No effect of reducing visual realism on motion sickness in virtual reality

Simulated navigation in virtual reality often causes motion sickness. We tested whether motion sickness is modulated by the amount of visual realism of the simulated environment. Visual realism has been found to affect sense of presence, and some evidence suggests that motion sickness in virtual reality depends on presence. The only previous study that tested for a relationship between visual realism and motion sickness was small and did not control low-level motion information. We performed two experiments that compared motion sickness in conditions with different levels of visual realism but matched motion information. In the baseline realism condition, subjects navigated through a virtual town constructed of moderately realistic models with detailed texture maps. For the low realism conditions, the virtual town was modified to appear more cartoon-like without affecting motion information: complex 3D models were replaced with simple geometric shapes and detailed texture maps were replaced with simple patterns (Experiment 1) or solid shading (Experiment 2). Subjects performed both conditions in counterbalanced order. In each condition, subjects reported motion sickness (SSQ) and sense of presence (PQ) after navigation. We found no effects of visual realism. There was no difference between the motion sickness reported in the baseline and low realism conditions of either Experiment 1 (N=30) or Experiment 2 (N=24), and no differences in presence ratings. The lack of effect on presence might have been due to the limited realism of our baseline condition and the fact that subjects performed an active navigation task. Our results demonstrate that visual realism of a virtual environment can be varied over a wide range without affecting the tendency to produce motion sickness.