Bicolor processing tips

[astrovienna]

Any basic advice on how to get good color in a bicolor image?  I think mine are a bit washed-out and pastel-like compared to some, and I've never been able to get the blue tones down right.  This is my most recent bicolor, of Abell 21:

http://www.pbase.com/image/167423502

I used a 50/50 Ha/OIII blend as luminance, and color was Ha/OIII/0.85*Ha+0.15*OIII in PixelMath.  I applied a 50% neutral green SCNR to shift a bit of the green to blue.

I'd very much like it if someone else tried their hand at the data.  The integrated Ha and OIII masters are here:

https://www.dropbox.com/s/b5li3yetyomgimk/Abell%2021%20NB%20Integrated.zip?dl=0

Sorry about the registration errors.  I had to clean that up with an artificial flat in PS.  I shot this over two years, with a minor equipment change in the middle.

Kevin

[pfile]

one thing you can try is to isolate the blue colors with ColorMask (either a Blue mask or Cyan or perhaps a combination of the two) and then use curves to adjust the blue. i'd smooth the mask a little bit to make sure you don't get super sharp transitions, and also if you boost some of the channels i'd pull the Lum down also just to keep the brightness similar before and after the curve application.

rob

[RickS]

I often do a simple HOO combination and then beat the crap out of it with ColorMask and CurvesTransformation   Removing the stars often simplifies colour processing as well.

Cheers,
Rick.

[Geoff]

H alpha and H beta are emitted from the same regions but Hb is in the blue and 25-30% weaker than Ha, so putting 25-30% of Ha into the blue channel (as well as the OIII) will do a good imitation of the Hb radiation and should boost the blue.

[RickS]

Hi Kevin,

Here's a very quick & dirty process that I did on my laptop while waiting for a few Drizzle Integrations to finish on my processing desktop.  Is this more like what you're looking for?  If so, I can describe what I did and perhaps it will give you some ideas...

Cheers,
Rick.

[astrovienna]

Hi Rick - Yes, that's more the color outcome I had in mind.  If you can give a few details, that would be great.  And I definitely need to spend some more time with your Color Mask script.

Kevin

Edit: I should add that I got a result fairly similar to this when I did the NB color merge before stretching, but I found it much harder to bring up the faint nebulosity this way.  In the image I linked to above, I stretched Ha and OIII separately, then did the color merge and used a 50/50 blend of the stretched NB images for the luminance.  Anybody have any thoughts on whether the merge should be done pre- or post-stretching?

[msmythers]

Kevin

One tool I like for bringing out faint stuff is the ExponentialTransformation tool. Here is a copy of your image(jpg) from the web. I just used an inverted L mask and the tool. Tweaking your masking can really help with this tool.


Mike

[astrovienna]

Thanks at lot, Mike.  I had been looking for a better way to bring out the surrounding nebulosity, and that looks like it should do the trick.  I've never used ET before.

Kevin

[RickS]

Hi Rick - Yes, that's more the color outcome I had in mind.  If you can give a few details, that would be great.  And I definitely need to spend some more time with your Color Mask script.

Kevin

Edit: I should add that I got a result fairly similar to this when I did the NB color merge before stretching, but I found it much harder to bring up the faint nebulosity this way.  In the image I linked to above, I stretched Ha and OIII separately, then did the color merge and used a 50/50 blend of the stretched NB images for the luminance.  Anybody have any thoughts on whether the merge should be done pre- or post-stretching?

Hi Kevin,

I didn't use ColorMask this time but I could have

Colour was: LinearFit Oiii to Ha, simple HOO combine, HT stretch, ACDNR, HT tweak, Hue curve pushing cyan to blue, clipped lightness mask: CT Saturation+, ColorSaturation curve to push blues, combine with separately processed Lum (noise weighted integration of Ha and Oiii with no rejection; MLT linear mask noise reduction; HT stretch.)  I wanted a bit more blue in the final result so I stretched and black clipped the Oiii, applied it as a mask and then used CT to boost the blue channel.

I didn't push the Lum too far but could have stretched it a bit further without losing too much colour, I think.  Mike's suggestion of ET is a good one.

Cheers,
Rick.