Waiting for clear skies since mid-October last year, I recently played around with different stacking methods in Deep Sky Stacker (DSS). To evaluate all the different stacking methods, I created this nice overview using the available data of my M42 image.

Interestingly, the result is similar to what I got some years ago, when I did a similar experiment. At least for my equipment, the chosen method doesn't really seem to matter as I don't see pronounced differences. Maybe its worth to mention that I took 113 images of M42 back then (see the M42 image in my gallery) and only used the best 64 for this exercise here. So the quality among those 64 images has been good and consistent. Furthermore, all columns base on exactly the same "x" best raw files. The only difference is the stacking method.

The major point that is handled differently by each stacking method is the outlier detection & rejection (if any). But this doesn't really count here as all the raw files are of good and very consistent quality. So the only thing that is left to do for the corresponding algorithms is to calculate the median or average values for a number of up to 64 R, G & B values per pixel. And in case there are no major outliers, the median and average values will even be very similar or the same.

So my takeaway is to not overthink the stacking methods in case you have consistent, high quality subs. My recommendation is to go with the most advanced median method (median kappa sigma clipping in case of DSS). This will give good results and also removes all plane & satellite trails that might be present in some of your subs. So no need to scrap any of them, not even if you had been bothered by a bunch of Starlink satellites.

-----------------------

After stacking with 42 different settings in DSS (version 4.2.5), I did the following processing steps in Pixinsight for each image.

- Dynamic crop (700x700 pixel)

- STF AutoStretch (Boosted)

- Curves Transformation

- Histogram Transformation

- Saved as .jpg

Th original data was recorded in the following way (see also my M42 image).

Equipment

- Main scope: Bresser Messier 10" Newton (254/1270)

- Main camera: Nikon D3300 (unmodified)

- Guide scope: Orion AC 50/162 Deluxe Mini

- Guide camera: ZWO ASI 290MC

- Mount: Skywatcher EQ6-R

- Explore Scientific HR coma corrector

Acquisition

- 113 lights x 150 s = 4 h 42 min

- no dark, flat, bias frames

- no filters

- ISO 200

- Bortle 4