Best way to process these two Ha & Lum Horsehead stacks?

[billhinge]

Following on from an earlier post about processing out diffraction spikes I did an experiment in Ha but I'm not sure what to do next, anyone care to help?

(the naive plan of stacking the Ha and Lum didn't work as I expected, I can get rid of the ccd bloom in Luminosity but the diffraction spike remains slightly subdued as is the Ha detail)

I have Pixinsight plus Photoshop 3 available so I someone could advise could they advise on the method, I'm not an expert but I'm reasonably competent

My question is, what's the best way to process this, ideally I would like the detail of the Ha, the smoothness of the Lum , removal of the big diffraction spike above the horse head and the nebulosity around the bottom star.

Here are the 2 stacks, one in Luminosity (19 subs 75 min total ) and one in Ha (8 subs 60 min total )

thanks in advance, don't mind putting more detailed files in dropbox if it helps

[rfdesigner]

well I'm sure this isn't the best way to process.. but it is a way to process...

luminence to G & B. Ha to R

Ha G and B zero set to minimum pixel value in horse neb, to ensure it's good and black. (a bit of a thuggish thing to do, there must be more delicate ways to treat it)

G & B then multiplied so as to roughly match the Ha in the saturated stars.

combine them into one image.

Ideally I'd want to match as many stars (bright and dim) to make as many as possible white... with the clipped (but very nice) mono images it's a little hard to match the various brightnesses.

[billhinge]

Thanks , I should also have added that I will be planning on taking some RGB bin 2 colour subs in future (assuming thats a good idea?)

Alternatively I guess I could also make a feature of the big diffraction spike by making a mosaic with the flame nebula and Alnitak

[rfdesigner]

Propper RGB should help a lot.

Had another go.. this time I removed the stars from the luminance image, added them to the Ha image (so they'd end up being very similar brightness to the luminance image stars) then combined them into on RGB image. Also I balanced the dark regions based on the large expanse of dark in the bottom right, then boosted contrast accordingly.

anyway.. I think it looks much better now, others no doubt would do much better.

And my recepie for star removal in IRIS:

1. apply ring_median to image. (I used ring_median 4)

2. save as 'temp'

3. use 'Substitute' command between origional image and 'temp', tune threshold to optimise balance between star removal and protecting the nebula.

Derek