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

More integration time=more nebula?

  • Please log in to reply
16 replies to this topic

#1 Cancington42

Cancington42

    Lift Off

  • -----
  • topic starter
  • Posts: 23
  • Joined: 10 Feb 2019

Posted 21 February 2019 - 01:51 AM

My setup is as follows:

star adventurer

full spectrum D7200,

rokinon 35mm 1.4 with hoya uv/ir cut for astro (same filter as astro mod) and hoya red intensifier for LP. Focal length on crop is 52mm. Aperture set to F/4

 

I’ve been wanting to do a portrait of orion ever since I started doing astrophotography a couple years back. I got this lens to be able to get the framing I wanted. There was the making for a clear night, so I setup, got everything running with 2 minute exposures, iso 1600, F/4. Once I was setup, I retreated from the cold winters night to hang out by the fire.. by the time I went to go check outside the clouds rolled in so I packed up. I only got 3 clean exposures. I was kind of **** off, but the weather was due to be clear the following night. So I setup again, and this time I got 137 2 minute exposures. 4.5 hours of relatively good data. I collected my darks and bias and went to bed. 

 

I processed the 3 exposures from the failed night just to see what I got, and ended up with a half decent image with a nice showing of nebulosity. Nothing spectacular, but a good example of orion. I was really excited to see what more data would show. upon collecting 4.5 hours of data and spending the day calibrating, debayering, registering, and integrating, I thought that more time integrated equals more nebulosity, to bring out faint details. 

 

Once I finished integrating the image, it didn’t seem as if there was a deeper red, or more data at all. I processed the 4.5hrs exactly the same as 6min and both seemed comparable. 

 

Am I missing something in my integrating process? 

 

I’m using pixinsight with the following settings

 

Image integration:

combination: average

normalization: additive with scaling

weights: noise evaluation

Scale estimator: iterative k-sigma/biweight midvariance

check boxes checked:

generate integrated image

evaluate noise

 

pixel rejection (1)

rejection algorithm: linear fit clipping

normalization: scale + zero offset

check boxes checked:

generate rejection maps

clip low pixels

clip high pixels

 

pixel rejection (2)

linear fit low - 5.0

linear fit high - 7.0

 

anyway.. any help would be appreciated in explaining what’s going on.

 

thanks

 

 



#2 whwang

whwang

    Vanguard

  • *****
  • Posts: 2493
  • Joined: 20 Mar 2013

Posted 21 February 2019 - 02:27 AM

It will be helpful if you can show us the actual images.
  • Jim Waters and Dynan like this

#3 Cancington42

Cancington42

    Lift Off

  • -----
  • topic starter
  • Posts: 23
  • Joined: 10 Feb 2019

Posted 21 February 2019 - 02:52 AM

It will be helpful if you can show us the actual images.

 

here are the two images. they have been integrated and stretched.

 

based on the settings i'm using, does that look like an appropriate way to integrate the amount of images I have?

 

https://www.dropbox....tUYgu_vZYa?dl=0



#4 Schleppafahrer

Schleppafahrer

    Explorer 1

  • -----
  • Posts: 96
  • Joined: 06 Feb 2018

Posted 21 February 2019 - 03:25 AM

First of all, nice shots! You already got a good amount of data!

You need to work on star masking and bringing out the nebulosity without clipping the highlights. I don't know anything about pixinsight but after 4.5 hrs you should be able to get A LOT more detail. Could you perhaps post the tif files?

Also, why only 2 minute exposures? Even with a very rough Polar alignment you & should be able to shoot 4+ min exposures at a lower iso.
  • Cancington42 likes this

#5 whwang

whwang

    Vanguard

  • *****
  • Posts: 2493
  • Joined: 20 Mar 2013

Posted 21 February 2019 - 03:26 AM

Hi,

 

I don't see big problems in either of the two images, nor in your PI setup (except for the part about linear fit, which I am less familiar with).

 

What looks obvious to me is that your 4.5-hr image has much higher S/N than the 6-min image.  If I am going to further process one of the two to get rich nebulosity, I will definitely start with the 4.5-hr one instead of the 6-min one.  

