Just another bad night with Moon. I stoped the run after some time because of clouds and fog. Target? GR Tau again. Following Google I discovered interesting and useful thing. Google archive (the link archive in search results) shows me text and images of papers (astronomy & astrophysics for example) which are not public yet and you have to login (and pay of course). Viva La Googl ;) Now I'm experimenting with dome flats and performed also a simple test of camera linearity. Having constant (?) light source (two light bulbs) I changed only exposure duration and estimated mean background value of the flat filed. I got this:

Gnuplot: mean background value on flat vs exposure time (seconds).

Here are differences between sky and dome flat fields in our filters. I made 30 individual flat exposures (with 20 individual darks), made mean of all images and subtract them. Here we go:

B filter (flat_dome-flat_sky) [16bit]
B filter (flat_dome-flat_sky) [32bit]

Maximal difference is about 120 in intenzity range (0-65000). The difference is probably caused by (except other things) Moon shining above the horizon...
grtauR/flatR_sky - grtauR/flatR_dome