On old-fashioned film photos, the brighter a star on an exposure of a given length of time, the greater the apparent diameter of the image on the film. This explains why many astronomical photos have "big" and "small" stars on them, although in reality the diameter of the object in the sky is essentially zero, or a point.

A skilled astronomer can probably do a good job estimating the magnitude of any object in such an exposure if he knew the magnitude of one of the objects on the plate.

As far as digital photography goes, I'm stumped. But I wonder if there's some way to use a computer to take the raw data from the CCD and do something similar in comparing magnitudes. Just guessing. Not my field.