Purpose: A dark video is capturing all the electrons coming from the chip which are not due to visible or near IR light. It's meant to correct for what thermal noise has been layered on top of the actual signal from the sky. Therefore the dark video must be done as follows:
1. At the same Temperature as the camera was during the event
2. At the same exposure circumstances: That means you must use...:
-- same exposure setting (e.g. 1x, 2x, 4x, etc)
-- same Gain setting,
-- same
sharpness,
-- same
gamma, and
-- same for
any settings additionally applied to from the IOTA VC 2.4 software to maximize the visible image of the star.
Therefore, the best way to take dark video is to just get your event, and then cap the telescope and take another ~10-15s of video, w/o touching anything else. If you forget, you'll have to get your dark later under the same temperature (very important. Temperature very strongly affects dark frames. 5 C of temperature difference approximately doubles the thermal noise). And again, if you forgot to take your dark video and do it later instead, then you'll need to have the IOTA VC 2.4 brightness and contrast settings identical too. It's a good idea to take a photo of the recording screen right after the event, including the settings panel which will show the BRIGHTNESS and CONTRAST bars. On IOTA VC 2.4, you shouldn't be changing the other settings, I believe. They're not helpful for B/W video anyway.
Purpose: The flat field purpose is to determine the individual differences between the pixel sensitivities. And it is also to determine the way the optics may change the light level from the sky reaching the chip. Vignetting especially is possibly an issue. And it is also to correct for dust motes, which will show as little donuts on your video if you stretch the contrast and look carefully. It's a good idea to use a photographer's dust rubber bulb air blower to give a few good whifts while the Watec cap is off and the opening is pointed down, so gravity can cause dust to leave and drop from the window in front of the chip.
What we're correcting for is the RELATIVE difference in pixel sensitivities. Dust motes too. That means you should not use the dust bulb between the time of the event and the time of the flat field. If there was dust on the chip, you want that dust to still be there when you take your flat field video. You also want the Watec mounted on the 8SE just as it was during the event. The real hassle is that you can't have any stars on your flat field video. That means you can't do your flat field right after or right before the event. You'll have to do it in twilight on a flat sky. I've found that in daylight it's just too darn bright. Even at 1/100,000 of a second, the exposure is too bright and pixels are saturated. You could, in theory, just lower your gain setting, but I'm not really certain that changing the gain won't also change the relative brightness between pixels. It's not supposed to, but I worry anyway. You could also set up your scope in a dark bedroom and aim the scope at a clean white wall under even illumination. I was able to get a dark video in daylight by going inside my carport and aiming at a very out of focus very close piece of wall not lit enough to saturate pixels.
== ideally, don't change gain, but if you must, it might be better than not getting the flat at all
== Be careful you're looking at true flat field. It's fine on your monitor to be seeing some vignetting, but if you're seeing blurry smudges that you know are because you're e.g.using last night's dinner paper plate you fished out of the trash - that's not good enough!
== Use whatever short exposure you need in order to have the result look roughly medium gray
== Do NOT adjust gamma! As a rule, I leave Gamm=1 for any occultation that I can get away with. However, some recommend lowering Gamma from 1 to some smaller value in order to make dimmer things brighter vs the high value that you're looking at. I've not found adjusting gamma to be helpful, myself. And it complicates interpretting the signal if Gamma is not 1.0 since now pixel value isn't easily related to actual light flux.
=== Again, take flat video for about 10-15s should be plenty.