I can't complain about light pollution for this object! This is an HDR made using 30s subs for the Trapezium, and 5 minutes for the outer regions. All up it's only about 5 hours worth of LRGB data, but of course this is one of those objects where the challenge is in the processing, and not the image acquisition! I'll blame the seeing for the lack of resolution of the E and F stars, but you can see them very clearly in the linear data (along without about 30 more little ones). Maybe it's time to learn about star masks in PI!
Anyway, hopefully this might stimulate some discussion about the dos and don'ts of processing this magnificent object.
Before the saturation addicts jump in - I've chosen to keep the colouring subtle, because I can see these hues visually in my 25", and cranking up the colour just looks wrong to me.
Anyway, I welcome comments and thanks for looking.
Hi res version here: https://www.astrobin.com/325894/B/
I think you are right about processing being the challenge with M42! I have only attempted it once and that was back when I was just starting out with PixInsight, and I don't have any great words of wisdom to share. Let me know if you figure it out You've done a good job even if E & F aren't visible.
If you deconvoluted the 30 sec stack, maybe some tweaking of the global dark and light in deringing may remove the dark rings and dark centre of the stars which is visible in e hi-res version
Agree with Peter - a fine rendition and I think you've handled the dynamic range very well. Your ringed stars are a distraction at full res though. Overall, it's a bit soft for me too - I definitely think your data is good enough to show some more contrast / sharpness in the neb.
PS: Saturation looks spot on. Just for fun though I just had a bit of a fiddle to fix the stars, bring out some more faint detail and sharpen a bit ... WOW! You've got some great data here Andrew! IMHO this great image can be even greater!
Last edited by marc4darkskies; 19-12-2017 at 02:52 PM.
Some especially beautifully presented passages off toward the right and bottom right.
Thanks Mike and Trish - I spent ages just tracing some of these features to try and figure out how they could have formed.
Quote:
Originally Posted by Paul Haese
That's a great image Andrew. I really think the detail in the core is very nice. Lovely colour too.
Thanks Paul!
Quote:
Originally Posted by strongmanmike
Yes, seems easy... but definitely a challenging object to process but you have done a pretty good job here Andrew, the soft pastels are pleasing too
Mike
Thanks Mike! And I was trying so hard to keep the decon below your irritation threshold
Quote:
Originally Posted by gregbradley
A very fine M42 Andrew.
Greg.
Thanks Greg!
Quote:
Originally Posted by marc4darkskies
Agree with Peter - a fine rendition and I think you've handled the dynamic range very well. Your ringed stars are a distraction at full res though. Overall, it's a bit soft for me too - I definitely think your data is good enough to show some more contrast / sharpness in the neb.
PS: Saturation looks spot on. Just for fun though I just had a bit of a fiddle to fix the stars, bring out some more faint detail and sharpen a bit ... WOW! You've got some great data here Andrew! IMHO this great image can be even greater!
ArrgH! I need more details on what you've done there! I couldn't squeeze much more detail out without creating too much noise.
Quote:
Originally Posted by Peter Ward
A fine M42. Quibbles about sharpening rings on some, but not all, stars aside, the dynamic range is really well handled. Nice one.
Thanks Peter! It sure is tricky to get decon that doesn't show up against such bright nebulosity.
Quote:
Originally Posted by vlazg
Very nice image Andrew.
If you deconvoluted the 30 sec stack, maybe some tweaking of the global dark and light in deringing may remove the dark rings and dark centre of the stars which is visible in e hi-res version
Thanks - I've been able to get the E and F stars to pop out by masking the decon out of the shorter sub region, but using the global settings then makes the decon ineffective outside the brighter area. Definitely one for the local support feature I think.
I think you are right about processing being the challenge with M42! I have only attempted it once and that was back when I was just starting out with PixInsight, and I don't have any great words of wisdom to share. Let me know if you figure it out You've done a good job even if E & F aren't visible.
Cheers,
Rick.
Hi Rick - funny thing was the E and F stars were clear before I blended the Lum in, and the decon has done more harm than good there. Time to pull it all apart and start again!
...
ArrgH! I need more details on what you've done there! I couldn't squeeze much more detail out without creating too much noise.
...
Let me reiterate - this is a very fine image Andrew! Also, I'm not a PI person, I do all my post calibration and RGB combine processing in PS with layers and masks using home grown and commercial tools.
In any case, FWIW (and because you asked), below is what I did to your image (your original is on the left). It's 20% full scale. Noise was kept at the same level as the original (when viewed at full res). Stellar profiles are larger but generally within the margin of your decon artefacts. Didn't touch your colour or saturation - they're spot on IMO.
Caveats:
Because I'm not operating on raw data I can't get optimal results so this is only a quick and dirty done for illustrative purposes only!
If I was using raw data I'd spend much more time on stars, detail and noise processing.
Finally, detail and depth are, of course, a personal taste issue!
Thanks Marcus - I did ask, and in my opinion it's posts like yours where someone takes the time to explain a few processes and share some techniques that make this forum so valuable! I tried PS and a few carefully tuned runs of the smart sharpen and was able to bring out a similar amount of detail. Very cool, thanks! I just need to figure out the PI equivalent so I'm not cheating .
thanks again!
Andrew.
Quote:
Originally Posted by marc4darkskies
Let me reiterate - this is a very fine image Andrew! Also, I'm not a PI person, I do all my post calibration and RGB combine processing in PS with layers and masks using home grown and commercial tools.
In any case, FWIW (and because you asked), below is what I did to your image (your original is on the left). It's 20% full scale. Noise was kept at the same level as the original (when viewed at full res). Stellar profiles are larger but generally within the margin of your decon artefacts. Didn't touch your colour or saturation - they're spot on IMO.
Caveats:
Because I'm not operating on raw data I can't get optimal results so this is only a quick and dirty done for illustrative purposes only!
If I was using raw data I'd spend much more time on stars, detail and noise processing.
Finally, detail and depth are, of course, a personal taste issue!
You could try Local Histogram Equalisation. I use it with a variety of mask to bring out small and large detail respectively. I'm not sure that is right for that tool although it seems to work well.
You've done a good job even if E & F aren't visible.
