Bubble nebula - modified SHO, first NB image

[vlaiv]

It's obvious that I have no idea what I'm doing since this is my first "complete" NB image, but here it goes (click on image for full size version):

This is sort of preliminary version, I need more practice and probably need to restack my data before another processing attempt (a lot of issues like dew on secondary, OIII not being that immune to LP so gradients are still there, ....)

Tech specs:

RC 8" F/8, ASI1600, Baader NB filters, Heq5 (OAG+ASI185), 64 x 240s per filter, total 12h 48m
Captured with SGPro, data reduction ImageJ, processing Gimp

 

Commends & Critique welcome as always, thanks for looking.

[carastro]

Looks good, just need to adjust the colours. 

Carole

[vlaiv]

11 minutes ago, carastro said:

Looks good, just need to adjust the colours. 

Carole

Hi Carole, I appreciate your input, but could you please be a bit more specific? Adjust colors in what way? (saturation, hue shift, you think choice of palette is wrong?)

[wimvb]

In pixinsight, I would use scnr on green, which works wonders. Unfortunately in GIMP, that luxury doesn't exist. But I believe there are colour manipulation processes that let you adjust individual colours (not just rgb, but magenta - green, cyan - red and yellow -blue). You might want to play with those.

[wimvb]

Here's your image with scnr green applied in PI

Btw, there's a lot more to extract from the image, but the jpeg version is of too low quality to do this.

[vlaiv]

14 minutes ago, wimvb said:

Here's your image with scnr green applied in PI

I get your point, but that kind of looses SHO palette feel to it, doesn't it? Ha is mapped to green in SHO and Ha is by far the strongest signal in the Nebula

[Allinthehead]

Great detail in that bubble Vlaiv. I agree with Carole and Wim. A little colour balance adjustment and and you've got those beautiful gold and blue colours you get with SHO.

[wimvb]

5 minutes ago, vlaiv said:

I get your point, but that kind of looses SHO palette feel to it, doesn't it? Ha is mapped to green in SHO and Ha is by far the strongest signal in the Nebula

Have you tried this process (PhotoShop, but should be similar in GIMP)?

http://bf-astro.com/hubblep.htm

Scnr green removes the green colour cast, but doesn't alter the intensity. This means that you get rid of the green, but keep the detail/lightness from the Ha.

[wimvb]

Here's a version with scnr only 50% strength. then DBE to remove the blue cast on the right, and a little contrast enhancement + star reduction.

Remember that the image is false colour anyway. Removing the green cast but keeping the detail from the green channel, doesn't really destroy anything, it just maps it differently. The original Hubble image of the Pillar of Creation also has very little green in it.

personally I like the version with less green/yellow and more reds better.

[Adreneline]

2 hours ago, vlaiv said:

It's obvious that I have no idea what I'm doing

I cannot disagree more. The detail is very good with lots of structure clearly visible. The image has depth and there is a real sense of 3D to the bubble.

I would be very happy if that was my image. As for the colours? Colours are very subjective - detail is very objective - it's there or it's not and in your case IMHO it's there in spades.

Don't beat yourself up vlaiv - you should be well pleased with your first NB image.

Adrian

[carastro]

I think the others have covered my explanation regarding colour.  Good data and whilst Ha is in the green channel, it is not typically shown as green in the final image, using yellow and red tones.

I have had a tweak of your image in Photoshop.

There is more data in there than you think.

Good data and combining of the channels as well.   Bit of a dark blob NE of nebula, si this a dust bunny?

Carole 

 

 

[Ouroboros]

I can only reiterate what others have said. Your data obviously has a lot of potential. You can have lots of fun bringing it out now. ?  I'll be interested to see you make progress with it in future posts if you choose o process the data again.  This is one of my favourite objects, and something I mean to have a go at when and if I ever go mono.  My unmodified DSLR really can't do it justice.  I've tried. But you've obviously cracked the technical issues of taking the data.  

[carastro]

Sorry had to rush out. Will have another look at my process when l get back to make sure that black blob is not something l did as l did do a gradient exterminator on it once l stretched it. 

