavedash, dash→run afterwards : 5 frame window for shine 2.
2.6 • 1SDI+ASDI with 180°, 197° wavedash, walkfast afterwards: ~5 frame window for shine 2 depending on the exact precurring action. If shine 1 was hit from a walk, it will whiff.
2.7 • No SDI by Marth, 225° wavedash, walkfast afterwards: ~5 frame window for shine 2 depending on the exact precurring action. If shine 1 was hit from a walk, it will whiff.
2.8 • 1SDI+ASDI with 180°, 225° wavedash, walkfast afterwards: ~1 frame window for shine 2 depending on the exact precurring action. If shine 1 was hit from a walk, it will whiff.

Shine 2 (after frame 5 of 1.2)
Because reverse hits make you move left while Marth is moving to the right, only dash→run will reach him in time. Walking out of the (backwards) wavedash is not possible because it requires the entire turn animation (11 frames) to complete before it can start. Turning around earlier during shine takes way too long.

Waveshining Marth indefinitely
1. Start the waveshine with 1.2.
2. Continue with 2.3 to ensure you don’t hit reverse.
3. Repeat 2.3 until your next waveshine is the last that ends before the edge.
4. Do 1.2 again and hit the shine quite late during your shine to force a reverse hit.
5. Do a backwards wavedashrunshine and repeat steps 2-4 to the other side.
6. Repeat forever and finish with an upsmash.

This is inescapable. Marth can’t even change hit orientation by SDIing away from you when you need to hit reverse.
Edit: Because I tested this with debug mode’s frame advance, damage staling might make a difference.
Edit2: It seems move staling only changes the amount of hitlag and the damage of Fox’s shine. Hitstun and knockback are unaffected.
A completely stale shine has 1 frame advantage from electric hitlag opposed to 2 when fresh. This should reduce the frame window for wavedashrunshine from 5 to 4 – nothing dramatic.

 

[Strong Badam]

Since I'm significantly less interested in Melee (I only ever research it for PM purposes), I'm officially changing the title to be non-specific. I'm still subscribed so I might pop in here or there, but otherwise Kadano should be able to handle most requests. Great work, Kadano! If there's something that's stumping you feel free to shoot me a PM and I'll see if I can tackle it.

 

[HyugaRicdeau]

How long is ledge invincibility and when is the first actionable frame after grabbing the ledge? Is this the same for all characters?
Does the game require you to move from neutral to down in order to ledge drop or can you hold down from the non-actionable ledge-grab frames?

 

[Superspright]

Can Ganon invincibly ledge-stall by jumping backwards and regrabbing if done fast enough? And also, how much leeway does he have for an invincible ledge-stall when he simply drops and jumps back up?

 

[Kadano]

How long is ledge invincibility and when is the first actionable frame after grabbing the ledge? Is this the same for all characters?
Does the game require you to move from neutral to down in order to ledge drop or can you hold down from the non-actionable ledge-grab frames?

33 frames for Link, 37 for everyone else. 5 for Link, 9 for everyone else (inputs read 1 frame earlier).
No, it can’t be buffered.
More information: http://www.ssbwiki.com/Ledgestall_(SSBM)

Can Ganon invincibly ledge-stall by jumping backwards and regrabbing if done fast enough? And also, how much leeway does he have for an invincible ledge-stall when he simply drops and jumps back up?

Yes, but depending on how far he held the control stick to cause a backwards jump, he might need to hold towards the ledge afterwards so that he doesn’t miss it.
Zero leeway on the drop, 6 frames on the jump. ~5 on the fastfall, depending on the jump timing.

 

[DRGN]

I'm liking the wiki page. Nice work, Kadano!

 

[Sashimi]

On what frame does Yoshi's grounded Down+B leave the ground?

How many frames of invincibility does a player get when they break out of Egg Lay?

 

[Kadano]

On what frame does Yoshi's grounded Down+B leave the ground?

On frame 8.

How many frames of invincibility does a player get when they break out of Egg Lay?

14.

 

[dRevan64]

Man, this thread doesn't get any posts anymore and that's a damn shame.
2 questions:
How fast do you have to mash b to gain height with mario, doc and luigi's down b?
What is the timing necessary for a vududash? I know it's a perfect waveland but it feels like at some spacings I won't get the acceleration.

 

[Varist]

What is the frame window for Marth to escape a perfectly timed Falco pillar combo starting at 0% (if there is one)? What moves can he use to escape with? Is it only fair or can he fit counter in there too? At what percent can he finally escape?

 

[Bones0]

What is the frame window for Marth to escape a perfectly timed Falco pillar combo starting at 0% (if there is one)? What moves can he use to escape with? Is it only fair or can he fit counter in there too? At what percent can he finally escape?

Counter comes out just as fast as fair, as does up-B. You can check the frames for all of the moves in the hitbox thread.

The % he can escape at is going to depend on where you are on the stage, how much you SDI, and whether or not he guesses your DI correctly.

 

[TKD]

What's the start-up for blocking? Don't just assume it's active on frame 1; I want confirmed information from anyone that's tested it or read a thread about it. Just to make sure.

 

[Ripple]

your shield comes out frame 1.

it doesn't get to its maximum size until like frame 5 however, so you could be poked during start up.

reflecting a projectile happens frame 1-2 of shield. but the window for this is based on your remaining shield. less shield = harder to reflect.

powershielding a hit is 1-4 of shield

 

[Kadano]

More in-depth information: www.ssbwiki.com/Perfect_shield#Chronological_frame_data

 

[Varist]

Does anyone have a picture of some kind for moonwalking that shows the exact movements for the longest, best possible moonwalk and # of frames allowed for each movement?

I'm thinking just a big circle with the deadzone circle in the center and then just red lines for the stick movement, including the motion to link two moonwalks together

 

[WHA?]

