Have a look at:
AstroPhoto Insight Feb / Mar 2008 which you can find here:
http://www.astrophotoinsight.com
Quote follows:
"One way to test the quality of the flat field frame light source is to create a master flat field frame and create a second master flat field frame with the telescope or light
source rotated 90 degrees. The first master flat field frame is used as the test image
which is calibrated by the second rotated master flat field.
A good flat field calibrated test image is a uniform noisy image with no dark or light areas. A good test image histogram is a normal probability distribu
tion. The normal probability distribution is caused by at least three types of noise that are in the master flat field frame and the test image. The noise sources are photon noise, dark current noise and readout noise. A nonuniform light source causes dark or light areas in the test image."