M16 (Pillars of Creation) blended Ha + OIII

[lazjen]

I've managed to grab some OIII data over a couple of nights (6 x 30 min), some not exactly ideal that I hope to eventually ditched if I get more.

With this data though I've tried to make a blended Ha + OIII image.

I mapped as follows:

L: Ha
R: Ha
G: OIII
B: OIII

I've got some gray stars in the resulting image that I'd like to make white without wrecking anything else. I haven't worked out how to do that yet, so I'd like to hear of ways to do it, preferably in PixInsight. My StarMask attempts haven't been successful, but that could be that I'm not getting the right combination of settings to get it to work.

185698

Astrobin: http://astrob.in/197456/0/ - larger version

OIII only astrobin, with details: http://astrob.in/197381/0/
Ha only astrobin, with details: http://astrob.in/195475/0/

[RickS]

Looks good, Chris.

It can take quite a bit of tweaking to get a good star mask, but that would be a good option for fixing the gray stars. Another way to get nice white stars in NB is to use JP Metsavainio's tone mapping technique, or at least the important bits of it (process Luminance and colour separately with stars removed from the NB masters used for the colour combine.)

Cheers,
Rick.

[rustigsmed]

nice close up Chris, and thanks for posting the OIII data, looks like there is a fair bit there.

cheers

Russ

[lazjen]

Sometime I'll have another go at the star mask and try a few more variations to the settings to see if it can do what I want.

And I'll read up on the tone mapping technique too.

I managed to get some SII data last night, although I fear some of it is even more dodgy than the OIII (I'm having some guiding issues I need to sort out). If the data is ok, I might pause on this image and go for a full combine instead. I'm sure there will be a cloudy night sometime where I can revisit this... :)

[lazjen]

No worries Russ. I don't mind showing what data or steps I'm going through to get to the end product. Maybe it will help someone else in the future. I know I like seeing the details of the process when I see other images.

[Rex]

Nice Chris. Love the detail in the pillars. Out of the three my favourite would be the HA only version. looks stunning.

[lazjen]

Thanks Rex, and the Ha data is definitely the best data at this stage, so it's not surprising you would find that the best. From memory all 7 subs were good, while for the OIII, 50% are not ideal and were only used for this version to "make up the numbers". Still that's better than my SII I looked at earlier tonight - only 33% are ok.

[lazjen]

I did some more processing on the image tonight. I was trying to get a better Star Mask. Eventually I gave up up Star Mask process and tried a different approach. It was a mix of mostly Colour Extraction and MultiscaleMedianTransform to create masks. With the right options, I got pretty good Star Masks to "fix" a good number of the gray stars. I iterated 3 times through this process to pick up more stars. With more care and time, I think I could have "fixed" the stars better, however it was an interesting experience.

Anyway, I've uploaded a few more versions to astrobin:

http://astrob.in/197456/B/ - The original with saturation boost
http://astrob.in/197456/C/ - Mask applied and stars whitened
http://astrob.in/197456/D/ - Same as previous, but saturation boosted

Edit: Meant to add I looked at JP Metsavainio's tone mapping technique that Rick mentioned, but didn't progress to trying it out.