Jump to content

Banner.jpg.b83b14cd4142fe10848741bb2a14c66b.jpg

NGC7635 in Hubble Palette

swag72

By swag72
in Imaging - Deep Sky

Recommended Posts

swag72

swag72

Posted
Posted

Following on from a thread here http://stargazerslounge.com/topic/156142-narrowband-imaging-deciding-on-how-many-subs/ I have been wondering how much OIII and SII is needed to complement Ha data. I originally had 9 hours of Ha and was not sure how much more OIII and SII I'd need. So I ended up with the following amounts

Ha 14x1800s

OIII 13x1800s

SII 23x1800s

25 hours in total

I am far from happy with the background and so wonder if you good folks have any thoughts on whether additional OIII would smooth that out at all. It's clearly rather short of OIII data - Does this show? How would more change the image? I would so love some thoughts on amounts of data for NB capture - As if 25 hours isn't enough on this already, I'd think it should be.

7598689886_313ea1eff2_c.jpg

NGC7635 Bubble nebula in Hubble palette by swag72, on Flickr

Link to comment
Share on other sites
  • Replies 29
  • Created
  • Last Reply
AndyUK

AndyUK

Posted
Posted

Hi Sara, the more data you have, the more you can get away with stretching it without introducing noise (when the histogram starts becoming ragged and/or comb-like). Your Ha was definitely much stronger than the SII/OIII, which meant that when stretching to maintain balance between the channels, you had to stretch the SII/OIII more, So effectively, in order to maintain the balance between all 3 channels (histogram widths) the amount you could stretch without blowing the histogram was limited by the SII and OIII data. If you had yet MORE OIII and SII, you'd potentially be able to push that limitation further and thus get wider histograms for each channel (and thus a potentially deeper looking image).

(I bet if you looked at you looked at just the Ha channel within this image and compared it to your single Ha processed image you'd see that the Ha was much deeper...)

So to answer your question, yes, more data "would" help, but that's not taking anything away from this image at all - It's an absolute belter!!! :smiley:

Link to comment
Share on other sites
swag72

swag72

Posted
  • Author
Posted

Thanks Andy - And thank you so much again for all the help you gave me yesterday. Martin has suggested a ratio of 1:2:3 for Ha:OIII:SII, so I am going to give that a go for starters. I know that it will be different for each target, but it at least gives me a decent starting point. So that means that if I do one run per night of approximately 5 hours, that's 1 night Ha, 2 nights OIII and 3 nights SII - Blimey, that's a lot of nights and probably in excess of 30 hours worth of data. I see so many excellent images with much less data, that perhaps I should just use a couple of hours of Ha and then work the ratio from there. It will be trial and error to ascertain a good baseline.

Link to comment
Share on other sites
Martin-Devon

Martin-Devon

Posted
Posted

Congratulations Sara, this is a beautiful image, nicely processed too. You can go on adding data forever, but your image is already rich with detail (25 hours worth) - do you really think you'll benefit significantly by adding even more to this?

Link to comment
Share on other sites
swag72

swag72

Posted
  • Author
Posted

Thanks everyone for your comments. I was thinking along similar lines to your Simon about stacking my existing data into a 1:2:3 ratio. I may well have a go and see if there is much difference. Alternatively, I've moved onto something else now - Perhaps I just do the 1:2:3 ratio with another target!

@Zul - Hubble palette is just a way of describing a way of mixing the colours into different channels to get the blue / yellow effect. In Hubble palette the channels are mapped as SII to red, Ha to green and SII to blue. With narrowband the possibilities really are endless!

Link to comment
Share on other sites
davew

davew

Posted
Posted

Hi Sara,

I think the answer lies in what the target has in the way Ha, [O111] and [s11] and how you want the final image to look. Take for example an image that is made up of 80% Ha, 15% [O111] and 5% [s11]. There wouldn't be much in the way of blue and even less in red. However, it appears that most people like the image to look as if the Ha and [O111] are almost 50/50 with a bit of red chucked in for interest. Nothing wrong with that, you make the image look like what you want to. There are some imagers that are taking the [O111] channel, removing the stars, stretching it to near breaking point, running a noise reduction on the background and then putting the stars back. Another tactic I've noticed is Binning the [O111] and [s11]. All false of course but that's what they want.

