Self-initiated motion causes less sickness than imposed motion

What it is

The same visual motion, pushed through on a stick, usually sickens less than watching that same road from the seat behind. The split is active versus passive motion: the active side has a motor command and a predicted consequence; the passive side has only the picture. Cutscenes, being towed by a vehicle, watching someone else’s first-person recording, all impose motion on the head. What differs in sickness is not what the flow looks like; it is whether this motion was ordered.

The comparison is authorship, not stand versus sit, and not how many degrees of retina the flow covers.

Why it happens

A movement goes out with an efference copy, used to predict which way the retina should wash in a few milliseconds. When the prediction matches, residual error is small and can be treated as the external world. With no initiation there is no such prediction; the same ground-wash is a surprise, and compensation lags.

Yoked experiments isolate the difference: one person drives, another wears the same picture. The driver sickens less, the passenger more, path and flow identical. In games the peak often sits in a “watch this” cutscene that then hands control back — not because the cutscene is busier, but because control was confiscated.

Prediction needs a little temporal alignment. Delay large enough to miss the prediction window shrinks the active advantage — the wearer is still initiating, but the consequence is no longer the order they placed. Control that is only start/stop, with the path fully interpolated by the system, is active in name and close to passive in fact.

Studying it

The standard paradigm is yoked control: an experimenter or one participant drives, another receives the same visual trajectory and cannot change it. Alternatively the same person walks actively, then watches a replay of their own trajectory.

Independent variables: active versus passive, control delay, continuous versus on/off control, whether a takeover is announced. Dependent variables: in-session discomfort, vection, dropout, the active-minus-passive difference.

A replay condition must either replay head motion too or leave head motion with the participant; otherwise the comparison is “whole pose imposed” versus “only translation imposed”, which are not one kind of passive.

Where it stops holding

Scenes with no visual self-motion cannot show the split. Seasickness and the back seat of a real car are already passive; whether the headset offers a stick, the body is being imposed upon, and any active advantage then depends on virtual motion lining up with bodily motion. Some people drive harder when they have control, so the active condition has stronger flow and the between-group difference cancels — lock the path with yoking. Extremely short cutscenes do not have time to accumulate, and the difference will not be measured.

Applying it

• If a path can be walked by the wearer, do not rewrite it as a forced first-person tow. When a tow is required, announce it and keep it short.
• Do not put acceleration and whip-turns in cutscene cameras. If a cinematic beat is needed, leave first person, or leave a sliver of control (look, brake).
• When watching someone else’s view, label it as view-only, and offer an immediate take-back or a way to kill translation.
• How to check: record one path, one group drives, one watches the replay. In-session ratings should be systematically higher in replay. If the groups match, first check whether the replay still allowed head turns, then whether the path was already too mild to show a ceiling.

Related

• Same group: N1.13.1 The area and speed of optic flow determine sickness intensity · N1.13.3 Standing is more prone to imbalance and sickness than sitting · N1.13.4 Discomfort must be sampled continuously in-experience with comparable scales
• Nearby: N1.06 Cybersickness · N4.02 Continuous Locomotion
• Search terms: active motion · passive vection · efference copy