I dont know if this applies

but how long do sheik's needles last when you stick them on the ground??? (like if i wanna set up a trap for pillaring falcos)

And do they disappear when i charge up more?

 

[Ripple]

I dont know if this applies

but how long do sheik's needles last when you stick them on the ground??? (like if i wanna set up a trap for pillaring falcos)

And do they disappear when i charge up more?

2 full seconds. and no

 

[Kadano]

What is the frame window for Marth to escape a perfectly timed Falco pillar combo starting at 0% (if there is one)? What moves can he use to escape with? Is it only fair or can he fit counter in there too? At what percent can he finally escape?

I’d really like to answer these questions, but it’s not that simple as there are so many assumptions about DI, Falco’s move and movement choices and much more to be made. If you give me a video link where a Falco did a pillar combo that seemed near-inescapable to you, I can look into it.

Does anyone have a picture of some kind for moonwalking that shows the exact movements for the longest, best possible moonwalk and # of frames allowed for each movement?

I'm thinking just a big circle with the deadzone circle in the center and then just red lines for the stick movement, including the motion to link two moonwalks together

1. Press forward (0°) for 1 frame.
2. Hold backwards (180°) for [dash duration − 1] frames.
Repeat steps 1 and 2 indefinitely.

This only applies for fresh moonwalks, ie not done out of dashdance / turn. For those, you need to do go below the dead zone like you described it, although it’s technically a dead square and not a circle. Anyway, so far the best motion I’ve discovered for turn moonwalks is:
1. Press forward (0°) for 1 frame.
2. Hold back- and downwards (225°) for 2 frames.
3. Hold backwards (180°) for [dash duration − 3] frames.

 

[Bones0]

You can moonwalk with just left and right cardinal directions? O_o

 

[TKD]

How much hitstun does crouch canceling a move give you? I'm kinda tired of losing strictly to Fox's nair.

And this is off-topic, but can you SDI a move you crouch canceled?

 

[Metal Reeper]

This is the best thread I've ever seen. Is there a difference between Fox's dash>shine. And Fox's dash Canceled shine? I'd assume dash canceling would get the shine out faster. And maybe less distance? If you could somehow gif both of those <33333333333333333

 

[Kadano]

Is there a difference between Fox's dash>shine. And Fox's dash Canceled shine? I'd assume dash canceling would get the shine out faster. And maybe less distance?

The term “dash-canceling” is obsolete and kind of incorrect so it’s been replaced with “run-canceling”. Anyway, you can answer your question with the data provided in this post.
The table there lists a initial dash frame window of 11 for Fox. Thus, your run will start on frame 12. Down-B is listed as an option available straight out of Run, so you can shine on frame 13.
In that post and in the one I wrote two entries below, run-canceling is explained as well. So if you were to run-cancel perfectly, you would spend frame 13 in RunBrake (triggered by holding down) and frame 14 in Squat (crouch, triggered by holding down during RunBrake). Your shine would start on frame 15.

The run-cancel shine would cover less distance, but by simply running for two additional frames before interrupting Run with shine, you cover even more distance.

 

[SAUS]

I noticed that it is possible to put yourself in tumble by wavelanding/wavedashing off of a platform or the edge of the stage and shielding before you fall (I only did it with marth and captain falcon). I think it would be completely pointless to do, but is it possible to walltech this off the edge to do derpy wall jump edge gaurds with marth? I'm wondering if you can actually jump out farther with this (if it is even possible).

 

[Bones0]

I noticed that it is possible to put yourself in tumble by wavelanding/wavedashing off of a platform or the edge of the stage and shielding before you fall (I only did it with marth and captain falcon). I think it would be completely pointless to do, but is it possible to walltech this off the edge to do derpy wall jump edge gaurds with marth? I'm wondering if you can actually jump out farther with this (if it is even possible).

I don't think you can walltech or ceiling out of tumble, only hitstun, but you can test it on MKII.

 

[Kadano]

I noticed that it is possible to put yourself in tumble by wavelanding/wavedashing off of a platform or the edge of the stage and shielding before you fall (I only did it with marth and captain falcon). I think it would be completely pointless to do, but is it possible to walltech this off the edge to do derpy wall jump edge gaurds with marth? I'm wondering if you can actually jump out farther with this (if it is even possible).

During DamageFall – the animation you enter when you slide off an edge while shielding – hard collision is only possible against floors, not walls. (Is there a better term for “hard collision”, by the way? I know there’s “knockdown“ for floors, but this term doesn’t really make sense for walls and ceilings.)
Without a hard collision, you cannot tech, and without a tech you cannot walljumptech. Therefor, what you thought of isn’t possible.

 

[Sashimi]

Is there a way to find out how fast a character is moving on a particular frame (specifically during a jump)? I found that little red bar in the debug menu that tells you whether you're moving up or down, but it only gives a general idea. Is there some way to get the exact number?

 

[Bones0]

Is there a way to find out how fast a character is moving on a particular frame (specifically during a jump)? I found that little red bar in the debug menu that tells you whether you're moving up or down, but it only gives a general idea. Is there some way to get the exact number?

Speed is change in distance over time. You can't measure speed on any one particular frame. This is Physics 101. lol

If you want to measure jump speed, you can simply look at how far your character travels from one frame to the next. You can probably get the exact number through some more intense methods, but I wouldn't know how.

 

[Ripple]

get dolphin. run melee. look at the code being read. find x, y, and z positions of your character. advance 1 frame.

done

 

[Kadano]

get dolphin. run melee. look at the code being read. find x, y, and z positions of your character.

Someone did a tutorial on how to read Melee code in Dolphin, but I can’t find it anymore. Could you please give me the link to it if you have it? It would be so great to be able to do that for me.