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

Post your Pixinsight Processing Flow!

  • Please log in to reply
112 replies to this topic

#51 starhunter50

starhunter50

    Viking 1

  • *****
  • Posts: 564
  • Joined: 07 Sep 2010
  • Loc: Tilbury,Ontario 42N , 82W

Posted 20 May 2019 - 04:06 PM

@cfosterstars , do you have an updated or any Ha-L-RGB  workflow Docx ? just getting started on the way of imaging. 

Otherwise a good youtube Ha _ LRGB combination. i made a video but its not that accurate , im afraid.

This is for my personal imaging... 

 

Are you subscribed to Astrodude on Youtube?

 

Mitch


  • elmiko likes this

#52 cfosterstars

cfosterstars

    Mercury-Atlas

  • *****
  • topic starter
  • Posts: 2577
  • Joined: 05 Sep 2014
  • Loc: Austin, Texas

Posted 20 May 2019 - 04:36 PM

@cfosterstars , do you have an updated or any Ha-L-RGB  workflow Docx ? just getting started on the way of imaging. 

Otherwise a good youtube Ha _ LRGB combination. i made a video but its not that accurate , im afraid.

This is for my personal imaging... 

 

Are you subscribed to Astrodude on Youtube?

 

Mitch

Mitch,

 

I have a section on blending Ha data into LRGB data. In my opinion, it is the hardest thing to do in the processing. Others state it is easy but this is not my experience. I recently had much better success with it for galaxies than I have had in the past and am in the middle of a reprocess of a 48 hour data set on M101. I will do an update at that point. Basically, for me and that is what I have found from my own experience, is that you need a good mask for the areas that you want to enhance with the Ha in the emission regions. If you do a straight blend with the NbLRGBCombination process, you get way too much reddening of the image. I take the output of the NBLRGBCombinaiton, apply the Ha mask to the RGB, and then use the BLEND script or even pixel math to mix the two images. This prevents the NBLRGBCombo image from screwing up the over all color tone. The trick is to get an Ha mask that only has the Emission regions and excludes the stars if possible. I take a star mask that I use for Deconvolution and subtract it from the Ha mask to remove most of the stars. This works since its a mask. The blended combination in this manner really can enhance the Ha regions and not cause a reddening of the background or screw up the star color. I have a hard enough time with star artifacts as it is with the added pain from Ha blending. The section in my latest version is from when I was still experimenting and every image was basically a unreproducible track in phase space!!!. Even when I tried to do it the same way with the same data, it would come out completely different. Now, The method describe above has worked on several data sets and I am going back to old data - M101 for example - to see if I can improve on what I did previously. I know that this is a bit scetchy, but I have not had time to really do a good job on writing it down. 

 

By the way, I have not found - even with looking extensively - a good and solid video for Ha blending that was easy to follow. There was one for this from Richard Block I think, where he did something like this with M31, but Its not quite what I do. Its the only one out there that I think is on the right track. I keep hoping that Adam Block will do a video on his subscription on how he does this. Its been requested several time, but he has a long list of requests. 

 

Let me know if this make any sense to you.

 

chris


Edited by cfosterstars, 20 May 2019 - 04:38 PM.

  • starhunter50 likes this

#53 starhunter50

starhunter50

    Viking 1

  • *****
  • Posts: 564
  • Joined: 07 Sep 2010
  • Loc: Tilbury,Ontario 42N , 82W

Posted 20 May 2019 - 09:20 PM

Hey Chris, looks like i have my work cut out for me. 

will try your suggestions. see what comes out of it.

 

thanks , Mitch



#54 cfosterstars

cfosterstars

    Mercury-Atlas

  • *****
  • topic starter
  • Posts: 2577
  • Joined: 05 Sep 2014
  • Loc: Austin, Texas

Posted 20 May 2019 - 10:57 PM

Hey Chris, looks like i have my work cut out for me. 

will try your suggestions. see what comes out of it.

 

thanks , Mitch

I just went through the above blending with M101. It actually does work. The key as I stated above is in the blending mask. I just spend two hours getting the mask the way I wanted. However, I really like the results so far. The HII regions in the spiral arms really POP. What I still have trouble with is how to enhance the blue star forming regions. I dont have any OIII data and they look rather pale at this point.

 

I will re-write the section that I have in the workflow since this method seems to be rather robust.


  • starhunter50 likes this

#55 starhunter50

starhunter50

    Viking 1

  • *****
  • Posts: 564
  • Joined: 07 Sep 2010
  • Loc: Tilbury,Ontario 42N , 82W

Posted 21 May 2019 - 09:13 AM

Hi Chris,  after your done can you forward your PDF to me..  

take your time.  we are all grateful for this huge time consuming hobby that you prepare for others!

