StarTools

Chunkles wrote: ↑Sun Nov 22, 2020 12:26 am using the L-enhance filter:

use things like the Hubble palette

Hi Scott,

I'm afraid that is not physically possible. A HST-palette image consists of three bands, namely S-II, H-alpha and O-III, mapped to red, green and blue respectively (e.g SHO:RGB).

The L-enhance filter records Ha, no S-II and records H-beta and O-III lumped into one band. E.g. you are short one band, while to O-III band will be mixed with H-beta signal as well.

The best(and most popular thing to do) you can do, is to create a bi-color as suggested in this thread.
A red/cyan rendering HOO rendering is the most popular as it shows where Ha emissions are, and where O-III/H-beta emissions are, while also roughly corresponding to their location in the visual spectrum.

So (in 1.7) load your data like this in the Compose module; Process your image (in mono), then, once you hit the Color module, use the Bi-Color preset and tweak to taste. Note you can use the Matrix option to flip between other popular channel mappings besides HOO (a feature unique to StarTools thanks to Tracking and being able to do color calibration so late in the workflow). Be prepared for a lot of blue stars as stars don't show up well in Ha (but do so in the O-III through Hb part of the spectrum).
Shrink the stars if you want and/or push them back with the Super Structure module.
Hope this helps!

Hi Ivo,

Please let me do a newbie question.
What's the point of composing an image from a rgb one? I intuitively get it, but when I think it twice I get confused. When shooting with a mono camera with filters, I get it; you have forcibly to compose and mount the image. But when you are dealing with a color image taken with the l-enhance filter, you already have the image composed with a red channel "filled" with Ha data (wavelengths roughly matches), green channel with Hb info, and blue channel with Oiii (last two ones more or less overlaps). Is right this approximation or it's a nonsense? If it makes more or less sense, which is the advantage? Build the Illuminance channel? Which is the logic behind?

Thanks!

mgutierrez wrote: ↑Wed Nov 25, 2020 8:46 pm What's the point of composing an image from a rgb one? I intuitively get it, but when I think it twice I get confused. When shooting with a mono camera with filters, I get it; you have forcibly to compose and mount the image. But when you are dealing with a color image taken with the l-enhance filter, you already have the image composed with a red channel "filled" with Ha data (wavelengths roughly matches), green channel with Hb info, and blue channel with Oiii (last two ones more or less overlaps). Is right this approximation or it's a nonsense?

Indeed, you are correct that the channels already roughly correspond to a red/cyan bi-color. Importing through the Compose module formalises that this is a narrowband dataset (in addition to giving some more flexibility with regards to compositing). With the application knowing that this is a narrowband dataset (and not a visual spectrum dataset), it knows - amongst other things - what to show you in the Matrix option in the Color module for example. A visual spectrum dataset from a DSLR will show vendor and model-specific color matrices, while a narrowband dataset will allow for extensive channel remappings (as the color rendition is regarded as false color).

Coloring will also definitely differ between a dataset whose coloring is left as RGB versus a dataset whose coloring is forced to R(G+B)(G+B). The latter will yield a bi-color image where the coloring is comparable predictable and identical, no matter whether you use a duoband filter, tri-band filter or quad-band filter, and no matter what the response of your sensor may be to the different bands (e.g. whether your camera is, for example, Ha modded or not). And as you know, replicability and consistency is a highly sought after attribute in photography and science!

Hope this helps and makes sense!

Thanks for the explanation, Ivo; it makes sense! One last question. Do I need to double the green exposure compared to red and blue when composing?

Thanks Ivo! That completely makes sense and is very helpful!

mgutierrez wrote: ↑Thu Nov 26, 2020 8:52 am Do I need to double the green exposure compared to red and blue when composing?

In 1.7 you don't, as long as you choose the long-but-descriptive 'L + Synth L From R(2xG)B, R(GB)(GB) (Bi-Color from OSC/DSLR)' option. The 'R(2xG)B' bit before the comma denotes luminance is re-weighted for you to incorporate 2x the green channel. So, just select this option, load the same dataset in each of the slots and you should be good to go. The Compose module does all the dirty work.

admin wrote: ↑Fri Nov 27, 2020 12:50 am In 1.7 you don't, as long as you choose the long-but-descriptive 'L + Synth L From R(2xG)B, R(GB)(GB) (Bi-Color from OSC/DSLR)' option. The 'R(2xG)B' bit before the comma denotes luminance is re-weighted for you to incorporate 2x the green channel. So, just select this option, load the same dataset in each of the slots and you should be good to go. The Compose module does all the dirty work.

got it!!! thanks ivo!

So i worked on the Rosetta..

With L-enhance , is this the way it should look?
Red HA on the outsides and the blue/cyan OIII

With L-enhance , is this the way it should look?

This definitely seems like a very plausible rendition! (and it looks quite nice too )

The Optolong L-eNhance passes H-alpha (red) and H-beta + O-III (blue)

Your image indeed shows red H-alpha dominance in all the right areas at your chosen signal balance, so anything else must be H-beta + O-III (because that is all else the filter passes). There are no artifacts here.

As you probably know, you can perfectly linearly throttle the H-alpha vs H-beta + O-III signal in the Color module by using the red (=H-alpha) and green (=H-beta + O-III) bias (you can also use the blue bias instead of the green bias; blue and green will do the same thing if a bi-color matrix is selected, exactly as it is supposed to).

In other words, you can perfectly choose the balance between Ha vs H-beta + O-III you wish to show. This is a somewhat subjective choice (there is no one right balance when showing false color narrowband images, as long as the relationship between the two spectrum bands remains perfectly linear, which StarTools ensures). I very much like your current rendition and balance though!

fmeireso wrote: ↑Wed Feb 24, 2021 6:58 pm So i worked on the Rosetta..

congrats for the result! I also tried this object last week, with the same filter, but I like more your processing than mine: