NGC 6888-Ha-OIII

[Rodd]

Need some advise on this one.  I was planning on acquiring 40 20 minute subs in both Ha and OIII.  However, this image has 40 20 min subs of Ha, but only 14 20 minute subs of OIII, and the Ha is getting faint, the OIII is drowning it out I think--curves and saturation boosts do not help--they only turn the ends bright red--not the middle where the blue overlies the red.  Not sure what the problem is.  The Ha stack is very bright, but when the OIII is added, the Ha becomes subdued.  Any suggestions?  Shot with C11Edge at f7.

Thanks---Rodd

[johnrt]

The OIII really forms like a shell over the Ha when you get really deep, and I think all your are seeing in the results of the make up of the nebula. You really are hard on yourself, just knock back that green with SCNR set to green and you have a cracking image there.

[Rodd]

7 minutes ago, johnrt said:

The OIII really forms like a shell over the Ha when you get really deep, and I think all your are seeing in the results of the make up of the nebula. You really are hard on yourself, just knock back that green with SCNR set to green and you have a cracking image there.

Thanks--here is a version with a little mote saturation just to the blue envelope (done in Pixel math with masks) and the SCNR applied--maybe too blue?

[ollypenrice]

It's good. Of course, you shoot yourself in the foot by using a mask rather than layer based programme! 

If you used layers you could apply the filters to different colour channels in different layers (really OIII to blue on one and to green in the other) and adjust the balance while seeing in real time what the result would be.

But it's still good!!!

Olly

[Rodd]

5 minutes ago, ollypenrice said:

It's good. Of course, you shoot yourself in the foot by using a mask rather than layer based programme! 

If you used layers you could apply the filters to different colour channels in different layers (really OIII to blue on one and to green in the other) and adjust the balance while seeing in real time what the result would be.

But it's still good!!!

Olly

Thanks Olly--I did forget about the Blue in the green channel--that's why there is a little green in the last image--I used SCNR with a mask subtracting green the blue channel--the red and green were hidden.  I need to do it to the green channel too!

[ollypenrice]

Just now, Rodd said:

Thanks Olly--I did forget about the Blue in the green channel--that's why there is a little green in the last image--I used SCNR with a mask subtracting green the blue channel--the red and green were hidden.  I need to do it to the green channel too!

You need Photoshop...

lly

[Rodd]

Here's a version with the green removed from the OIII veil in the green channel as well as the blue.  Did it in PI--Perhaps not as well. But its all I have.  Maybe I should try the Gimp2

[Rodd]

6 hours ago, johnrt said:

The OIII really forms like a shell over the Ha when you get really deep, and I think all your are seeing in the results of the make up of the nebula. You really are hard on yourself, just knock back that green with SCNR set to green and you have a cracking image there.

 

6 hours ago, ollypenrice said:

You need Photoshop...

lly

Here we go--starting to get there.  I was trying to decide whether I should capture more OIII (only got 14 20 min subs compared to 40 for the Ha.  This rendition suggests it wouldn't be a bad idea perhaps--that way I can keep the saturation boost down and let the data do the talking.  This version might be over saturated for many folks--perhaps for me as well--but it shows more of the Ha network beneath the OIII shell, which is what I was going for.  Also a bit of star reduction.

[Rodd]

And here with a bit of sharpening to the nebula only.  Not sure which I prefer.

[PatrickGilliland]

Hi Rodd,

I actually think your first version is more accurate - OIII is the outer shell and this should be a more teal like colour than blue.  Having reviewed some spectroscopy data on this area i now agree with that more.  Of course artistic license means it's up to you.

I have worked on three version of this and recently discussed here at SGL.  My first two version i was not happy with being unable to maintain the colours i wanted. Eventually i decided on this version as the best i could muster http://www.astrobin.com/full/254474/0/ 

For me it required a different approach - colour was the issue and having played with it for days i could not quite get right.  So the other element that impacts colour is the luminosity of the data.  Summarising i found that Ha and OIII worked against each other especially around the core of the crescent so the challenge was to decrease separate the channels better.  Step in some RGB data.  I used the channels and blended with NB data this actually provided less saturation and more defined detail in each channel making it easier to define the red and teal out.  The distinction between the two increased and i could then isolate and enhance. 

All that said still takes some work - the object shows red ha shell thrown off some time ago being overtaken but a more excited OIII shell from a secondary event as WR136 dies.  So the teal is in front of the red in many places.  All you can do here is gently keep tweaking your red until you find the right balance.  Finally more OIII will reinforce the outer shell and allow the teal colour to show through better.

Your last version excluding the purple hue is showing nice ha though so long as you can maintain that more OIII may be all you need.

 

Paddy

 

[aparker]

I think I agree with Paddy - the first image is probably closer to a balance representation of what's actually out there in terms of fluorescing gas.  If you really want to do a deep dive into the delicate structure of the shell, I think a mono H-alpha-only presentation is the way to go.

[Rodd]

2 hours ago, PatrickGilliland said:

Hi Rodd,

I actually think your first version is more accurate - OIII is the outer shell and this should be a more teal like colour than blue.  Having reviewed some spectroscopy data on this area i now agree with that more.  Of course artistic license means it's up to you.

