However, rather than simply relying on scattershot reflections, McDonald and Jongejan used hundreds of structured light scans to capture the volumetric position of every pixel being projected by each of the three projectors. The pair then used SketchUp to predict the reflected pixel positions.