The assumption that they are similarly bright is questionable. First, not all of the light from the brightest stars is counted because parts of the CCD that collects the light become saturated. But more importantly, the images you see are heavily processed. Usually the interesting stuff in an astrophoto is dimmer than the brightest stars, and so the image is "stretched." The pixels on the bright end are all crowded together at similar brightnesses, and on the dim end, pixels that formerly had similar brightnesses end up with dramatically different brightnesses. Check out this link for a description of the processing that goes on even just to show you a preview image of "raw" Hubble data: http://archive.stsci.edu/hst/preview...orrection.html
. Go here http://archive.stsci.edu/cgi-bin/hst...name=J8JY04060
and try viewing the image of V838 Mon using several levels of Pixel Correction (GIF Only), starting at "None." Using the None setting, you won't even see V838 Mon, one of the most fascinating objects in Hubble History.