Five-speed runs
Same distance, increasing speed
0 counts
Run 1 / 5
Hold the button, move the target distance very slowly, release. Then repeat faster each time.
Travel ratio vs speed (dashed line = 1:1)
Record at least three speeds to see the curve.
Mouse Acceleration Curve Test
This mouse acceleration curve test plots counts against speed. Set the distance and your DPI, then do five runs of that distance from very slow to very fast, holding the left button each time. Every run becomes a point; the line between them is your acceleration curve.
Run history
Plot the curve
1. Measure the distance
Tape marks 10 cm apart on the pad keep every run honest.
2. Go slow to fast
Very slow, slow, medium, fast, very fast. The tool labels each run.
3. Read the line
Flat at 1.0 means no acceleration. Rising means the OS or driver is scaling fast movement.
What the curve shows
Travel ratio
Counts measured divided by counts expected from DPI and distance.
Speed axis
Each run's average speed in cm per second.
1:1 reference line
Dashed line so deviation is obvious.
macOS-aware
The macOS curve is strongly non-linear; this is the clearest way to see it.
Curve terms
- Acceleration curve
- The relationship between hand speed and cursor speed.
- Travel ratio
- How far the cursor went compared with a 1:1 expectation.
- Linear tracking
- A flat curve where the ratio is 1 at all speeds.
- Scaling factor
- The macOS term for its pointer speed and acceleration setting.
Curve questions
My ratio is 0.9 at every speed. Is that acceleration?
No. A constant ratio is just DPI or distance error. Acceleration changes with speed.
The curve bends down at the top. Why?
Sensor skipping at high speed or hand error on the fastest run.
Can I export this?
Screenshot the chart; the run list has the numbers.
Flatten the curve
Rising curve on Windows
Untick Enhance pointer precision.
Rising curve on macOS
Disable scaling via Terminal or use a tool such as LinearMouse.
Rising curve in one game only
That game has its own acceleration; look for a raw input option.