Building a Dark Library

[Somerled7]

I've only tried a couple of imaging sessions so far, but already I'm thinking that taking dark frames at the end of each imaging session is eating into useful imaging time (or sleeping time!!) especially in the summer when nights are short.  So, what if I put my camera in a fridge, then take a series of dark frames as the camera warms up?  I'm using a Canon EOS80D and Backyard EOS which records the sensor temperature, so I could build up a library of different exposure/ISO/temperature combinations.  I assume that as the frames are supposed to be dark, the camera does not have to be connected to any lens/telescope, so taking them with the body cap on is OK?  If this is a viable approach, two questions:-

1) Would there be an 'expiry date' on the Dark Frames - I can imagine that pixels may fail from time to time giving a different hot/dead pixel array.

2) What would be a sensible temperature range that could be used when processing an image? 

Any advice welcome. 

[CloudMagnet]

Dark frames can last for months but will lose the effectiveness the longer time goes on. Also, it wont be a good idea to take darks as the camera is heating up. This defeats the purpose as darks need to match temperature with themselves and also the temperature of your imaging night. I never though it would be a good idea to put a camera in a fridge to be honest, but if you are going down that route, then you would need to take the darks inside the fridge as well.

The temperature range you take the darks over really depends on how close you want them to match with the outdoor conditions. Bear in mind that your camera might drop 5 degrees when you are just out imaging normally over a few hours. This means that the thermal noise in the first picture will be very different from the noise in the last picture later in the night. Because of this, I dont think there is much to gain by having darks for every single degree change in temp. Really that precision matters more with a cooled camera where temperature control is possible.

[Somerled7]

Well I've had a quick go and answered my own question - it doesn't work!!  It's surprising how quickly the camera warms up - 1st image at +5degC, 2nd at +9degC, so there is probably quite a temperature difference in each frame.  So, if the important thing is to get stable ambient conditions, I have a conservatory which gets quite cold at night.  I can safely leave the camera to run over-night without having to monitor it.  It won't quite match the outside temperature, but a cold night in the conservatory might match a warmer night outside.  Incidentally, I don't see a problem with putting the camera in a fridge - I've shot landscape photographs down to -17degC, which is well below my fridge temperature!    

[CloudMagnet]

4 minutes ago, Somerled7 said:

  Incidentally, I don't see a problem with putting the camera in a fridge - I've shot landscape photographs down to -17degC, which is well below my fridge temperature!  

Probably not, but its just never felt right to me!

Also worth making a note of what temperatures the picutes have been taken at (if it isnt recorded) just so you dont lose track of it.

[Dinglem]

I wrapped my camera in tinfoil with my intervalometer and put it in the fridge until the battery ran out a few times. It gave me a range of temperatures that I sorted into folders 8-12C went into 10C folder, 13-17C into 15C etc. then I stacked what I had and made masters for each temp range.

[rickwayne]

I know some DSLR astrophotographers who've tried it both ways and find that they're not really getting visible noise reduction from dark frames. Pentax K DSLRs are notorious for this, they just have low noise to begin with.

That's another experiment you could try, actually -- how many dark frames does it take to actually improve your images? You can run it with an existing set; say you had 16 dark frames, try it with all of them, 8, 4, 2, 1, and none. Recall that dark frames are intended to subtract the thermal noise signal, so in principle you just need one. Of course there's a noise component to the dark frames themselves so in practice you need to average, just as with light frames. But the tradeoff is integration time versus less noise in the thermal-signal reduction frames; at some point, your overall quality will be better if you spent the in-the-field dark-frame time just on shooting more lights.

I live on Easy Street with a cooled astro cam, so I can set a temp, tell it "shoot 100 darks at these combinations of gain, temperature, and time" and walk away. Two days later I have 20 GB of darks to integrate but at the end I have butter-smooth dark frames. Of course since I'm shooting with a 183 I absolutely HAVE to use dark frames, or all my images would display this lovely pattern:


 

 

Edited May 17, 2020 by rickwayne

[jiberjaber]

With my DSLR I don't have any temperature measurement (Nikon) so at the end of a session, I just wrapped it in a drybag outside with the intervalometer and cover on the camera and left it to run through taking darks. This gave me a good number of darks at circa 6degrees of 10min over an hour and a bit.  I just use them within pixinsight for any sub up to 10mins (pixinsight will 'optimize' the dark to match the sub duration).