Method

CPI Benchmarking Methodology

temppal.gg measures gaming mouse CPI fidelity by comparing motion-capture-reconstructed physical sensor motion against raw HID reports. The method is designed to test whether a mouse turns intended hand motion into predictable computer input across a broad motion range.

Reported quantities

CPI bias

Signed fitted base-CPI error. It is computed from c0 in the fitted CPI model, so zero means the fitted base CPI matches the nominal CPI.

CPI jitter

Residual variation between HID reports and motion-capture ground truth after model fitting. Lower means less report-level variation.

CPI drift

Signed effective-CPI shift with translational speed and angular speed. Closer to zero means CPI changes less with motion.

Model used for public reporting

For a nominal CPI setting, temppal.gg summarizes effective CPI using the fitted model CPI = c₀ + c₁v + c₂ω, where v is translational speed and ω is angular speed. Public records are tested over a broad range of approximately 0-1 m/s and 0-5 rad/s.

The public pages report the fitted quantities and benchmark summaries, while internal implementation details such as exact sampling and processing choices remain part of the measurement system.

How the aggregate score is combined

The aggregate score is a dataset-relative summary, not a separate measurement. Each condition row is standardized against the current public benchmark distribution. The score combines -z(|CPI bias|), -z(CPI jitter), and -z(|CPI drift|).

CPI bias has weight 0.5 because users can often compensate a small base-CPI offset after measurement by adjusting the selected CPI, including through temppal.gg's CPI Match tool. CPI jitter and CPI drift have weight 2 each because report inconsistency and motion-dependent drift cannot be corrected as cleanly with a simple CPI setting. Higher aggregate score is better. In benchmark tables, the Bias, Jitter, and Drift columns show the weighted contribution added by each term, so those three displayed values sum to the final Score.

Open CPI Match

Measurement pipeline

1. Track the physical mouse body and reconstruct the optical sensor location.
2. Record raw HID mouse reports under controlled CPI, pad, polling-rate, and motion-sync conditions.
3. Align the physical motion trace and HID trace in time.
4. Fit the CPI model and derive CPI bias, CPI jitter, and CPI drift.
5. Publish a public snapshot with mouse-level and condition-level summaries.