 

rcfmitch @ gmail . com 


  • elmiko likes this

#56 cfosterstars

cfosterstars

    Mercury-Atlas

  • *****
  • topic starter
  • Posts: 2577
  • Joined: 05 Sep 2014
  • Loc: Austin, Texas

Posted 21 May 2019 - 09:36 AM

Hi Chris,  after your done can you forward your PDF to me..  

take your time.  we are all grateful for this huge time consuming hobby that you prepare for others!

 

rcfmitch @ gmail . com 

Will do. I started to re-write the section of Ha/Nb blending with RGB data last night. I should finish in a few days.



#57 starhunter50

starhunter50

    Viking 1

  • *****
  • Posts: 564
  • Joined: 07 Sep 2010
  • Loc: Tilbury,Ontario 42N , 82W

Posted 21 May 2019 - 10:06 AM

Will do. I started to re-write the section of Ha/Nb blending with RGB data last night. I should finish in a few days.

Thank you sir.



#58 cfosterstars

cfosterstars

    Mercury-Atlas

  • *****
  • topic starter
  • Posts: 2577
  • Joined: 05 Sep 2014
  • Loc: Austin, Texas

Posted 24 May 2019 - 09:02 PM

I have an update to my process flow. As previously, I have done a full review of all sections for typos and clarity. I have update several parts of the preprocessing including registration to a single master reference and using low noise references for localnormalization. I also added a new scoring formula with variables that is much easier to use. The major changes are:

 

1) how to use the PSFImage script as part of deconvolution

2) a completely updated section of blending narrowband data into RGB data. This includes a detailed section on how to build a mask to isolate only the narrowband enhancements. This section is the major addition.

3) more detail for some of the final post processing

4) lots more screen captures

 

 

https://drive.google...dg2AXsNvBExcC_R

 

 


  • H-Alfa, R Botero, starhunter50 and 9 others like this

#59 starhunter50

starhunter50

    Viking 1

  • *****
  • Posts: 564
  • Joined: 07 Sep 2010
  • Loc: Tilbury,Ontario 42N , 82W

Posted 24 May 2019 - 09:09 PM

This Workflow deserves some kind of Pixinsight tutorial  "AWARD"

100 s of hours can be  wasted trying to get 1 image to look decent enough to show your friends.

Now you can follow an outline and get real results....

This Pixinsight lesson  is "PriceLess " AAA++

 

Mitch. 


Edited by starhunter50, 24 May 2019 - 09:10 PM.

  • elmiko, bsavoie and NuclearRoy like this

#60 cfosterstars

cfosterstars

    Mercury-Atlas

  • *****
  • topic starter
  • Posts: 2577
  • Joined: 05 Sep 2014
  • Loc: Austin, Texas

Posted 24 May 2019 - 09:15 PM

This Workflow deserves some kind of Pixinsight tutorial  "AWARD"

100 s of hours can be  wasted trying to get 1 image to look decent enough to show your friends.

Now you can follow an outline and get real results....

This Pixinsight lesson  is "PriceLess " AAA++

 

Mitch. 

Thanks so much. The Nb blending method that I added has taken me man-weeks to figure out. I found it basically impossible with many methods and it was a pain to get anything decent. So far, this method has been very solid, but you really need patience. It is NOT straight forward, but I think it really works. I have done it now on several targets and just went back to some old data set to redo with it. I just posted my M101 with really nice Ha data for H II regions just popping and no ugly stars or red cast. 

 

M101LHaRGB_NL_LessSharp_Crop_SM.jpg


Edited by cfosterstars, 24 May 2019 - 09:16 PM.

  • NMCN, elmiko, starhunter50 and 5 others like this

#61 niccoc1603

niccoc1603

    Ranger 4

  • -----
  • Posts: 318
  • Joined: 16 Aug 2018
  • Loc: Firenze, Italy

Posted 25 May 2019 - 02:56 PM

I have an update to my process flow. As previously, I have done a full review of all sections for typos and clarity. I have update several parts of the preprocessing including registration to a single master reference and using low noise references for localnormalization. I also added a new scoring formula with variables that is much easier to use. The major changes are:

1) how to use the PSFImage script as part of deconvolution
2) a completely updated section of blending narrowband data into RGB data. This includes a detailed section on how to build a mask to isolate only the narrowband enhancements. This section is the major addition.
3) more detail for some of the final post processing
4) lots more screen captures


https://drive.google...dg2AXsNvBExcC_R

Thank you for the awesome tutorial. A perfect complement to Keller's book!

#62 cfosterstars

cfosterstars

    Mercury-Atlas

  • *****
  • topic starter
  • Posts: 2577
  • Joined: 05 Sep 2014
  • Loc: Austin, Texas

Posted 25 May 2019 - 03:07 PM

Thank you for the awesome tutorial. A perfect complement to Keller's book!

