I use a ZWO ASI 1600 pro Mono CMOS camera for photometry this is a 12 bit sensor.
I’ve observed 7 exo-planets – as part of EXO-Clock without any problems – I tend to use darks that are the length of the exposure but the HOPS software will scale them anyway to create darks for the flats.
The sensor I’m using is very linear – as far as I can tell – this is looking at the peak pixel in a 100 pixel radius near the centre of the field.
I get similar results on star tests.
I’ve also noticed if you look at integrated counts – as photometry software does – and use the right apature then the integrated counts remain linear even when some of the pixels are saturated – however I’m told to stay within the linearity of the chip – so I do – I’m a beginner when it comes to photometry so perhaps someone can tell me why this is so…