Pict vs. rtrace runtime


I noticed that when comparing run-times on a single core, with equivalent input flags, rpict is significantly faster than rtrace, possibly by a factor of 2 or more.

Why is that? I though it may have to do with rtrace incurring an overhead from reading rays from stdin, but I think there may be more to it.

Any ideas?

Thanks in advance

It’s probably due to image-plane sampling, which rtrace doesn’t do because it has no idea how the requested samples are related or might be interpolated. If you run rpict with the “-ps 1” option, then the performance should be similar.