Official Ask Anyone Frame Things Thread

[dRevan64]

Oh, I get what you mean. I don't think you can ever followup really if they full DI behind you, but maybe vs. spacies. Again though, it'll depend a lot on the character and DI so there's no simple answer. I'd just recommend testing in Training Mode on no DI vs. various characters to get a feel for around when it will work.

I've been screwing around in 20xx using it and I think there are some percents where it's guaranteed regardless of DI but it's an awkward thing to test because I think I kinda suck at the input, which is my own problem. It does feel like a severely underused tool though. As I test it I realize how good it could be after a shine on spacies at highish percents where you can't jab reset or they tech it the wrong way.
Also it looks cool.

 

[Zappdos]

if performed frame- perfect, how many frames does it take for jiggs to ledge drop, hop, and 0 charge rollout to ledge grab?

how many frames does it take for jiggs to turn around when jumping? ik if you attack while jumping+ switching directions the turn is interrupted and your still facing the same way after the aerial.

 

[Kadano]

if performed frame- perfect, how many frames does it take for jiggs to ledge drop, hop, and 0 charge rollout to ledge grab?

Too many to be of use. The lowest I got was 14 vulnerable frames, which if added to the 37 intangible frames sums up at 51 per iteration.
For simple rejump-ledgestalls, the lowest I got was 8 vulnerable frames, thus 45 per iteration. Plus you don’t die this way if someone edgehogs you.

how many frames does it take for jiggs to turn around when jumping? ik if you attack while jumping+ switching directions the turn is interrupted and your still facing the same way after the aerial.

You are turned around starting on frame 5 of your rejump, which is just the same amount as with tilt turns for Jigglypuff.

 

[Zappdos]

So for jiggs to have an invincible ledge stall, youd have to find a way to ledge grab 8 frames faster than it takes to naturally grab the ledge, and if rollout doesn't work, im sure sing wont either.

So hhaxdashing is probably the only way to accomplish that.

 

[Kadano]

So for jiggs to have an invincible ledge stall, youd have to find a way to ledge grab 8 frames faster than it takes to naturally grab the ledge, and if rollout doesn't work, im sure sing wont either.

So hhaxdashing is probably the only way to accomplish that.

Yes, I’m pretty sure that ledgedashstall is the only way for Jigglypuff to refresh ledge intangibility without any vulnerability in-between.

 

[Zappdos]

Cool. I sistill want to try a few things (if hugging FDs cliff, if reverse sing can ledge grab faster than normal, but im not too optimistic. I may ask yiu to check something out if it looks fast enough though If you dont mind. :3

 

[schmooblidon]

Yoshis is your best bet, due to the slant, it makes the ledgesnap area higher up then the actual ledge. You can get 6 vulnerable frames from a DJ stall on Yoshis.

 

[Zappdos]

Noted thanks.


Im really surprised the rollout thing is slower. It aesthetically looks faster. ha.

 

[Kyu Puff]

What effect does analog L/R have on powershielding? If a lightshield comes out for 1 frame before the fullshield, can you still get a powershield?

I'm trying to figure out why this doesn't work when analog shield is disabled.

 

[Zappdos]

Ledge drop reverse sing may? (Damn close either way) have a faster ledgesnap than simple jumpstall on yoshis, but its still not enough. Any other ideas that arent rollout/sing cancelling?

 

[schmooblidon]

What effect does analog L/R have on powershielding? If a lightshield comes out for 1 frame before the fullshield, can you still get a powershield?

I'm trying to figure out why this doesn't work when analog shield is disabled.

If you lightshield, you must digital press on the second frame to still be able to powershield. This is called ADT-PS (analog to digital). You g et a Z-PS sized powershield but it does not PS physical attacks on the first 2 frames. This method sucks because you have a big non powershield lightshield on frame 1 which will eat the projectile. It also breaks yoshis shield/parry.

The Z-PS works when you hit Z and a Digital Press on the same frame with A buffered. This only works with Z because the game doesn't take priority of one input, but instead certain aspects of each. It gets the big shield from the Z and the PS from the digital press.

Ledge drop reverse sing may? (Damn close either way) have a faster ledgesnap than simple jumpstall on yoshis, but its still not enough. Any other ideas that arent rollout/sing cancelling?

I got it to be the same because it still requires jiggs to reach the apex of the doublejump, and not anything to do with sing.

 

[Bones0]

During the ground bounces from missed techs, spiking the opponent usually sends them back into the ground, but there is a small window where they are considered grounded and get knocked upwards. Can anyone find out the window/timing for this effect and whether it's affected by landing facing up/down or by which character it is?

