There was some analysis posted when the article originally appeared on HN: http://news.ycombinator.com/item?id=1227936 http://news.ycombinator.com/item?id=1227

There was some analysis posted when the article originally appeared on HN: http://news.ycombinator.com/item?id=1227936 http://news.ycombinator.com/item?id=1227936 As I understand it, the trig calls are just to establish the camera position and direction. The rest of the calculation (the two inner loops) is marching along a ray to find a depth value.