The author built a device to measure end-to-end input latency on Linux, aiming to verify optimization claims. The device, based on the OSLTT and other open-source projects, uses an Adafruit QT Py RP2040 to simulate mouse clicks and measure the time until a screen brightness change is detected.

Testing focused on display server (X11 vs. Wayland), Variable Refresh Rate (VRR), and a low-latency DXVK fork. Results from Diabotical on a 500 Hz monitor indicated minimal differences (< 0.72 ms) between most configurations, with all medians falling between 4.21 ms and 4.93 ms.

However, XWayland added a significant 3.13 ms of latency compared to native Wayland. In uncapped scenarios, the dxvk-low-latency fork reduced latency by 0.84 ms, demonstrating its effectiveness in absorbing frame time fluctuations.