Processing
Use stacking software and a ‘bushfire’ colour palette to enrich detail
My winning entry in the ‘Stars and Nebulae’ category at the 2020 Astronomer Photographer of the Year competition is entitled ‘Cosmic Inferno’. It features the emission nebula NGC 3576, one the most intriguing nebulae in the southern skies. It has striking features, with great loops of ionised gas that stretch over 100 lightyears into the vacuum of space. Indeed, astronomers have identified polycyclic aromatic hydrocarbons within NGC 3576, hydrocarbons being an essential component for the formation of the earliest life on Earth. And 3.7 billion years after life first appeared on our planet, the bushfires of 2019–20 would raze the ecosystem across vast swathes of the Australian landscape of life – something alluded to in the title I gave to the image.
My original intention was to take a colour image of NGC 3576 using my Alluna RC16 telescope fitted with a monochrome 4k x 4k SBIG CCD camera, a combination giving excellent image resolution and depth. A colour image requires red, green and blue filtered exposures, but I also like to add hydrogen-alpha (Ha) data to enhance the structures of faint nebulae. The bushfires that engulfed much of southeastern Australia at the time had a significant effect on my attempts to capture any image data. Time after time, smoke from these fires made imaging impossible, shrouding not just my small observatory, but also thousands of square kilometres around it; indeed, NASA satellite data later showed the aerosols completed an entire circuit of Earth.
With attempts to capture colour data thwarted, my attention was focused on the four hours of Ha exposures I had managed to acquire. Mid-evening news reports of the fires showed vast arcing and turbulent fireballs – structures that were uncannily similar to my Ha data. In this article I will explain my workflow and how my ‘Cosmic Inferno’ took shape.
To begin, I calibrated the RAW image
files (see top right image on the opposite page) with suitably matched dark and flat frames; this essential step removes thermal noise (unwanted artefacts), vignetting and the shadows from any dust motes. For image registration and stacking I use PixInsight software because of its sub-pixel registration accuracy and noise-reducing algorithms.
Getting started
To start the registration process in PixInsight, click on ‘Select > Process > ImageRegistration > StarAligment’ to bring up the ‘StarAlignment’ window (see Screenshot 1, above) and then click ‘File’ in the top right drop-down option and choose a reference image. Next, click on the ‘Add Files’ button to add your sub-exposure files for registration, and then click on the folder icon to the right of the ‘Output directory’ and select a folder to store the registered images. Now, click on the ‘Apply Global’ icon and you’re done (see Screenshot 1). Once registered, the image data needs stacking and this is done in the ‘ImageIntegration’ window (see Screenshot 2), which is found by clicking ‘Process > ImageIntegration > ImageIntegration’.
Click the ‘Add Files’ button to add previously registered image data. I use the ‘Median’ combine setting and ‘Winsorized Sigma Clipping’ to reduce noise. Clicking the ‘Apply Global’ button stacks the images. The combined image file can now be brightness-stretched to your liking; the key is to use subtle changes to bring out shadow details without burning out the highlights. To mimic a maelstrom of swirling flames I needed to remove the stars from the image data and show just the nebulosity.
To do this I select the ‘StarNet’ function in PixInsight, opened by clicking ‘Process > Object Recognition> Star Net ’( see Screenshot 3). Next, I click on the square ‘Apply’ button, saving the resulting star-less image file for further processing.
The ‘StarNet’ star-removal routine is not perfect and it leaves bright star halos and diffraction spikes in the image, but these can be removed in Adobe Photoshop by using ‘Patch’ or ‘Clone’ tools, as well as applying Photoshop’s noise-reduction filter (‘Filter > Noise > Reduce Noise’) to smooth out the nebulosity.
The last step in the process is to “colourise” the image, or map a colour palette to the grayscale image in Photoshop. A new layer can be added by selecting ‘Layer’ > ‘New Adjustment Layer’ > ‘Gradient Map’. By clicking on the gradient icon, you can select and define a range of colours to map brightness changes within the selected grayscale image (see Screenshot 4). By moving the adjustment sliders you can also vary the brightness levels with different colour transitions.
Despite plenty of tinkering with my grayscale image, I found that using only shades of red gave an unconvincing result. However, I discovered there was no better colour guide than in the actual news footage images of the bushfires. By using Photoshop’s colour selection tool, I sampled the purple-black, red, yellow and orange tones from these images and placed them into the gradient of the adjustment layer. The final result was perfect (see main picture, opposite), and gave something of an Earthly relevance to the vast cosmic structures far beyond our home planet.