• Support SmashBoards and get Premium Membership today!

  • 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!

The intricacies of frame canceling and why it happens


Smash Journeyman
Apr 19, 2015
Side 3
So we have all known about frame canceling for a while now, how to do it, combos it opens up, etc. What we have not known, is why it happens. I think I may have the answer.

The hypothesis came to me when reading Source Gaming's Famitsu translation "Thinking about hitstop" but until recently I have been unable to test it.
I am not going to post any of the translation here, so ill post the link if you would like to read it. http://sourcegaming.info/2015/11/11/thoughts-on-hitstop-sakurais-famitsu-column-vol-490-1/

The article shows Sakurai's thoughts on hitlag (hitstop) in Smash 4. In the article, he discusses some of the ways that he has emphasized hitstun in the new smash title. One of the methods he discusses is vibration.

This is where frame canceling is explained. During hitlag, not only does the opponent vibrate, but the attacker also vibrates, side to side on the ground, and up and down in the air. This mechanic was not present in previous titles.

Sakurai tells us that to avoid inconsistencies in hitbox detection, the characters hurtboxes do not move during the hitlag. Interestingly enough the characters lcs seems to still vibrate with the character. This in turn, cause the character to interact with a surface and trigger landing.

The sharp reader may respond with, "What about the fact that attacking character does not actually stop during hitstun and still moves slightly?" I would respond with, this component is largely irrelevant but still carries some effect to the timing of frame cancels. Because the amplitude of the vibration decreases over time by about proportional to the remainder of hitlag, it has an actual effect on only the first frame of downward momentum and in most cases the movement caused on the first frame is incredibly small.

So what does this mean?

We can draw a few helpful conclusions from this.
1. A character does not have to fast fall to trigger a frame cancel.

2. The window for a successful frame cancel will fluctuate and is not always one frame. Ill discuss this more later.

3. Frame canceling is not always possible.

4. Frame canceling may be able to be preformed traveling up through platforms! I have been unable to find an instance of this but it should be possible. Unless of course the game makes exceptions to landing detection based on the velocity angle of the character.

Lets talk about how to increase the window for frame canceling. The easiest way is to play with more people.

You heard that right. Since the amplitude of vibration corresponds inversely to the zoom of the camera, by increasing the number of people and thus decreasing the zoom of the camera, the vibration will increase for visual effect.

There is another way we can manipulate this trait. The camera is not often fast enough to keep pace with play and as a result by changing the distance between you and your opponent quickly you can increase the amplitude of vibration that results from a connected attack because the camera will be wider that its final position when the characters are a set distance apart.

Other ways to increase the window exist and are employed to different effectiveness by floaty and fastfalling characters.

By not fast falling, a character can abuse its "floatyness" to increase the window by increasing the number of frames the character is close enough to the ground to trigger a landing. If you are having trouble understanding, imagine how far fox falls between single frames and then compare that to Zelda. Now understand that the vibration amplitude is the same for both characters. The only example i have been able to produce and test reliably is Zelda's forward air. It have been able to confirm the frame cancel is possible on two distinct frames when hitting from the length of a battlefield platform without fastfalling.

Alternatively fastfallers can try to fast fall to increase the distance traveled in "slow motion" during hitlag.

Fastfallers also will encounter instances where on the frame before landing, they are too high to preform a frame cancel and thus are unable to cancel their lag.

Another niche way of causing a frame cancel is to utilize a rising platform. This rising platform will behave like an increase in fall speed, except to a greater extent.


Determining the frame advantage gained by using a frame cancel is quite simple. Simply add the number of hitlag frames to an already known frame advantage and subtract 2.

Actually im not sure what number to subtract. So far I have been unable to determine If the first direction of movement is random or not. Sometimes my testing has indicated down and sometimes up. That will make this number fluctuate between 0 and 2.


Characters do not vibrate in shield lag :(

Thanks for reading i know it was long. If anyone finds anything wrong please post it. Also if this was already known please forgive my arrogance. I did not see this anywhere. When i get back from vacation ill get to work on a vid showing my test footage.
Last edited:
Top Bottom