A fixed reference frame supplies a stable visual anchor
Aliases: rest frame · virtual nose · cockpit frame
What it is
A cockpit window frame, a virtual nose, a grid at the feet — still nailed in the view while the world streams backward. An independent visual background, more colloquially a rest frame, is a stable structure that does not follow virtual translation, giving vision an anchor that says “the world is not dragging me.” It does not narrow the field of view. It inserts a motion-decoupled referent inside it.
A tunnel hides optic flow. A rest frame recodes some of that flow as “the world outside the cabin is moving; the cabin is not.”
Why it happens
Vision uses large, low-spatial-frequency structure to judge self-motion. If texture streams across the whole retina, the conclusion is “I am moving.” Overlaying a stable surface locked to the head or body — a cabin, a nose, a horizon grid — supplies visual evidence that matches the vestibular report of no acceleration, and layers it against the streaming outside: the stream is parsed as a moving environment rather than a moving self.
The anchor has to actually be stable. Micro-jitter with the head, latency, or a slow drift of its own turns it into a second conflict. Coverage is a trade: too small (a speck of virtual nose) and the anchor is weak; too large (a heavy cockpit eating half the view) and the task scene is occluded, mitigation becoming blockage. Small central objects such as a nose work for some people because they sit near the fovea and on the same side as facial proprioception, not because they hide peripheral flow.
Studying it
On one continuous-locomotion scheme, overlay different rest frames: none, a virtual nose, a cockpit frame, a world-locked distant horizon (the last is not an independent background). Take discomfort with FMS/SSQ, and take whether the referent was parsed as a frame with IPQ or a single “did you feel inside a cabin.” Eye tracking can show whether people use the frame as an anchor or treat the nose as an ignorable occluder.
Manipulate lock mode (head / body / world), area, opacity, and whether the frame fades with motion. A world-locked “stable” horizon streams with virtual motion and cannot be a rest-frame control level.
Where it stops holding
In on-foot first-person content, a sudden cockpit is semantic noise; comfort from the anchor can be interrupted by “what is that.” Under optical see-through the real room is already the strongest rest frame, and a virtual cabin is usually surplus. If the anchor yaws with virtual yaw, it is no longer independent. People averse to near-face objects will reject a virtual nose rather than stabilize on it. Still viewing with no translation does not need the anchor.
Applying it
- In content that already has vehicle semantics, use a cabin or windscreen frame. For walking content, prefer a weak, disableable virtual nose or a grid at the feet; do not force a cockpit.
- Lock the anchor to the head or body, not to a world that rotates with virtual translation.
- Ship a toggle and an opacity control. Default pale, so the frame can be “there” without being “in the way.”
- How to check: the same stick-translation segment with and without the anchor. FMS should drop. If people report staring at the frame or nose, or task time jumps, the anchor is too large — cut area, do not fill it solid.