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

First DSO Image - Q's

  • Please log in to reply
21 replies to this topic

#1 fastfiveoh

fastfiveoh

    Mariner 2

  • -----
  • topic starter
  • Posts: 225
  • Joined: 23 Apr 2018

Posted 27 May 2024 - 04:24 PM

Hey All,

 

Thanks for all the help.  Still working out issues, and just started learning PixInsight today.  

Stats:

 

Pelican Nebula

 

30m Luminance (30x60s)

15m Red (15x60s)

15m Green (15x60s)

11m Blue (11x60s)

 

Wasn't able to get any Ha data because my AF routine wasn't working/taking a ton of time last night.  (Although Everything seems pretty sharp even with AF failures from my initial Bhatinov manual focus for the L filter)

 

I ran them through batch processing with my calibration frames, followed some videos on Youtube (VisibleDark and one other guy).  However I pretty sure my stars are blown out, nothing can get them under control in PixInsight.  When I do anything in PixInsight that kinda matches the Youtube videos, even the minute adjustment makes the colors / masks go nuts.  Also if you apply any sharpening, no matter how minor the stars go either haloes or black rings around them. 

 

Also no Ha data yet, im thinking about capturing that tonight, however I want to retake the LRGB data probably as well.  So my question is:

 

In NINA how do I see correct star exposure, I keep searching and answers are all over the place? 

 

 

Anyways first image, be gentle.   Again thanks for everyones help even getting me this far.

 

PelicanNebula.jpg

 

 

Mike


Edited by fastfiveoh, 27 May 2024 - 05:16 PM.

  • mikewayne3, TXLS99, CowTipton and 2 others like this

#2 John Berger

John Berger

    Apollo

  • *****
  • Posts: 1,238
  • Joined: 30 Jun 2023

Posted 27 May 2024 - 05:38 PM

geee, you got all that equipment and you just took your first astrophoto?

 

you went from nothing to monochrome, which is a brave move.

 

Have you tried Starnet?

 

IMO,

 

While your LRGB image is linear, do an Autostretch and do Background Normalization, and then ColorCalibration; this way your nebula and stars will have balanced colors, ready for saturation.

 

Afterwards, when your LRGB image is still linear, run Starnet, then stretch your starless as much as you like (I think you can bring out some more color; try Arcsinh stretch, a little), then stretch your stars separately, but do a much gentler stretch; I would first start with Arcsinh, maybe once, then a small Curves Transformation stretch as the last stretch for the stars.

 

Recombine with Pixemath: Stars + Starless - Stars * Starless

 

You should have a cleaner image now

 

Your data will speak to you; learn it's language; understand what it needs; experiment as much as you can. grin.gif

 

I recommend getting the RC-astro tools if you don't have them already


Edited by John Berger, 27 May 2024 - 05:42 PM.

  • fastfiveoh likes this

#3 John Berger

John Berger

    Apollo

  • *****
  • Posts: 1,238
  • Joined: 30 Jun 2023

Posted 27 May 2024 - 05:55 PM

I think you could use Gradient Correction too; there is a light blue blob at the top left of the image


Edited by John Berger, 27 May 2024 - 05:55 PM.

  • fastfiveoh likes this

#4 fastfiveoh

fastfiveoh

    Mariner 2

  • -----
  • topic starter
  • Posts: 225
  • Joined: 23 Apr 2018

Posted 27 May 2024 - 05:55 PM

geee, you got all that equipment and you just took your first astrophoto?

 

you went from nothing to monochrome, which is a brave move.

 

Have you tried Starnet?

 

IMO,

 

While your LRGB image is linear, do an Autostretch and do Background Normalization, and then ColorCalibration; this way your nebula and stars will have balanced colors, ready for saturation.

 

Afterwards, when your LRGB image is still linear, run Starnet, then stretch your starless as much as you like (I think you can bring out some more color; try Arcsinh stretch, a little), then stretch your stars separately, but do a much gentler stretch; I would first start with Arcsinh, maybe once, then a small Curves Transformation stretch as the last stretch for the stars.

