Database and charts for the latencies of input devices

1 - Latency charts 2 - 8000 Hz Polling rate - real benefit or voodoo? 3 - Test system and methods

The LDAT measurement setup: NVIDIA LDAT V2, Sharkoon SGH50 and the ASUS ROG Chakram X as a reference

The NVIDIA LDAT V2 is used in my setup to precisely record the end-to-end latency times of mice. The system is complemented by the Sharkoon Skiller SGH50 headset, whose microphone is used to acoustically detect the click noise. The microphone is positioned at a constant distance of around 4 to 5 centimeters directly in front of the mouse button so that the mechanical click can be recorded reproducibly and with a clear timing. LDAT then correlates this signal with the time of the first change in brightness on the screen, enabling a complete measurement of the end-to-end latency – i.e. from the mouse click to the visible reaction.

The ASUS ROG Chakram X serves as a reference device for the temporal classification of the measurement results. This mouse supports both wired and wireless operation at a high technical level and, with its uniform click mechanism and stable signal processing, provides a reliable basis for comparative measurements. This reference model was chosen in close consultation with Fritz Hunter, who provided me with the Chakram X to ensure a methodically identical measurement approach. He also checked the plausibility of my measurement results and confirmed their consistency in comparison with his own surveys. This provides a technically validated and verified database.

A central element of the LDAT measurement is the display component. While Fritz Hunter conducts his analyses on an ASUS ROG Swift PG279QM (IPS, 240 Hz), I use the AOC AGON Pro AG276QZD2 with QD OLED panel for my tests. The pixel response time of this OLED monitor with a declared 0.1 ms gray-to-grey is significantly below that of the IPS panel from Fritz Hunter, which requires around 2.5 ms in the range relevant for LDAT (RGB 64 to 192). This difference has a direct effect on the measured overall latency: For example, the latency measured with the Chakram X in my setup is 12.1 ms, while Fritz Hunter determined a value of 14.5 ms – the delta of 2.4 ms corresponds almost exactly to the difference in response time between the two displays.

However, to ensure comparability across systems, we agreed on a general correction of the values by 2 ms. This allows the relative differences between the tested mice to be correctly mapped without the display technology creating a systematic advantage. In addition, small deviations of less than one millisecond are not considered critical in terms of measurement technology, as they are within the range of normal tolerances – caused, for example, by slight changes in the sensor position, USB jitter or ambient noise.

This coordinated approach results in a robust, reproducible measurement setup that ensures both internal comparability within the company’s own database and external comparability with existing LDAT measurements. Many readers will already be familiar with LDAT and its possibilities. For those who have not yet dealt with the topic, I recommend the detailed article by Fritz-Hunter, which I would like to link to here. It not only explains the functionality in detail, but also illustrates the practical use of the measuring system:

Kabellose vs. kabelgebundene Mäuse – wir messen die Latenzunterschiede! | NVIDIA LDAT V2 im Praxistest