Someone plz...

 

[Kyu Puff]

If you lightshield, you must digital press on the second frame to still be able to powershield. This is called ADT-PS (analog to digital). You g et a Z-PS sized powershield but it does not PS physical attacks on the first 2 frames. This method sucks because you have a big non powershield lightshield on frame 1 which will eat the projectile. It also breaks yoshis shield/parry.

The Z-PS works when you hit Z and a Digital Press on the same frame with A buffered. This only works with Z because the game doesn't take priority of one input, but instead certain aspects of each. It gets the big shield from the Z and the PS from the digital press.

Just to be clear

Frame 1: Analog press
Frame 2: Digital press + Z

yields a fullsized lightshield, but doesn't reflect projectiles until frame 2, and doesn't reflect physical attacks until frames 3-4?

 

[schmooblidon]

Just to be clear

Frame 1: Analog press
Frame 2: Digital press + Z

yields a fullsized lightshield, but doesn't reflect projectiles until frame 2, and doesn't reflect physical attacks until frames 3-4?

I cant be 100% sure but you will likely just get a ADT-PS, assuming you buffer A. You might change the PS size to match a Z-shield if the analog press value is different which it likely would be. Perhaps I wasn't super clear, an ADT-PS goes like this.

1 - Light Shield
2 - Big Powershield (GuardReflect 1) cannot PS physical attacks
3 - Big Powershield (GuardReflect 2) cannot PS physical attacks
4 - Small Powershield (GuardReflect 3) cannot PS projectiles
5 - Small Powershield (GuardReflect 4) cannot PS projectiles
6 - Small Shield

 

[Kadano]

During the ground bounces from missed techs, spiking the opponent usually sends them back into the ground, but there is a small window where they are considered grounded and get knocked upwards. Can anyone find out the window/timing for this effect and whether it's affected by landing facing up/down or by which character it is?

DownBound = Non-Teched Floor (Back & Stomach)
Animation length: 26 (DownboundD and DownBoundU 0-25)
Character is airborne during frames 3-22
Character is grounded during frames 1-2 and 23-26

Just to be clear

Frame 1: Analog press
Frame 2: Digital press + Z

yields a fullsized lightshield, but doesn't reflect projectiles until frame 2, and doesn't reflect physical attacks until frames 3-4?

Yes, see the table in this post (linking for easy reference, I know you know that post already). Frame 1 lightshield size depends on the analog level you input, and the Z input PSS size can be used for ADT-PS as well. I guess I’m gonna include this as ADT-Z-PS.
I also just edited it again to include the information @ schmooblidon just provided.
(I still believe ADT is something you want to avoid altogether, except for PSing projectiles; for that it’s somewhat acceptable as effectively one frame of Z-PS digital L/R input leniency, just with the one frame unwanted non-PS lightshield.)

 

[schmooblidon]

Just to be clear

Frame 1: Analog press
Frame 2: Digital press + Z

yields a fullsized lightshield, but doesn't reflect projectiles until frame 2, and doesn't reflect physical attacks until frames 3-4?

Ok my b. I think this you might of discovered a new variation of ADT-PS, unfortunately it's still terrible.

So let me try to display this nicely.

Case 1: Hold Z for one frame, hold Digital Press for more than one frame
Case 2: Hold Z for more than one frame, hold Digital Press for more than one frame
Case 3: Hold Z for one frame, hold Digital Press for one frame
Case 4: Hold Z for more than one frame, hold Digital Press for one frame

Case | PSS-Projectile Size | PSS-Physical Size | PSA downtime
1 | Dependent on Analog Light Shield | Small | GuardReflect 1-2 (total - frame 2-3)
2 | Dependent on Analog Light Shield | Huge | GuardReflect 1-2 (total - frame 2-3)
3 | Dependent on Analog Light Shield | Dependent on Analog Light Shield | GuardReflect 1-2 (total - frame 2-3)
4 | Dependent on Analog Light Shield | Huge | GuardReflect 1-2 (total - frame 2-3)

The only potential use for this I can see is doing a small lightshield into a case 2 or 4 for powershielding physical attacks, but honestly it's really hard to perform and not really worth it.

I am using pause buffering so take this with a grain of salt, but I think this is accurate. Apologies for disregarding it and claiming you would grab, Z stuff is tedious to test.

edit: released there are more than 2 cases, tried to make the table more clear

 

[Kadano]

You must release analog press on frame 2.