 

At the current level of your processing, they may appear to have comparable amount of nebulosity for two reasons:

 

A. Both have sky gradient, which limits the amount of stretch you can apply.  Further stretching the current images will only amplify the sky gradient, rather than faint nebulas.  So to go from here, you will need to first remove the sky gradient.  You may try ABE in PI, and limit the background model to 1st order or 2nd order.  Once the sky background is removed, you will get fainter nebulas, and you will see the true potential of the 4.5-hr image.

 

B. Both have many many stars. This also limits the amount of stretch you can apply.  Even after you remove the sky gradient, a strong stretch may simply blow up the stars and prevent you from seeing the faint nebulas underneath the stars.  If you see this, you can try to shrink the stars.  I believe this can be done in PI, and I usually do it in Photoshop with the minimum filter plus a star mask.  

 

If you can solve both A and B, I believe you will see that the 4.5-hr version gives you more (if not much more) nebulas.

 

One subtlety is that I feel the 4.5-hr images have bigger stars.  This can be caused by being slightly out of focus, or simply an illusion of mine.  I can't be certain without doing quantitative analyses.  Nevertheless, like in (B), big stars can prevent you from getting fainter nebulas.  One will have to deal with it in post processing, and will need to be careful about focusing during the imaging.  It's not uncommon that you will need to refocus for a couple of times during a long session of 4 to 6 hours, to compensate the temperature change.

 

Hope this helps, and good luck.

 

Cheers,

Wei-Hao


  • Cancington42 likes this

#6 Cancington42

Cancington42

    Lift Off

  • -----
  • topic starter
  • Posts: 23
  • Joined: 10 Feb 2019

Posted 21 February 2019 - 03:37 AM

First of all, nice shots! You already got a good amount of data!

You need to work on star masking and bringing out the nebulosity without clipping the highlights. I don't know anything about pixinsight but after 4.5 hrs you should be able to get A LOT more detail. Could you perhaps post the tif files?

Also, why only 2 minute exposures? Even with a very rough Polar alignment you & should be able to shoot 4+ min exposures at a lower iso.

Thanks a lot!

 

I noticed that when I was doing 4 minute exposures I had to throw away a bunch, and my success was much lower. Even at 2 minutes I had to throw away about 40 minutes worth of images. I am pretty good at polar alignment, but wind and periodic error are the issues.. I’ve been looking into the az-gti to use in eq mode, go-to, and could use a polemaster to align.. the star adventurer is basic, would probably run better with a guider.

 

I really should be using an auto guider, or better yet using my proper setup.. eq6-r with qhy183m.. but the weather is so fickle here that I’ve setup my big setup and more often than not I’m having to take it all in because the clouds roll in. 

 

The listed setup for this image is my grab and go, simple and easy to use, no computer. Just an eq mount and a camera.. I’m just waiting to build an observatory as well.. one day.

 

i would upload the tif, but I live in a rural location, and it’s going to take several hours to upload. I’ll do that before I go to bed, and upload the link tomorrow!! 



#7 Cancington42

Cancington42

    Lift Off

  • -----
  • topic starter
  • Posts: 23
  • Joined: 10 Feb 2019

Posted 21 February 2019 - 03:46 AM

Hi,

 

I don't see big problems in either of the two images, nor in your PI setup (except for the part about linear fit, which I am less familiar with).

 

What looks obvious to me is that your 4.5-hr image has much higher S/N than the 6-min image.  If I am going to further process one of the two to get rich nebulosity, I will definitely start with the 4.5-hr one instead of the 6-min one.  

 

At the current level of your processing, they may appear to have comparable amount of nebulosity for two reasons:

 

A. Both have sky gradient, which limits the amount of stretch you can apply.  Further stretching the current images will only amplify the sky gradient, rather than faint nebulas.  So to go from here, you will need to first remove the sky gradient.  You may try ABE in PI, and limit the background model to 1st order or 2nd order.  Once the sky background is removed, you will get fainter nebulas, and you will see the true potential of the 4.5-hr image.

 

B. Both have many many stars. This also limits the amount of stretch you can apply.  Even after you remove the sky gradient, a strong stretch may simply blow up the stars and prevent you from seeing the faint nebulas underneath the stars.  If you see this, you can try to shrink the stars.  I believe this can be done in PI, and I usually do it in Photoshop with the minimum filter plus a star mask.  

