Caldwell 4 - Iris Nebula
NGC7023 or Iris Nebula - Click here for full resolution
NGC7023 (R.A.: 21h 02m 03.30s, Dec: +68º 17’ 10.0”) is an open star cluster in the constellation Cepheus at a distance of 1,400 lightyear from Earth. It is mostly known from the surrounding reflection nebula LBN487, better known as the Iris Nebula, named after the similarly intense blue-coloured flower. At the heart of the nebula is a bright young star (SAO 19158), lighting up the surrounding gaseous and nebular material.
Sky-plots with a FoV of 50º (left) and 5º (right). Click to enlarge
Conditions
Images were taken in sessions during March 2020 as well as two years later, in March 2022. All images were taken from the backyard in Groningen, The Netherlands (53.18, 6.54). Moon was at around 27-44% illumination. This target is visible from this location all year around and crosses the Northern skies in the March periods, with altitudes between 30 and 50 degrees.
Visibility charts showing 22:00h altitude throughout the year (left) and throughout the session on March 07, 2022 (right).
Weather was generally good, with temperatures around freezing during the three sessions. For further details, see below table.
Capturing
The image was captured using the Takahashi TOA-130 in combination with the ASI1600MM-Pro camera. The area of dark nebulosity around the Iris Nebula is quite extended, but the reflection nebula itself is fairly small at around 18 arcmin in diameter. This combination of scope and camera has a perfect Field of View to capture a lot of detail. Further details of the setup used in the last session are as follows:
Telescope
Mount
Camera
Filters
Guiding
Accessoires
Software
Takahashi TOA-130 + FL67 flattener, Sesto Senso 2
10Micron GM1000HPS, Berlebach Planet
ZWO ASI1600MM Pro, cooled to -15 ºC
Astrodon 1.25” LRGB mounted, ZWO EFW 8-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
The image was captured using LRGB broadband filters. The frames taken in each session that made it to the final image are:
Image
In 2020 half of the frames were taken and, with current insights, with far from optimal exposure settings. The 300s exposures at gain 139 is far too much for this target. Not only the center star, but also many other stars are clipped in the highlights. The dataset was a bit forgotten, and when the weather was so good in February/March this year, it was a perfect opportunity to complete the set. The focus was on keeping everything as similar as possible to the existing set. With hindsight it might have been better if the exposure per frame had been reduced to get some more details in the highlights. On the positive side, with the long exposures, the dark nebulae probably came out a bit better.
After a little crop to straighten the edges, the final image has a resolution of 4500 x 3000 pixels, or 13.5 Megapixels. It covers a field of view of almost one degree horizontally.
Annotated image showing other deep sky objects, stars brighter than mag. 9 and the orientation.
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 weighting parameters.
The luminance was developed by removing the gradient and then performing a deconvolution using the EZ Decon script. In the past EZ Decon has been tried without much success, but the script is now in version 2.7 and turned out to work quite well. In the EZ Decon script, the stars are not affected much, but the structure in the nebulosity comes out with just a bit more detail. The script builds a star mask using Starnet++. That algorithm has now been updated to Starnet2, but EZ Decon does not leverage the increased performance of Starnet2 yet. So outside of the script, Starnet2 was used to create a star mask, which was then loaded into the script. The script allows any of the masks to be either generated by the script, or loaded in from an external file. In this particular case, Starnet2 picked up a few stars in the nebula that Starnet++ had missed. Noise was removed using the MMT-method, which was very effective.
Stretching was done with a completely new method, a script called the Generalised Hyperbolic Stretch (GHS). In regular stretching tools such as HistogramTransformation, the midpoint is moved, always creating the same type of hyperbolic curve. In the GHS script, there is much more control over the shape of the hyperbolic (so not hyperbolic anymore) curve. A symmetry point can be set at the brightness value where most contrast is required. Highlights can be better protected by flattening the right part of the curve. Contrast in dark areas can be maintained by flattening the curve between black point and symmetry point. There is a colour stretch option that much better maintains the colours in the image. Besides the new GHS technique, also familiar stretching techniques such as HistogramTransformation and ArcsinhStretch are available in the tool, both benefiting from some of the enhanced control as explained above.
From a practical point of view, the script is very easy to use. There is a constant real-time preview screen open that shows the effect of the selected stretching parameters. Also the zoom and pan function in the histogram view work very well to precisely place symmetry points, black points, etc. A logarithmic representation of the histogram further enhances the insight into the histogram and allows better fine-tuning of smaller adjustments.
The script has so many parameters to change that finding the right set of settings can be quite challenging. It is certainly a script that benefits from a lot of trial and error and taming it will probably give a much finer control over the stretching process than can typically be achieved using STF and HT tools.
For anyone wanting to try the script, it can be added to PixInsight by adding the following repository: https://www.ghsastro.co.uk/updates/ followed by running Resources/Updates/Check for updates. Try restarting PixInsight if it gives a problem. For information on the script itself, including several tutorial videos on how to use it, visit the website.
The Generalised Hyperbolic Stretch (GHS) script offers a wide range of controls to fine-tune the stretching process.
The exposures of 300s at gain 139 were too long. Many stars had clipped center points. That also resulted in the stars being a bit more bloated than necessary. Since EZ Decon did not address this, the MorphologicalTransformation tool was used to reduce star size overall. Contrast was enhanced using CurvesTransformation. In the corners there were some annoying dark edges, which looked like the flats had not fully calibrated out. This is a recurring issue with the filters used here and does not come out, even when flats are taken very carefully. The current theory is that it is due to some reflections somewhere in the filterwheel, most likely somewhere in the mounted filters themselves. A gentle crop took care of it in this image.
Each of the colour channels was treated with the usual steps of removing any gradient in the background, cropping to the size of the luminance image and noise reduction using the MMT method. The individual channels were combined to an RGB image. After color calibration of the image using PhotometricColorCalibration, still a bit of a green cast was present. A mild SCNR correction (amount 0.4) helped out. Also the RGB image was stretched using the GHS script, after which the luminance was added to give the image a lot more brightness and contrast. Combining the channels had let to the introduction of some noise, which was reduced using ACDNR, under protection of an inverted Luminance mask, similar to the one used in the MMT method.
The resulting image needed a bit more fine-tuning. The bright parts of the nebula could do with a bit more contrast and brightness. The dark nebulas would benefit from some extra contrast in the darker areas of the image. But the most sticky issue was a green cast in the background signal. This turned out to be very difficult to neutralise without affecting the other colours. A lot of back and forth between CurvesTransformation, GHS script, HistogramTransformation and ColorSaturation resulted in the final image. In the process, the core of the nebula was worked on under an elliptic mask prepared using the GAME script. One of the enhancements in the nebula was a bit more definition using the LocalHistogramEqualization.
The green in the background was removed as a consequence. One could argue that the overall colour balance of the image might be leaning a bit towards the blue.
All the fine-tuning resulted in some additional noise, which again was effectively removed using ACDNR.
The final image represents 17.5h of exposure on the Iris Nebula with data collected two years apart. It was a great experience to see that even after such a long time, when documented properly, datasets can be easily combined.
Processing workflow (click to enlarge)
This image has been published on Astrobin.
