Work Log by Kevin Beason

Backdate: 3/20/2008

Fixed quaterion julia intersector. Turns out all I needed to fix the precision problems I was having was to put a cap on the maximum distance the sphere tracing could advance. Estimates far from the intersection were prone to overflow and accuracy issues, so adding a maximum stepsize fixed this. I also added a rudimentary glossy reflection (the phong one in S12.2.3 of Shirley’s RRT 2nd ed.), a class of reflection long overdue for me. This image was path traced with monochromatic emissive sky and ground. Don’t have rendering time handy but it was a number of hours on a quad core Q6600 for the 2048×2048 original.

This was from 3 months ago, hence the “backdate”.