NVIDIA Video Details Foveated Rendering Techniques

    Prior to the SIGGRAPH 2016 conference this week NVIDIA announced a partnership with eye tracking technology specialists SensoMotoric Instruments (SMI). The companies held a session at the event to showcase a new perceptually based foveated rendering technique they had been working on. NVIDIA has released a demonstration video detailing how the process works.

    Using eye tracking sensors foveated rendering is a process where software renders only the part of an image that the user is actually looking at, while the peripheral parts of the image are reduced in quality. It’s a similar technique to how eyes work and for computers this helps to reduced the burden on graphic cards, as they don’t need to constantly render the entire image in HD quality.

    The video below shows in an exaggerated form how the process works, which VRFocus got a hands-on with at SIGGRAPH 2016. As David Luebke, NVIDIA vice president of graphics research, and Aaron Lefohn, NVIDIA director of research explain in the company’s blog: “Our researchers used SMI’s prototype eye-tracking HMD to perform a careful perceptual study of what people actually see in their peripheral vision in VR. Our researchers then used those insights to design a new rendering algorithm that enables much greater foveation, or reduction in rendering effort, without any discernible drop in visual quality.”

    Foveated rendering isn’t likely to be appearing in any headsets just yet, but its certainly a technology that has useful applications for the future of virtual reality. For all the latest VR news keep reading VRFocus.