Over the past few days i have made a few improvements. I have focused a bit more on the performance side of things and implemented (fairly bad) CPU multi threading as it was running single threaded before. This resulted in a 2-3x performance increase even though I went from 1 to 8 threads. I believe this is due to my bad code and the fact I create 8 new Thread objects every frame.

Other than performance optimisations I also expanded the functionality of the renderer. I added Lambertian scattering for more accurate matte surfaces. I gave metal a fuzziness propitiate that allows for more surface options. I also started to implement glass (which is an absolute pain) but it is still broken at the moment (see screenshots) :) I also did some general housekeeping and cleaned up the code making it easier to read.

The code is now visible on github for anyone who is interested.