The additive effects of the two types of oscillation on vection

Visually induced illusions of self-motion in depth (i.e., vection) can be facilitated by adding simulated viewpoint oscillation to the optic flow (Palmisano et al., 2011). In principle, this visual oscillation could be added to the display either as: 1) additional camera motion (oscillating the position or orientation of the camera); 2) global scene motion (adding global perspective oscillation to the environmental objects). While the visual consequences of camera and global scene manipulations appear similar, they can result in different strengths of vection (Sato et al., 2020). Here we examined the effects of adding both types of oscillation on the experience of vection in depth. There were 6 different display conditions, each of which simulated self-motion through a 3-D cloud of dots arranged in a regular or curved grid array. These were constructed by combining 3 different simulated camera motions (forward camera translation, vertical oscillatory camera translation facing forward, vertical oscillatory camera translation and rotation) with 2 different scene motions (backward scene translation, or vertical oscillatory scene translation). The dots appeared to move up and down with vertical undulation and the simulated forward self-motion. Latency, duration and magnitude were obtained. Results showed that vection in depth was stronger in all oscillating (compared to non-oscillating) conditions. This oscillation advantage for vection was further increased when simulated camera oscillation was combined with simulated scene oscillation. The strongest vection was found in conditions which also contained simulated oscillatory camera rotation (similar to the view during a rollercoaster ride – as opposed to the other 5 conditions where the observer always appeared to look forward). The oscillation effects were robust, not dependent on the type of oscillation, and had additive effects.