• Welcome to Smashboards, the world's largest Super Smash Brothers community! Over 250,000 Smash Bros. fans from around the world have come to discuss these great games in over 19 million posts!

    You are currently viewing our boards as a visitor. Click here to sign up right now and start on your path in the Smash community!

L-cancel speeds appear to be constants which vary dramatically

Antomeq

Smash Rookie
Joined
Aug 3, 2014
Messages
7
So I was TASing about with dolphin, when I noticed that captain falcon's aerials have very peculiar L-canceling speeds.

Not only is the L-cancelled animaiton playing at around 4X speed, that speed is varying by a considerable, non-integer amount depending on the aerial.

So I decided to also check what happens when I do not L-cancel..

The landing lag still appears to be reduced? And again with the significant variations in speed, from ~1-3X, none of the integer multipliers.

Does anyone have an explanation for this? My assumption is that every aerial has some decimal or floating point multiplier applied to it.

I also tested some of foxes and ganondorf's aerials to see if there was similar variation, and there is, but I didn't bother making gifs. If you want I can probably make some tomorrow.

Also note that this is TAS, so everything is being performed frame-perfect.
 
Last edited:

Stratocaster

Smash Ace
Joined
Oct 6, 2007
Messages
672
Location
Knoxville, TN
You're just looking at the current animation frame. They are non integer normally without l canceling. This is an effect of modifier values being applied later in development giving odd values. It's hard to explain but it has nothing to do with l canceling. Compare to not l canceling your landings and count up the actual frames that pass and it'll be a factor of 2 different.
 

Antomeq

Smash Rookie
Joined
Aug 3, 2014
Messages
7
In the second GIF, I am not L-canceling my landings. I'm simply assuming it's auto cancelling, because it's still going faster than normal. I edited the OP to clarify this.

You may be right though, the landing lag could be given a rather odd value, which is being doubled by the L-cancel. The fact that the multipliers are in cases so close to whole numbers bewilders me, though.
 
Last edited:

Yeroc

Theory Coder
BRoomer
Joined
Feb 28, 2004
Messages
3,273
Location
In a world of my own devising
@Magus420

I think what's happening is that the animations are all the same length or close to it (~30 frames seems to be a common number for landing animations) and the landlag values are set by character attribute (this is how it works in Brawl). Then the L-cancel mechanic doubles the rate that the animation plays, which was already a non-whole number because of the attribute.
 

hectohertz

Smash Ace
Joined
Aug 20, 2006
Messages
800
Location
Brooklyn, NY
yeah most of the people here got it correct. "normal" animations don't necessarily play at 1X speed, each animation plays at a different speed (this was likely done so that they could tweak speeds of animations later in development without having to redo the whole animation). this is accomplished via a multiplier. you'll notice that the frame-counter is a 32-bit float.

l-cancelling just doubles the multiplier for the landing animation. auto-cancelling causes the normal (as if you landed without doing an aerial) landing animation to play
 
Last edited:

Antomeq

Smash Rookie
Joined
Aug 3, 2014
Messages
7
Thanks for the informative comments, guys. I guess I didn't expect them to apply different speeds to the landing animations, especially considering the game mostly has a locked framerate!
 
Top Bottom