I saw the public data pool of Richard Muhlack's data, decided to have a go at processing his data, so a big thanks for sharing this brilliant data.

Processing Summary

First I stretch each channel with FITS Liberator, I use arcsinh(x) and click auto scaling a few times until it looks right,

I put the black level a little left of the end of the curve, to not clip anything and the white level all the way right.

I then import all 3 tif files into 1 colour image in Photoshop (easiest way is to open one and go to image-mode-rgb and paste each tif in the channels)

Then I add a curves adjustment layer and change the R,G and B seperately until the background has no or very little tint and the stars are distinctly yellow and blue (you may need to use a vibrance adjustment layer on high saturation to help with this). You can always use an image online to compare the colour to.

Most of the time you may need to use the HLVG plugin for any green tint that is left after this, so no data is lost, you should make a copy of the flattened/merged image (you can undo the merge after you have copied it) and paste this layer ontop. Then set the blending mode of this layer to colour, and run HLVG on strong. This works well for everything except nebulas with green parts (you can always use a mask to exclude the nebula.

Once the colour is correct, you can remove any remaining colour in the background by creating a vibrance adjustment layer and setting the saturation to -100. Click on the mask, zoom in to an area with few stars, then go to select-colour range, set the sample size at the top to 101x101 average, then click an area with no stars and change the fuzziness to around 30.

You can then use a normal curves adjustment layer to stretch the image a little, but making sure not to clip the background or the highlights, I normally use a mask on the stars and the brighter regions, although make sure not to overdo this (normally you only have to do it on the stars).

I occasionally use the selective colour adjustment layer with a mask to get the colour right and to remove any remaining tint.

I then normally create a copy of the merged layer like before and run a smart sharpen on it (but only on the object itself, not the background/stars), I normally sharpen the whole image and use a mask to only sharpen the object, you can create two different layers with different sharpening radiuses and use masks to sharpen large features with the large radius and detail with the small radius. It is normally best to set the sharpened layers blending mode to luminance only, as it adds a lot of colour noise. Photoshop CC also has an updated version of smart sharpen, which has a lot less noise.

I also create another copy of the layer which I denoise with XiDenoiser, the best mode is NLM with learn radius on max and compare radius on lowest, enhanced enabled, spatial sigma and conservative mode depends on the image. It is best to use a mask to only apply to the background (a lot of the time you have to exclude smaller stars too). It is quite slow but gives impressive results.

Final tweaks I use vibrance and selective colour adjustment layers to finish off the colour (use another image online to see if they are roughly correct) and a curves adjustment layer with a mask to bring out some of the fainter parts of the object. I also occasionally run a unsharp mask with a medium radius (I think I used 4/5px for this image) on a low percentage to increase the contrast a little (use a mask again to only apply to object).

The main things you have to avoid are; clipping background or highlights, oversharpening or sharpening parts other than the object, too much denoising that the image looks 'gloupy', colour tint on object or background, too much colour saturation (you can zoom in and see certain colours clipping) and too much stretching that ruins the contrast in the image (unless the purpose is to reveal faint objects).