 

Recombine with Pixemath: Stars + Starless - Stars * Starless

 

You should have a cleaner image now

 

Your data will speak to you; learn it's language; understand what it needs; experiment as much as you can. grin.gif

 

I recommend getting the RC-astro tools if you don't have them already

I could not get Starnet working in, was getting memory address errors when I tried to install it (downloading the latest from their site and putting the files extracted in the /bin folder) Some other forum said its not compatible with the latest PixInisight?

 

I did background division then subtraction on the master lights for each channel (after some cropping), then combined the channels so far, then the stretch.

 

Thanks for all the other advice!

 

P.S.  buy once cry once grin.gif


Edited by fastfiveoh, 27 May 2024 - 05:59 PM.

  • John Berger likes this

#5 fastfiveoh

fastfiveoh

    Mariner 2

  • -----
  • topic starter
  • Posts: 225
  • Joined: 23 Apr 2018

Posted 27 May 2024 - 05:56 PM

I think you could use Gradient Correction too; there is a light blue blob at the top left of the image

I had some very weird vignetting in my Red and Green channels, and cropped most of that out.  Still working through that as well.   Might retake my calibration frames to be sure.


  • John Berger likes this

#6 John Berger

John Berger

    Apollo

  • *****
  • Posts: 1,238
  • Joined: 30 Jun 2023

Posted 27 May 2024 - 05:59 PM

StarXterminator gives cleaner results than Starnet, and it'll be sure to work in PI

 

I recommend SXT over Starnet


Edited by John Berger, 27 May 2024 - 05:59 PM.

  • fastfiveoh likes this

#7 fastfiveoh

fastfiveoh

    Mariner 2

  • -----
  • topic starter
  • Posts: 225
  • Joined: 23 Apr 2018

Posted 27 May 2024 - 06:00 PM

StarXterminator gives cleaner results than Starnet, and it'll be sure to work in PI

Oh awesome will take a look!  When I was trying to stretch colors further the stars started doing seriously wonky things, so they need addressing.  Would clipped/blown out starts cause something like that?



#8 imtl

imtl

    Cosmos

  • *****
  • Posts: 8,928
  • Joined: 07 Jun 2016
  • Loc: Down in a hole

Posted 27 May 2024 - 06:13 PM

What sharpening are you trying to do?
  • fastfiveoh likes this

#9 fastfiveoh

fastfiveoh

    Mariner 2

  • -----
  • topic starter
  • Posts: 225
  • Joined: 23 Apr 2018

Posted 27 May 2024 - 06:17 PM

What sharpening are you trying to do?

I was just following along with some Youtube video, he added some sharpening to his and when I did it exactly how he did it it blew every star up with a black halo ( sorry away from the comp I am blanking on the tool name, but it can also do noise reduction (linear transform something i think )



#10 imtl

imtl

    Cosmos

  • *****
  • Posts: 8,928
  • Joined: 07 Jun 2016
  • Loc: Down in a hole

Posted 27 May 2024 - 06:30 PM

Here is my general suggestion. Take it as you please.

Try to focus on basics. Like proper calibration, combinning channels and stretching. Master that first. Try not to follow all kinds of youtube recipes but rather understand what each step you are doing actually does and how to control it. As usual with social media influencers, they show only stuff that promotes a certain agenda. Whatever that might be.
  • steveincolo and gsaramet like this

#11 John Berger

John Berger

    Apollo

  • *****
  • Posts: 1,238
  • Joined: 30 Jun 2023

Posted 27 May 2024 - 06:39 PM

Oh awesome will take a look!  When I was trying to stretch colors further the stars started doing seriously wonky things, so they need addressing.  Would clipped/blown out starts cause something like that?

Wonky things? What do you see? How were you trying to stretch colors?



#12 fastfiveoh

fastfiveoh

    Mariner 2

  • -----
  • topic starter
  • Posts: 225
  • Joined: 23 Apr 2018

Posted 27 May 2024 - 06:44 PM

Wonky things? What do you see? How were you trying to stretch colors?

