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

Combining HOO OSC final image with broadband OSC image in Pixinsight

  • Please log in to reply
13 replies to this topic

#1 stockhatch

stockhatch

    Explorer 1

  • *****
  • topic starter
  • Posts: 53
  • Joined: 03 Dec 2023
  • Loc: Raleigh NC

Posted 16 April 2024 - 03:07 PM

Hey all. IDK if I am phrasing this question properly, but I have a few hours of broadband OSC data, and some narrowband data of the same target that I shot with a dual band filter (Player One Anti Halo Pro).

 

I have processed the narrowband data via HOO and narrowband optimization, and now I want to combine that data with the broadband data. Should I process the broadband data fully as well and then use pixelmath to combine the final images, or should I somehow map the narrowband data into the unprocessed broadband stack and process it all together?

 

If the first option is best, how would I align the frames? I assume plate solve them and use that to align them somehow? 

 

I'm still pretty new to Pixinsight, so please be kind if this is some common process. I can't seem to find a tutorial on it.

 

TIA, and clear skies!


Edited by stockhatch, 16 April 2024 - 03:12 PM.


#2 acrh2

acrh2

    Mercury-Atlas

  • *****
  • Posts: 2,730
  • Joined: 16 Mar 2021

Posted 16 April 2024 - 03:14 PM

Aligning different data is easy - Star Alignment process in Pixinsight, which is pretty self-explanatory.

But what is the point of combining your OSC data with your dual narrowband data? What are you shooting?


  • stockhatch likes this

#3 stockhatch

stockhatch

    Explorer 1

  • *****
  • topic starter
  • Posts: 53
  • Joined: 03 Dec 2023
  • Loc: Raleigh NC

Posted 16 April 2024 - 03:18 PM

Aligning different data is easy - Star Alignment process in Pixinsight, which is pretty self-explanatory.

But what is the point of combining your OSC data with your dual narrowband data? What are you shooting?

Okay, I'll check that out. I'm shooting IC443, and wanted the additional nebulosity data from the narrowband to see what, if any, difference it makes in the final result.



#4 Rydeen 98

Rydeen 98

    Mariner 2

  • *****
  • Posts: 206
  • Joined: 06 Feb 2018
  • Loc: New Mexico

Posted 16 April 2024 - 03:36 PM

I'm following this to see if there is a better way.

 

As for me, I've been combining my data before stretching.  Register all images to the same reference frame.  Then stack the color and narowband seperately.  For both images I perform background extraction -> BXT correct only -> color calibration -> BXT -> NXT and additionally Bill Blanshans SCNR on the broadband data.  From there remove the stars from both images with SXT with screened stars.  Save the broadband stars to add back in later.  Split the channels of the narrowband data and extract out the Ha and Oiii (there's a few ways of doing this but the simplest is probably just to use the red channel as Ha, and the sum of the green and blue channel for OIII, there are also pixel math formulas for more accurate representations if you search the forums).

 

Here's where I may be using an outdated method but I still use the old NBRGBCombination script to generate a combined image using the broadband and the extracted Ha and OIII data.  There's a really old video from Galactic Hunter on this if you can find it on youtube.  I do another round of color calibration on this image and proceed to stretch from there.  When satisfied with the stretch I process my color star image that I extracted earlier and add it to the final image using pixel math.

 

This probably seems like a lot but it's a super condensed explanation of a complicated process.  I've never seen a great start to end tutorial for this type of composite image.  I learned it from watching mishmash of many tutorials.

 

Maybe someone has a better easier way of doing it these days.  I haven't kept up with processing changes on this front.


Edited by Rydeen 98, 16 April 2024 - 03:50 PM.


#5 acrh2

acrh2

    Mercury-Atlas

  • *****
  • Posts: 2,730
  • Joined: 16 Mar 2021

Posted 16 April 2024 - 03:39 PM

Okay, I'll check that out. I'm shooting IC443, and wanted the additional nebulosity data from the narrowband to see what, if any, difference it makes in the final result.

That's primarily an emission nebula, so unless you have a super dark sky, I would recommend using the OSC data for stars only. Otherwise, adding noisy OSC data to narrowband data would most likely degrade the overall quality of the image.

 

But if you really have you heart set on it, you could use the narrowband data as luminance for OSC colors. You would process those separately, as you have done, and then use LRGBCombination process. The procedure to apply luminance to OSC data is simple - open the LRGBCombination process, select the luminance view under L, uncheck the other three RGB boxes, and drop the process triangle (New instance) on top of the OSC view. 

If you feel like your OSC data has some important features that the luminance data is missing, you can use Pixelmath to make a mix of the two and use that as luminance. 



#6 stockhatch

stockhatch

    Explorer 1

  • *****
  • topic starter
  • Posts: 53
  • Joined: 03 Dec 2023
  • Loc: Raleigh NC

Posted 16 April 2024 - 03:46 PM

I'm following this to see if there is a better way.

 

As for me, I've been combining my data before stretching.  Register all images to the same reference frame.  Then stack the color and narowband seperately.  For both images I perform background extraction -> BXT correct only -> color calibration -> BXT -> NXT and additionally Bill Blanshans SCNR on the broadband data.  From there remove the stars from both images with SXT with screened stars.  Save the broadband stars to add back in later.  Split the channels of the narrowband data and extract out the Ha and Oiii (there's a few ways of doing this but the simplest is probably just to use the red channel as Ha, and the sum of the green and blue channel for OIII, there are also pixel math formulas for more accurate representations if you search the forums).

