Measuring Audio-Visual Latencies in Virtual Reality Systems

In virtual reality (VR) systems, delays may occur while the signal passes through hardware and software components, thus causing asynchrony or even cybersickness, as a result. To better understand and control the role of delays in VR experiments, we tested an accessible method for measuring audio and visual end-to-end latency between game engines (Unity and Unreal), head-mounted displays (HMD) (Oculus Rift and Oculus Quest 2) and vertical synchronisation (V-sync) setting (on/off). The measuring setup consisted of the microcontroller, a dedicated serial port, a microphone, a light sensor and an oscilloscope. The measurements showed that Unreal Engine with Oculus Rift had ≈ 16 ms less visual delay and ≈ 33 ms less audio delay than Oculus Quest 2. The Unity Engine with Oculus Rift had ≈ 22 ms less visual delay and ≈ 39 ms less audio delay than Oculus Quest 2. These values may differ between systems, but they are above the discrimination thresholds. No differences were found for the V-sync on/off parameter. Compared to the Unity Engine, the Unreal Engine showed much lower visual latency performance and significantly lower audio latency. In addition, Oculus Rift's audio and visual latency performance had lower delays than Oculus Quest 2; therefore, using Oculus Rift is advisable in VR research where lower latencies are essential, even though Oculus Quest 2 is a newer version of the HMD. Our approach provides a convenient way to measure audio and visual end-to-end latency in VR without a strong engineering background.