I have worked on three version of this and recently discussed here at SGL.  My first two version i was not happy with being unable to maintain the colours i wanted. Eventually i decided on this version as the best i could muster http://www.astrobin.com/full/254474/0/ 

For me it required a different approach - colour was the issue and having played with it for days i could not quite get right.  So the other element that impacts colour is the luminosity of the data.  Summarising i found that Ha and OIII worked against each other especially around the core of the crescent so the challenge was to decrease separate the channels better.  Step in some RGB data.  I used the channels and blended with NB data this actually provided less saturation and more defined detail in each channel making it easier to define the red and teal out.  The distinction between the two increased and i could then isolate and enhance. 

All that said still takes some work - the object shows red ha shell thrown off some time ago being overtaken but a more excited OIII shell from a secondary event as WR136 dies.  So the teal is in front of the red in many places.  All you can do here is gently keep tweaking your red until you find the right balance.  Finally more OIII will reinforce the outer shell and allow the teal colour to show through better.

Your last version excluding the purple hue is showing nice ha though so long as you can maintain that more OIII may be all you need.

 

Paddy

 

Thanks Paddy--Your image is awesome.  I always did feel the bright blue was a bit off--How do you combine all those filters? All at once or do you create NB and BB images then combine them? I had a version of this target shot with the TVnp101is that had 10 hours of SII data, but the SII seemed to ruin the image-probably the way I combined it.  Anyway--this is a good example of my processing ruining my image.  If only I was satisfied with the first image I wouldn't need to process!!

[PatrickGilliland]

9 minutes ago, Rodd said:

Thanks Paddy--Your image is awesome.  I always did feel the bright blue was a bit off--How do you combine all those filters? All at once or do you create NB and BB images then combine them? I had a version of this target shot with the TVnp101is that had 10 hours of SII data, but the SII seemed to ruin the image-probably the way I combined it.  Anyway--this is a good example of my processing ruining my image.  If only I was satisfied with the first image I wouldn't need to process!!

TBH with a lot of experimenting i think the final ratios were around r = r*.7 + ((sii*.76+ha*.24)*.3), b = =b*.7+((ha*.85+sii*.15)*.3, b = b*.7 + Oiii*.3.  I did however create many versions and blend and re-blended at different ratios until happy.  Nothing ruined fine image - but if ever we think we have cracked it we are probably getting ahead of ourselves

 

Paddy

[johnrt]

I think you have veered off in to purple in the last 2, I prefer the OIII to be a little bluer than in the first image, but I think you've gone a little far where you have used the SCNR. Your background is starting to get too noisy in the last couple of images too.

[Davey-T]

Hi Rodd, still in search of perfection I see, as said an image such as this never seems to "be finished"endless fiddling is possible.

If it will make you feel better I'll PM you one of my Crescent attempts, wouldn't dare put out in public 

Dave

[Rodd]

14 hours ago, Davey-T said:

Hi Rodd, still in search of perfection I see, as said an image such as this never seems to "be finished"endless fiddling is possible.

If it will make you feel better I'll PM you one of my Crescent attempts, wouldn't dare put out in public 

Dave

I am sure it is not as bad as you think Dave  The Crescent is summing up to be a tough little nut.  It reels you in by promising a beautiful image with only 2 filters and whamo--3 weeks later you are going crazy.  I'm afraid this is another good example of my processing ruining an image (or making it worse--never was great).  Its amazing but I cropped and enlarged an image of the Crescent I took with the np101is and it is almost identical to the one shot using the C11 Edge HD.  Comparison was made in Ha so processing didn't interfere.  Makes me wonder why I bought 2 scopes (well--I need it for the Moon and planets)-but I was expecting more detail out of the C11 Edge than the Televue on smaller DSOs.  Not really the case once you crop and enlarge.

[Rodd]

15 hours ago, PatrickGilliland said:

TBH with a lot of experimenting i think the final ratios were around r = r*.7 + ((sii*.76+ha*.24)*.3), b = =b*.7+((ha*.85+sii*.15)*.3, b = b*.7 + Oiii*.3.  I did however create many versions and blend and re-blended at different ratios until happy.  Nothing ruined fine image - but if ever we think we have cracked it we are probably getting ahead of ourselves

 

Paddy

Just when I thought I had me an easy 2 filter target!

[Rodd]

On 7/28/2016 at 15:58, johnrt said:

The OIII really forms like a shell over the Ha when you get really deep, and I think all your are seeing in the results of the make up of the nebula. You really are hard on yourself, just knock back that green with SCNR set to green and you have a cracking image there.

 

23 hours ago, Davey-T said:

Hi Rodd, still in search of perfection I see, as said an image such as this never seems to "be finished"endless fiddling is possible.

If it will make you feel better I'll PM you one of my Crescent attempts, wouldn't dare put out in public 

Dave

 

On 7/29/2016 at 08:41, aparker said:

I think I agree with Paddy - the first image is probably closer to a balance representation of what's actually out there in terms of fluorescing gas.  If you really want to do a deep dive into the delicate structure of the shell, I think a mono H-alpha-only presentation is the way to go.

 

On 7/29/2016 at 06:23, PatrickGilliland said:

Hi Rodd,

I actually think your first version is more accurate - OIII is the outer shell and this should be a more teal like colour than blue.  Having reviewed some spectroscopy data on this area i now agree with that more.  Of course artistic license means it's up to you.

I have worked on three version of this and recently discussed here at SGL.  My first two version i was not happy with being unable to maintain the colours i wanted. Eventually i decided on this version as the best i could muster http://www.astrobin.com/full/254474/0/ 

For me it required a different approach - colour was the issue and having played with it for days i could not quite get right.  So the other element that impacts colour is the luminosity of the data.  Summarising i found that Ha and OIII worked against each other especially around the core of the crescent so the challenge was to decrease separate the channels better.  Step in some RGB data.  I used the channels and blended with NB data this actually provided less saturation and more defined detail in each channel making it easier to define the red and teal out.  The distinction between the two increased and i could then isolate and enhance. 

All that said still takes some work - the object shows red ha shell thrown off some time ago being overtaken but a more excited OIII shell from a secondary event as WR136 dies.  So the teal is in front of the red in many places.  All you can do here is gently keep tweaking your red until you find the right balance.  Finally more OIII will reinforce the outer shell and allow the teal colour to show through better.

Your last version excluding the purple hue is showing nice ha though so long as you can maintain that more OIII may be all you need.

 

Paddy

 

 

On 7/28/2016 at 16:20, ollypenrice said:

You need Photoshop...

lly

I had to post this comparison--the following 2 images were captured using a Televue np101is (4" refractor) and the C11Edge HD (11" corrected Cass).  The TV im age has been cropped so the image scale is the same.  Please do not trouble yourselves with the quality of the processing, which is not really the point of the posting.  All I will say about that is both images were afforded the same effort, however misguided it was.  I included a more recent version of the original image (image #3) from this post for the heck of it.  The first image is the original image from this post, the second image is the Televue cropped, and the 3rd image is the most recent version of the original image.  If I wasn't into imaging the Moon and planets, I would think real hard about purchasing a 10-14" scope.  A 4-5" refractor may be all that is required.

Whats the verdict?

 

[Davey-T]

Guess it depends on the imaging scale / sampling if they're similar then the images will be similar.

Dave

[ChrisLX200]

I think you're correct Rodd, I'm yet to be convinced my purchase of a 12" scope is offering a marked improvement in recoverable detail compared to a fine 5" refractor.

ChrisH

[Rodd]

11 minutes ago, ChrisLX200 said:

I think you're correct Rodd, I'm yet to be convinced my purchase of a 12" scope is offering a marked improvement in recoverable detail compared to a fine 5" refractor.

ChrisH

How about on those nights of exceptional seeing, when FWHM is at 1-2 (That's exceptional here--think .5-1 if you want).  Then would the .57 arcsec/pixel I get with the C!! at F7 show more than the 2.06 arcsec/pixel I get with the TV?  I did not think of this

[swag72]

You have a central obstruction with the SCT, nothing with the refractor.....I do believe that a good example of a refractor can outperform many larger scopes.

[ChrisLX200]

2 hours ago, Rodd said:

How about on those nights of exceptional seeing, when FWHM is at 1-2 (That's exceptional here--think .5-1 if you want).  Then would the .57 arcsec/pixel I get with the C!! at F7 show more than the 2.06 arcsec/pixel I get with the TV?  I did not think of this

What theory predicts and what I get in practice seem to be at odds - in such cases I always go with empirical data :-) Like you, if I look closely at images taken with both a 5" refractor and my 12" ODK I don't see more detail despite the larger aperture. On the other hand I cannot say with confidence I'm getting all the performance out of the ODK12 it is (theoretically) capable of delivering.

ChrisH

[Rodd]

2 hours ago, ChrisLX200 said:

What theory predicts and what I get in practice seem to be at odds - in such cases I always go with empirical data :-) Like you, if I look closely at images taken with both a 5" refractor and my 12" ODK I don't see more detail despite the larger aperture. On the other hand I cannot say with confidence I'm getting all the performance out of the ODK12 it is (theoretically) capable of delivering.

ChrisH

Makes sense.  The only question i have then is when you crop a DSO in an image taken with the refractor--like I did with the Crescent--do you have to re-sample first (down or up) in order to avoid presenting the image at a ratio of something like 5:1.  Some feel 1:1 is the most one can reasonably take data, and I think I agree--so how do I equate that with a sever crop using a 4" refractor?

[aparker]

As far as detail on bright objects like NGC6888 goes, I think you are right.  Unless you live someplace that's blessed with exceptional seeing on a regular basis, you are going to be seeing-limited at focal lengths well below that 11" SCT.  For really dim objects, aperture always wins for any fixed pixel scale.  A Hyperstar on that C11 would put you at roughly the same general focal length as the TV101, and would pull in an equally bright image much, much faster (albeit probably not quite as flat and crisp a field as the Televue).