The Bubble neb is made up of tons of Ha. If you see a mono image of the region is it Ha or [O111] that's used ? Also you may notice that many of the Bubble images you see are close in and biased towards the blue area.

I think you have more than enough data already. I wouldn't use less of the Ha because when you stretch it you'll just add noise that you spent all that time trying to get rid of ! I may also be a good idea to hold the highlights back a touch. There is a patch of Ha on the edge of the Bubble that's gone white.

Just play around with each channel until it looks something like :o)

Warning...... Don't confuse me with someone who knows what they're talking about.

Dave.

Link to comment
Share on other sites
swag72

swag72

Posted
  • Author
Posted

Martin, your edits are great - Can I ask what you did to the first one? Also regarding the tone mapping, did you follow the tutorial written by J-PM? I have tried to follow this without success, but then I am a photoshop virgin and this is my first ever process fully in Photoshop!

Link to comment
Share on other sites
swag72

swag72

Posted
  • Author
Posted

Cheers for that link Martin - It is the J-PM tutorial that I tried to follow rather unsuccessfully. I will give it another go with my next batch of data and see if I can do better now I have a vague idea with Photoshop! I got lost in the 'sticking it all back together' stage!

Link to comment
Share on other sites
squeaky

squeaky

Posted
Posted

Um... I did four iterations of lifting levels and curves for RGB, then a single iteration to lift Red only, followed by one to reduce Blue only... So is this more of what you're looking for or have I done it the wrong way round?

post-23222-0-37872500-1342729527_thumb.p

EDIT: Forgot to say... GREAT image :)

Oh, a thought...I agree (mostly) with the earlier suggestion that you could try using the best of your subs with 6, 12, and 18 to see what the balance then looks like. But I said "mostly" because sometimes you can get a better image from less subs if you are first pretty ruthless about culling them. So set aside any sub that you think is below a desired standard. Not below standard for star count because you'll cover that easily - but any that are below standard for the amount of detail they have in the nebula, or have rubbish backgrounds. Otherwise you'll be stacking "rubbish" so you'll GET rubbish, If you see what I mean?

Then, if you can't do 6, 12, and 18 you could do 5, 10, and 15, or even 4, 8, and 12.

That keeps the ratio at 1-2-3 and then after that you can start following whichever PS tutorial(s) you think will help you bring out the best.

Good luck :)

Link to comment
Share on other sites
martin_h

martin_h

Posted
Posted

Forgot to say, for my original version I just brought the black point up a bit in Levels, the JPEG looked a bit washed out, and, using selective colour reduced the yellow in the cyan channel, a little tweek in curves to bring up some feint areas and loads of noise reduction - stretching JPEGS is not a good idea!

Link to comment
Share on other sites
squeaky

squeaky

Posted
Posted

Forgot to say, for my original version I just brought the black point up a bit in Levels, the JPEG looked a bit washed out, and, using selective colour reduced the yellow in the cyan channel, a little tweek in curves to bring up some feint areas and loads of noise reduction - stretching JPEGS is not a good idea!

Yep :)

It should all look a lot better when the work is done on the original tiff.

Not that it's exactly bad now. It's a darn good image.

Link to comment
Share on other sites
swag72

swag72

Posted
  • Author
Posted

Thanks for the info Martin. When you sat 'bought the black point up a bit in levels' - What does that mean? Moving the left slider towards the main histogram? Don't forget, a Photoshop virgin!!

@Richard, your edit is a little too red for me - I like yellow and turquiose!

Link to comment
Share on other sites
squeaky

squeaky

Posted
Posted

Fine :)

While I aimed at red and blue - you can aim at any single colours you like to ease the overall colour to where you want it.

When in adjust levels - there's a dropdown box with "RGB" in it. You can pick red green or blue from there.

You can aim at yellows and cyans etc using the Hue adjuster, and you can aim at more using - Image >> Adjustments >> Selective Color.

And yes, moving the left hand slider part way towards the start of you histogram in levels will help. This is a good youtube tutorial on curves and levels:

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

Go to topic listing

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue. By using this site, you agree to our Terms of Use.