How different do conditions have to be to need to build new dark frames

The night time temperature was 5C/41F and my exposure was 3sec at ISO 200 when I took both 15 dark frames and quite a few light frames. However, I also took some photos at 5sec and 8sec and the temperature warmed up through the night to 9C/48F. At what point do I decide if I need to collect new dark frames?

It’s the temperature of the sensor you’re fighting, and the camera itself can generate a lot of heat as electricity courses through the circuits. I think that becomes significant with rapid-fire sequential exposures or significantly long exposures.

I’ve never done this, so take with a grain-o-salt: I think what would help the most is to take the dark frames right after the capture of the image sequence, when the sensor is soaked with the temperature inflicted by both the environment and its operation…

You might take the question to DPReview’s Astrophotography forum; a number of adepts hang out there, including a few professional astronomers.


Dark signal, as Glenn pointed out, is a function of sensor temperature. Astronomical cameras are cooled by liquid nitrogen and thus the temperature is always constant. Dark frames for these cameras are just much longer in exposure time than science frames and many of them are avaraged and scaled to the expsure time of the science frame (if correction is needed at all).

For consumer cameras the principle question is the following: Does the dark pattern stay constant with temperature and only scales with temperature or does the pattern change with temperature. Both is possible and I would first make tests to see which is the correct answer for your camera. But note that environmental temperature alone is not the only factor, due to the camera warming up during use.

The answer may be different for the overall pattern and the behaviour of hot pixels, however. I.e. the pattern may scale, hot pixels may not. But pixels permanently dead or hot have to be treated separately anyhow. I am not sure if commercial cameras suffer from this or not. I would test it with a series of long dark exposures with constant exposure time but varying ambient temperature (giving the camera enough time to adjust in temperature).


1 Like