Very nice, like the depth effect u got there, funky colours too hehe, but u get that doing it that way.
I think the second version is better for my eyes
Great work Paul. M16 is a complex object well suited to narrowband imaging. I got a tip from Marc on one of my images a while ago on reducing the pink star halos. You may want to check it out. http://www.iceinspace.com.au/forum/s...49&postcount=8
The original purple star version looks better IMO. I don't have a problem with the purple stars really, they are in Hubbles narrowband shots so we are all familar with the look. The vibrancy of the original image is quite striking. Best when you can substitute in RGB stars, just removing the purple halos like this can make the image look a little lifeless and pasty.
I tend to agree with Mike on the image, though I struggle to accept purple stars these days. Even with Hubble images as they become a distraction if not managed correctly.
Paul, correct me if I'm wrong but I think you've applied the purple halo reduction globally across the image. You need to use masking to ensure you keep the nebula saturation intact.
You can either use the basic colour range tool, feather the selection and apply the setting. The colour range tool makes it hard to get the real faint stars however. My preferred method is to create a selection from the real data. This is very precise as you'll be able to select all bright and dim stars though more involved. I think this came from Russ Croman.
Make a grayscale copy of the image. I call this image #2.
High-pass filter image #2 with a radius of one pixel.
Apply a Gaussian blur to image #2 with a radius of one pixel.
Invoke Image->Adjust->Threshold.
Adjust the Threshold Level one click at a time until just the stars are white and everything else is black.
In the original image, in the Channels Palette, create a new channel. Name it "Stars." Choose "color indicates masked areas."
Paste image #2 into this channel.
Make just the RGB channels visible (i.e. make the Stars channel invisible).
Discard image #2.
In the original image, invoke Select->Load Selection. Choose the Stars channel you just created.
Invoke Select->Expand and expand the selection by a few pixels (e.g., three).
Done. You are now ready to selectively apply any fixes to the stars - bright or dim.
The original purple star version looks better IMO. I don't have a problem with the purple stars really, they are in Hubbles narrowband shots so we are all familar with the look. The vibrancy of the original image is quite striking. Best when you can substitute in RGB stars, just removing the purple halos like this can make the image look a little lifeless and pasty.
Always a good area to image though
Mike
Mike - thanks for comments. I agree that the first image is better. I don't have an issue with the purple stars, but if there is a way to remove then at least there are two versions available. Its all preference I suppose.
Yes, my previous RGB of this area is difficult to get right, I prefer NB for this target! (Plus the moon was high and bright!)
I tend to agree with Mike on the image, though I struggle to accept purple stars these days. Even with Hubble images as they become a distraction if not managed correctly.
Paul, correct me if I'm wrong but I think you've applied the purple halo reduction globally across the image. You need to use masking to ensure you keep the nebula saturation intact.
You can either use the basic colour range tool, feather the selection and apply the setting. The colour range tool makes it hard to get the real faint stars however. My preferred method is to create a selection from the real data. This is very precise as you'll be able to select all bright and dim stars though more involved. I think this came from Russ Croman.
Make a grayscale copy of the image. I call this image #2.
High-pass filter image #2 with a radius of one pixel.
Apply a Gaussian blur to image #2 with a radius of one pixel.
Invoke Image->Adjust->Threshold.
Adjust the Threshold Level one click at a time until just the stars are white and everything else is black.
In the original image, in the Channels Palette, create a new channel. Name it "Stars." Choose "color indicates masked areas."
Paste image #2 into this channel.
Make just the RGB channels visible (i.e. make the Stars channel invisible).
Discard image #2.
In the original image, invoke Select->Load Selection. Choose the Stars channel you just created.
Invoke Select->Expand and expand the selection by a few pixels (e.g., three).
Done. You are now ready to selectively apply any fixes to the stars - bright or dim.
Jase - You are correct - I took your tip and did a quick and dirty without masking. This tutorial is excellent! I will grab some more data on this target and then I will definately apply this technique. That way i will ahve two versions to post.
Narrowband has no rules ultimately in terms of colour, and this debate about colour inevitably comes up when someone posts a NB image. But to my view green is a colour that usually doesn't sit well in an astroimage. I guess virtually nothing in astro LRGB images is ever green unlike terrestial shots.
The green comes from the fact your Ha is assigned to green and Ha is by far the dominant signal.
The best way to handle this dominance is to do pixel maths on your red and blue channel S11 and O111. Usually say 8X the S11 and perhaps a bit less for the blue.
When you add S11 as a layer it should be set to lighten mode. Now you can push the red using curves so it starts to look more balanced and you have the deeper red/browns even yellowish hues which to my view are a lot more pleasing.
If the end result is a pleasing image for others to view of an astro object then I think the balancing of the colours similar to the above is definitely one way to do it.
That plus RGB stars and layer them in to get rid of the purple ones.
So really to me, NB consists of either Bicolour Ha red O111 Blue (Ha+O111)/2 green with RGB stars layered in or Ha green 8X S11 4-6X O111 blue with RGB stars layered in. I think that is the pleasing formula.
A similar argument is there for whether to layer in some Ha as luminosity on top or not. Doing so washes out colour but may pick up a bit of detail. It may supress delicate blue reflection nebula but sometimes it adds a bit of vibrance and detail to the image. Usually in very small doses.
No doubt Ken Crawford has some excellent tutorials on this as his NB images are among the very very best.
You got the data there, you are 50% there. The tricky fine points of advanced Photoshopping is the remaining 50%. We're lucky to have Jase posting here as they are some golden processing tips. You should write a book Jase or do some of those DVD tutorials you've got a lot of know how there.
Greg.
Last edited by gregbradley; 15-08-2011 at 07:30 AM.
Great result with so little time and plenty of tips for processing so I am taking notes. I don't mind pink stars provided they are not too severe. You have good strong blues but you know the S2 is the one that needs most subs - great to see what you can do once the Moon goes away.
That looks really nice Paul, I think you have treated the colours right.
Trent
Thanks Trent. After all the discussion on nb processing in this thread I have learnt a great deal about processing. But I agree with you that the colours are pleasing. I think that the astronomik filters I use are a great deal more "natural" as they are 12 nm and also seem to have a different spectral graph than other Brands which when you research the graphs allow for overlapping wavelengths which " blends " the colours giving zm earthy tone. This is most prominent with their rgb filters.
