I am looking at different Lab/Lch colorspaces (OKLab/SR2Lab/...) - One issue they share is that the RGB color you get for very dark L-values becomes negative. Seems non-sensible to me.

Comments