 

If you can solve both A and B, I believe you will see that the 4.5-hr version gives you more (if not much more) nebulas.

 

One subtlety is that I feel the 4.5-hr images have bigger stars.  This can be caused by being slightly out of focus, or simply an illusion of mine.  I can't be certain without doing quantitative analyses.  Nevertheless, like in (B), big stars can prevent you from getting fainter nebulas.  One will have to deal with it in post processing, and will need to be careful about focusing during the imaging.  It's not uncommon that you will need to refocus for a couple of times during a long session of 4 to 6 hours, to compensate the temperature change.

 

Hope this helps, and good luck.

 

Cheers,

Wei-Hao

Thank you so much for the information!! I really do appreciate it! I used a bahtinov mast to start the session of, i guess I should have check on focus periodically, but the temp didn’t seem to change much over the time, it was -14c to maybe -17c, would a change like that really affect focus that much? I’ll keep an eye on it next imaging session. 

 

I’m going to mess around with it a bit more and see what I can get out of it. I usually use DBE and get good results, I was just surprised that in a 6 minute integration time I got that much nebulosity, and figured in 4.5 hours it would really pop. I’m going to give it another process tomorrow and I’ll upload the results.

 

As the for the stars being blown out. Should I be underexposing a bit more. Maybe go to f/5.6, keep 2 minutes, and drop to iso 800? Or do you think my camera settings are decent? I almost feel like I’m over exposing the stars and that’s one of the problems..

 

thanks again for the help!!



#8 whwang

whwang

    Vanguard

  • *****
  • Posts: 2493
  • Joined: 20 Mar 2013

Posted 21 February 2019 - 04:54 AM

Hi,

 

Based on your 6-min image, I feel the lens you use is reasonably sharp.  There is probably no need to further stop it down.  Changing ISO or sub exposure time will not change the star profile either.  I think the key is good (and frequent) focusing and post processing.

 

Cheers,

Wei-Hao


  • Cancington42 likes this

#9 Rudy Pohl

Rudy Pohl

    Ranger 4

  • -----
  • Posts: 336
  • Joined: 27 Dec 2016
  • Loc: Ottawa, Ontario Canada

Posted 21 February 2019 - 07:31 AM

Hi Cancington42,

 

Here is what I was able to tease out of your 4.5-hour jpeg. As Wei-Hao mentioned, learning how to remove gradients and especially how to do star management (removal and reduction) is critical to processing an image like this. There are so many stars that they completely overwhelm the other structures. The star reduction/removal method I used for this image was done in Photoshop.

 

M42_widefield_PS_PI_FINAL.jpg

 

Best regards,
Rudy


Edited by Rudy Pohl, 21 February 2019 - 08:09 AM.

  • elmiko, whwang, 2ghouls and 4 others like this

#10 Remaxman

Remaxman

    Vostok 1

  • -----
  • Posts: 111
  • Joined: 16 Jul 2017
  • Loc: Pickering Ontario

Posted 21 February 2019 - 11:20 AM

I’m no pro at this but the first thing I noticed was the focus. The centre of the nebula was blown out. I’m using the gradient removal plugin for PS. It’s expensive but worth it. I’m also using Astronomy Tools action set which is 20 bucks. Only reason I bought these is I’m really new at this and learning. I have zero affiliation with these tool providers. But doing things over and over and trying to remember all the steps is daunting. But things others have said like the masks and shrinking the stars will really help. 


  • Cancington42 likes this

#11 Cancington42

Cancington42

    Lift Off

  • -----
  • topic starter
  • Posts: 23
  • Joined: 10 Feb 2019

Posted 21 February 2019 - 01:36 PM

Thanks for taking the time to process the image Rudy! I’ve been learning pixinsight over the last year, and it’s such a powerful piece of software, I’m still learning it. Did you just use photoshop for your edit?? 

 

Remaxman, I did use a bahtinov mask when I started the session, and I think there was focus shift throughout the 4.5 ours, and when stacking things got a bit fuzzy. 



#12 Rudy Pohl

Rudy Pohl

    Ranger 4

  • -----
  • Posts: 336
  • Joined: 27 Dec 2016
  • Loc: Ottawa, Ontario Canada

Posted 21 February 2019 - 03:14 PM

Thanks for taking the time to process the image Rudy! I’ve been learning pixinsight over the last year, and it’s such a powerful piece of software, I’m still learning it. Did you just use photoshop for your edit?? 

