I really took my time processing this image and I’ll attempt to be detailed in my approach.   I’ve been working on dynamic channel combinations which opens up a whole new world of potential as well as complexity.  Channels are individually stretched to taste prior to being used to boost a weighting determined by a specific channel and it’s own relative level of stretch.  In this case O3 is used as the reference channel while Ha and S2 are combined dynamically based on the local pixel strength of O3.  It’s fairly straight forward and you can learn the foundation over at The Coldest Nights website.  What is not straight forward is how to tailor the individual channel stretches to achieve your goals when the relative strength of the emissions from the target varies greatly and uniquely.   This requires extremely careful processing of channels and subsequent delicate handling of the combined data.  Brute force manipulation of the data is no longer necessary with this approach as it is entirely based on relative signal strength.  The downside to this approach is that you do lose a lot of details in the initial channel combination because everything is based primarily on O3, which is a weaker signal and with less structural detail than HA.   Fortunately, you can simply apply the already stretched HA channel as luminance to “pour” the details back in.  No further manipulation of HA is required as it already perfectly matches the data set.  In my usage it is better not to manipulate HA prior to using it as a lum, as you end up primarily contributing details only without the desaturated or washed-out side effect that luminance can so often have on a color image.

Here is one thing that I found remarkable about this approach.  The only mask I used in the image was a tiny photoshop created mask to selectively increase the saturation for the blue filament in IC405.  Otherwise, I didn’t use a single mask for anything!  Which really surprised me.   I’m not saying that you can get away with not using masks if you process with these techniques, but in this case I was able to.  Masks are wonderful tools but they also require careful usage to avoid artifacts or other undesirable effects.  Anytime the data and speak for itself it is just so much easier to achieve goals. 

Here was my workflow:

For each channel prior to combination:
-BlurX
-StarX
-DBE
-LinearFit (Using H as Reference)
-HT Stretch

For the Stars:
-Combined with a custom pixel math equation to make the stars more RGB like.
-Curves Saturation
-SCNR

For the RGB Image:
-Combined Channels using a Dynamic Pixel Math Expression
-Applied Ha as Luminance
-Curves to balance color
-Saturated the Filament (Only time a mask was used)
-Very light application of BlurX to sharpen the nebula
-NoiseX
-Exponential Transformation
-Very light LHT
-Curves Saturation
-PM to add the stars back in
-SCNR