Explanation
The time elapsed between an action and the corresponding system response.
Real-world example
The delay between moving your head and seeing the image follow — critical in VR.
Practical applications
- VR comfort: latency below 20 ms is necessary to avoid nausea
- Responsiveness: interactions must feel instantaneous
- Presence: delay breaks the illusion of "being there"
- Competition: pro gamers optimize every millisecond
Sources of latency in VR
Tracking latency
- Sensor read time (IMU, cameras)
- Generally 1–5 ms on good headsets
- Sensor fusion can add delay
- Motion prediction partially compensates
Example: Fast IMUs in VR headsets provide rotation data in under 1 ms
Rendering latency
- Time to draw a frame
- Depends on scene complexity
- Powerful GPU = reduced latency
- Target: 11 ms for 90 FPS
Example: A simple scene renders in 5 ms; a complex one in 15 ms
Display latency
- Screen refresh time
- Rolling update adds variable delay
- Scanout time depends on position on the screen
- OLED is generally faster than LCD
Example: The bottom of the screen displays roughly 5 ms after the top
VR scenario
You quickly turn your head to the right. If total latency is 50 ms, you will see the world "follow" with a visible lag — as if the universe were made of jelly. At 20 ms, the motion feels natural. At 11 ms (the target for high-end headsets), it is indistinguishable from reality. Every millisecond saved improves comfort and presence.
Why it matters in professional VR
- Physiological threshold: above 20 ms, the brain detects the delay
- Nausea: latency is the primary cause of VR sickness
- Continuous optimization: engineers chase every millisecond
- Quality/latency trade-off: less beautiful graphics but more responsive