How to test your mouse DPI (without drivers)
DPI is the foundation every sensitivity number sits on: get it wrong by 20% and every conversion you make is wrong by 20% too. Here are the three reliable ways to find your true value, cheapest first.
Published 07 Oct 2026 · By the SensForge team
Method 1, Read it from your driver (30 seconds)
If your mouse came with software (Logitech G Hub, Razer Synapse, SteelSeries GG, etc.), open it and check the active profile's DPI setting. Two traps: (1) some drivers show per-profile values, and the stored on-board profile may differ from the on-screen one; (2) DPI-cycle buttons change the value without updating the window you were looking at. Read it, note it, then verify with Method 2. The two should agree.
Method 2, The count method (2 minutes, works for any mouse)
- Turn off acceleration. Windows: Mouse settings → Additional options → Pointer Options → uncheck "Enhance pointer precision". Also check your driver for any acceleration or angle-snapping feature.
- Open the DPI analyzer. It counts sensor steps (mouse counts) your system reports while you drag the mouse in a straight line.
- Drag a measured distance. Place a ruler on the desk, drag slowly along it (a consistent 20–30 cm), press stop.
- Enter the distance. The tool divides counts by the distance in inches:
DPI = counts ÷ inches. Repeat three times and average; a good result repeats within ±3%.
Why this works: with acceleration off, every inch of movement produces a fixed number of counts. That rate is your effective DPI, measured post-driver and post-OS.
Method 3, cm/360 cross-check (indirect, needs a game)
In any supported game, do a slow full 360° turn and measure the physical distance it took
(see cm/360 explained). Then solve
DPI = 360 ÷ (cm/360 ÷ 2.54 × sens × yaw), or just enter your values in the
cm/360 calculator and adjust DPI until the number matches what
you physically measured. This method checks the whole chain (sensor → driver → game) at once,
which makes it the best final validation but the worst first measurement (a single wobble skews it).
Accuracy expectations
| Method | Typical accuracy | Best for |
|---|---|---|
| Driver readout | Depends on profile state | A quick start; verify it |
| Count method + ruler | ±2–3% | Confirmation and new mice with no software |
| cm/360 cross-check | ±3–5% single try | End-to-end validation before tournaments |
Any of these beats guessing, and all three agree when your setup is clean.
Common DPI mistakes
- Confusing DPI with sensitivity. DPI is hardware; sensitivity is the game multiplier. They combine into eDPI, see what is eDPI.
- Chasing huge DPI numbers. 3200+ DPI only helps if your game supports raw input and you convert sensitivity down properly; otherwise it just amplifies noise.
- Testing with acceleration on. The single most common cause of "my measurement is 30% off" reports.
- Forgetting polling rate exists. Polling rate (1000 Hz etc.) affects input latency, not DPI. Don't mix them up when troubleshooting.
Frequently asked questions
What is DPI actually measuring?
DPI (dots per inch) is how many sensor counts your mouse reports per inch of physical movement. A 800 DPI mouse reports 800 counts per inch, the operating system then turns those counts into cursor movement, and your game turns them into camera rotation.
Why measure instead of trusting the sticker?
The number on the box is a nominal value: driver profiles override it, dual-mode mice switch it with a button, and sensors vary ±1–3%. Measuring takes one minute with a ruler and removes all doubt, especially after a driver update or profile import.
Does DPI higher than 1600 make my aim better?
No. Once raw input is on, DPI and in-game sensitivity trade off exactly: 1600 DPI at half sensitivity equals 800 DPI at full. Higher DPI only reduces sensor smoothing at the low end or lets you use finer slider steps. Pick a clean value and convert properly.
Does Windows sensitivity change my DPI?
It changes the cursor speed, not the mouse's report rate. For games with raw input (all competitive shooters today), the game reads the counts directly, but leave Windows at 6/11 and acceleration off, because some overlay and menu paths still use the OS path.