Latency test

How does the test work?

The Report rate mode times the gap between successive reports from your controller. Nobody reacts to a cue, so the result depends only on the controller and its connection: the rate in Hz, the time between two reports in milliseconds, the jitter (how much that gap varies) and the share of dropped reports.

Two methods feed it. WebHID (Chrome and Edge, one-time permission) receives every real HID report, so its reading is exact, and on a DualSense it also reads the controller's own clock to separate the pad's timing from the delay added by the connection. The Gamepad API works everywhere, but the browser re-samples it on its own schedule, so its reading can never exceed the browser's rate; the tool labels it as approximate. On Windows, Xbox controllers go through XInput and are usually not reachable over WebHID.

The Wired vs wireless mode is a reaction test: press a button when the box turns blue. Your score is mostly your own reaction time, so it only means something as a comparison. Run it wired, save the result, switch to wireless and run it again; your reflexes stay roughly constant, so the gap between the two runs comes from the connection. The tool calls a gap meaningful only once it is larger than the statistical noise of both runs.

How do I read my results?

Start with the report rate. The time between two reports is the longest an input can wait before the controller sends it, and on average a press waits half of that.

Then check that the stream is steady:

• Jitter under 1 ms is a stable link. Above 5 ms, something is interfering, and that unevenness is what makes an on-stream overlay stutter.
• Dropped reports above 3% point to the cable, the port or a USB hub when wired, and to distance or 2.4 GHz congestion when wireless.
• In the reaction mode, a median of 150 to 220 ms is normal. A spread of ±40 ms or more usually means fatigue or distraction, so collect more trials before comparing setups.

At 8000 Hz the browser clock becomes the limit: performance.now() resolves to about 0.1 ms, close to the 0.125 ms between reports. The tool then computes the rate over the whole run and shows jitter as n/a instead of rounding noise. The DualSense clock analysis has no such limit, because the controller timestamps each report to about a third of a microsecond.

How fast are popular controllers?

The figures below come from Gamepadla, which measures latency with a robot that physically presses each button and records when the PC receives it (±1 ms). Each result is one unit, on Windows 11, tested in 2025 or 2026.

Three things stand out. Raising a DualSense from its stock 250 Hz to 1000 Hz over USB halves its latency, from 7.8 to 3.9 ms. Xbox controllers report at about 125 Hz on every connection, so the official adapter is nearly as fast as a cable while Bluetooth roughly doubles the delay. And the Switch Pro Controller is the slowest pad in the list, wired or not.

Button latency includes the controller's own processing, so it is always longer than the time between reports, and only a hardware rig can measure it. This page measures the report rate and its stability, which you can check on your own pad in a minute.

Is a cable always faster?

No. It depends on the controller. An Xbox controller is fastest on a cable or its own adapter, and Bluetooth adds about 6 ms. PlayStation controllers behave the other way round: over Bluetooth they report several times faster than their 250 Hz USB default, and their average latency drops accordingly. Bluetooth keeps one drawback on every pad, though: occasional spikes. The DualShock 4 peaked at 20 ms over Bluetooth against 13 ms on a cable.

For your own setup, run the Report rate mode on each connection for a minute and compare the jitter and the dropped reports, not only the rate. For competitive matches, pick the connection with the steadier stream.

What affects controller latency, and how do I lower it?

• Choose the right connection. Use a cable or the official adapter for Xbox controllers, and compare Bluetooth with an overclocked cable for PlayStation controllers.
• Raise the polling rate carefully. On PC, a polling overclock tool can take a wired pad from 125 or 250 Hz to 1000 Hz. Test again afterwards: some controllers become unstable or disconnect at rates they were not designed for.
• Plug into the computer directly. A cheap or unpowered USB hub can add polling delay and drop reports compared with a port on the motherboard.
• Keep the PC free. A busy CPU, many browser tabs or a throttled background tab skew both the measurement and the game.
• Mind the game. Frame rate, V-Sync and the game's own input buffering often add more delay than the controller does.

Is it controller lag or overlay lag?

If the game reacts instantly but your stream overlay looks late, the controller is not the cause. Overlay delay almost always comes from a Browser Source setting in OBS or Streamlabs, and it is easier to fix. Start with the overlay delay checklist, then check the scene against the recommended OBS settings.

What cannot a browser measure?

The full delay between your finger and the game. That would need to know when the button was physically pressed, and a web page only knows when the input reached it. Labs solve this with hardware: a robot or microcontroller that presses the button and records the moment, as in the table above, or a 1000 fps camera filming the pad and the screen together. A web page that claims an absolute hardware latency on its own is mostly measuring your reflexes.

FAQ