RGB + H-Alpha Galaxy Star Formation Knots

9 replies901 views
Mau_Bard avatar
Hi All,
I am asking for your advice. I have recently elaborated an image of NGC2403, having RGB data and H-Alpha plus OIII. I used the NBRGBC process in PixInsight to combine the NarrowBand and RGB data. This is the image https://astrob.in/pqncwq/B/

One passage I was not able to manage properly was to isolate in the HA image the star-formation-"knots", in order to avoid that the full image RGB image gets a red tone from the HA. I tried also to develop a "Knot mask" extracted by the HA image, with no success.
The fact is that I could not get rid of the core of the galaxy that had a similar luminosity of the star forming regions.

I would appreciate your suggestions and experiences about techniques you use to extract a clean "H-alpha star formation knots" image.

I wish you a nice holiday season, and clear skies into 2022!

Mau
Respectful Engaging
David Nozadze avatar
Hi Mau!

This is a short video, showing how I processed my M31 session. 
Please turn on the captions, to see the explanation of each step. Hope this helps. 

https://youtu.be/ofAF4f7HAP8


CS

David
Well Written Concise Supportive
Wei-Hao Wang avatar
Just add H-alpha to the R channel in the initial linear stage, and dial down the R during color balance.  Then the red knobs will naturally stand out.
Well Written Concise
Mau_Bard avatar
Thanks David for your video, and Wei-Hao for your suggestion.
Adding H-a to RGB before the color calibration makes all the process way simpler! I will reprocess my image using the sequence you suggest for sure.

I express to you both all my appreciation - many thanks.

And on the occasion, I wish you a nice holiday season and a healthy new year!

Mau
Well Written Respectful Supportive
Wei-Hao Wang avatar
BTW, you don't even need to do a straight R + Ha.  You can do things like R + 0.2*Ha.  (0.2 is just an example.  You can try different factors and find out what you like the most.)  

My personal experience is that just a little additional Ha (like the 0.2 factor) plus subsequent saturation boost after color re-balancing is sufficient to give you very nice red nebulas along the spiral arms.  A factor that's smaller than 1 can also reduce the impact of noise in the Ha image, if the exposure time for Ha is not very long.  (I typically give Ha a total exposure time that's only about 1/3 of R.)

Cheers,
Wei-Hao
Well Written Helpful Insightful Concise
Mau_Bard avatar
Dear Wei-Hao thanks for your valuable additional advice! It is key in post processing steps to have good staring parameters and the conceptual background as well!

Cheers,
Mau
Fluorine Zhu avatar
Mau_Bard avatar
Thanks Fluorine, and congratulations for your M31 image: it is really gorgeous!
Well Written Respectful Supportive
Leonardo Landi avatar
I use another method, well explained in the book "mastering pixinsight". I use pixelmath with a max expression.  In other words, it replaces the r pixel with the ha pixel if the latter has a highter value. I do not remember the expression but I have an icon in my set. Give me your email and I'll send you.
Mau_Bard avatar
Thank you Leonardo, grazie.
I have just sent you my mail address in a private message here in astrobin.

The idea to pick the maximum is brilliant.
Probably the H-a's and Red image's medians have to be equalized before maximization, but I suppose that the PI formula takes care of that.

Following the interesting discussion here I was landing on averaging the R and H-a with Pixelmath, but definitively the Maximum is a more direct solution.
I am looking forward to see the math expression.

The second point I take from your post, is that the book "Mastering Pixinsight" is worth to be read. I enjoyed a lot W.A. Keller's "Inside Pixinsight", and it looks that Rogelio's book is giving additional perspective.

Thanks again, and I am looking forward to reading from you.
Happy new year! Ciao,

Maurizio
Respectful Supportive