Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

Should I remove stars in Ha subs before blending it with RGB?

  • Please log in to reply
13 replies to this topic

#1 Dwight J

Dwight J

    Gemini

  • *****
  • topic starter
  • Posts: 3,433
  • Joined: 14 May 2009
  • Loc: Lethbridge, Alberta, Canada

Posted 08 December 2019 - 02:23 PM

I have finished taking Ha subs using my modded camera and Canon 200 mm at F4 on the Rosette and California nebulae.  I now have to obtain some RGB subs to complete the colour.  My question is:  should I remove stars from the Ha stacked image prior to blending it in the Red channel?  I have read conflicting procedures with latest being that star colours will be adversely affected if you blend in the stars from the Ha frames.  I really like the dust fine Ha stars though but I would like better star colour if possible.  



#2 bobzeq25

bobzeq25

    Hubble

  • *****
  • Posts: 18,430
  • Joined: 27 Oct 2014

Posted 08 December 2019 - 03:19 PM

I don't.  I use NBRGBCombination in PixInsight.  The processing is not trivial.



#3 Poochpa

Poochpa

    Surveyor 1

  • *****
  • Posts: 1,904
  • Joined: 12 Feb 2005
  • Loc: Buffalo Burbs, NY

Posted 08 December 2019 - 03:59 PM

As shown below, on my most recent project, for the first time, I removed the stars in the Ha image before combining with the rgb. I think it produces a better result. Specifically, it seems to lessen the bright color star halos I get after combining Ha and rgb star images. 

Mike

https://www.cloudyni...658_1080522.png


Edited by Poochpa, 08 December 2019 - 04:02 PM.

  • psandelle and elmiko like this

#4 gatsbyiv

gatsbyiv

    Viking 1

  • *****
  • Posts: 703
  • Joined: 29 Mar 2009
  • Loc: Doylestown, PA

Posted 08 December 2019 - 04:13 PM

Since the stars in Ha are so much sharper than RGB, another option is to keep luminance from your Ha stars and get their color from the RGB data.  Since it sounds like you are doing an HaRGB image, the easiest thing to to simply add your Ha image as luminance to your RGB image, but there are lots of other approaches.


  • elmiko likes this

#5 Dwight J

Dwight J

    Gemini

  • *****
  • topic starter
  • Posts: 3,433
  • Joined: 14 May 2009
  • Loc: Lethbridge, Alberta, Canada

Posted 08 December 2019 - 09:56 PM

Yes, an HaRGB image was what I had in mind.  I hadn’t thought about using Ha as a luminance frame though and that could work for star colour and size.  I use Photoshop and Astro Tools for processing.  AstroPixelprocessor would be my next step up for processing images.  I can also use Maxlm DL to for some steps if needed.  



#6 Dwight J

Dwight J

    Gemini

  • *****
  • topic starter
  • Posts: 3,433
  • Joined: 14 May 2009
  • Loc: Lethbridge, Alberta, Canada

Posted 08 December 2019 - 10:08 PM

An additional question popped up for me; when I had a Starlight Xpress MX7C, the non debayered image was used as the luminance and the coloured images were blurred as they did not add to the detail, that was done by the luminance.  Would it matter if the RGB “layers” were blurred or not.



#7 the Elf

the Elf

    Vanguard

  • *****
  • Posts: 2,046
  • Joined: 06 Sep 2017
  • Loc: Germany

Posted 09 December 2019 - 11:09 AM

Short answer: yes!

Long answer: https://youtu.be/OxX1aC6kiwM

The process takes gatsbyiv's point into account, the sharp stars from Ha are used as well. For quick reference here is a diagram taken from the video:

 

workflow.jpg


  • Dwight J likes this

#8 Dwight J

Dwight J

    Gemini

  • *****
  • topic starter
  • Posts: 3,433
  • Joined: 14 May 2009
  • Loc: Lethbridge, Alberta, Canada

Posted 09 December 2019 - 07:27 PM

Short answer: yes!

Long answer: https://youtu.be/OxX1aC6kiwM

The process takes gatsbyiv's point into account, the sharp stars from Ha are used as well. For quick reference here is a diagram taken from the video:

 

attachicon.gif workflow.jpg

I don't have PI but most of those steps can be mimicked in Maxlm DL which I do have.  



#9 Samir Kharusi

Samir Kharusi

    Mercury-Atlas

  • *****
  • Posts: 2,613
  • Joined: 14 Jun 2005
  • Loc: Oman

Posted 10 December 2019 - 03:43 AM

There are a million ways of combining this stuff, none being particularly more "correct" than others. I'd offer a couple of possibly useful comments:

 

1. Stars are much, much brighter than your nebulae. You do not have to go through the rigors of taking "proper" subs and stacking just to capture their colors. Single, short exposures may suffice. All you need is color on the brightest stars. Life is short...