My findings contradict that. In Dolphin, pressing analog L or R (“or”, not “xor”, so pressing both works too) on frame 1 and holding it through the DP L or R plus Z input does not decrease PSS size or reduce frame windows.

You can input:
1. light press L on frame 1, followed by DP (digital press) L and Z on frame 2
2. light press L on frame 1, followed by DP R and Z on frame 2
3. light press R on frame 1, followed by DP L and Z on frame 2
4. light press R on frame 1, followed by DP R and Z on frame 2
5. light press L and R on frame 1, followed by light press R and DP L and Z on frame 2
6. light press L and R on frame 1, followed by light press L and DP R and Z on frame 2

All of these produce ADT-Z-PS.

 

[schmooblidon]

My findings contradict that. In Dolphin, pressing analog L or R (“or”, not “xor”, so pressing both works too) on frame 1 and holding it through the DP L or R plus Z input does not decrease PSS size or reduce frame windows.

You can input:
1. light press L on frame 1, followed by DP (digital press) L and Z on frame 2
2. light press L on frame 1, followed by DP R and Z on frame 2
3. light press R on frame 1, followed by DP L and Z on frame 2
4. light press R on frame 1, followed by DP R and Z on frame 2
5. light press L and R on frame 1, followed by light press R and DP L and Z on frame 2
6. light press L and R on frame 1, followed by light press L and DP R and Z on frame 2

All of these produce ADT-Z-PS.

Thanks for clearing that up. It's not something I tested because my hands were getting pretty full. Having to press L lightly + R fullpress + Z + X + DpadUp then hit Start 1 frame later, then do it again but hit Start 2 frames later, was a bit too much lol.

But does it not outprioritize Z for GuardReflect 3+ size in case 2?

EditL Wait no I'm mixing it up, Z outpriortizes the triggers for size. So case 2 shouldn't be affected right?

 

[Kadano]

Thanks for clearing that up. It's not something I tested because my hands were getting pretty full. Having to press L lightly + R fullpress + Z + X + DpadUp then hit Start 1 frame later, then do it again but hit Start 2 frames later, was a bit too much lol.

But does it not outprioritize Z for GuardReflect 3+ size in case 2?

EditL Wait no I'm mixing it up, Z outpriortizes the triggers for size. So case 2 shouldn't be affected right?

I am not sure what you are asking for. In any case, all of the six cases (and there are probably a few input combinations that I left out which work the same) produce a Z size (0.35) powershield.

 

[Magus420]

Finished that develop mode mod thing I mentioned in my previous post that displays useful data and such.

http://smashboards.com/threads/physics-input-etc-display-in-develop-mode.391956/

 

[flieskiller]

What are the inputs/actions needed to (instantly?) turn after a ledge cancel after an aerial? I know I can do that a few times inconsistently (I play with G&W), but knowing it and doing it consistently would help my platform game a lot.

let's say I do b-air from the right to the left of the platform, I want to ledge cancel, turn myself very fast and do a f-air on the opponent who is at the right side of the platform.

 

[schmooblidon]

It took me until today to realise that the horizontal shield tilt shielddrop isn't actually changing the shielddrop vertical input value, it's just stopping you from rolling. Can't believe I didn't work that out before.

This new tool is the ****. Shielddrop has 3 vertical input values, there are 80 values from neutral to full down, that's crazy!

Can anyone explain why the hell the other shai drops work? It just seems so random and intricate of a design choice.

 

[tauKhan]

@ schmooblidon I think all the shai drops work like this: The shielddrop input is active for 6 frames, while spotdodge for 4 frames. Because spotdodge overrides drop, you need to time shield after spotdodge input is already gone. So to drop you need to shield on frame 4-5 after starting to press down, because then only the shieddrop is active on frame 5-6 and you drop.

Also the design choice is in line with all the other control stick actions that require "smash input": the frames "count" even when you can't do the actions, for example you can input dash on the 2nd last frame of landing lag and still dash afterwards.

 

[schmooblidon]

@ T tauKhan ah that makes a lot of sense thanks. but what about the frame perfect shielddrop? spotdodge and a standard platform drop (crouch) have identical ranges. why are there 3 values that trigger a drop only when in shield? is there even a reason?

 

[tauKhan]

Don't know, I don't personally agree at all with the design choice on the highest possible y-value for shielddropping being so close to spotdodge. I guess they just intentionally made it very hard to do, maybe they felt it would be too powerful tool otherwise. Though if the drop area started higher, it would probably be difficult to angle shield down without dropping, maybe that's one reason.

