Gathered some LRGB last night to go with the HOS data previously gathered. Spacing seems to be sorted on the flattener now, so the LRGB is less crap than the NB was.
Noise reduction is a lot more subtle than in the NB version, and I've also applied some deconvolution this time.
Overall pretty happy with it now. As always, constructive criticism welcome!
Well that came out ok, huh? Looks great, nice separation of the core stars there and some lovely gaseous extensions. If your intention is to have it look like a natural RGB though, my only suggestion is that it is missing some green, otherwise top stuff Lee
No bad at all Lee! Not sure about lacking green though, as Mike says. The colours seem reasonable to me. It does lack some brightness & contrast though IMO. It also looks like the background sky is brighter than the faint HII areas in some places - most notably on the left and top left of frame. I'd darken the background sky a little.
It's fun to compare with your NB only version, and see the extent of the enveloping blue reflection nebulosity in the combo that gets almost totally blocked by an OIII filter. On the other hand, your NB version shows the Herbig-Haro jet.
Edit: Whoops, I just noticed that your NB version used a nonstandard mapping. But what I said about the reflection nebulosity vs OIII is still true when I compare it to other NB images.
Well that came out ok, huh? Looks great, nice separation of the core stars there and some lovely gaseous extensions. If your intention is to have it look like a natural RGB though, my only suggestion is that it is missing some green, otherwise top stuff Lee
Mike
Thanks Mike! I think the background (where there's no apparent faint nebulosity) is neutral. Maybe it's just too saturated?
Quote:
Originally Posted by marc4darkskies
No bad at all Lee! Not sure about lacking green though, as Mike says. The colours seem reasonable to me. It does lack some brightness & contrast though IMO. It also looks like the background sky is brighter than the faint HII areas in some places - most notably on the left and top left of frame. I'd darken the background sky a little.
Cheers, Marcus
Thanks Marcus, appreciate the feedback.
The lack of contrast is probably due to me using Masked Stretch in PI, which does lead to somewhat dull results. I did up the contrast a bit after, but maybe not enough. I've uploaded a newer revision on Astrobin with a bit more contrast. Still not quite right though I don't think so I'll keep tinkering with it. May yet get some more data too.
That background nebulosity basically doesn't exist in my super luminance. Probably due to me combining everything using noise evaluation though I did briefly try and unweighted median and it still didn't show up.
Not sure if there's better ways to combine NB and BB data.
Quote:
Originally Posted by Placidus
Very pleasing, Lee.
It's fun to compare with your NB only version, and see the extent of the enveloping blue reflection nebulosity in the combo that gets almost totally blocked by an OIII filter. On the other hand, your NB version shows the Herbig-Haro jet.
Edit: Whoops, I just noticed that your NB version used a nonstandard mapping. But what I said about the reflection nebulosity vs OIII is still true when I compare it to other NB images.
Thanks Mike! There's gotta be a way to get the best of both worlds, I'm just not sure how to do it yet... any tips would be appreciated :-)
Quote:
Originally Posted by gregbradley
Very nice. A slightly different look to a familiar object.
These short exposure strategies pick up some nice star shapes and colours.
Greg.
Thanks Greg :-) I think the main thing for the star shape and colour here was the masked stretch, which I don't usually like to use. That, combined with the creation of "superR" (r+sii+ha), "superG" (oiii+g) and "superB" (oiii+b) helped I think.
The broadband data actually had heaps of clipped stars due to the low dynamic range of the ASI 1600 when used at unity gain.
Thanks Greg :-) I think the main thing for the star shape and colour here was the masked stretch, which I don't usually like to use. That, combined with the creation of "superR" (r+sii+ha), "superG" (oiii+g) and "superB" (oiii+b) helped I think.
Interesting colour formula. I should try that. Is it your idea?
The red patch on the left seems to have sudden edge. Is that correct or an artefact or from a mask with hard edges?
Interesting colour formula. I should try that. Is it your idea?
The red patch on the left seems to have sudden edge. Is that correct or an artefact or from a mask with hard edges?
Greg.
The colour blend was my idea, but I don't think it's a novel approach. I had a quick look around and found an article with someone describing a similar approach, though they didn't use SII in the red. Just seemed like the most logical way to do it I'm convinced it's not the best way though, there's way more dusty nice stuff in the super red than you can see in this image, it just gets lost in the "ultra super luminance"
That red patch is real nebulosity. It's lost in the luminance so it looks a bit weird. Attached is a 100% crop of the same area (but rotated in its original orientation) from my colour data (i.e. no lum).