I have used both editions of Keller's book. Its really good, but not really step by step and he is light on CMOS specific issues although volume 2 is much better than volume 1. My strongest recommendation is that after you get somewhat confortable with PI, subscribe to Adam Blocks tutorials. Its a big time investment, but unless you are willing to go to one of the live class instructions, that is the best option. 


  • bsavoie likes this

#63 H-Alfa

H-Alfa

    Ranger 4

  • -----
  • Posts: 386
  • Joined: 21 Sep 2006
  • Loc: Spain

Posted 02 June 2019 - 04:39 AM

This Workflow deserves some kind of Pixinsight tutorial "AWARD"
100 s of hours can be wasted trying to get 1 image to look decent enough to show your friends.
Now you can follow an outline and get real results....
This Pixinsight lesson is "PriceLess " AAA++

Mitch.

+1

Thank you Chris for this magnific document. 127 pages is in fact a book!
I will take a look to your updates asap.

Enviado desde mi ANE-LX1 mediante Tapatalk

#64 cfosterstars

cfosterstars

    Mercury-Atlas

  • *****
  • topic starter
  • Posts: 2577
  • Joined: 05 Sep 2014
  • Loc: Austin, Texas

Posted 29 June 2019 - 09:26 PM

Here is the latest version of my process flow. I have added a section on forming mosaics and a section on determining the gain and read noise of the you camera.

 

https://drive.google...Ku8aaKpLKTu-iug

 

I am struggling with a two year long project of a four-panel mosaic that I would like to finish. I will probably have to update my flow several times on how to get this correct. There is also a new method for merging NB data that I have just found out about that I will test out and add later also. This is a never ending project in and of itself.


  • elmiko likes this

#65 cfosterstars

cfosterstars

    Mercury-Atlas

  • *****
  • topic starter
  • Posts: 2577
  • Joined: 05 Sep 2014
  • Loc: Austin, Texas

Posted 03 July 2019 - 09:59 PM

Lots of updates to the Mosaic processing section. I finished my Sagittarius Mosaic and learned a bunch in the process. Here you go:

 

https://drive.google...SO0ZbHXyPpEDQF-


  • bobzeq25, astronate, Dynan and 1 other like this

#66 bobzeq25

bobzeq25

    Hubble

  • *****
  • Posts: 16931
  • Joined: 27 Oct 2014

Posted 03 July 2019 - 11:11 PM

Lots of updates to the Mosaic processing section. I finished my Sagittarius Mosaic and learned a bunch in the process. Here you go:

 

https://drive.google...SO0ZbHXyPpEDQF-

Wow.  Fantastic information on the details of using PI.


  • bsavoie likes this

#67 pfile

pfile

    Fly Me to the Moon

  • -----
  • Posts: 5186
  • Joined: 14 Jun 2009

Posted 04 July 2019 - 09:34 AM

i hate to say this in the face of so much documentation and work, but why not use MosaicByCoordinates to come up with the input panels for GradientsMergeMosaic? it takes care of all the synthetic star field generation behind the scenes. all you have to do is run ImageSolver on all the input panes and save them to disk, then run MosaicByCoordinates.

 

i made this 12-panel mosaic using that technique and it was pretty painless.

 

https://www.astrobin...304367/?nc=user

 

rob


  • cfosterstars likes this

#68 cfosterstars

cfosterstars

    Mercury-Atlas

  • *****
  • topic starter
  • Posts: 2577
  • Joined: 05 Sep 2014
  • Loc: Austin, Texas

Posted 04 July 2019 - 09:49 AM

i hate to say this in the face of so much documentation and work, but why not use MosaicByCoordinates to come up with the input panels for GradientsMergeMosaic? it takes care of all the synthetic star field generation behind the scenes. all you have to do is run ImageSolver on all the input panes and save them to disk, then run MosaicByCoordinates.

 

i made this 12-panel mosaic using that technique and it was pretty painless.

 

https://www.astrobin...304367/?nc=user

 

rob

Rob,

 

I had never heard of this method nor this new script. It is not one of the methods that I had found looking for way to process mosaics. If it is better that what I did, that is great, because it was enough to make me rethink mosaics in general. What I did was not easy or even optimized. I will check this out.

 

However, most of the work was not in generating the input panels. Although a bit time consuming, it was rather straight forward. The difficult part was in the blending with GMM. That was 80% of the time I spent.

 

Thanks,

 

Chris


Edited by cfosterstars, 04 July 2019 - 09:51 AM.


