Perceptual model for adaptive local shading and refresh rate

When the rendering budget is limited by power or time, it necessary to find combination of parameters, such as resolution and refresh rate, that could deliver best quality. Variable-rate shading (VRS), introduced in last generations GPUs, enables fine control quality, which each 16×16 image tile can be rendered with a different ratio shader executions. We take advantage this capability propose new method for adaptive local rate. The analyzes texture content, on-screen velocities, luminance, effective suggests rate VRS state map maximizes quality animated content under budget. based on content-adaptive metric judder, aliasing, blur, derived from psychophysical models contrast sensitivity. To calibrate validate metric, we gather data literature also collect measurements motion variable rates, velocities motion, display capabilities, persistence, angular resolution. proposed implemented game engine plugin. Our experimental validation shows substantial increase preference our over fixed an existing motion-adaptive technique.