I'm trying out OctoEverywhere's Gadget AI print failure detection.
It's been working surprisingly well, but I wondered if there was an ideal camera position, image format, or whatever to ensure it could always catch failures. Does anyone have any idea?