The intensity of vection is affected by many factors:
- Percentage of field of view: If only a small part of the visual field is moving, then people tend to perceive that it is caused by a moving object. However, if most of the visual field is moving, then they perceive themselves as moving. The human visual system actually includes neurons with receptive fields that cover a large fraction of the retina for the purpose of detecting self motion [35]. As VR headsets have increased their field of view, they project onto a larger region of the retina, thereby strengthening vection cues.
- Distance from center view: Recall from Section 5.1 that the photoreceptors are not uniformly distributed, with the highest density being at the innermost part of the fovea. Thus, detection may seem stronger near the center. However, in the cases of yaw and forward/backward vection, the optical flow vectors are stronger at the periphery, which indicates that detection may be stronger at the periphery. Sensitivity to the optical flow may therefore be strongest somewhere between the center view and the periphery, depending on the viewpoint velocities, distances to objects, photoreceptor densities, and neural detection mechanisms.
- Exposure time: The perception of self motion due to vection increases with the time of exposure to the optical flow. If the period of exposure is very brief, such as a few milliseconds, then no vection may occur.
- Spatial frequency: If the virtual world is complicated, with many small structures or textures, then the number of visual features will be greatly increased and the optical flow becomes a stronger signal. As the VR headset display resolution increases, higher spatial frequencies can be generated.
- Contrast: With higher levels of contrast, the optical flow signal is stronger because the features are more readily detected. Therefore, vection typically occurs with greater intensity.
- Other sensory cues: Recall from Section 6.4 that a perceptual phenomenon depends on the combination of many cues. Vection can be enhanced by providing additional consistent cues. For example, forward vection could be accompanied by a fan blowing in the user's face, a rumbling engine, and the sounds of stationary objects in the virtual world racing by. Likewise, vection can be weakened by providing cues that are consistent with the real world, where no corresponding motion is occurring.
- Prior knowledge: Just by knowing beforehand what kind of motion should be perceived will affect the onset of vection. This induces a prior bias that might take longer to overcome if the bias is against self motion, but less time to overcome if it is consistent with self motion. The prior bias could be from someone telling the user what is going to happen, or it could simply by from an accumulation of similar visual experiences through the user's lifetime. Furthermore, the user might expect the motion as the result of an action taken, such as turning the steering wheel of a virtual car.
- Attention: If the user is distracted by another activity, such as aiming a virtual weapon or selecting a menu option, then vection and its side effects may be mitigated.
- Prior training or adaptation: With enough exposure, the body may learn to distinguish vection from true motion to the point that vection becomes comfortable. Thus, many users can be trained to overcome VR sickness through repeated, prolonged exposure.
Due to all of these factors, and the imperfections of modern VR headsets, it becomes extremely difficult to characterize the potency of vection and its resulting side effects on user comfort.
Steven M LaValle
2020-11-11