2. With the best intentions you may never get around to do the star color captures. but you may still get satisfying images just by false-color post processing wink.gif

 

Since both your mentioned objects are primarily Ha, I offer examples below of

1. Using quickie star-color frames added to a pure Ha stack, background colored as per next image below, just superposed by brightening:

large.jpg

 

and 2. using a gradient mask on a pure Ha stack to make all the stars white (and also the background greenish blue):

large.jpg

 

Yes, you already have done the hard work of capturing your Ha stacks. Post processing can yank them away from being all red in case you never get around to getting those extra white light frames.



#10 Dwight J

Dwight J

    Gemini

  • *****
  • topic starter
  • Posts: 3,433
  • Joined: 14 May 2009
  • Loc: Lethbridge, Alberta, Canada

Posted 10 December 2019 - 02:49 PM

There are a million ways of combining this stuff, none being particularly more "correct" than others. I'd offer a couple of possibly useful comments:

 

1. Stars are much, much brighter than your nebulae. You do not have to go through the rigors of taking "proper" subs and stacking just to capture their colors. Single, short exposures may suffice. All you need is color on the brightest stars. Life is short...

2. With the best intentions you may never get around to do the star color captures. but you may still get satisfying images just by false-color post processing wink.gif

 

Since both your mentioned objects are primarily Ha, I offer examples below of

1. Using quickie star-color frames added to a pure Ha stack, background colored as per next image below, just superposed by brightening:

large.jpg

 

and 2. using a gradient mask on a pure Ha stack to make all the stars white (and also the background greenish blue):

large.jpg

 

Yes, you already have done the hard work of capturing your Ha stacks. Post processing can yank them away from being all red in case you never get around to getting those extra white light frames.

 

I have already used the false colour action in PS Astro Tools and yes, the images turned out nice.  I have since added more Ha to both but haven’t stacked all of it.  I plan to get maybe 60 minutes worth of RGB or less depending on the weather which is continuous cloud.  



#11 Samir Kharusi

Samir Kharusi

    Mercury-Atlas

  • *****
  • Posts: 2,613
  • Joined: 14 Jun 2005
  • Loc: Oman

Posted 11 December 2019 - 12:30 AM

The point I was making in 1. is that you do NOT need 60 minutes for star colors. 2 minutes may be enough. You do not even have to set up proper tracking/autoguiding. These star colors came from 10 second subs:

large.jpg

 

The stars above are very dim, M13, but I was using a C14 at f11. Just try things out. You can in fact capture a decent spectrum for any star visible to the naked eye, i.e. brighter than Mag 7, with a single 30 second exposure at f2.8. Push the boundaries and you may be surprised at how short a single exposure can deliver all the star colors you wish to overlay on your Ha stack. Good hunting!



#12 Dwight J

Dwight J

    Gemini

  • *****
  • topic starter
  • Posts: 3,433
  • Joined: 14 May 2009
  • Loc: Lethbridge, Alberta, Canada

Posted 11 December 2019 - 02:04 AM

The point I was making in 1. is that you do NOT need 60 minutes for star colors. 2 minutes may be enough. You do not even have to set up proper tracking/autoguiding. These star colors came from 10 second subs:

large.jpg

 

The stars above are very dim, M13, but I was using a C14 at f11. Just try things out. You can in fact capture a decent spectrum for any star visible to the naked eye, i.e. brighter than Mag 7, with a single 30 second exposure at f2.8. Push the boundaries and you may be surprised at how short a single exposure can deliver all the star colors you wish to overlay on your Ha stack. Good hunting!

 

I will certainly do that.  Now the moon is big so a short exp, will minimize any gradients induced by it.  With my luck the only clear sky will be at full moon.  



#13 Dwight J

Dwight J

    Gemini

  • *****
  • topic starter
  • Posts: 3,433
  • Joined: 14 May 2009
  • Loc: Lethbridge, Alberta, Canada

Posted 13 December 2019 - 04:22 AM

Here is what I have so far in Ha for NGC 1499.  12 nm clip in filter in a modded Canon T3i on a Canon 200 mm lens stopped to F4.  3.9 hours via 3 min. subs.  

 

3D294D70-0EB1-4FD5-A1FC-ED6168416718.jpeg

 

 


  • elmiko likes this

#14 the Elf

the Elf

    Vanguard

  • *****
  • Posts: 2,046
  • Joined: 06 Sep 2017
  • Loc: Germany

Posted 13 December 2019 - 04:10 PM

Lucky you! I got a single sub and then the clouds rolled in.

Looking good. You probably have to wait until the moon is gone to get some color data.




CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Recent Topics






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics