Gamma Cygni in HaSHO with RGB Stars

[Rodd]

I decided on a new thread because often times additions understandably overlooked.  This is an HaSHO composition with 53 5min Ha, 74 5min OIII, and 70 5min SII, captured with the FSQ 106 with .6x reducer and ASI 1600 with Astrodon 3um filters.   The RGB stars gave me some trouble.  I used the RGB data to compose an RGB image--lightly stretched it so the stars were not too big, then masked out everything nut the stars and used Pixel Math to replace.  But there must be something about adding RGB data to NB images that is funky--the star were not replacing with much color--even if saturation in RGB stars was high,  It took allot of finagling.  I know there is a purple halo around Sadr--I think the increased halos I have been experiencing might be related to F3 imaging--for 3um NB imaging that makes sense--but not for LRGB, so maybe there is no truth in it.

 

[Sunshine]

My gosh it really is beautiful, so perfectly contrasted against the background of space, very beautiful blue's, awesome effort and it shows!

[Rodd]

16 minutes ago, Sunshine said:

My gosh it really is beautiful, so perfectly contrasted against the background of space, very beautiful blue's, awesome effort and it shows!

Thanks Sunshine.  Most of the effort went toward data collection ( quite a few number of nights) and processing the RGB data.  Once I captured the SII the SHO image wasn't to bad.  Broadband has always been more difficult for me.

Rodd

[gorann]

That is a very beautiful image indeed Rodd! Lovely colours and detail.

The stars look very good and tight to me without disturbing rings or halos. There is only one thing that distracts me (and you also comment upon it): the purple halo around Sadr. I think a red halo would look better and make a nice colour contrast with the blue and yellow in the rest of the image. At least in PS it is easily fixed by locally (using layermasks) suppressing the blue channel around Sadr.

 

EDIT: Just checked Sadr and Wikipedia says that it has the characteristic yellow-white hue of an F-type star. So, it should neither be violet or red. I assume it could also be tweaked towards yellow with some processing. Since you used RGB for the stars I expect that you would like them to have "natural" colours.

[wimvb]

Excellent image, Rodd.

Barry Wilson commented in another thread on how to remove magenta halos. Here's the essentials his method:

Create a colormask, targeting magenta (pixinsight script, press button Magenta). This should isolate the offending region. Tweak the mask to get a good definition of it.

Invert the image and apply the mask. Since the opposite of magenta is green, the halo will now appear bright green.

Use scnr green. Play with the best settings, and repeat if needed.

Remove the mask and invert the image back.

[wimvb]

11 hours ago, Rodd said:

The RGB stars gave me some trouble.  I used the RGB data to compose an RGB image--lightly stretched it so the stars were not too big, then masked out everything nut the stars and used Pixel Math to replace.  But there must be something about adding RGB data to NB images that is funky--the star were not replacing with much color--even if saturation in RGB stars was high,  It took allot of finagling. 

Can you give more details? Eg, masked stretch of the rgb image, rather than histogram stretch, keeps better star colour. Just an idea.

[Rodd]

5 hours ago, gorann said:

That is a very beautiful image indeed Rodd! Lovely colours and detail.

The stars look very good and tight to me without disturbing rings or halos. There is only one thing that distracts me (and you also comment upon it): the purple halo around Sadr. I think a red halo would look better and make a nice colour contrast with the blue and yellow in the rest of the image. At least in PS it is easily fixed by locally (using layermasks) suppressing the blue channel around Sadr.

 

EDIT: Just checked Sadr and Wikipedia says that it has the characteristic yellow-white hue of an F-type star. So, it should neither be violet or red. I assume it could also be tweaked towards yellow with some processing. Since you used RGB for the stars I expect that you would like them to have "natural" colours.

Thanks Goren.  I will try and get rid of the halo.  

Rodd

[Rodd]

4 hours ago, wimvb said:

Excellent image, Rodd.

Barry Wilson commented in another thread on how to remove magenta halos. Here's the essentials his method:

Create a colormask, targeting magenta (pixinsight script, press button Magenta). This should isolate the offending region. Tweak the mask to get a good definition of it.

Invert the image and apply the mask. Since the opposite of magenta is green, the halo will now appear bright green.

Use scnr green. Play with the best settings, and repeat if needed.

Remove the mask and invert the image back.

I will give this a try--but I find teh color mask hard to work with.  Sometimes it works but most of teh time I find it cumbersome.  Maybe in this case it will be perfect as there is only 1 bright magenta area--how could it miss?

Rodd

[Rodd]

4 hours ago, wimvb said:

Can you give more details? Eg, masked stretch of the rgb image, rather than histogram stretch, keeps better star colour. Just an idea.

Well--its just that my NB image has white stars,  If I create a nice fitting star mask so the stars are exposed and use pixel math to replace the stars with the RGB stars of the RGB image--the colors do not transfer.  The RGB stars in the RGB image look very colorful.  It does not matter how saturated I make them in the RGB image, when I replace them into the NB stars--teh color is mostly cone.  Maybe I need to linear fit the 2 images first?  How do most people do it--this is my first attempt and I thought it would be easy.

Rodd

[Rodd]

5 hours ago, gorann said:

EDIT: Just checked Sadr and Wikipedia says that it has the characteristic yellow-white hue of an F-type star. So, it should neither be violet or red. I assume it could also be tweaked towards yellow with some processing. Since you used RGB for the stars I expect that you would like them to have "natural" colours.

This is the halo we are discussing--the star itself is fairly color free (I am having trouble getting RGB stars to transfer to the NB image for some reason--working on it).  I suspect the colored halo has more to do with the surrounding nebula than the star.

