Sulfur in RGB pallet - Forums

Topic starter

Sep 15, 2022

Sorry if this has been asked before, I tried searching but couldn't find anything.

I am in the beginning stages of acquiring data for a new project. I'd like to have a more natural pallet but still use narrowband data for color and detail. I am thinking about a LHASORGB pallet, as in Ha = red, O = blue but would sulfur really be green? Seems it should be yellow but I am not 100% certain, in that case I'd assume it would be 50/50 between red and green. Has anyone processed in this pallet before? I'd be interested to see how it looks before I commit telescope time to gathering sulfur data.

Engaging

Björn Arnold

Sep 15, 2022

I haven’t processed this palette before but placing [SII] in green is certainly bringing you further away from natural colors.

Using narrowband filters improves contrast and helps to bring out faint nebula detail. If you want natural colors, you need a very dark sky and capture LRGB only. You‘d probably be disappointed as many large nebulae will have a pink/purple color (here are two recent examples from my own gallery: one, two).
Regarding color tones, both Ha and [SII] are red (Hue = 0 in HSL).

Now some (hopefully) more constructive idea:
I‘d rather try to use the NB data to highlight the nebula areas by adding just a bit of color but mostly use it in the luminance in order to emphasize the nebulae and reduce star sizes and increase contrast.

I‘m curious to see your results — and of others of course.

Björn

Well Written Helpful Insightful Respectful Engaging Supportive

Blaine Gibby

Topic starter

Sep 15, 2022

placing [SII] in green is certainly bringing you further away from natural colors.

Yes this is what I was afraid of.

I have a secondary setup that consists of a Redcat51 with an ASI2600mm, and I use it mainly for somewhat experimental runs while my primary setup is going. Sometimes it produces great results, sometimes its a good learning opportunity!

I am really looking to get the detail of narrowband with more natural color nebulosity and stars. NB as luminosity would certainly do that, but I am hoping to get more blue tones than what typically comes through in RGB nebulae. I may just omit the sulfur in the end but I'll probably still acquire it and try blending it in as yellow because for some reason that makes sense

Engaging

kuechlew

Sep 15, 2022

You may check: https://www.youtube.com/watch?v=cu76vNPq4Tc

There are endless discussions in various fora what the "real colors" of an object are. The video refers to the proper RGB values of the absorption/emission lines of H, S and O.

There is nothing wrong from an artistic or scientific viewpoint to assign "wrong" colors, you'll find a lot of SHO, HSO or HOO images here in the gallery. There are more elaborate formulae in use, e.g. PixInsight : Dynamic narrowband combinations with PixelMath (thecoldestnights.com) and (3) The Forax Narrowband Palette - YouTube

So have fun to experiment.

Clear skies
Wolfgang

Helpful

Roger Nichol

Sep 15, 2022

If you wanted to keep it natural, then Sii at 672nm is slightly further into near infra-red than Ha, at 656nm. The human eye responds up to about 700nm, but the efficiency at Ha and Sii wavelengths is low - maybe 10-15%. So the natural colour of Sii and Ha would be a deep red, but quite dim. Since the Ha and Sii have a very similar 'natural' colour, it is normal to remap to other colours so that they can be distinguished from each other.

Well Written Helpful Insightful Concise

Tim Hawkes

Sep 15, 2022 · Edited: Sep 15, 2022

As stated above a 'naturalistic pallete' would have both the SII and HA contributing to the red with O3 to green and blue. Optionally to allow for H beta - the presence of which is implicit if HA is present - can also, for example, add 10% of the HA into the blue channel.

Personally - depending on the object - I usually prefer a naturalistic pallete to using the Hubble even if that means not adding any (normally rather weak) SII at all. It just depends what you want the picture to show. For e,g, Hubble depictions of Type II HA nebulae the SII and OIII contributions are normally greatly addiitionally stretched to match the much stronger HA before being recombined into the final NB image. This stretching also amplifies the noise - and so often a lot of OIII and SII data has to be collected to get good SNR. Hubble palette mages have the scientific value of providing a good depiction of the relative geographic locations of the three emissions - normally OIII near the centre closest to the hottest stars and SII near the ionization edge but are highly misleading in terms of their relative strengths.