Quote:
Originally Posted by Atmos
Looking good Lee, the detail in the central dusty regions is showing up nicely
With the red I meant the sharp border being an artefact rather than the red nebulosity. Nebulosity has a gradual border rather than a harsh edge. That's what I was referring to. Perhaps using the blur tool on the edges to duplicate the edge shown in that crop would help.
I think one of the reasons that SII isn't usually added is because it is quite weak. In narrowband the only reason SII usually shows up nicely is because of colour calibration, i.e. making everything NOT green
Thanks Greg, I'll have a bit more of a play with it. To actually answer your question, I didn't use any hard masks or anything like that in processing this image.
All masks used were based off the stretched super luminance. I used a low contrast variant for 1st pass NR, a low contrast higher attenuation variant for 2nd pass NR, and stretched lum with clipped stars subtracted for deconvolution... those were the only masked I used (auto magic masked stretch excluded 'cause god knows what that's doing under the hood).
I think one of the reasons that SII isn't usually added is because it is quite weak. In narrowband the only reason SII usually shows up nicely is because of colour calibration, i.e. making everything NOT green
Yeah, fair enough. I just added it in because I had limited data and more data = better SNR.
I've never used the Masked Stretch in PI, what do you think of it?
Usually not much haha. It tends to create really flat looking images, and if you have any blown cores you get these really abrupt transitions that look pretty ugly. These can be overcome by adding a bit more contrast back in, and using morphological transformation with a range mask containing the clipped stars.
I rarely use it... in fact, I don't think I've ever used it in an image that I've posted anywhere before this one. I think it does work better on nebulae than galaxy shots though.
Having said that, it does get you some nice saturated colour, and keeps the colour in stars (maybe too much colour even). I think when used carefully in conjunction with other stuff it can be a good tool.
Thanks Greg, I'll have a bit more of a play with it. To actually answer your question, I didn't use any hard masks or anything like that in processing this image.
All masks used were based off the stretched super luminance. I used a low contrast variant for 1st pass NR, a low contrast higher attenuation variant for 2nd pass NR, and stretched lum with clipped stars subtracted for deconvolution... those were the only masked I used (auto magic masked stretch excluded 'cause god knows what that's doing under the hood).
I am not sure either what causes that effect. I have had it happen to me several times. The only thing I can put it down to is the data is too weak and needs more. If stretched too hard it breaks down and for some reason forms abrupt borders instead of a gradual feathering out. I think its a form of posterising which is caused by lack of depth in the data.
that edge looks like it is a real feature - it shows up in other images of M20.
It is a very fine and subtle image Lee. Agree with Mike that a bit more green might give more "normal colour, but even so, the enhanced background structure adds a lot to the overall effect. Like it a lot. regards ray
Detail looks good Lee but like others I am inclined to think it needs some green. There are a few yellow/red stars in that area and your present view has those colours quite muted.
I am not sure either what causes that effect. I have had it happen to me several times. The only thing I can put it down to is the data is too weak and needs more. If stretched too hard it breaks down and for some reason forms abrupt borders instead of a gradual feathering out. I think its a form of posterising which is caused by lack of depth in the data.
Greg.
Yeah I'm not sure exactly what the issue is yet, but I'll play around with it a bit more. As can be seen in the crop I posted above, it looks ok in the colour data, and looks even better when stretched more, but when combined with the super lum something weird happens.
I'm also going to try and get some more data tonight so maybe that will help (assuming my hardware plays nice for once, a rare thing!).
Quote:
Originally Posted by Shiraz
that edge looks like it is a real feature - it shows up in other images of M20.
It is a very fine and subtle image Lee. Agree with Mike that a bit more green might give more "normal colour, but even so, the enhanced background structure adds a lot to the overall effect. Like it a lot. regards ray
Thanks Ray! I'll have a another crack at the colour.
Quote:
Originally Posted by Paul Haese
Detail looks good Lee but like others I am inclined to think it needs some green. There are a few yellow/red stars in that area and your present view has those colours quite muted.
Cheers Paul! Consensus is definitely that it needs more green. I'll give it another crack. Of course I no longer have my working images so it'll probably come out completely different this time
Looks great, nice separation of the core stars there and some lovely gaseous extensions. If your intention is to have it look like a natural RGB though, my only suggestion is that it is missing some green, otherwise top stuff Lee 
I'm convinced it's not the best way though, there's way more dusty nice stuff in the super red than you can see in this image, it just gets lost in the "ultra super luminance"
