PixInsight Forum (historical)
PixInsight => Image Processing Challenges => Topic started by: mcbbcn on 2012 October 21 18:24:22
-
Hi all,
I have the following image which has been DBE, RGB combined with Ha, and then background calibration, color calibration, HDRMT, and finally combined with Luminance. I believe the image is pretty well balanced from a color standpoint, but the blue signal is weak. I've tried to saturate the image during L + RGBHa but it gets a fair amount of good signal in the Red/Ha channel, but the blue is not great. I'm familiar with the effect of increasing the signal for just one color channel in Photoshop, but usually that gets all the colors out of whack. I have a pretty good star mask in Pixinsight, so I know I can protect the star colors, but I'm worry that if I increase the blue for the nebula, it'll take the Red/Ha channel out of whack for the whole nebula, so I wanted to ask for your opinion or any tips you may have...
By the way, I expose from a Bortle 8 location, so it's difficult enough to get any signal at all given all the LP, but if you have any tricks up your sleeve, I would love to hear.
Thanks for your help,
Miquel Casas
Portland, OR
-
Hi, you can selectively saturate using curves, saturation tool or curve interchannel script.
The easier being the second in my opinion: here you can increase the saturation in the blue only.
you can be even more selective preparing a mask: i.e split R,G, B and create a luminance mask from the B (or B&G).
By the way, may be you pushed your ha too much and that overwhelmed the rest.
I don't know if it is the same but I would build the LRGB first, then split color channels, blend Ha and red, and last recombine HaR,G,B.
Ciao,
Andrea
-
Hi Andrea,
Thanks for the tips. This is very useful. By the way, regarding your mask suggestion, are you implying a mask around the nebula? If yes, I've never done that, I have experience building star masks though...which tool would you use to select just the nebula?
Thanks for your help,
Miquel Casas
Portland, OR
-
I think you have two options if you want to increase saturation a bit on the nebula only.
-create a luminance mask and subtract your star mask
-create a luminance mask and use a tool like atwt to eliminates small scales containing the stars.
I would try to lower the ha weight first and see the results: it would be interesting to see your lrgb.
it doesnt seem a saturation issue to me
Be careful with dbe, it can eat your data if not done properly. It happened to me on the same field.
Andrea
-
Hi Miquel!
Combining all images in Linear status is the way we did it to keep the colors right.
After separate DBE and Deconvolution we did a LinearFit to all channels.
Than the Combines -
First R-G-B, than L to RGB, than
using HaRGBCombine Script to integrate Ha into the LRGB, it keeps the star colors nicely.
Than using HaRGBCobine Script to mix OIII into the Green and Blue Channel.
(Mixing OIII into the NGC6992 gets the colors of the nebula nicely and reduces the unreal dominance
of Ha.)
Than stretching the linear image with MaskedStretch Script.
The HaRGBCombine Script is an excellent tool for combining Ha etc into the RGB (LRGB) image.
We got the best results when using the HaRGBCombine with linear images.
If you have a basic understanding of german language i have produced a tutorial video (30 minutes)
to show how we did this. (with final basic stretching)
http://dl.dropbox.com/u/57910417/Video03_Kombination_OIIIHaLRGB_END.wmv (http://dl.dropbox.com/u/57910417/Video03_Kombination_OIIIHaLRGB_END.wmv)
The final image is on:
http://www.werbeagentur.org/oldwexi/gallery_guest.html (http://www.werbeagentur.org/oldwexi/gallery_guest.html)
Aloha
Gerald
-
Hi Andrea,
Thanks a lot for the tips, I'll give it a try.
M.
-
Aloha Gerald,
I don't have an OIII or any other narrow band filter other than the Ha filter, but you've given me some goo tips to try, so I'll give them a try too.
...by the way, that is quite an impressive image. :P
Aloha,
M.
-
Hi Andrea,
I've been trying to lower the Ha signal with no success. I was wondering if you could give me some specific steps on how to do that. I've tried to do it using the HaCombination script, and even when I set the multiplier to zero, it's still too much signal.
Also, when you say substract the luminance mask what do mean by that? Apply the mask or do you mean something else?
Thanks for your help,
M.
-
Hi Miquel,
You have also the possibility to play with the filter bandwith.
I wonder if the ha image is too strong for the rgb, given the poor location if I understand correctly.
I would check this at the beginning and follow o good process like the one Gerald is suggesting, ensuring you don't clip any data.
I would start completing the rgb and lrgb first: from there you can understand if adding ha is feasible/add value. If you don't see the teal nebula parts here, you have no chance to get them after the ha combination.
Also, I would try not to push the processing too much.
For the mask, this is detailed in a tutorial/processing example on the PI site: you can subtract the star mask from the lightness mask and get a new mask to act only on the nebula portion. In the case of this image everything is complicated by the dense star field.
But again, check that you have luminance support of the teal areas first.
Andrea
-
Hi Andrea,
Thank you so much. I checked on doing an LRGB, and I do have the teal areas. The Ha exposure is so beautiful and has so much detail that it feels wrong to throw it away, so I'll Geralds approach and see what happens.
By the way, Pixinsight has several tutorials, which one are you referring to regarding mask subtraction?
Thanks for all your help, I really appreciate it.
M.
-
Hi,
Check "Processing a H? CCD Image with PixInsight " in processing examples.
Ciao,
Andrea
-
Hi Andrea,
I own you a tone, as I do to many of the people who have been helping me grow in this challenging hobby. Thanks a tone to you & everybody else for all of your support. :)
M.
-
I am happy to know you found the discussion interesting.
Best,
Andrea
-
Hi Gerald,
I tried to follow your video guide, and although I don't speak German, I got a better understanding of the LinearFit process....and somehow, later on in your video, I got lost with all the channel extractions and combinations, specially, because I don't have an OIII exposure, so it's not a 1 for 1 process. I'll keep experimenting and see if I can come up with a process that the Ha signal does not overwhelm the teals in the image.
Also, I wanted to take a moment to express my sincere appreciation for sharing this video.
Best regards,
M.
-
Hi,
Reporting progress to the community. This is the next try @ L_HaRGB...so far applying LinearFIT, and then adjusting parameters in the HaRGB, and doing some curvestransformation...
I was wondering if there is a way to protect the stars when doing HaRGB...has anybody had any luck with that? I've not been able to do it. I think I'm going to have to combine Geral's method with Andrea's method. I think Andrea's had a way to protect the stars when applying the Ha signal.
It feels that I'm doing better, but I'm not totally satisfied yet...however, I've been going at it for a few hours, and I need to step out for a while, so I'll give it another try this weekend...
Your feedback is more than welcome...
Thank you all,
M.
-
Well...I could not walk away, I had to try to combine Andrea's tips with Gerald's tips... ;)
I ended up subtracting the stars from the Ha signal using Pixelmath (my first time using Pixelmath, I'm quite a newbie), but when I did that, then I had to skip applying linearfit from the Ha signal, otherwise the overall color balance was off.
Now, I'm a bit happier with this one because the star colors are a bit more balanced.
I'm going to leave it here for now, and as always, your comments are more than welcome.
M.
-
Hi all,
I'm pretty happy with the below result. I ended up discarding the luminance, and using my Ha as luminance. Then, I applied some of Andrea's and Gerald's techniques for linearfit, and I really like where I ended up. The blues in nebula are back, and although it's not the amazing results that I've seen when using an OIII filter, it's actually pretty good for a blue filter from a light polluted pollution.
Thanks everybody for your help,
Miquel Casas