 

[Kadano]

@ T tauKhan ah that makes a lot of sense thanks. but what about the frame perfect shielddrop? spotdodge and a standard platform drop (crouch) have identical ranges. why are there 3 values that trigger a drop only when in shield? is there even a reason?

I would bet 100$ that HAL did it like that on purpose. During development, they obviously programmed shield drops into the game. When they worked on setting the input action ranges during shielding, they decided to put it below the shield angling range to still enable you to do quick downward shield angling. The shield drop area actually goes down from y=75 to y=0, but the area [0;72] is occupied for the downward smash input range on the first four frames like tauKhan explained, effectively taking priority.

Them leaving three values on the first frame was intended, I guarantee you. It’s like an easter egg they threw in for dedicated players. When you look at it from the programmer's perspective, shield drop mechanics make a lot of sense. As with 99% of things we have in this game, which is why I love it so much (unlike the sequels).

They could probably have extended the range from [73;75] to [73;80] or something. But the white shield angling area might actually be an overriding input coded at a deeper level, and I kind of like this technique being hard to execute perfectly.

In fact, I don’t think it’s possible to execute frame 1 shield drops reliably. Not because of human hand precision limitations, but because the Gamecube polls for inputs less frequently than our hands (if you can even call it that, haha). To do it consistently, you would need to have your finger inputs synced to the Gamecube’s input polling timing. (And I really don’t know how you would do that.)

(Take this with a grain of salt, I actually don’t know how exactly Melee processes control stick input and what it does if two polls during a frame have different values, so this is only “educated guessing” here.)

 

[Bones0]

We really need a Sakurai meme for all this pro-gramming.

 

[dRevan64]

We really need a Sakurai meme for all this pro-gramming.

That pun was inexcusable.

 

[King Mob]

What's the velocity of the air dodge animation?

 

[schmooblidon]

What's the velocity of the air dodge animation?

Depends which way you airdodge. And it decreases each frame.

It starts at 2.79 though.

 

[Meru]

Not exactly a frame thing, but I wanted to find out the sizes of the stages, and I couldn't quite find it in the smash wiki. I wanted to be able to use the stage size data to figure out the exact percentages that Usmashes/upward trajectory attacks kill certain characters.
On that note, I also had a question about the percentage calculations. Due to staling factors, damage values can be given as decimals, so I was wondering how melee treated those values. I did some rudimentary testing myself, and when it comes to adding the values together, melee certainly(?) doesn't just floor the values for damage before adding them (say, you have an attack that does 10%, then you use it again, so it does 9.1%. It doesn't just floor the decimals before adding it to the victim's total damage taken, according to the calculations I did with falco's forward tilt staling - you can forward tilt someone to 41%, but if it floored every value first it would be 40%). So, I was then wondering how melee added these things -> Right now I'm thinking that the percentage value it shows is just a floored value, but it adds together the exact decimals "behind the scenes" anyway. I'm asking this because it would potentially make a small difference if melee calculated hitstun/KB/Velocity for only the given whole number value of the character's damage, or if it just showed the whole number value but really used the exact decimal values for the calculations of hitstun, KB, etc. Also, hitstun values are just floored right? And are hitstun and velocity values just kept exact?
sorry for the kinda lengthy post

 

[Bones0]

Not exactly a frame thing, but I wanted to find out the sizes of the stages, and I couldn't quite find it in the smash wiki. I wanted to be able to use the stage size data to figure out the exact percentages that Usmashes/upward trajectory attacks kill certain characters.
On that note, I also had a question about the percentage calculations. Due to staling factors, damage values can be given as decimals, so I was wondering how melee treated those values. I did some rudimentary testing myself, and when it comes to adding the values together, melee certainly(?) doesn't just floor the values for damage before adding them (say, you have an attack that does 10%, then you use it again, so it does 9.1%. It doesn't just floor the decimals before adding it to the victim's total damage taken, according to the calculations I did with falco's forward tilt staling - you can forward tilt someone to 41%, but if it floored every value first it would be 40%). So, I was then wondering how melee added these things -> Right now I'm thinking that the percentage value it shows is just a floored value, but it adds together the exact decimals "behind the scenes" anyway. I'm asking this because it would potentially make a small difference if melee calculated hitstun/KB/Velocity for only the given whole number value of the character's damage, or if it just showed the whole number value but really used the exact decimal values for the calculations of hitstun, KB, etc. Also, hitstun values are just floored right? And are hitstun and velocity values just kept exact?
sorry for the kinda lengthy post

THIS just happened to be on the first page of MD.
Edit: Nvm, I thought you meant the actual stages, not the blastzones.