#69 Peregrinatum

Peregrinatum

    Ranger 4

  • *****
  • Posts: 397
  • Joined: 27 Dec 2018

Posted 04 July 2019 - 10:23 AM

U da man, Chris!  Thanks so much for your time and effort to share your wealth of knowledge, enabling PI within the grasp for the rest of us bow.gif waytogo.gif


  • bobzeq25 likes this

#70 pfile

pfile

    Fly Me to the Moon

  • -----
  • Posts: 5186
  • Joined: 14 Jun 2009

Posted 04 July 2019 - 10:40 AM

Rob,

 

I had never heard of this method nor this new script. It is not one of the methods that I had found looking for way to process mosaics. If it is better that what I did, that is great, because it was enough to make me rethink mosaics in general. What I did was not easy or even optimized. I will check this out.

 

However, most of the work was not in generating the input panels. Although a bit time consuming, it was rather straight forward. The difficult part was in the blending with GMM. That was 80% of the time I spent.

 

Thanks,

 

Chris

 

yeah i agree, the problem is usually figuring out how to eliminate those 'pinched' stars. or dealing with a reflection that's in one pane and not another.

 

rob



#71 PilotAstronomy

PilotAstronomy

    Explorer 1

  • *****
  • Posts: 97
  • Joined: 25 Sep 2017

Posted 28 July 2019 - 11:17 AM

Chris,

 

I've been using the various versions of your guide for a while now but as it's grown I've started to get lost as I go. Usually I'm doing straight LRGB only or SHO only and your guide gets my wires cross a little bit -- I get confused on if they step I'm on it applicable or when I should be doing what because you mix the NB and RBG processing in your guide (for mixed images) or the terminology when you'er writing sometimes (or I'm just dumb) (especially when doing NB SHO only). Just some feedback! I do love your guide a lot and it's helped me understand the processes so much better.


Edited by PilotAstronomy, 28 July 2019 - 11:23 AM.


#72 cfosterstars

cfosterstars

    Mercury-Atlas

  • *****
  • topic starter
  • Posts: 2577
  • Joined: 05 Sep 2014
  • Loc: Austin, Texas

Posted 28 July 2019 - 09:02 PM

Chris,

 

I've been using the various versions of your guide for a while now but as it's grown I've started to get lost as I go. Usually I'm doing straight LRGB only or SHO only and your guide gets my wires cross a little bit -- I get confused on if they step I'm on it applicable or when I should be doing what because you mix the NB and RBG processing in your guide (for mixed images) or the terminology when you'er writing sometimes (or I'm just dumb) (especially when doing NB SHO only). Just some feedback! I do love your guide a lot and it's helped me understand the processes so much better.

I guess I could try and split thing up, but there are really only a few things that are unique to Nb. Other than the section on tonal mapping for Nb or the stuff on OSC, I treat LRGB and NB basically the same. Would a different font or something help? I could bold text on NB or something. What would you suggest?


  • RossW likes this

#73 PilotAstronomy

PilotAstronomy

    Explorer 1

  • *****
  • Posts: 97
  • Joined: 25 Sep 2017

Posted 28 July 2019 - 11:19 PM

Hmmm, I think the spot where I get lost is when you start color calibrating, deconvolution, and going non linear. I get lost a bit. Your focus is on LRGB w NB possibly mixed in. I'm not sure what steps apply to narrowband only or not. Font color might help. Or better a simplified summary of steps for NBLRGB, LRGB, and SHO separately at the very end. Can follow the steps and refer to the detailed description of them in your document. Would read like a checklist or to do list.

Edited by PilotAstronomy, 28 July 2019 - 11:20 PM.

  • starhunter50 likes this

#74 starhunter50

starhunter50

    Viking 1

  • *****
  • Posts: 564
  • Joined: 07 Sep 2010
  • Loc: Tilbury,Ontario 42N , 82W

Posted 29 July 2019 - 09:46 AM

Hmmm, I think the spot where I get lost is when you start color calibrating, deconvolution, and going non linear. I get lost a bit. Your focus is on LRGB w NB possibly mixed in. I'm not sure what steps apply to narrowband only or not. Font color might help. Or better a simplified summary of steps for NBLRGB, LRGB, and SHO separately at the very end. Can follow the steps and refer to the detailed description of them in your document. Would read like a checklist or to do list.

That's a good idea, either color or Fonts, the PDF i passed onto others are a bit overwhelming for them also , but the attention to detail is great. 

 

Astro Dude, 



#75 cfosterstars

cfosterstars

    Mercury-Atlas

  • *****
  • topic starter
  • Posts: 2577
  • Joined: 05 Sep 2014
  • Loc: Austin, Texas

Posted 29 July 2019 - 10:14 AM

That's a good idea, either color or Fonts, the PDF i passed onto others are a bit overwhelming for them also , but the attention to detail is great. 

 

Astro Dude, 

Well PI is overwhelming period. But I will do this on my next revision. I am working on adding a section on MUREDenoise, which is a marvelous tool. I will do this at the same time. 


  • RossW 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