[RCC] Looking for feedback on my first SHO narrowband image - Forums

Topic starter

Jul 3, 2017

I am looking for constructive criticism from the Astrobin community on my image of the N. America and Pelican Nebulae. You can see all the equipment and acquisition details by clicking through to the image below. For processing I used both PixInsight and Photoshop. Here are the processes I used in both:
PixInsight:
SubframeSelector, Blink, StarAlignment, ImageIntegration, DrizzleIntegration, Dynamic Background Extraction, MaskedStretch, (Tone maps created in Photoshop), SHO-AIP script (R: 50/50 SII and HA, G: 50/50 HA and OIII, B: OIII, L: All three using Ha as master for stars), Curves, Resample to 70%, ICCProfileTransformation, (To Photoshop for Final Adjustments)
Photoshop:
Tone maps: Dust and scratches filter and Healing brush to make starless versions
Final adjustments: Curves and Selective Color Adjustment layers, save for web [PNG]

QUESTIONS:
What would you do differently? What do I still have to work on as a beginner? Looking at my different revisions on Astrobin, do you like this version best or am I pushing the data too far?
Link to full screen version of latest revision for evaluation: http://www.astrobin.com/full/301117/D/

Thank you!

Respectful Engaging

Slawomir

Jul 5, 2017

Hi 2ghouls,

A beautiful image of a very interesting area that I have not been able to image yet. I like Version C the most becuase of most pleasing colour palette, but colour preferecne can be quite subjective.

I think detail in nebulosity is very good given limited integration time. What would I do differently? Although the image is very nice to look at, at full resolution stars shapes are not ideal. I believe lens is here to blame, and perhaps Masked Stretch makes them look a bit hard too. You have not listed any noise reduction tools, but it looks like data has been a bit distorted in the darker areas - it looks like an artefact from noise reduction. But these are really minor things that I would try to imporve if it was me processing this data. I would probably try to double integration as well if I was going for a greater wow factor.

All in all, a great image

Helpful Insightful Respectful Engaging Supportive

Victor Van Puyenbroeck

Jul 5, 2017

I agree with Slawomir, the stars look a little flat and the background has dark 'spots' in areas with low SNR. There's also some tilt in the bottom left (defocused stars). The overall color and nebulosity looks great.

Maskedstretch should give you better stars, but their shape might have changed during tonemapping or while making the synthetic L layer.

Helpful Insightful Respectful

Andy 01

IOTD Staff

Jul 7, 2017 · Edited: Jul 7, 2017

Hi Nico, Nice image and colour palette - however in my opinion it's too smooth, like the NR is overdone resulting in loss of detail.
Are you following J-P Metsavainio's tone mapping method? If so there appears to be an issue with your star removal that has lost the detail in the luminosity layer.

Helpful Insightful Concise

Nico Carver

Topic starter

Jul 7, 2017

You have not listed any noise reduction tools, but it looks like data has been a bit distorted in the darker areas - it looks like an artefact from noise reduction.

Victor Van Puyenbroeck:
Maskedstretch should give you better stars, but their shape might have changed during tonemapping or while making the synthetic L layer.

Andy:
like the NR is overdone resulting in loss of detail.
Are you following J-P Metsavainio's tone mapping method? If so there appears to be an issue with your star removal that has lost the detail in the luminosity layer.

Thank you for your replies! They were all very useful. I don't remember using any NR, but I think I may have "messed up" the tone mapping process (yes, J-P Metsavainio method as Andy guessed) by not making them carefully enough in Photoshop, stretching them too far, and then using these "not so great" tone maps in addition to the Ha Master to make the Lum, which resulted in the artifacts that look like NR blockiness. I am going to start over with my processing, and really take my time, maybe try out "Straton" software for making the starless versions.

As for the stars themselves. I started this project before I had really secured the setup very well, so I think there is more tilt in the Ha data then there is in the OIII and SII. Not sure about how to fix them being "flat" as Victor points out. Just saw Mike Oates IOTD of this region and he used RGB stars, which I think might be the answer?

Thanks again for the helpful constrictive criticism! I really appreciate it.

Respectful Supportive