Hi @mostapha and @C.Bech
That was so 2014 ! Actually the Four-Phase method is the way to go.
There is some amount of “magic” in Andy’s solution as it requires you to switch between probabilistic and deterministic sampling in the same workflow:
To be fair, his solution was far ahead of it’s time because this was long before the Four-Phase Method was “invented”. After understanding the Four-Phase Method, I had tried something a couple of years ago just to see if it could be made to work(The images in the bottom row are from the four-phase simulation):
Anyway, during last years workshop Taoning presented a short validation of the Four-Phase Method with regards to light-pipes.
The “facade” in the context of the Facade matrix is the light pipe. Depending on how comfortable you are with commandline Radiance, the simplest approach would be the F1 type Four-Phase Method (https://www.radiance-online.org/learning/tutorials/matrix-based-methods
, page 58). And you should be able to make it work on windows. At least computationally this method is not very intensive.
If you face any issues in setting up the simulation through commandline, let me know.