Hi there,

 

Yes, everything was done in Photoshop CS5.

 

Luckily I have the pleasant advantage of having used Photoshop professionally in our graphics business for about 25 years so it comes fairly naturally to me. My wife and I closed the business and retired a couple of years ago but I still use Photoshop every day. The star reduction/removal method that I used on your image is one that I stumbled upon just in the last year and which I've tweaked a bit further since then. It does a decent job on most subjects.

 

Here is an image  where the stars were removed with this method: https://www.flickr.c.../in/dateposted/ 

 

Here is an image where the stars were not removed, but reduced in size with this method: https://www.cloudyni...ebulae-m42-m43/

 

Best regards,
Rudy


Edited by Rudy Pohl, 21 February 2019 - 04:35 PM.

  • Cancington42 likes this

#13 Cancington42

Cancington42

    Lift Off

  • -----
  • topic starter
  • Posts: 23
  • Joined: 10 Feb 2019

Posted 21 February 2019 - 05:00 PM

Hi there,

 

Yes, everything was done in Photoshop CS5.

 

Luckily I have the pleasant advantage of having used Photoshop professionally in our graphics business for about 25 years so it comes fairly naturally to me. My wife and I closed the business and retired a couple of years ago but I still use Photoshop every day. The star reduction/removal method that I used on your image is one that I stumbled upon just in the last year and which I've tweaked a bit further since then. It does a decent job on most subjects.

 

Here is an image  where the stars were removed with this method: https://www.flickr.c.../in/dateposted/

 

Here is an image where the stars were not removed, but reduced in size with this method: https://www.cloudyni...ebulae-m42-m43/

 

Best regards,
Rudy

I'm really curious to know what the Method is! is it an action? i'm impressed that you were able to do that in photoshop to that degree with M42. 



#14 Rudy Pohl

Rudy Pohl

    Ranger 4

  • -----
  • Posts: 336
  • Joined: 27 Dec 2016
  • Loc: Ottawa, Ontario Canada

Posted 21 February 2019 - 05:16 PM

I'm really curious to know what the Method is! is it an action? i'm impressed that you were able to do that in photoshop to that degree with M42. 

Hi again,

 

No it's not an action, it's a series of steps in which you make your own "star reduction template" of all the stars in the image and then you reduce them using other steps.

 

I'll be happy to share this method with you and anyone here who is interested, but it will take a little time for me to write it all out as I'm crazy busy for the next few days. I'll let you know when I've finished and posted it.

 

Cheers,
Rudy


Edited by Rudy Pohl, 21 February 2019 - 05:20 PM.

  • VincentD, OldManSky and Cancington42 like this

#15 Cancington42

Cancington42

    Lift Off

  • -----
  • topic starter
  • Posts: 23
  • Joined: 10 Feb 2019

Posted 22 February 2019 - 03:00 AM

Hi again,

 

No it's not an action, it's a series of steps in which you make your own "star reduction template" of all the stars in the image and then you reduce them using other steps.

 

I'll be happy to share this method with you and anyone here who is interested, but it will take a little time for me to write it all out as I'm crazy busy for the next few days. I'll let you know when I've finished and posted it.

 

Cheers,
Rudy

that's amazing!! i'd be really curious to know your method. take your time, there's no rush!!

 

thank you!!



#16 Alen K

Alen K

    Viking 1

  • -----
  • Posts: 650
  • Joined: 25 Nov 2009

Posted 22 February 2019 - 12:40 PM

that's amazing!! i'd be really curious to know your method. take your time, there's no rush!!
 
thank you!!

+1

#17 Kevin_A

Kevin_A

    Explorer 1

  • -----
  • Posts: 74
  • Joined: 22 Aug 2018
  • Loc: London, Ontario

Posted 22 February 2019 - 02:59 PM

Personally I would drop the lens back down to F2.0-F2.8 and drop your ISO to 800 with that filter setup and try to get 3min subs. I have a D7200 (non mod) and its really lousy anywhere past ISO800... too noisy and lacking dynamic range... just a thought.




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