Rodd

[gorann]

7 minutes ago, Rodd said:

This is the halo we are discussing--the star itself is fairly color free (I am having trouble getting RGB stars to transfer to the NB image for some reason--working on it).  I suspect the colored halo has more to do with the surrounding nebula than the star.

Rodd

My guess is that the halo is just the normal phenomenomenon created by diffreaction in the optics - we all usually get more or less brigh halos around bright stars. An exception would be of course if Sadr is actually embedded in the nebulosity. I expect it just happens to be somewhere inbetween us and that nebulosity - maybe someone knows.

It was easily fixed in PS but I let you do it in PI without being disturbed by my attempt.

[Rodd]

1 hour ago, gorann said:

My guess is that the halo is just the normal phenomenomenon created by diffreaction in the optics - we all usually get more or less brigh halos around bright stars. An exception would be of course if Sadr is actually embedded in the nebulosity. I expect it just happens to be somewhere inbetween us and that nebulosity - maybe someone knows.

It was easily fixed in PS but I let you do it in PI without being disturbed by my attempt.

I guess I am a disturbed individual--got it!  I also saturated the stars a bit and "fixed them" a bit--and made a few tweaks,  

 

5 hours ago, wimvb said:

Can you give more details? Eg, masked stretch of the rgb image, rather than histogram stretch, keeps better star colour. Just an idea.

Got it Wim--Not sure what I did differently, other than its a new day and I am not struggling to keep my eyes open after processing for 12 hours.  Here is the fix.  I also brightened up the image just a little to give it some depth.

 

 

[wimvb]

2 minutes ago, Rodd said:

Not sure what I did differently, other than its a new day and I am not struggling to keep my eyes open after processing for 12 hours

Sometimes that's all it takes.

[gorann]

1 hour ago, Rodd said:

I guess I am a disturbed individual--got it!  I also saturated the stars a bit and "fixed them" a bit--and made a few tweaks,  

 

Got it Wim--Not sure what I did differently, other than its a new day and I am not struggling to keep my eyes open after processing for 12 hours.  Here is the fix.  I also brightened up the image just a little to give it some depth.

 

Well, PI-guys,

sorry, but there is still a halo (=optic artefact) there, even if it is now dark blue, and Sadr is not really yellow-white. In PS I could easily remove the halo completely with the powerul colour tools of that mighty programme and give the star a more realistic colour? (I did it on your first version Rodd and I have not touched anything but Sadr and the halo)

[Rodd]

10 minutes ago, gorann said:

Well, PI-guys,

Do not think that I have represented what PI CAN do......that would be a mistake.  A true PI aficionado can do what ever you can do with PS.  I did not take it far enough.  Here is the best I can do without a full reprocess--which I am in the middle of by the way.  But, Goren--I do have PS now--have ot installed it yet.  Got a copy from Davey-T.  Any pointers would be welcome.

[gorann]

I assume you (like me) are lucky enough to get one of the older versions without monthly payments. So, welcome to the club of PS-PI co-owners. You are about to get the best of two worlds! (and as a bonus double the confusion and frustration). Wim @wimvb talked me into getting PI, Lets see if he one day would give in to PS.... But serioulsy, I think it is clear that both have their pros and cons. Unfortunatley they are so different in many ways.

[Rodd]

8 minutes ago, gorann said:

I assume you (like me) are lucky enough to get one of the older versions without monthly payments. So, welcome to the club of PS-PI co-owners. You are about to get the best of two worlds! (and as a bonus double the confusion and frustration). Wim @wimvb talked me into getting PI, Lets see if he one day would give in to PS.... But serioulsy, I think it is clear that both have their pros and cons. Unfortunatley they are so different in many ways.

Quire frankly I am afraid to start--I tried GIMP a while back and got nowhere.  Its not like there is a manual.   I think the best way to do it, ill be to process an image in PI then wgen I get to something I can't accomplish with PI--try PS just for that task.  After all--I will use PI mostly--supplemented by PS, so why not learn it as needed.

Rodd (ps--I am still pleased with this image despite its issues)--EDIT--not the lt one--I just noticed I killed the blue in other places.  I will go with the previous--halo and all

[Stub Mandrel]

Sadr sits in a small blue reflection nebula.

[Rodd]

I managed to get the blue back in remoter areas.  Sadr is definitely better than the original post--though perhaps not up to Goren's standards.  Maybe the next process!

[wimvb]

13 minutes ago, Stub Mandrel said:

Sadr sits in a small blue reflection nebula.

Yes, and it would be a feat to preserve this. I think that the culprit here is diffraction from the microlenses of the ASI1600.

Unfortunately I don't have access to PixInsight this weekend, or I could have a try.

[gorann]

Yes Rodd that is the way I play it (although the other way around) by being in my (near) comfort zone in PS and making occational expeditions into PI. Fortunately for you I think it is a general consensus that PS is more intuitive and easier to learn than PI without a university degree in physics and mathematics (like Wim)

[Allinthehead]

Great detail in that Rodd. Nice work.

[Rodd]

Final Tweaks--Improved Sadr and found more depth in brightest areas.

[gorann]

2 hours ago, Stub Mandrel said:

Sadr sits in a small blue reflection nebula.

Well, great with someone who knows their astronomy! Would that nebula correspond to what we thought was a halo? If that is the case Rodd you have just accomplished a nice rendation of it by turning it from purple to blue, and please disregard my tweak!

[Rodd]

9 hours ago, Allinthehead said:

Great detail in that Rodd. Nice work.

Thanks Richard.  I am always happy when I collect the SII data--which sometimes I am tempted not to in favor of bicolor.  But I am never disappointed when  I do.

Rodd