Carina Nebula - which is better?

[Jeff]

I'm trying to get the hang of PixInsight by processing Carina Neb data from January. Have used quite different workflows for the following.
Which do you prefer ....and why?

Full resolution(original):
https://www.astrobin.com/full/334623/F/

Full resolution (reprocessed):
https://www.astrobin.com/full/334623/G/

Any processing tips from PI black-belts out are most welcome. :prey:

Jeff

[Atmos]

The data looks really good Jeff. I'd say that version G is better but I'd also say that the colour lies somewhere between these two :)

As an example (https://www.astrobin.com/full/330272/0/) this shows that although there is a lot of hydrogen in the area, there is other colour in there as well.
F shows colour variation but not the right colour. G looks closer but it's more of a heavy red colour cast and lacks the colour variation within the central nebula region.

[RickS]

Jeff,

I prefer G too but agree with Colin's assessment (colour is not quite right and more colour variation would be an improvement.)

If you are willing to describe your workflow(s) then I'd be happy to comment if there's anything that I think might help. How did you do the colour calibration? That's a good place to start!

Cheers,
Rick.

[Jeff]

Thanks Colin and Rick.

Appreciate your comments ... and agree that a more pleasing rendition is probably part way between the two. I often see heaps of red in Carina Neb pics, but tend to see more depth/structure with paler shades of pink.

My original PI workflow:
1. Batch Pre-Processing
2. Dynamic Crop
3. ABE ... to address light gradient from 82% moon
4. MLT (5/.5/2, 3/.45/2, 2.5/.4/2, 1.5/.4/2, .5/.35/2), without masking
5. HT from STF ... with some tweaks
6. LHE to increase contrast slightly
7. SCNR to help reduce green (especially halos around bright stars)
8. CT (non-linear stretch)
9. Resampled to 50% size and rotated

Reprocessed PI workflow:
1-2-3. Re-used original output from BPP/DC/ABE
4. Gentle MLT (3/.5/2, 2/.5/2) with masking ... protecting neb details
5. AH for linear stretching (instead of using HT from STF)
6. HT (manual tweak of channel white balance and dark point clipping)
7. CT (non-linear stretch)
8. CS to crank up colour saturation (too far...matching my red wine)
9. MLT again (5/.5/2, 3/.45/2, 2.5/.4/2, 1.5/.4/2, .5/.35/2) with mask
10. Resampled to 50% size and rotated

Cheers,
Jeff

[Atmos]

It’s making sense now. What you need to do is use the ColorCalibration script. What it’ll do is a white balance on the stars in your image.

ABE is removing the background cast so you’re getting a black background but you haven’t run any white balance.

[el_draco]

I looked at it straight from the detail/contrast point of view. In both cases G is better, more nebulosity and better contrast. Stars in the other one are less visible when located in nebulous areas. I like seeing other colour combinations in images, but its the detail that counts.

My2Cw

[Jeff]

Thanks Colin and Rom - I appreciate the guidance while I get the hang of things.

Did another reprocess in PI, adding a couple more steps early in the piece to improve star white balance:
- Background Neutralisation (defaults)
- Colour Calibration (defaults with a "starfield" region of interest)

Has significantly reduced the colour tweaking effort later in the workflow.

Updated crop attached.

High resolution is here:
https://www.astrobin.com/full/336123/0/

Still working through a few reference materials.
But throwing the racquet at a few images along the way with a cut-down workflow initially certainly aids the absorption and retention of info.

Thanks again.
:prey2::thanks:

[RickS]

Colours look at lot more natural in the new rendition, Jeff. Perhaps a little muted. A clipped luminance mask and a bit of the CT Sat curve would add a bit more zing.

Cheers,
Rick.

[Jeff]

Thanks Rick .... I'll give that a whirl.
Cheers.

[gregbradley]

Yeah that looks not bad. As Rick suggested its a bit muted as Carina is a very dynamic area and its not showing it full there. Close though and you are on the right track.

Greg.