Adjusting pinkish color to red?

[magnusl]

Hi!

On this image, I feel that the color on Ha-intensive regions are too "pinkish" and not as deep red as I would like. Why is this so, and is there something I can do about it? I've red about the pinkish color on Ha in several places, not been able to understand why it is so, nor how to adjust it (if possible).

I've used a DSLR, a Canon 1000D, that is not astro modified. In all there are 163 subs of 300 secs here, taken with a CLS clip-in filter. All processing done in PI. Telescope is a C8 with a f6.3 reducer.

png and xisf-version of the image is here:

https://www.dropbox.com/sh/fs744ujp8oed79d/AACfnztnt-E98N8km8Tv2dUBa?dl=0

Grateful for any advice!

Magnus

[RickS]

Hi Magnus,

You'll often see that pink thing happening when Ha is added to Luminance as well as the Red channel.  It's a result of strong lum overpowering the colour.  A quick fix would be to make a mask for the red areas (ColorMask script) and then use CurvesTransformation to tweak the colour.  You can boost the red channel, reduce the green channel, boost saturation, reduce luminance, etc.  If you're feeling brave you can get dramatic effects with the CIE a* and b* channels.

Cheers,
Rick.

[magnusl]

Hi!

Thanks! In this case, I did NOT add Ha, that is partly why I'm a bit surprised.

I was guessing I could do something with the color channels in Curves, yes. Is there a guide somewhere to understand them better? I feel that they are not that intuitive.... I play around with the curves, with increasing, decreasing, makeing S-curves and the like, but I really feel like I'm on thin ice or very very deep water here ... Any resources for understanding the logic here would be warmly appreciated.

Best,

Magnus

[aworonow]

Mask revealing just pink, then reduce G and B. see attached

Alex

[magnusl]

Hi!

Great!! But how do you mean, mask revealing just pink? The color mask script, or how?

Magnus

[RickS]

Great!! But how do you mean, mask revealing just pink? The color mask script, or how?

ColorMask is how I'd do it, but I'm a little biased    There are other ways but I can't think of a simpler one.

Thanks! In this case, I did NOT add Ha, that is partly why I'm a bit surprised.

I was guessing I could do something with the color channels in Curves, yes. Is there a guide somewhere to understand them better? I feel that they are not that intuitive.... I play around with the curves, with increasing, decreasing, makeing S-curves and the like, but I really feel like I'm on thin ice or very very deep water here ... Any resources for understanding the logic here would be warmly appreciated.

Adding Ha is often the culprit but any time you stretch colour data hard you'll see a bleaching effect.

You don't need advanced curves for this purpose.  Just pulling up or down the midpoint is enough to get started.  I don't know of a good reference offhand but I expect some of the PI tutorial sites would have useful content.

Cheers,
Rick.

[magnusl]

Hi!

Great, thanks!! Seems to make a difference

Magnus

[Warhen]

Hey Magnus, A few thoughts, and this is only my opinion. Pink is good, salmon isn't. It's not that I don't like salmon (if wild caught ;>), but salmon can be the result of diluted reds by too strong an L and/or Ha contribution to chrominance. Now, pink is good, because Ha/HII is not blood red, it's a bold pinkish color, and your result is not far off at all. Simply increasing saturation, or Alex's excellent idea will get you a bit more reddish pop! Having been around long enough from the era of film, all we got were blood reds from emulsion, but as we see in the modern age, hydrogen and helium emission, mixing with reflection nebulosity can yield gorgeous pink, red, and even magenta and purple. So, don't be afraid of the variety. But by all means, feel free to tweak to make things speak for you aesthetically.