Quick answers before you test
How to check mouse DPI without software
Mark a known distance on the desk, lock the pointer in this page and move the mouse across it. Counts divided by inches is the observed DPI; three runs and the median make it a measurement instead of a guess. No driver, no download.
DPI analyzer vs labeled DPI
The analyzer estimates DPI from browser-observed movement and a measured desk distance. The mouse label or driver setting remains the configured value, not proof of the observed result.
Why does my mouse read 1580 instead of 1600?
A few percent of deviation is normal sensor tolerance and also includes your ruler error. It matters for conversions: use the observed value in the eDPI calculator or the sensitivity converter so cm/360 is real.
Use the observed DPI in the eDPI calculator →What should I do with the estimate?
Record the median and variation, then use the estimate as an input to a sensitivity conversion rather than treating it as a hardware certificate.
Open the sensitivity converter →What the result actually measures
The analyzer locks the pointer and adds up the mouse's movement counts, then divides them by the distance you measured on the desk. In Chrome and Edge it asks for the raw sensor counts, before the operating system applies its pointer speed and acceleration; each run is labelled with which kind it got. Multiple runs produce a median estimate, range and variation percentage.
Measurement boundary: Where the browser cannot give raw counts (the run says OS-adjusted), the operating system's pointer speed and acceleration are inside the number: on Windows keep pointer speed at 6 of 11 with Enhance pointer precision off, and on a Mac expect acceleration to skew it. Your desk measurement is the other error source.
How to interpret the output
Median estimate
The middle run resists one accidental overshoot better than selecting the cleanest-looking result.
Variation
A large run-to-run spread usually means the physical procedure or software path is not repeatable enough.
Raw or OS-adjusted
Raw sensor counts are what the mouse reported. OS-adjusted counts have been through the pointer speed slider and acceleration, so they only equal the DPI at Windows speed 6 of 11 with acceleration off.
Troubleshooting unexpected results
Estimate is far from the label
Look at the run label first. OS-adjusted runs are scaled by the pointer speed slider and acceleration; switch to Chrome or Edge for raw counts. Then re-measure the desk distance before concluding the mouse is off-spec.
Pointer lock is denied
Click the test control from a focused top-level page and allow pointer lock; some embedded or restricted browsers block it.
Frequently asked questions
Does Windows pointer speed or Enhance pointer precision change the result?
Not in Chrome or Edge, which give this page the raw sensor counts and label each run raw. In a browser that cannot, the run is labelled OS-adjusted and both settings scale it; keep pointer speed at 6 of 11 and Enhance pointer precision off there.
Is this a raw DPI certificate?
No. Even with raw counts, the result is only as good as your desk measurement, which is why it keeps every run and reports the spread.
Why use a longer distance?
A one-millimeter measuring error becomes a smaller percentage of a longer run.
Does browser zoom matter?
It can affect some delivered coordinate paths, so keep zoom at 100% and record the environment.
Can I use this as a DPI tester for a Logitech, Razer or budget mouse?
Yes, any mouse the browser can lock the pointer on. Set the driver DPI you want to check, run three passes over the same distance and compare the median with the driver value; a gap above about 5% is worth a second measurement before you blame the sensor.
What is the difference between a DPI meter and a DPI analyzer?
Both estimate the same thing: counts delivered per inch of real movement. A meter usually shows one number; this analyzer keeps every run, reports the spread and tells you when the environment rather than the mouse is the problem.
Why does the sensor lab exist separately?
DPI is one number. The mouse sensor lab checks acceleration, angle snapping, rest noise and tracking skips at speed, which no DPI figure can show.