 

Here's where I may be using an outdated method but I still use the old NBRGBCombination script to generate a combined image using the broadband and the extracted Ha and OIII data.  There's a really old video from Galactic Hunter on this if you can find it on youtube.  I do another round of color calibration on this image and proceed to stretch from there.  When satisfied with the stretch I process my color star image that I extracted earlier and add it to the final image using pixel math.

 

The probably seems like a lot but it's a super condensed explanation of a complicated process.  I've never seen a great start to end tutorial for this type of composite image.  I learned it from watching mishmash of many tutorials.

 

Maybe someone has a better easier way of doing it these days.  I haven't kept up with processing changes on this front.

Great explanation of the workflow, thank you!



#7 stockhatch

stockhatch

    Explorer 1

  • *****
  • topic starter
  • Posts: 53
  • Joined: 03 Dec 2023
  • Loc: Raleigh NC

Posted 16 April 2024 - 03:47 PM

That's primarily an emission nebula, so unless you have a super dark sky, I would recommend using the OSC data for stars only. Otherwise, adding noisy OSC data to narrowband data would most likely degrade the overall quality of the image.

 

But if you really have you heart set on it, you could use the narrowband data as luminance for OSC colors. You would process those separately, as you have done, and then use LRGBCombination process. The procedure to apply luminance to OSC data is simple - open the LRGBCombination process, select the luminance view under L, uncheck the other three RGB boxes, and drop the process triangle (New instance) on top of the OSC view. 

If you feel like your OSC data has some important features that the luminance data is missing, you can use Pixelmath to make a mix of the two and use that as luminance. 

Thanks! It may not be adding anything valuable. I'm just curious, and have the data already so I figured I would learn how to combine it as an experiment. I appreciate the help!



#8 Rydeen 98

Rydeen 98

    Mariner 2

  • *****
  • Posts: 206
  • Joined: 06 Feb 2018
  • Loc: New Mexico

Posted 16 April 2024 - 04:00 PM

Thanks! It may not be adding anything valuable. I'm just curious, and have the data already so I figured I would learn how to combine it as an experiment. I appreciate the help!

It's definitely a worthwhile process on several galaxies that have Ha activity.  It makes all of the areas of extragalactic emission nebulosity pop out with red speckles.  I've never seen much come out of the additional OIII data but since we've collected the data I combine it in anyway. 

 

It's also worth collecting broadband in the case of nebulae so you can have true color stars.  Narrowband produces either all white or worse magenta stars.  Having nice broadband color stars makes a huge difference in the final image.



#9 acrh2

acrh2

    Mercury-Atlas

  • *****
  • Posts: 2,730
  • Joined: 16 Mar 2021

Posted 16 April 2024 - 04:25 PM

If you want to add Ha to galaxies, here's a pretty good guide:

https://www.nightpho...and-combination



#10 stockhatch

stockhatch

    Explorer 1

  • *****
  • topic starter
  • Posts: 53
  • Joined: 03 Dec 2023
  • Loc: Raleigh NC

Posted 16 April 2024 - 04:56 PM

I tried CombineRGBandNarrowband and also tried LRGBCombination, but neither align the images, even if I leave all of the stars in. The rotation and alignment are very close, as it's the same framing and imaging train just different filter, but stacking artifacts etc have cause the images to be slightly different sizes, which causes them to be misaligned when combined.


Edited by stockhatch, 16 April 2024 - 04:56 PM.


#11 nwcs

nwcs

    Apollo

  • -----
  • Posts: 1,113
  • Joined: 12 Nov 2013
  • Loc: Tennessee

Posted 16 April 2024 - 05:01 PM

What I do is split out the Ha from the dual band image. Then I remove stars from it and from the rgb. I’ll use Jurgen’s toolbox script to do continuum subtraction from the Ha and then use his Combine RGB with Ha to add it back into the rgb. Works great and is straightforward. I’ve tried other approaches and some are better than others but this works well. I know there’s a script Mike Cranfield worked with Adam Block on that can do the same thing but it’s not as streamlined.


  • stockhatch likes this

#12 Rydeen 98

Rydeen 98

    Mariner 2

  • *****
  • Posts: 206
  • Joined: 06 Feb 2018
  • Loc: New Mexico

Posted 16 April 2024 - 05:02 PM

I tried CombineRGBandNarrowband and also tried LRGBCombination, but neither align the images, even if I leave all of the stars in. The rotation and alignment are very close, as it's the same framing and imaging train just different filter, but stacking artifacts etc have cause the images to be slightly different sizes, which causes them to be misaligned when combined.

This is why I register them all to the same reference image at the beginning.  If you have to crop out stacking artifacts, use dynamic crop to make the same crop to both the broadband and narrowband images.  If you didn't do that you can still align the images with dynamic alignment selecting a bunch of common stars.  This will generate a registered image out of one of them that can be combined with the other.


  • stockhatch likes this

#13 stockhatch

stockhatch

    Explorer 1

  • *****
  • topic starter
  • Posts: 53
  • Joined: 03 Dec 2023
  • Loc: Raleigh NC

Posted 16 April 2024 - 05:13 PM

Thanks all! I got them all lined up with DynamicAlignment. :)



#14 Juggernaut

Juggernaut

    Viking 1

  • *****
  • Posts: 571
  • Joined: 17 Feb 2013

Posted 27 April 2024 - 11:37 AM

Hi,

I figured out how to do this with Siril.

https://youtu.be/Mqo079GOOxI

 

I hope this helps!

Paul


  • stockhatch likes this


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