M43, color struggle

[devonshire]

Ivo,

I'm wondering if you can take a look at my M43?

https://www.amazon.ca/clouddrive/share/ ... 3PHJxvz8t3

I've tried this with 1.6.386 and have no real difficulty bringing out interesting detail in Luminance, but when I get to Color, it's pretty much all-one-thing-or-another, not much shading or gradation. This was taken with an Optolong CLS-CCD filter, so yes, it's dual-band, but those bands are not narrow and I'd expect more shading than I'm seeing.

The stack does benefit from cropping, both due to meridian flip misalignment, and a new 0.8 reducer I'm testing, which looks like it needs some spacing work done. That's just mechanics - it's the color I was bumping up against.

Thanks in advance for any help with this!

- Bob

[admin]

Hi Bob,

I'm not having any particular trouble with this dataset (aside from the obvious missing parts of the spectrum due o the CLS filter).

Screenshot-1.jpg (184.9 KiB) Viewed 6829 times

Some CLS filters can be particular harsh with cutting the yellow part of the spectrum. Nevertheless, you should still be able to recover the blue reflection nebulosity in the Running Man and M42, as well as the red Ha emissions and the teal green OIII emissions in the core of M42.

Is this what you expected? If not, can you help me understand what might be going wrong when you hit the Color module?

[devonshire]

Thanks, Ivo!

I guess what I was expecting to see was more variation in shading, particularly of the reds, and more relative brightness (and possibly variation) in the blues.
Here's a link to a roughly comparable image, taken with Astronomik's CLS-CCD filter, which delivers almost the same spectrum slices as the Optolong.
I'm just having trouble believing that there isn't more spectral variation within those slices than is showing up in the coloration.

https://www.cloudynights.com/topic/5697 ... 977-orion/

-Bob

[admin]

Hi Bob,

This is - IMHO - pretty much the most accurate you can expect in terms of true colour and emissions from this target in the visual spectrum, as filtered by a wide CLS filter.
When evaluating colouring, please always bear in mind what would be causing colour hues, and then work backward from there. Trust the data and don't try to impose what you feel an image "should" look like. Shades are remarkably uniform across different objects, even far apart, even when comparing features in entirely different galaxies. This is because they are largely caused by the same processes, same chemical makeup.

When it comes to M42 and the Running Man, things to look out for are a teal green M42 core (dominant OIII emissions - a rarity- , excited by the young O and B giants in the core), red Ha emissions (excited by the young O and B giants in the core), blue reflection nebulosity (fine dust reflecting the light of the young O and B giants in the core). NGC1977 should roughly match M42's hues, as they are roughly the same thing; HII areas with star formation driven by blue giants and associated reflection nebulosity (e.g. the reflection nebulosity in NGC1977 is not significantly greener than the reflection nebulosity in M42).

Of course, for obvious reasons, a CLS filter will cut some smooth blends and transitions from one shade to another if those blends are made up of any of the colors in the filtered bands. Anywhere white light contributes to the colouring of an object (for example stars or dust lanes) will have a compromised hue, devoid of any spectral data for those wavelengths, thereby being "segmented" into the two wide bands (red and blue) the CLS filter does pass through.

Hope this helps!

EDIT: You could experiment with the 'Legacy' preset, 'LRGB Method Emulation' and Bright/Dark saturation parameters. This will yield a less faithful/strict rendition, depending on the settings, which may be more to your liking.

[whixson]

Hi Bob! I’d be interested in more details - the camera and lens/telescope and the subframe image quantity and duration. Thanks!

Wayne

Ivo,

I'm wondering if you can take a look at my M43?

https://www.amazon.ca/clouddrive/share/ ... 3PHJxvz8t3

I've tried this with 1.6.386 and have no real difficulty bringing out interesting detail in Luminance, but when I get to Color, it's pretty much all-one-thing-or-another, not much shading or gradation. This was taken with an Optolong CLS-CCD filter, so yes, it's dual-band, but those bands are not narrow and I'd expect more shading than I'm seeing.

The stack does benefit from cropping, both due to meridian flip misalignment, and a new 0.8 reducer I'm testing, which looks like it needs some spacing work done. That's just mechanics - it's the color I was bumping up against.

Thanks in advance for any help with this!

- Bob

[devonshire]

Thanks, Ivo!

[devonshire]

Wayne,

Imaging train:
Nikon D5300 -> 0.8 Reducer -> Optolong CLS-CCD filter -> Stellarvue 120EDT (f6.1, approx f4.8 net w/reducer). Still testing the reducer - its spacing probably still needs work.
Tried mixed exposures to bring out detail without blowing out the center:
19 @ 30s
69 @ 90s
11 @ 180s
Plus bias and flats.
Conditions not great - seeing just ok, transparency not great, but at least there was no Moon. Gotta love this weather

HTH...

- Bob

[admin]

Nikon D5300
19 @ 30s
69 @ 90s
11 @ 180s

Don't mix exposure lengths when you pass data to your stacker. It will cause the signal-to-noise ratio to vary all over the place and also wreaks havoc on point spread functions (causing decon to fail and/or create artifacts). If you must do HDR, stack just two stacks, process them separately and blend them with the Layer module.

If you are using the latest version of DSS, make sure to keep it from white-balancing your data. Then, if you use 1.6 and import your data as such, you can select your D5300's camera matrix to mimic colour response as defined/optimised by the manufacturer. This may also really help produce a markedly different result to the standard colour response.

E.g. this is your stack colored using the Legacy preset, some tweaks and the D5300 profile (sadly that causes the core to show as white, even though it definitely has a distinct teal/green coloring due to the strong O-III emissions).

Master_Stack_16.jpg (342.67 KiB) Viewed 6773 times

[devonshire]

Ivo,

Thanks!

I'm pretty sure I tried the D5300 profile + cycling through all of the presets, and I was on the beta. That's a more artistic result than I was able to achieve. It may not have the color accuracy that you favor, but with the wider color differentiation, the texture of the nebulosity shows better. Tradeoffs...

- Bob

[admin]

Ivo,

Thanks!

I'm pretty sure I tried the D5300 profile + cycling through all of the presets, and I was on the beta. That's a more artistic result than I was able to achieve. It may not have the color accuracy that you favor, but with the wider color differentiation, the texture of the nebulosity shows better. Tradeoffs...

- Bob

No problem Bob!
Theoretically though, if luminance and colour is processed separately (as it is), texture should be wholly determined by luminance and any colouring should not modify that one bit. If it does, a screen calibration may be in order...
Colour differentiation in this version is also much less (for example, it is much harder to discern between blue reflection nebulosity and red Ha emissions, while O-III is completely lost). Perhaps I am misunderstanding what you are after?