Help processing my data - specifically star colors

3 replies391 views
Deepan Vishal avatar
Hi everyone, 

I am processing my first broadband target m81 and m82 which I shot in the last couple of months and I am looking for help to process my star colors. 

My initial 120s bin2 LRGB subs appeared too saturated and I again imaged a set of 60s bin1 subs to process stars out of it.  

My Work flow -
1. Process 120s LRGB + Ha separately 
2. Starless image - Remove stars with starXterminator 
3. Stars processing - Use 60 s subs 
  3.1. ABE 
  3.2. Linear fit blue channel to Red and Green channel
  3.3. Channel combination 
  3.4. Color calibration 
  3.5. Repaired HSV separation - Separate HSV and combine H, S, and unrepaired V channel . 
  3.6. Masked stretch - target mean 0.14 iteration 100
  3.7. Extract stars using starnet2 
  3.8. Color saturation 
4. Combine stars from step 3 to the starless image in step 2

Here's my results - 

Testing TS Optics 130/910 Photoline on M81



I am not happy with the star colors and I feel that it needs to be better.
Can some one point me what I am missing here?
Am I making mistakes in any processing steps here? Or is my data just bad?
Should I limit my exposure length?


You can find the preprocessed subs in the below location. 
https://drive.google.com/drive/folders/1ez6_dKoC-kgno0dPTQEkqbJdxgEWmwJv?usp=sharing

I'll be happy to hear your feedbacks and suggestions. 

Thanks,
Deepan
Helpful Engaging
SemiPro avatar
For star colours, I usually split them off after colour calibration. Then I use SCNR to remove all the green, and then I use the "correct magenta stars" script in PixInsight, even for broadband. Crank up the saturation a bit and bam. You might have to use some ACDNR to smooth out some of the pixelation.



M13 From Bortle 8


That is a picture of mine using such a technique.

Your image seems a bit too blue overall. So maybe something went wrong during colour calibration? 

Amazing Crescent Nebula by the way!
Helpful Engaging Supportive
SemiPro avatar


Took a quick stab at it, and yeah something was definitely funky during the colour calibration part
Deepan Vishal avatar
For star colours, I usually split them off after colour calibration. Then I use SCNR to remove all the green, and then I use the "correct magenta stars" script in PixInsight, even for broadband. Crank up the saturation a bit and bam. You might have to use some ACDNR to smooth out some of the pixelation.



M13 From Bortle 8


That is a picture of mine using such a technique.

Your image seems a bit too blue overall. So maybe something went wrong during colour calibration? 

Amazing Crescent Nebula by the way!

Thanks for your suggestions and kind words!
The posted M13 is a great example and that's the kind of colors that I am looking to get.
Well Written Respectful
Related discussions
Processing competition to win time on the Chile CDK 17”
Hey everyone! So last year I had the amazing opportunity to purchase a setup and send it to the Obstech observatory down in Chile. Teaming up with @Deep Sky West (Lloyd) and @Tolga (all credit goes to these two, as everyone knows they are extremely t...
Author processes competition data; may find peer processing techniques helpful for their own work.
Mar 20, 2021
Looking for guide/tutorial/ressources for star layer processing
Dear fellow star processors, I recently turned unsatisfied with the faint halos I get on medium stars, between the diffraction spikes. See the following crop : Thinking that this was an acquisition issue, I hopened a thread in the corresponding forum...
Directly addresses star layer processing and color/halo issues on stars.
Feb 22, 2024
A silk purse from a sows ear? A PI and BlurXt 'fix' for poor colour as well as star shape using an old refractor
I am the owner of an old F7.0 WO 110 mm 'apo' triplet. It is a great scope for visual - and I am fond of it – but when taken down to F 5.5 using a FLAT III6A 0.8X reducer/ flattener it doesn't yield an entirely flat field for astrop...
Discusses star color correction and processing techniques for improving star appearance.
Jan 23, 2024