This is the start of a write-up of my current PixInsight calibration workflow. I’ve spent quite a bit of time getting to understand how PixInsight calibration works, and this is what I’ve come up with after much experimentation and helpful insights from several people on the PixInsight forums. This workflow now produces results that are significantly better than I was getting with my previous tools (such as Maxim).

What we’re trying to do is to integrate sets of bias, dark and flat frames to create, respectively, master bias, dark and flat frames, and then use those master calibration frames to calibrate each of the light frames. Somewhere during that process we also need to calibrate the dark frames with the master bias frame, and calibrate the flat frames with the master bias and master dark frames. There are multiple different ways we can do this, depending on whether the calibration of the calibration frames happens before or after the integration step.

To illustrate these options, consider first what I’ve labelled as “Calibration Workflow I”. This is the simplest workflow, consisting of these steps:

1. Integrate the bias frames B_i to create the master bias frame B_m
2. Integrate the dark frames D_i to create the master dark frame D_m
3. Integrate the flat frames F_i to create the master flat frame F_m
4. Calibrate each of the light frames L_i with the masters B_m, D_m and F_m

In this case, the calibration of the calibration frames (dark with bias, flat with dark and bias) happens within the single calibration step. That is, we calibrate the darks by subtracting the master bias from the master dark, and we calibrate the master flat by subtracting the master bias and a scaled version of the master dark. The PixInsight calibration tool has checkboxes that allows selection of which of these post-integration calibration steps we want to use.

A more complex version would calibrate each of the frames in the dark and flat frame sets prior to integration. This version I’ve called “Calibration Workflow III” and consists of these steps:

1. Integrate the bias frames B_i to create the master bias frame B_m
2. Calibrate each of the dark frames D_i to create a set of calibrated dark frames D_c
3. Integrate the calibrated dark frames D_c to create a master dark frame D_m
4. Calibrate each of the flat frames F_i to create a set of calibrated flat frames F_c
5. Integrate the calibrated flat frames F_c  to create a master flat frame F_m
6. Calibrate each of the light frames L_i with the calibration masters B_m, D_m and F_m

In this case, we now have 3 calibration processes, but in each case we don’t need to do any calibration of the master calibration frames because that all gets done prior to integration.

So why would we want the more complex version? If the integration is an unweighted combine, the results are almost identical. However, by calibrating the individual frames, we separate out the component of the dark frames that is due to dark current and the component of the flat frames that is due to photon-dependent effects prior to the integration. By doing that, we can introduce a scaling during the integration process. This will be useful if there is a variation in temperature of the dark frames or a variation in brightness across the set of flat frames (which will generally be the case unless the flats are being captured with a flat box with a very stiffly regulated power supply). In the case that we are using a regulated cooler in the camera, we don’t need to worry about variation across the set of darks, but we might need to take account of variation within the set of flats. That leads us to an intermediate version that I’ve labelled “Calibration Workflow II”:

This includes the simplification that we don’t need to calibrate dark frames prior to integration, but we do calibrate flats prior to integration. In this way, the integration of flats (step 4) can scale the flats prior to combining. The steps are now:

1. Integrate the bias frames B_i to create the master bias frame B_m
2. Integrate the dark frames D_i to create a master dark frame D_m
3. Calibrate each of the flat frames F_i to create a set of calibrated flat frames F_c
4. Integrate the calibrated flat frames F_c  to create a master flat frame F_m
5. Calibrate each of the light frames L_i with the calibration masters B_m, D_m and F_m

That’s the basic outline. I’ll explain in detail how to carry out these steps with PixInsight in a follow-up post.