Carole

[carastro]

I re-checked it but it is still there.

Carole 

[vlaiv]

36 minutes ago, carastro said:

I re-checked it but it is still there.

Carole 

Not sure if it is in the data, I calibrated subs properly, and none of dust doughnuts (nor vignetting) is showing in single images. Might be artifact of the fact that you were working on 8bit already (poorly) processed data.

I need to do data reduction again, and do it "properly" this time. I identified couple of points that can be improved, so first step is to again get linear stacks prior to any processing attempt.

Need to finally implement SNR based weighted stacking (I believe there will be actual visible difference in stack due to difference in SNR on single frames - both LP and dew on secondary created quite different subs in terms of SNR), and see what I can do about OIII gradient - I've got idea how to implement its removal (sigma upper reject to get "background" pixels only, then do least squares fitting of linear gradient function and subtract it in the end).

When I'm finished with data reduction, I'll post results so anyone fancying having a go at processing can have proper data to work with.

[Jkulin]

Hi Vlaiv,

I took your image into PI and then into PS and did just a few tweaks, there some nice data in there and I like the fact that you got some star colour in there, not sure if you approve, but it does look quite nice: -

 

[Jkulin]

BTW when I took your data I never saw the black blob and so I never processed it out as I coudln't see it.

[vlaiv]

3 minutes ago, Jkulin said:

Hi Vlaiv,

I took your image into PI and then into PS and did just a few tweaks, there some nice data in there and I like the fact that you got some star colour in there, not sure if you approve, but it does look quite nice: -

 

I like it a lot. I did not do any RGB color for stars, so any star color is just simply NB coloration - ratio of Ha, OIII and SII

I'm not still sure I buy into this sort of palette though. It looks like people are sort of used to it, and I wonder why (maybe because it is offered in one of most widely used processing suits?). It does feel "more balanced" than plain old SHO or any modifications thereof, but what I'm having trouble with - I can't seem to be able to distinguish SII from Ha regions in such palette. It is maybe present in the image, and maybe it's just my brain but I could not tell which areas are SII rich in this image.

[Jkulin]

The nice thing about Astronomy is that you process it how you want and not to please others, if it isn't for you then process it the way you like.

ATB

[vlaiv]

2 minutes ago, Jkulin said:

The nice thing about Astronomy is that you process it how you want and not to please others, if it isn't for you then process it the way you like.

ATB

Well I'm still learning here, so can't really say what would be to my best liking. My idea is to both create visually pleasing and balanced image, while keeping important information contained in image "easy to access".

[gorann]

Looks like very nice data to me. As someone said here, even if you put Ha in green it seems customary to later tweak that towards yellow as most think yellow looks nicer than green for images of the universe, and I am one of them. There are tutorials for how to do that in PS. Here is one, and even if you do not have PS you can probably do it in a similar way in other programs:

http://bf-astro.com/hubblep.htm

[MartinB]

Well done Vlalv.  The colour is false so there is no "right" colour.  The Hubble palate does a very nice job of showing dust.  However, there has been a tendency to eliminate green all together and to go for reddish orange and blue.  This doesn't make sense to me since the palate should still show where concentrations of Ha are high.  The famous "Pillars of Creation" is loaded with green.  Your initial image shows some yellowing of the ionisation  zones which indicates the presence of SII and the bubble has a blue tinge from the OIII.  Bob Franke uses the selective colour tool in PS to very good effect http://bf-astro.com/hubbleP.htm  but I don't know if there is an equivalent in GIMP.  The colour you have presented is not wrong but might be more pleasing to the eye if reddened a bit but don't get rid of the green altogether!

[vlaiv]

Two more contenders for acceptable palette

Or

Would you call any of these an improvement?

[MartinB]

Yes!!! You will find the jury split on which is preferred but I love the colouration of the second version.

[RolandKol]

@vlaiv I could not hold my self without inserting my 2p  tried to bring out the dark clouds on the bubble

a bit overcooked