I tried with the Histogram and curves.  I could barely move the curves or the histogram arrows very far without it blowing completely out.  I had to stretch the historgram across 3 screens to get granularity enough to be able to move the histogram blacks up and the mids down.  Ill take a pic when I get home


  • John Berger likes this

#13 fastfiveoh

fastfiveoh

    Mariner 2

  • -----
  • topic starter
  • Posts: 225
  • Joined: 23 Apr 2018

Posted 28 May 2024 - 07:38 PM

Attempt 2:

 

Added in 55 minutes of Ha data, and learning more stuff in PixInsight. 

 

Slowly getting better. 

 

attempt--2.jpg



#14 John Berger

John Berger

    Apollo

  • *****
  • Posts: 1,238
  • Joined: 30 Jun 2023

Posted 28 May 2024 - 08:56 PM

one thing I want to mention

 

assuming you have a stars removal software:

 

you should never incorporate H-alpha stars with your RGB stars

 

the H-alpha stars need to be discarded, and I would recommend you discard the Lum stars too

 

look at my processed images: LRHGB (starless) with RGB (stars) https://www.astrobin...ction&nce=26109

 

blend red starless with h-alpha starless; you could use (Rstarless + Hstarless) * 0.5, or you could experiment with other values, like 0.4 * Rstarless + 0.6 * Hstarless

 

then once you get the H-R hybrid image, you add that as Red in LRGB combination

 

so:

for Lum use Lum

for Red use your HR image

for green, green

for Blue, blue

 

 

(those are all starless)

 

so now you have a starless LRHGB image; do the background neutralization now, and color calibration if need be, and gradient correction.

 

then, get your RGB stars image by using RGB combination with your RGB stars.

 

then, now combine the two using the screen stars method: stars + starless - stars * starless

 

it should turn out good

 

I can clarify anything you want to know

 

this is my method, but it works.


Edited by John Berger, 28 May 2024 - 08:57 PM.

  • fastfiveoh likes this

#15 John Berger

John Berger

    Apollo

  • *****
  • Posts: 1,238
  • Joined: 30 Jun 2023

Posted 28 May 2024 - 08:59 PM

I processed this also, LRHGB with RGB stars https://www.astrobin...ction&nce=26109

 

this free data from DeepSkyWest; you could practice with that data too.


Edited by John Berger, 28 May 2024 - 08:59 PM.

  • fastfiveoh likes this

#16 fastfiveoh

fastfiveoh

    Mariner 2

  • -----
  • topic starter
  • Posts: 225
  • Joined: 23 Apr 2018

Posted 28 May 2024 - 09:04 PM

one thing I want to mention

 

assuming you have a stars removal software:

 

you should never incorporate H-alpha stars with your RGB stars

 

the H-alpha stars need to be discarded, and I would recommend you discard the Lum stars too

 

look at my processed images: LRHGB (starless) with RGB (stars) https://www.astrobin...ction&nce=26109

 

blend red starless with h-alpha starless; you could use (Rstarless + Hstarless) * 0.5, or you could experiment with other values, like 0.4 * Rstarless + 0.6 * Hstarless

 

then once you get the H-R hybrid image, you add that as Red in LRGB combination

 

so:

for Lum use Lum

for Red use your HR image

for green, green

for Blue, blue

 

 

(those are all starless)

 

so now you have a starless LRHGB image; do the background neutralization now, and color calibration if need be, and gradient correction.

 

then, get your RGB stars image by using RGB combination with your RGB stars.

 

then, now combine the two using the screen stars method: stars + starless - stars * starless

 

it should turn out good

 

I can clarify anything you want to know

 

this is my method, but it works.

Awesome advice, I just started messing with Starnet v2.  Will be trying it!!


Edited by fastfiveoh, 28 May 2024 - 09:44 PM.

