Color saturation - how to increase safely

[starman99]

Hi - Newbie here, trying to wrap my head around PI.  Amazing program - lots to learn!

I have a decent set of RGB data of M83 galaxy, each channel with 75-90 minutes of data. This is with a 6" refractor at f/7, mono CCD camera.  Each set of files was calibrated, aligned, and integrated in PI - no problems yet.  I performed DBE on each stack separately and then did LinearFit before integrating them using ChannelCombination.

The problem is that I just can't seem to get decent color, no matter how I try to boost saturation.  The channels all seem pretty even so the result is white.  I'm attaching a screen shot so you get the idea.

Is this normal, or have I made a mistake somewhere?

Any help appreciated!

[gvanhau]

Hello
I think you don't have to "linear fit" before combination. (I've never done it this way).
Just combine after DBE or jou can also combine and then DBE.
Next you have to do color calibration...

Regards
Geert

[RickS]

Have you tried boosting saturation using a clipped luminance mask?  Something like:

• extract luminance from your colour image (ideally set a 1:1:1 RGB working space first but it's not critical)
• use HistogramTransform to black clip and stretch the mask
• apply the mask to your colour image
• use CurvesTransform to boost saturation by pulling up the Saturation curve (use a real time preview to judge the strength of the effect)

You'll probably find there is colour there.  It's not unusual for the colour to need a lift.

Cheers,
Rick.

[starman99]

OK, so I'm able now to get some color by increasing saturation while linear (with a mask of course  ).  Then once the image is stretched I  increased saturation again in small steps.  I'm finding I have to treat the galaxy separately from the stars.  The color is extremely noisy and very pixelated, so I'm trying a MLT blur on wavelets 1-4 set to just the chrominance channel.  Is there a tutorial on this that you guys can recommend? Or what is 'best practice' as far as blurring the chrominance?

[RickS]

Or what is 'best practice' as far as blurring the chrominance?

Don't know if it is best practice but I still use ACDNR on non-linear data for chrominance noise reduction.

[Alejandro Tombolini]

I also like ACDNR for noise reduction in Chrominance on non linear images. Some processing example using it here, here, here, here, here, here, here, here, here, and here.

Saludos, Alejandro.