M33 - Triangulum Galaxy

08 September 2021

I already shot a bunch of galaxies:

But despite its size, I did never capture the Triangulum Galaxy. I guess that there are two main reasons for that: Without a goto-mount finding objects is a bit of a pain and the Andromeda Galaxy is way easier to find and afterward much easier to process. Well, that ends now and this time I took the long-overdue image of Messier 33.

The image itself was taken throughout two nights.

In the remainder of this blog entry, I want to present the processing steps for this particular image.

Stacking the data

Since I need to save my money to invest it into more gear (e.g. a Newtonian telescope), I am stuck with using (mostly) free software. This is not a bad thing by all means, as there is a plethora of incredible and powerful free and open-source software available.

For the stacking of my sub-exposures, I like to use DeepSkyStacker. It is reasonably simple to use and I can tweak many settings to get an optimal result.

180 sub-exposures of two imaging sessions stacked with DeepSkyStacker

When saving the image I make sure that it will be saved as a 32-bit tiff file.

Cleanup, color calibration and initial stretching

After the initial stacking is done I switch over to Siril which I like for two main reasons:

  • It contains an algorithm to reduce or eliminate gradients. This enabled me to abstain from using my light pollution filter which ultimately lead to better overall colors in my images.
  • It can perform a photometric color calibration which - when it works - is much easier than tuning the colors manually.

Last but not least I am using Siril to perform a first and slight stretch to my image.

In detail the steps I perform within Siril are:

  1. Remove gradient
  2. Color calibration (photometric if possible)
  3. Increase saturation
  4. Asinh transformation
  5. Histogram transformation
  6. Reduce green noise

Preprocessing with Siril

The result of this process will also be saved as a 32-bit tiff file.

Cropping and basic editing

As visible in the last image, my flat field images seem to be slightly off and there is a vignette visible. So the next step is to crop the image. For this I use Affinity Photo with is an adequate and quite affordable substitute for Adobe Photoshop.

After cropping the image to the center portion I applied some basic edits that start to bring out the galaxy:

  1. Curves adjustment
  2. Levels adjustment
  3. Noise Reduction
  4. Cleanup of a bad spot using the inpainting tool on a new layer

Basic edit with Affinity Photo
Adjustment Layers
Blend ranges for denoise-layer

As one can see there are not that many adjustments. Perhaps the most interesting layer is the denoise layer, which makes use of Affinity Photo's blend ranges feature. This results in a cleaner background while still preserving the details within the galaxy.

The result of these basic edits will be saved as a 16-bit uncompressed tiff file, which will be required as input for the next step.

Creating a starless version

The next step of my processing is to create a starless version of the image produced so far. This can be achieved using Starnet++. This is a small command-line tool that takes a 16-bit uncompressed tiff file and creates a starless version. Using it is as simple as typing

rgb_starnet++.exe Autosave.tif Autosave_s.tif 64

in the command line prompt of choice.

Starless version of the base image

The starless version of the image will serve two purposes:

  • First, I will use it to reduce the stars of the original image to ensure that the focus goes right to the main subject of the image.
  • Second, the starless version will be utilized to boost the colors of the galaxy and make it pop a little bit more.

As a result of this step, I will have two versions of the same image: The image containing the basic processing from the last step and a starless version of the very same image. The next step is to combine them to get the best possible image.

Final editing and composing

The first step I like to do when composing the final image is to create a star-reduced version of the image which will help draw the focus to the main subject, in this case, the Triangulum Galaxy. Since at this time I have two versions of M33 at my disposal, one with and another one without stars I like to take advantage of this and combine them.

Therefore I simply place the starless version on top of the version with stars and utilize the blend ranges to let only the brightest areas (e.g. stars) shine through the starless version.

Base image
Starless base image
Star-reduced image
Blend ranges for the starless layer

In the last screenshot, there is an additional pixel layer that is only there to fix some minor issues with sensor spots.

The next step is to create a layer that boosts the colors of the galaxy. To achieve this I start with the starless version of the image and apply various adjustments and filters until I have a clean background and a very saturated and bright galaxy. The noise reduction layer will again be applied only to the darker areas of the image using the blend ranges.

Color boost layer
Blend ranges for noise reduction filter

When the color boosting layer is ready it's time to combine the two images. This time I opted for combining them using a mask.

Original without color boost layer
With color boost layer
Mask for color boost layer

Last, but not least I added some final global adjustments to create an image I am happy with.

Final adjustments

Full Screen Image