Is there any quality advantage to doing dark frame subtraction in camera?
It seems to me that doing it manually in software can’t be any worse, and can be better by using a dark frame that is the mean or median of several dark frames to reduce random errors in a single dark frame.