Narrowband image processing tutorial
Learning how to create a vibrant deep sky images by merging narrowband images can be a challenging task. However, in this tutorial, we will explore a highly effective approach using PixInSight and Photoshop, which provides greater image control compared to alternative methods.
The initial stage of this processing commences at the telescope, as no processing technique can replace good quality data. To obtain the best image quality, it is advisable to allocate more exposure time to the fainter SII and OIII channels and less to the H-alpha channel as it is likely the brightest one. By doing so, the final image's overall noise will be reduced as the signal to noise ratio is more even across the channels.
In addition we will need some short exposures in RGB to add stars to the nebulas later. Around 20 minutes per channel is adequate to bring out colours of the stars. Make sure your sub-exposure times are short enough that you wont blow out the brightest stars in the field-of-view.
After calibrating and stacking all the channels separately in PixInSight using WeightedBatchPreProcessing script, it is recommended to blink all master frames to make sure they are aligned with each other. Then run your preferred operations that still keep the images linear such as DynamicBackgroundExtraction and BlurXTerminator for each channel. Emphasis on keeping the backgrounds of each channel even to avoid gradients and making sure your stacks are sharpened equally for RGB channels to avoid colour fringing later on.
Then combine your RGB channels using LRGBCombination tool and again use ArcsinhStretch to bring out the stars. You can add more highlights to brighter stars using HistogramTransformation tool. For the RGB image try not to over-stretch the image, the nebulosity in it should be just barely visible. We only want the stars from it, so there is no benefit from over-stretching the image. Save your RGB image as 16bit .tiff -format and open it along with SHO channels in Photoshop. You can now close PixInSight.
In Photoshop select your SII channel image and press Ctrl + A to select the entire image. Then press Ctlr + C to copy the image dimensions to your clipboard. Press Ctrl + N to create new image and use the copied pixel dimensions to create a blank canvas. This will be your final image file, now copy your SII channel to the red channel, Ha to green channel and OIII to blue channel. Toggle all channels visible and you’re left with starless SHO image. The cool thing about this is that now you can adjust each channel separately, run noise removal or use curves tool to fine tune your images tones and colours as desired. You can also use selective colour to adjust the palette as you wish. After you’re satisfied with the starless image drag your RGB image on top of the starless image as a layer and select the blend mode to screen. Your image can appear slightly washed at this point but you can use curves on the starfield to darken the background, effectively creating a “starmask” that just clips background value to 0 - displaying only the starless SHO image with RGB starfield on top of it. If stars appear to lack colour use saturation tool to bring out colour to your liking.
Benefits of this processing workflow is that you’re able to control stars and each narrowband channel separately to the end of the image processing. Here are some of the images that are processed with this technique by astrophotographer Ville Puoskari. The results are amazing!