Abstract
Virtual Reality (VR) has emerged as a groundbreaking tool in science. It is particularly appealing because it allows realistic scenes to be created in a controlled manner. This is critical for a variety of research questions, especially those dealing with perception and action. Outside of science, VR today primarily occupies a niche in the gaming and entertainment markets, but it has the potential to rapidly evolve into new application areas such as business and social media, which will embed more and more real-world content into VR. Therefore, the need for realistic rendering will increase, both for scientific and non-scientific VR applications. A fundamental aspect of realistic rendering is accurate color reproduction, which in VR depends on the type of head-mounted display (HMD), the graphics engine, and the materials and shaders used. We performed a comparison between different HMDs (HTC Vive Pro Eye, Pimax 8KX DMAS and Varjo Aero), systematically testing the effects of the graphics engine (Unity and Unreal) and different materials and shaders (Unlit and Standard). Tone mapping was disabled to avoid related non-linearities. Color measurements were made using a Konica Minolta CS-2000A for RGBW in 8-bit increments of 5. In addition, 27 predefined and 50 random RGB values were measured for validation. Based on these measurements, we provide the transformation for each HMD, graphics engine, and material to convert from XYZ to RGB color space. Color reproduction after calibration was most accurate for the HTC Vive Pro Eye using Unreal with standard material (dE2000 mean error = 0.5). Furthermore, the HTC Vive Pro Eye offers the largest color gamut of all HMDs tested. In summary, we successfully performed color calibration with various combinations of HMDs, graphics engines, and materials and shaders, which allows for accurate color reproduction in VR.