HDR VR

SIGGRAPH Emerging Technologies

Abstract

The human visual system operating in natural conditions can resolve luminance values that range from over a million candelas per meter squared (nits) to near zero, and is able to simultaneously resolve over 4 orders of magnitude without adaptation [Kunkel and Reinhard 2010]. While most traditional displays can only replicate a fraction of the smaller simultaneous range, high-dynamic-range (HDR) displays aim to support luminance and contrast ranges closer to perceptual limits [Reinhard et al. 2010]. HDR displays have achieved widespread commercial success across cinema, home theater, and personal use devices, but this popular technique remains largely unexplored in the context of Virtual Reality (VR) displays, typically limited to peak luminance values near 100 nits [Mehrfard et al. 2019]. The perceptual impact of HDR on the unique mix of controlled ambient, wide field-of-view, viewing optics, and immersive presentation typical of VR displays is largely unexplored. To address this, we present a high-dynamic-range virtual reality demonstrator with a display system comprised entirely of off-the-shelf parts, capable of peak luminances over 16,000 nits. We achieve this without reducing the field of view (FOV) or simultaneous contrast relative to commercially-available VR headsets. Furthermore, HDR demos in excess of 1,000 nits that support binocular and motion parallax depth cues have not been widely seen by our community. Consequently, our prototype has the potential to achieve a higher degree of perceptual realism than existing direct-view devices like HDR televisions and other high-luminance prototypes.

Latest Publications

Sustainable AI: Environmental Implications, Challenges and Opportunities

Carole-Jean Wu, Ramya Raghavendra, Udit Gupta, Bilge Acun, Newsha Ardalani, Kiwan Maeng, Gloria Chang, Fiona Aga Behram, James Huang, Charles Bai, Michael Gschwind, Anurag Gupta, Myle Ott, Anastasia Melnikov, Salvatore Candido, David Brooks, Geeta Chauhan, Benjamin Lee, Hsien-Hsin S. Lee, Bugra Akyildiz, Max Balandat, Joe Spisak, Ravi Jain, Mike Rabbat, Kim Hazelwood

MLSys - 2022