Quandary about calibrating lights

[Diane Miller]

I'm using a DSLR, which is low noise as DSLRs go:  either the Canon 7D2 or the Canon 1DX2.  When I get to the step for calibrating lights I find 3 scenarios given by trappedphotons: http://trappedphotons.com/blog/?p=693 about halfway down.

Jim Morse's cribsheet uses the master bias, master dark and master flat, but if I'm reading trappedphotons correctly, that scenario is for when you are scaling the master dark.  I don't know if my cameras provide sensor temperature data so I haven't tried to use dark scaling.  I just shoot the darks at as close to the same temp as I can.  (Temp will often drop 5-10 degrees during a run for me.)

So in that case of the darks, which of these three scenarios is the right one for me?

Thanks!

[pfile]

PI never relies on anything other than an analysis of the noise of the calibrated frame to scale the dark. there's a sticky here:

http://pixinsight.com/forum/index.php?topic=8839.0

that explains the whole thing - you don't need any temperature data (or time data) for PI to scale a dark.

the problem with DSLRs is that the camera's firmware is playing tricks behind the scenes to try to suppress dark current. this can lead to problems when scaling the dark as the CR2 files are never actually the raw data that the sensor collected. for this reason a lot of people don't try to scale DSLR darks and i've even heard that people have even stopped using darks with DSLR data, instead relying on dithering and lots of subexposures.

rob

[Diane Miller]

I apologize for not understanding this fully, but just want to use the correct workflow.  So with a DSLR, I haven't and won't try to scale darks, but instead will do my best to acquire them at the same temp as the lights. 

But at my level of understanding, I just need to know -- when I calibrate lights, should I use the scenario for:
only a master bias
a master bias and a master dark
a master bias, a master dark with scaling, and master flat

It sounds like the third scenario is not the right one for me.  Which of the other two is?  The second one, with the master bias and the master dark??

It should be obvious but I want to be sure.

Thanks for the indulgence!

[pfile]

i think using only a master bias would be a situation of last resort - you forgot to take darks or have no matching darks but want do at least do what calibration you can do.

all dark subs contain both the bias signal and the dark signal. so if you don't plan on scaling the darks, there's really no reason to use a master bias. just subtract an uncalibrated master dark from the lights and you're done.

so i'd add a 4th option: calibrate the lights with a master dark of matching temperature and duration, plus use a master flat. probably the best way to do this (if your flats are short-duration) is to first make bias-calibrated flat subs, then integrate them into a master flat. otherwise if the flats are longer than a few seconds, calibrate the flat subs with darks that match the flat duration. of course the master dark should be made from uncalibrated dark subs. then when you run IC on the lights, do not specify a master bias, specify the master dark, but leave "calibrate" and "optimize" unchecked, and specify the master flat, but leave "calibrate" unchecked.

rob

[Diane Miller]

OK -- thanks!