There have been around 20 images of this nebula so far on Astrobin and the most prominent have been SHO dominated. The Ha is quite strong and the OIII very weak - it reminded me of the California nebula in terms of the processing challenge.

Rather than a SHO approach, I opted for an RGB image that is enhanced with continuum subtracted S, H, and O. There is some broadband dust in the area, so I did a little emission subtraction on the RGB to create an emissionless broadband as well. While I did use the emissionless data to pump up the dusty areas a bit, it has a really minimal impact in the final image. My goal with the RGB+SHO approach was to keep a more natural palette and to avoid having to rely on an extremely pushed OIII channel. Aside from gradients, the only thing the OIII provided was the bluish/magenta area at the core. With the F3 DeltaRho and a dark site I felt I had an opportunity to rely on the RGB data more heavily than others. As always, play to your strengths!

The process of creating this SHO enhanced RGB image was something like this:

1) Create the RGB master, remove gradients and color calibrate
- I also make a copy of the RGB to create the stars layer later on
- I remove the stars for the next steps
2) Create starless continuum subtracted H, S, and O masters
- I used NightPhoton's PhotometricContinuumSubtraction script
- A combination of B and G was used to subtract from the O master
3) To eliminate the backgrounds in these continuum subtracted masters I sampled the dimmest areas of the master and subtracted that value instead of subtracting the median.
- I find this more accurate when working with nebulae that have signal across the entire field
- In Pixelmath that is: RGB/K: ContSubtractedHa - (sampled ADU count / 65535)
4) Still in linear, I then added the "cleaned" narrowband masters to the RGB color channels
- R: RGB + CleanH + CleanS
- G: RGB + CleanO
- B: RGB + CleanO + CLeanH*.1
5) Because the narrowband masters have been cleaned of stars, continuum signal, and background, they can be added directly without screwing up your color balance.
- I keep iterating, maybe putting multipliers on the narrowband masters till I get a mix I like
6) Now that all of the data in in one image, I run BlurX and NoiseX
- Because this is starless, I manually enter the PSF that is measured from the RGB master
7) I do the initial stretch and convert to .TIFF for Photoshop
8) Create a PsuedoLuminance that will be applied in Photoshop
- I extracted the Luminance from the RGB master and subtracted its background same as I described in step 3
- Blend the data in Pixelmath : RGB/k: Ha + Cleaned SII + RGB Luminance extraction
- The OIII was so low in signal I left it out
9) Run BlurX, StarX and NoiseX on the PsuedoLuminance
10) Stretch and convert to .TIFF for Photoshop
11) I now go back to my RGB master copy and do a light stretch just for the stars
- Run BlurX "correct only"
- I prefer lightly stretched stars over BlurX reduced stars
- After stretching, run StarX with unscreen checked
- Save as a .TIFF for Photoshop

Obviously, there are a lot of nuances to those steps and much to be done in Photoshop. But you see the conceptual order of things. Let me know if you have questions about why or how I did a particular step.