Perceptual Performance Impact of Gpu-based Warp & Anti-aliasing for Image Generators

In 2012 the U.S. Air Force School of Aerospace Medicine, in partnership with the Air Force Research Laboratory (AFRL) and NASA AMES, constructed the Operational Based Vision Assessment (OBVA) simulator. This 15channel, 150-megapixel display system remains one of the highest resolution displays ever built. One of the original goals for the simulator was to implement a distortion correction system that introduces “zero” frame latency into the overall system. This distortion correction was achieved using a combination of Scalable Display’s EasyBlend SDK and NVIDIA’s Warp and Intensity adjustment API. This paper describes the results of a collaboration between USAFSAM, Scalable, and NVIDIA to evaluate NVIDIA’s WARP 2.0 API, which allows for several user-selectable filtering techniques. These filters have the potential to improve the quality of the display warp and improve antialiasing performance without change to the low latencies already achieved. This paper provides a brief review of the different filtering techniques under investigation, as well as an assessment of their performance within a flight simulation environment. The evaluation has been conducted using psychometric methods to determine threshold performance of human observers on an operationally relevant aircraft orientation task conducted at an eye-limiting resolution (1 arcmin/lp).