For the damage question, I believe that's answered in Kadano's Melee Mechanics video about hitstun (at least about how the hitstun/KB calculation handles decimals).

 

[Kadano]

Not exactly a frame thing, but I wanted to find out the sizes of the stages, and I couldn't quite find it in the smash wiki. I wanted to be able to use the stage size data to figure out the exact percentages that Usmashes/upward trajectory attacks kill certain characters.
On that note, I also had a question about the percentage calculations. Due to staling factors, damage values can be given as decimals, so I was wondering how melee treated those values. I did some rudimentary testing myself, and when it comes to adding the values together, melee certainly(?) doesn't just floor the values for damage before adding them (say, you have an attack that does 10%, then you use it again, so it does 9.1%. It doesn't just floor the decimals before adding it to the victim's total damage taken, according to the calculations I did with falco's forward tilt staling - you can forward tilt someone to 41%, but if it floored every value first it would be 40%). So, I was then wondering how melee added these things -> Right now I'm thinking that the percentage value it shows is just a floored value, but it adds together the exact decimals "behind the scenes" anyway. I'm asking this because it would potentially make a small difference if melee calculated hitstun/KB/Velocity for only the given whole number value of the character's damage, or if it just showed the whole number value but really used the exact decimal values for the calculations of hitstun, KB, etc. Also, hitstun values are just floored right? And are hitstun and velocity values just kept exact?
sorry for the kinda lengthy post

Stage name |Left x boundary |Right x boundary |Upper y boundary |Lower y boundary |Side platform height |Top platform height |Left ledge x |Right ledge x
YS| -175,7 |173,6 |168 |-91 |23,45 |42 |-56 |56
FD |-246 |246 |188 |-140 ||| -85,5606 |85,5606
FoD |-198,75 |198,75 |202,5 |-146,25 |27,375 |42,75 |-63,35 |63,35
DL64 |-255 |255 |250 |-123 |30,2425 |51,4264 |-77,27 |77,27
BF |-224 |224 |200 |-108,8 |27,2 |54,4 |-68,4 |68,4
PS |-230| 230| 180| -111| 25 || -87,75| 87,75

Credits to Strong Bad Magus420 who collected all of these except for PS (for that, he kindly provided the necessary tools).

 

[Strong Badam]

@ Magus420 collected them, not me.

 

[Bones0]

All frame data can be traced back to Magus.

 

[King Mob]

Depends which way you airdodge. And it decreases each frame.

It starts at 2.79 though.

Huh. I didn't think it'd be so variable. Is there a thread with more information on the mechanics of the air dodge? Especially in regards to wavedashing.

 

[schmooblidon]

Huh. I didn't think it'd be so variable. Is there a thread with more information on the mechanics of the air dodge? Especially in regards to wavedashing.

Well it's the same in whatever direction, but you will be splitting the total velocity between horizontal velocity and vertical velocity depending on the angle.

I'm working in an in depth ground movement thread here. It has info on wavedashing but if it doesn't answer your questions then ask.

 

[King Mob]

Well it's the same in whatever direction, but you will be splitting the total velocity between horizontal velocity and vertical velocity depending on the angle.

I'm working in an in depth ground movement thread here. It has info on wavedashing but if it doesn't answer your questions then ask.

Dude, this is great. I've always been super curious about the nitty-gritty of movement in melee.

 

[schmooblidon]

Anyone know exactly how smash turns affect velocity?

I'm specifically looking for how it affects you when used within a dash, it seems to act differently in other states, but I'd also be interested in how it works in those ways.

I managed to figure out how to find the velocity of frame 1 turn and frame 1 dash (they are the same, because dash frame 1 just repeats the previous frame), but only when given a character specific value which I have no idea where it comes from.

The equation is:

( dash_lastframe_velocity_z - dash_lastframe_velocity_x ) / 4 + turn_velocity_x = turn_velocity_z

Luigi for example would be

( dash_lastframe_velocity_z - 1.1 ) / 4 + 0.25 = turn_velocity_z

So if I know one value, I can find any. But I don't know where the value comes from. Maybe I'm missing something but there doesn't seem to be any direct relation with initial dash velocity, dash acceleration, traction or max walk speed.

Frame 2 dash is also making little sense. Looking at doc it seems like it just adds the initial dash velocity and 1 frame of dash acceleration to frame 1 dash. But the same can't be said for the whole cast.

 

[the muted smasher]

I need the shield stun formal or stun from 3% to 11%