  • John Berger likes this

#17 fastfiveoh

fastfiveoh

    Mariner 2

  • -----
  • topic starter
  • Posts: 225
  • Joined: 23 Apr 2018

Posted 29 May 2024 - 04:58 PM

Attempt 3, went to basics and just did RGB.  Still learning as I go, but at least each iteration is getting better as I understand PI more.  I for sure need signficantly longer integration times.  Hoping to get out tomorrow and Friday.  This is like 60 minutes total between RGB

 

PelicanAttemp3.jpg


Edited by fastfiveoh, 29 May 2024 - 05:11 PM.

  • John Berger likes this

#18 John Berger

John Berger

    Apollo

  • *****
  • Posts: 1,238
  • Joined: 30 Jun 2023

Posted 29 May 2024 - 06:08 PM

big improvement; you're getting there

 

now all you need is background neutralization (because I see blue in the background), a run with Gradient Correction, a tiny black point increase, preferably with Arcsinh, and you got a decent image right there.


  • fastfiveoh likes this

#19 fastfiveoh

fastfiveoh

    Mariner 2

  • -----
  • topic starter
  • Posts: 225
  • Joined: 23 Apr 2018

Posted 29 May 2024 - 10:30 PM

big improvement; you're getting there

 

now all you need is background neutralization (because I see blue in the background), a run with Gradient Correction, a tiny black point increase, preferably with Arcsinh, and you got a decent image right there.

Also found the reason my AF wasn't working....helps if you screw the bolts down to the flat part on the EAF's drive bar.....man i feeel dumb.  I am going to discord all data I have gotten so far and chalk it up to learning.  The rig *should* work flawlessly tonight barring user error.  Going to get about 3h45m of RGB on Pelican tonight, and process that tomorrow. 

 

I am still on the fence about exposure times.  Trying not to clip the stars in a Bortle4.  Not worried about space so may do 30s subs (I did a mix of 20 and 60 on the image above.


  • John Berger likes this

#20 Chaser of Photons

Chaser of Photons

    Viking 1

  • *****
  • Posts: 896
  • Joined: 13 Aug 2020
  • Loc: California

Posted 30 May 2024 - 01:27 AM

Keep on going with integration time, may not even need to go and add Ha data.

 

Per what John recommended, here's your latest image with the following processing sequence applied:

 

1. Star extraction.

2. Gradient reduction.

3. Black point adjustment (for me, even a very modest adjustment with Arcsinh wiped out the dust region?), gently used histogram tool.

4. A touch of sharpening (0.5 in BXT.)

5. Rescreened stars back in.

 

post_reprocess.jpg


  • fastfiveoh and Chrisofweden like this

#21 fastfiveoh

fastfiveoh

    Mariner 2

  • -----
  • topic starter
  • Posts: 225
  • Joined: 23 Apr 2018

Posted 30 May 2024 - 11:58 AM

Keep on going with integration time, may not even need to go and add Ha data.

 

Per what John recommended, here's your latest image with the following processing sequence applied:

 

1. Star extraction.

2. Gradient reduction.

3. Black point adjustment (for me, even a very modest adjustment with Arcsinh wiped out the dust region?), gently used histogram tool.

4. A touch of sharpening (0.5 in BXT.)

5. Rescreened stars back in.

 

attachicon.gif post_reprocess.jpg

****! Thats awesome!  I am going to stack in 1h each  channel of 120s subs today I got from last night!.  *crossing fingers*.  I tested using the histogram and settled on this as it pulled the histogram off the left part of it so it was a distinct peak.  I also observed that my blue filter gathers much more light than R and G channels.  So much learning around every little thing.  Love it!


Edited by fastfiveoh, 30 May 2024 - 11:59 AM.

  • steveincolo likes this

#22 John Berger

John Berger

    Apollo

  • *****
  • Posts: 1,238
  • Joined: 30 Jun 2023

Posted 30 May 2024 - 07:53 PM

 

3. Black point adjustment (for me, even a very modest adjustment with Arcsinh wiped out the dust region?), gently used histogram tool.

 

sometimes you need to be very very careful with Arcsinh, but I always use it

 

the starless in this image was entirely stretched with Arcsinh https://www.astrobin...ction&nce=26109 and that means that I used Arcsinh black point as well


Edited by John Berger, 30 May 2024 - 07:59 PM.



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