Online Mouse Test
The instrument records button and wheel events, estimates the browser-delivered event rate while the cursor moves and flags unintended double clicks.
Open tool fullscreen →Interface and local controls remain in Portuguese.
How to use
- Move the cursor inside the sampling area in wide, continuous strokes for a few seconds so the polling reading settles.
- Click each button once, including the side buttons, and spin the wheel up, down and horizontally.
- Make single, well-spaced clicks and watch the suspicious double-click counter: any value above zero deserves attention.
When to use
- Record suspicious repeats before comparing in another application or device.
- Verify that side buttons and horizontal wheel are recognised by the system.
- Compare the event rate observed in the browser across different mice, ports and settings.
- Quickly test a second-hand mouse before closing the deal.
Limitations
The measured rate is capped by the event cadence the browser delivers, usually tied to the compositor. It is neither a direct firmware reading nor a DPI measurement.
Frequently Asked Questions
- My mouse is 1000 Hz but it measured much less. Is it broken?
- This page cannot answer that. The browser may batch, cap or discard events; the observed value is not the device USB report frequency.
- What does a repeated click below 60 ms mean?
- It is an event flagged by the tool heuristic. Frequent repeats deserve controlled comparison, but the interval does not prove a defect or identify a physical cause.
Technical Methodology & Display Science
Clicks are read through Pointer Events with preventDefault, which lets us capture buttons 3 and 4 (back and forward) that the browser normally consumes for history navigation. Every press stores a performance.now() timestamp per button.
The rate sums coalesced events returned by getCoalescedEvents on each pointermove. It describes events exposed by the browser and may differ from the device USB report frequency.
The 60 ms threshold is a triage heuristic for highlighting nearby events. Human cadence, software, firmware and hardware vary, so a flag should be repeated and compared.