Using WO refractors (FLT91 and Redcat 51), ASI294MC Pro and Optolong L-eXtreme filter. I can't get color in the stars that should have color. For instance, the Seagull Nebula is described as having hot blue stars.
How do I bring out the color in stars using PI
Well, for starters try CC and then ASINHSTRETCH. You processed those stars to death.
andrea tasselli:
Well, for starters try CC and then ASINHSTRETCH. You processed those stars to death.
Great info, but please elaborate. CC = ChannelCombination/ColorCalibration/CosmeticCorrection?? Also, at which point in the process do I use these tools, i.e.: linear/stretched, pre/post StarX? DavesView=1st grader 🤣
Engaging
We are discussing colors here so it's ColorCalibration. Just after you stacked and flattened the image and before anything else. Let's see what comes out of it first and we'll deal with the fine points later…
Hmm.. Isn't the L-Extreme a narrowband filter? That would hurt your star colors. If you want true color stars, you might try shooting some color subs separately. Process this stack separately and remove the stars. Then you can screen them back into a starless version of your L-Extreme stack. Mike Cranfield has a PI script that's available to help do this and documentation can be found here: Scripts – Cosmic Photons
CS
--Steve
CS
--Steve
Helpful Insightful
You can get plenty of colors in stars even with that NB filter, more so if you want to bring out the OB giants. No need to faff around with shooting a different set of images.
andrea tasselli:
We are discussing colors here so it's ColorCalibration. Just after you stacked and flattened the image and before anything else. Let's see what comes out of it first and we'll deal with the fine points later...
Using those tools, I can now zoom in on the stars and see color pixels surrounding them far better than I could prior. I'll have to play with it and go to youtube school for CC. My processing skills stink, but I see improvement with that bit of help. Thanks!
Hmm.. Isn't the L-Extreme a narrowband filter? That would hurt your star colors. If you want true color stars, you might try shooting some color subs separately. Process this stack separately and remove the stars. Then you can screen them back into a starless version of your L-Extreme stack. Mike Cranfield has a PI script that's available to help do this and documentation can be found here: Scripts – Cosmic Photons
CS
--Steve
I wondered if that was an issue. I'll pursue that path also. 👍
andrea tasselli:
You can get plenty of colors in stars even with that NB filter, more so if you want to bring out the OB giants. No need to faff around with shooting a different set of images.
Critique time again. Check out the revised and point me in the next direction. I carried the processing all the way through using the two tools. I see some improvement in the image overall, including the stars. Although the star colors just don't pop.
Like @andrea tasselli says, you don't necessarily need to shoot separate RGB stars (albeit that does makes things easier in terms of getting naturally colored stars). What @starfield suggests is valid: separating the stars and processing them to enhance color, and screening them back on. Make sure you don't over-stretch your stars, that is one of the issues in both the original version.
I use the Curves Transformation tool several times on the stars-only image, but in very "soft" increments on the stars only image on the "c" component and satuation (S) component, until I can bring out sufficient color:
Here's an example of a narrowband only image, with reasonably colorful stars using this technique:
https://www.astrobin.com/full/wx5igj/0/
Edit: Also, you seem to have some green stars in your image. Try running SCNR on the stars.
I use the Curves Transformation tool several times on the stars-only image, but in very "soft" increments on the stars only image on the "c" component and satuation (S) component, until I can bring out sufficient color:
Here's an example of a narrowband only image, with reasonably colorful stars using this technique:
https://www.astrobin.com/full/wx5igj/0/
Edit: Also, you seem to have some green stars in your image. Try running SCNR on the stars.
Helpful Engaging Supportive
Can recommend Arcsinh stretch for star colours. I usually do it at the full power of 1000, but my images are very dark so i can get away with it. With images that are not so dark, or with images that have strong colours the full thousand can be too much. Try to set it to a value where nothing in the nebula saturates and see how the stars look at that point.
You could always use StarXterminator first and do extra Asinh stretching on just the stars image (and combine later) if you want to push those further.
You could always use StarXterminator first and do extra Asinh stretching on just the stars image (and combine later) if you want to push those further.
Helpful
Ani Shastry:
Like @andrea tasselli says, you don't necessarily need to shoot separate RGB stars (albeit that does makes things easier in terms of getting naturally colored stars). What @starfield suggests is valid: separating the stars and processing them to enhance color, and screening them back on. Make sure you don't over-stretch your stars, that is one of the issues in both the original version.
I use the Curves Transformation tool several times on the stars-only image, but in very "soft" increments on the stars only image on the "c" component and satuation (S) component, until I can bring out sufficient color:
Here's an example of a narrowband only image, with reasonably colorful stars using this technique:
https://www.astrobin.com/full/wx5igj/0/
Edit: Also, you seem to have some green stars in your image. Try running SCNR on the stars.
Excellent reference picture! I'll give it a try.
DavesView:
Critique time again. Check out the revised and point me in the next direction. I carried the processing all the way through using the two tools. I see some improvement in the image overall, including the stars. Although the star colors just don't pop.
After CC, use SXT or SN++ V2 to separate stars from the rest. On the star-only image apply ArcsinhStretch to suit (between 30 and 300 on the slider). Then use HT to white-out the burnt cores of the brighter stars. Maybe an increase of saturation by 50% would add to the stars' colors.
On the nebula-only just use MaskedStretch, followed by ArcsinhStretch (you might not to do any changes to the default of 1), then HT to set you black and white point. Plus other stuff too long to list here...
Bring the two back together ~((~starless image)*(~star-only image)) with PixelMath.
Helpful