This is amazing, and very well explained, thank you! I have one question - does this algorithm take view-dependent information into account to also do specular GI? So far it seems to only focus on diffuse GI
The lack of consistent radiance is due to spatial parallax aliasing--something I am addressing in the next post.
This isn't a physics implementation, so a focus on things like multibounce, materials and wavelengths aren't a part, but can be implemented. The goal of these articles is to familiarize people with the core concept of RC and that's mostly it.
This is amazing, and very well explained, thank you! I have one question - does this algorithm take view-dependent information into account to also do specular GI? So far it seems to only focus on diffuse GI
The method is a form of discrete path tracing. If you want specular reflections or anything else you need to add that yourself.
Can RC handle it? Yes, but that's inherently separate from the algorithm and up to the developer to add themselves.
I noticed very obvious interference patterns, but it seems that the wavelength of light is not reflected in the algorithm. Why is this?
The lack of consistent radiance is due to spatial parallax aliasing--something I am addressing in the next post.
This isn't a physics implementation, so a focus on things like multibounce, materials and wavelengths aren't a part, but can be implemented. The goal of these articles is to familiarize people with the core concept of RC and that's mostly it.
This is very cool! Looking forward to part 2
Its coming! The optimizations are going to provide a lot to RC, but they're also slightly complicated to implement--easy to explain.
Part 2 coming on Friday/Saturday 7/12-13/2024.