Dark current subtraction¶
Note
It is possible that the dark current is negligible for your detector, and you can skip this step. Check the documentation of your instrument.
Dark current subtraction is not performed in any of the tutorial datasets.
The dark current estimation is implemented differently than the rest of the calibration steps, and does not need to be called explicitly.
Instead, if you have dark frames for your flat frames, science object frames and/or standard star frames, you have to create a directory named “dark” in the raw data directories and place the dark frames there.
For example, if your raw science frame directory is called:
/home/Documents/science
Then you have to create a directory:
/home/Documents/science/dark
And place the dark frames there.
This is only possible for the flat, science and standard star frames, as for the arc frames we are only interested in the position of the lines, and subtracting the dark current would not change it. Furthermore, most arc frames are exposed for a very short time, so the dark current is negligible. This is also mostly the case for the flat files, but the option is implemented if your detector has a significant dark current.
!IMPORTANT! The dark frames should have the same exposure time as the science, flat or standard star frames. This is also the reason why they have to be distributed in the same directories as the frames they are calibrating.
There are no parameters to set for the dark current estimation. The software takes a bias-subtracted median dark frame from the frames provided, and subtracts it from the corresponding frames. Technically, bias subtraction could be skipped for dark frames, and then the median dark could be treated as a combination of the dark current and the bias level, but we choose to subtract the bias separately, as we believe that the bias estimation from bias frames is more correct, as it is common that more bias frames are present than dark frames. Furthermore, treating bias and dark current separately allows us to invoke the overscan regions, allowing to correct for the bias drift between the individual frames (see bias documentation).
For users new to data reduction - short introduction to dark current¶
CCD detectors accumulate signal even when no light is present. This signal is called the dark current, and is caused by thermal excitation of the electrons in the detector. The dark current is usually very low, but can be significant for long exposure times, especially for non-cooled CCD detectors. The dark current is estimated by taking a series of dark frames - frames with no incoming light with the same exposure time as the observations. The dark current is then subtracted from the observations.
Bias subtraction ← Previous pipeline step
Next pipeline step → Combining arc frames