I think that in the most 'honest' naturalistic depictions of type II nebulae - in which the OIII is stretched only the same amount as the HA - i.e. it is not stretched up to match the HA - the resultant images are predominantly red with the location of the main OIII signal evident just as regions of pink or white (as opposed to the attractive blue that can result from stretching the OIII up to the level of the HA). The problem is perhaps that such images look less exciting - and there is really no point using an SII filter at all. Then, there are further complications such as the natural colour response of the human eye - on which basis it is legitimate to boost -maybe 2-3 fold? - the blue/green relative to the very deep reds of SII and HA. But probably all of these nebulae would be too diffuse to trigger colour vision at all anyway ? So in the end it is all a somewhat arbitrary choice.

(note that the above comments apply only to type II objects like the wizard and the pacman – but not of course to supernovae remnants, planetary nebulae etc. where HA is not dominant)

Tim

Helpful Insightful

Blaine Gibby

Topic starter

Sep 15, 2022

Thanks for all your very helpful input. It will be fun to experiment!

Well Written Respectful

kuechlew

Sep 15, 2022 · Edited: Sep 15, 2022

Tim Hawkes:
As stated above a 'naturalistic pallete' would have both the SII and HA contributing to the red with O3 to green and blue. Optionally to allow for H beta - the presence of which is implicit if HA is present - can also, for example, add 10% of the HA into the blue channel.

Personally - depending on the object - I usually prefer a naturalistic pallete to using the Hubble even if that means not adding any (normally rather weak) SII at all. It just depends what you want the picture to show. For e,g, Hubble depictions of Type II HA nebulae the SII and OIII contributions are normally greatly addiitionally stretched to match the much stronger HA before being recombined into the final NB image. This stretching also amplifies the noise - and so often a lot of OIII and SII data has to be collected to get good SNR. Hubble palette mages have the scientific value of providing a good depiction of the relative geographic locations of the three emissions - normally OIII near the centre closest to the hottest stars and SII near the ionization edge but are highly misleading in terms of their relative strengths.

I think that in the most 'honest' naturalistic depictions of type II nebulae - in which the OIII is stretched only the same amount as the HA - i.e. it is not stretched up to match the HA - the resultant images are predominantly red with the location of the main OIII signal evident just as regions of pink or white (as opposed to the attractive blue that can result from stretching the OIII up to the level of the HA). The problem is perhaps that such images look less exciting - and there is really no point using an SII filter at all. Then, there are further complications such as the natural colour response of the human eye - on which basis it is legitimate to boost -maybe 2-3 fold? - the blue/green relative to the very deep reds of SII and HA. But probably all of these nebulae would be too diffuse to trigger colour vision at all anyway ? So in the end it is all a somewhat arbitrary choice.

(note that the above comments apply only to type II objects like the wizard and the pacman -- but not of course to supernovae remnants, planetary nebulae etc. where HA is not dominant)

Tim

Very well explained Tim and I tend to agree. As a shameless self-promotion I suggest two of my own images as an example:
North America and Pelican Nebula (QHY268C, N.I.N.A, L-Extreme first light) ( kuechlew ) - AstroBin shows sort of a "natural" look with maybe a bit of a too bright red for Ha but at least OIII is not overdone.
Elephant's Trunk widefield (ZWO EAF first light) ( kuechlew ) - AstroBin is from my recent short colorful phase with OIII way over the top (Ha background nebulosity is way too strong either ...) .
As a beginner I'm currently experimenting with my development. From an artistic viewpoint there is nothing wrong with overpushing the colors - like I did in my second image. It just has nothing to do with a "natural look" at all.

Interestingly both images got about the same number of likes. So if you want to take the likes as a measure for the preferences of the community there seem to be adherents of both directions.

Clear skies
Wolfgang

Helpful Insightful Respectful Engaging