IC443 - Jellyfish Nebula
IC443 - Click here for full resolution
IC443 (R.A.: 06h 16m 36s, Dec: +22º 31.0’) is also known as the Jellyfish nebula. As Sh2-248, it is also known in the Sharpless 2 catalogue. It is located in the constellation Gemini. The object is a supernova remnant, about 5000 lightyears away. The supernova itself took probably place between 3,000 and 30,000 years ago. A neutron star, a bit off-center in the nebula is probably originating from that same supernova, where the stellar core collapsed into this super high density star. Using a wide-field telescope with a full-frame camera allowed capturing in the same image the hydrogen-rich emission nebula known as IC444, or Sh2-249.
Sky-plots with a FoV of 50º (left) and 5º (right). Click to enlarge
Conditions
Images were taken on three consecutive nights, starting on 26 February 2022 from the backyard in Groningen, The Netherlands (53.18, 6.54). Moon was absent. Best time to observe this target is between November and February. Most of the images were taken at altitudes between 30 and 60 degrees.
Visibility charts showing 22:00h altitude throughout the year (left) and throughout the session on 27 February, 2022 (right).
Weather conditions at each of the sessions were as follows:
Capturing
Telescope
Mount
Camera
Filters
Guiding
Accessoires
Software
Takahashi FSQ-106, Sesto Senso 2
10Micron GM1000HPS, Berlebach Planet
ZWO ASI6200MM Pro, cooled to -15 ºC
Chroma 2” LRGB unmounted, ZWO EFW 7-position
Unguided
MacMini 2018 (MacOS 10.14.6), Pegasus Ultimate Powerbox v2, Flip Flat
KStars/Ekos 3.5.6, INDI Library 1.9.3, Mountwizzard4 2.1.2, SkySafari 6.8.2, openweathermap.org, PixInsight 1.8.9
Frames
Exposure
Geometry
Annotated image showing the deep sky objects, and some of the brighter stars
Processing
All frames were calibrated with Bias (100), Dark (50) and Flat (25) frames and registered using the WeightedBatchPreprocessing script. Image frames were normalized and scaled using the NormalizeScaleGradient script and integrated using NSG parameters.
The narrowband image was created using the Hubble-palette, mapping SII to the Red, H-alpha to the Green and OIII to the blue channel. The resulting image had quite a bit of magenta in the background. That was reduced quite a bit using BackgroundNeutralization, and at a later step using the DynamicBackgroundExtraction tool.
For noise reduction, a new tool was applied, called NoiseXTerminator. This script is developed by Russell Croman. It is based on an AI algorithm to detect noise. In contrast to other AI-based noise reduction techniques, such as Topaz-denoise, this script is written with astrophotography in mind. Therefore it is much better able to keep star and nebula structures intact while reducing noise. The result is a much less ‘plastic’ look that Topaz denoise can suffer from quite badly. The tool is super easy to use, with essentially two sliders to experiment with. The first one determines the strength of the effect. The default value of 0.9 is a little too strong perhaps, but 0.8 or 0.85 gives a very pleasant result. The second slider determines the detail applied after the noise reduction, essentially an unsharp mask algorithm. The default is 0.25, but also here, increasing to for example 0.3 gives just a little bit more structure, making it look a bit more natural. Of all the noise reduction methods used so far, this is by far the easiest, and with very good results.
The effect of NoiseXTerminator noise reduction, with a before (left) and after (right) comparison.
After stretching the image, it was time to work on the colors and that could best be done on a starless image. StarNet2 is a much improved version of the original StarNet+ and produces a starless images with hardly any artefacts. In earlier Hubble palette images multiple iterations of applying selective color masks and changing individual color channels. This time, a more straightforward method was used, using Curves Transformation. The Hue panel allows for fairly precise color changes, applied to very specific color ranges. This tool was used to bring out the blue a bit better in the IC444 region, and to bring a bit more amber tones into the yellow areas. Perhaps color masks give you even more flexibility, but the current method worked remarkably easy. There were still some magenta tones left on the outer darker edges, which were removed with a mask specific for these dark areas.
IC443 and IC444 fit very well within the field of view of this telescope/camera combination. However, they were a bit out of center, so the whole image was reframed a bit using the DynamicCrop tool. Finishing touches were made using Curves Transformation and Local Histogram Enhancement tools.
Now it was time to develop the stars. This was done using a fairly straightforward process, by combining the Red, Green and Blue channels, remove a modest background gradient using DynamicBackgroundExtraction and applying color calibration. Unfortunately the broadband images were made with the target quite low in the sky. Due to refraction effects, not all the images were as pinpoint sharp as was hoped for. When combining RGB, this meant that not all stars had very uniform colors . In the stretching process, color was secured by using ArcsinhStretch, but overall colour was not boosted too much, to not introduce the artefacts into the final image too much.
Also for the star-image, NoiseXTerminator was used for noise reduction, after which the background was darkened a bit. The stars were reduced in size, using another new script, written by Bill Blanshan. This video shows a detailed explanation of the script, what it does, and how it can be applied. The scripts themselves can be obtained from Bill’s Google Drive folder. In essence the script applies different stretching to the stars. It requires a Starless image as a support image and StarNet2 was used to get that. There are three scripts, referred to as ‘transfer’ method, ‘halo’ method and ‘star’ method. For this image the star method was applied using 1 iteration and mode level 3 (smooth reductions).
With both the starless SHO image and the RGB star image ready, they could be combined into a final image. This was achieved using PixelMath. If both images originate from the same source, just adding the images together usually gives good results. In this case however, the images came from two distinctly different sources, and rather than just adding the images together, for each pixel, the maximum of the two images was used. The PixelMath function for this is ‘Max(‘SHO-image’,’RGB-star-image’)'. The resulting image was close to final. Boosting the contrast and reducing the saturation were some of the last tweaks done with CurvesTransformation.
Processing workflow (click to enlarge)
This image has been published on Astrobin.
