Direction, sprint, crouch and jump is four keys under one hand. Membrane boards routinely drop one of them, and the player interprets a hardware limit as bad timing.
Steam You will bullet jump ten thousand times. Measure one.
Crouch, then jump, while already moving — the input that carries every tileset in the game. Whether it flows or stutters comes down to the milliseconds between those two keys and to whether your keyboard reported all four at once. This bench records both, on your hardware, and tells you which of the two is actually costing you.
Nothing leaves this browser Physical key codes, any layout No invented launch window
KEYMAP / BULLET JUMP ARCLOCAL
fastreporting bandslow
Median gap—
Spread—
Attempts0
Bind it. Arm it. Jump the way you actually play.
Three separate facts, kept separate: whether the keys arrive, how long your crouch-to-jump gap is, and how much that gap varies. A wide spread is what makes chaining feel unreliable, and it is invisible without a measurement.
Tell the bench which keys you really use.
Defaults are Warframe's own. Click a key and press yours — the page stores the physical position, so a non-QWERTY layout is measured on the key your finger reaches.
Hold a direction. Crouch. Jump. Again.
Every crouch-then-jump inside one second is recorded as an attempt. Ten or more is where the spread starts meaning something.
Bench Bench idle
Press arm, then bullet jump the way you would across a room. Keep a direction held down.
WASD Shift Ctrl Space
Two things the gap alone cannot tell you.
Whether your keyboard can carry the whole manoeuvre at once, and whether the ability row is reachable while the same hand is holding movement.
Delivery
Direction, sprint, crouch and jump held together. Keyboards report a limited number of simultaneous keys, and the one they drop is usually the one you never suspected.
Press start, then hold all four together.
Ability row
Press 1, 2, 3, 4 in order with the hand that is also holding movement. The page times each step and names the one that does not land.
1234
Median step—
In order—
What your hands and your keyboard actually did.
Every line comes from a measurement on this page. Nothing is inferred from your frame, your mods or anybody else's settings.
Jump consistency — Out of 100, inside this model
Median gap—
Spread—
Inside band—
Deliverynot run
Ability rownot run
0 finding(s). Nothing on this page is uploaded; the export is a file you save yourself.
Why this one input The keyboard refuses The gap drifts The bind costs more than it looks
Warframe is a movement game with a gun in it.
Combat happens in bursts; movement happens constantly. A player crossing a tileset produces the same two-key sequence hundreds of times an hour, and three separate things can make it fail — only one of which is skill.
A median tells you what your hand usually does; the spread tells you whether it can be relied on. Chaining feels broken when some attempts launch and some do not.
Crouch on the left pinky, while that hand also holds sprint and a direction, is the most loaded finger on the board. Moving it is free and often worth more than practice.
Method and limits Measured Our band, not the game's Not claimed Repeatable
What is measured, and what is deliberately not claimed.
Milliseconds between the crouch key event and the jump key event, per attempt, with whatever else was held. Simultaneous delivery of the four movement keys. Order and timing of the ability row.
Warframe's launch window is not published. The band on the arc is this site's reporting band and is labelled that way everywhere it appears — including in the export.
Whether the game accepted a launch, what a frame's mobility stats do, network conditions, or anything about performance in a mission. None of that is observable here.
Same binds, same hand, same keyboard should land within a few milliseconds of the same median. Run it again after changing the crouch bind: that comparison is the point of the page.
Before you rebind crouch
Questions this bench gets asked
What counts as a bullet jump on this page?
A crouch key event followed by a jump key event within one second, recorded together with whatever else you were holding at the time. The bench reports the milliseconds between those two events on your own machine. A gap longer than a second is treated as a new attempt rather than a slow one, so a pause between runs is never recorded as a failure.
Is there a correct number of milliseconds?
Not one this site can publish. Digital Extremes has never released the launch window, so printing a threshold would be inventing a number, and this site does not invent numbers. The band drawn on the arc is our own reporting band and is labelled that way everywhere it appears, including in the export. The reading that actually helps is your spread — how much the gap varies — rather than any single attempt.
Why does it ask me to hold a direction?
Because a bullet jump in play always starts from movement, and holding a direction changes which fingers are free for crouch. A gap measured standing still is a different measurement from the one you make in a mission, so the report tells you when most of your attempts were made stationary and asks you to run it again properly.
My crouch never registers while I am sprinting. Is that Warframe?
Probably not. Direction, sprint, crouch and jump is four keys under one hand, and that is exactly where a keyboard matrix starts refusing to report one of them. The delivery check holds all four and names the one that never reached the browser. It is a hardware fact rather than a timing problem, and the fix is moving one bind to a key on a different row.
Does hold or toggle crouch matter?
It changes what your hand physically does, so it changes the gap it produces. Which one suits you is not something a guide can know, and the arguments about it online are made entirely without measurements. Run ten attempts under each and keep whichever gives you the narrower spread — that answer takes four minutes and belongs to you.
Where to go next
The bench reads your hands. These read the hardware they are working through.
Ghosting test Four movement keys under one hand is where a matrix limit shows up first. Open Chatter test A bouncing switch reads in game as a crouch that dropped halfway through a slide. Open Input latency test Separate the delay your hand produces from the delay the machine adds after it. Open Other game labs The same treatment for the rest of the library. Open
