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

To reduce or not to reduce [stars.]

  • Please log in to reply
39 replies to this topic

#1 Ballyhoo

Ballyhoo

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 5,738
  • Joined: 22 Jul 2011
  • Loc: San Diego

Posted 13 August 2020 - 02:17 PM

This is about 4 hours of LRGB- ha.   

 

But the field seems like far too many stars..    

 

Which version do you find more pleasing?

 

edit,

it is almost as if the morphological transformation reduced some of the light background. 

 

C& C welcome.

Attached Thumbnails

  • trifid aug 2020 reduced.jpg
  • trifid aug 2020.jpg

Edited by Ballyhoo, 13 August 2020 - 02:20 PM.


#2 pyrasanth

pyrasanth

    Gemini

  • *****
  • Posts: 3,114
  • Joined: 08 Jan 2016

Posted 13 August 2020 - 02:26 PM

I will be kind however I don't see the point of your question.

 

The number of stars is the number of stars- the only way you get more stars is by going deeper until you reach the limits of your optics or the background pollution levels. In a nutshell it is what it is- no more & no less. If I see more stars then I know I've gone a bit deeper than last time but I'm satisfied with either image provided they are well processed.

 

Fast optics go deep really quick with an apparent increase in star numbers but again that is what the sky is- nothing more or less.

 

I like any image with lots of stars...unless they have been added artificially!

 

Morphological transformation will always remove some of the stars- I've not found a way around that so I don't use it much. 

 

I like the first image better- the second looks like it was taken through a fog bank and the background is not as nice as it could be.


Edited by pyrasanth, 13 August 2020 - 02:39 PM.

  • Jim Waters, psandelle and RedLionNJ like this

#3 Ballyhoo

Ballyhoo

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 5,738
  • Joined: 22 Jul 2011
  • Loc: San Diego

Posted 13 August 2020 - 02:43 PM

I will be kind however I don't see the point of your question.

 

The number of stars is the number of stars- the only way you get more stars is by going deeper until you reach the limits of your optics or the background pollution levels. In a nutshell it is what it is- no more & no less. If I see more stars then I know I've gone a bit deeper than last time but I'm satisfied with either image provided they are well processed.

 

Fast optics go deep really quick with an apparent increase in star numbers but again that is what the sky is- nothing more or less.

 

I like any image with lots of stars...unless they have been added artificially!

 

Morphological transformation will always remove some of the stars- I've not found a way around that so I don't use it much. 

 

I like the first image better- the second looks like it was taken through a fog bank and the background is not as nice as it could be.

Well I suppose my op is a difficult one to answer because this is a subjective issue, or taste.  I am not sure what constitutes "fast" though. Do you mean Rasa  fast?

 

It seems to me, one byproduct of imaging nebula in the milkyway, is like this here: So much dense star fields, i think it has a tendency to detract from the target image. Well it does. It decreases the apparent contrast between the target object and background stars. 

 

That is why I ask. For me the star reduction image is more pleasing. I am wondering what others think about this. 

 

edit, I did not get the same effect using an L mask and curve transformation.  Well at least the effect was not merely so dramatic.


Edited by Ballyhoo, 13 August 2020 - 02:46 PM.


#4 pyrasanth

pyrasanth

    Gemini

  • *****
  • Posts: 3,114
  • Joined: 08 Jan 2016

Posted 13 August 2020 - 02:53 PM

Well I suppose my op is a difficult one to answer because this is a subjective issue, or taste.  I am not sure what constitutes "fast" though. Do you mean Rasa  fast?

 

It seems to me, one byproduct of imaging nebula in the milkyway, is like this here: So much dense star fields, i think it has a tendency to detract from the target image. Well it does. It decreases the apparent contrast between the target object and background stars. 

 

That is why I ask. For me the star reduction image is more pleasing. I am wondering what others think about this. 

 

edit, I did not get the same effect using an L mask and curve transformation.  Well at least the effect was not merely so dramatic.

I believe anything under F4.0 will be fast.

 

As for the sheer numbers of stars it depends if your looking for an accurate representation of the target. If you just want a pretty picture then remove the stars and replace with any nice star field. It is all down, ultimately, to what you want.

 

A mass of stars can look tremendous if the target is well processed- an example is M27-loaded with stars & it still looks great. There are I'm sure techniques to enhance the object leaving the stars to themselves.


Edited by pyrasanth, 13 August 2020 - 02:54 PM.

  • psandelle likes this

#5 Ballyhoo

Ballyhoo

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 5,738
  • Joined: 22 Jul 2011
  • Loc: San Diego

Posted 13 August 2020 - 02:57 PM

I believe anything under F4.0 will be fast.

 

As for the sheer numbers of stars it depends if your looking for an accurate representation of the target. If you just want a pretty picture then remove the stars and replace with any nice star field. It is all down, ultimately, to what you want.

 

A mass of stars can look tremendous if the target is well processed- an example is M27-loaded with stars & it still looks great. There are I'm sure techniques to enhance the object leaving the stars to themselves.

ok so I can do better whan with improving my technique and including all the stars.



#6 StarBurger

StarBurger

    Viking 1

  • -----
  • Posts: 578
  • Joined: 06 Mar 2018
  • Loc: North Country NY

Posted 13 August 2020 - 03:14 PM

I too prefer the first.

I only process using StarTools and am not familiar with a "morphological" transform in that process. Please inform what the equivalent is in ST!

I tend to mess around in ST until things look better and although I track my process as yet I have been unable to find a work flow that works every time.

Sometimes auto dev works well, sometimes manual dev, sometimes depends on the subject and/or sky conditions.

Even though I use the same DSLR with same settings and exposures it really depends on moon/no moon, hot night/cold night, humidity, wind, seeing, LP and whether there is a letter A in the month.

Sometimes I try some really unorthodox paths just to see  but usually the results end up the same no matter how I approach it.

I find I have to custom tweak just about every stack differently. Not complaining. In fact I enjoy the artistry involved.

Like your examples I can generate less large fuzzy stars or more fuzzy stars but don't ask me if there is any logic in all this.

The end result is the way I want it through trial and error and I am very tolerant of error!


  • psandelle and Aquawind like this

#7 OhmEye

OhmEye

    Viking 1

  • *****
  • Posts: 788
  • Joined: 15 Sep 2019
  • Loc: Western NY Southern Tier

Posted 14 August 2020 - 06:07 AM

The entire image looks very soft to me, as if the data is poorly focused. That's my first thought, and why I prefer the first image because it has higher contrast and less crop which call less attention to the lack of pinpoint stars.


  • Stelios, psandelle, RedLionNJ and 1 other like this

#8 Madratter

Madratter

    Voyager 1

  • *****
  • Posts: 12,673
  • Joined: 14 Jan 2013

Posted 14 August 2020 - 08:37 AM

There are times the number of stars is absolutely the point:

 

get.jpg?insecure

 

In general, I don't like the way MT changes the profile of stars and I don't use it often. That said, there are times especially with large aperture fast scopes, that the star field can get extremely crowded at the expense of the subject. Hyperstar and RA setups are especially prone to this.


  • psandelle and 06AwzIyI like this

#9 Huangdi

Huangdi

    Viking 1

  • *****
  • Posts: 745
  • Joined: 24 Jul 2019

Posted 14 August 2020 - 09:04 AM

The example images you uploaded are completely different so it's pretty much impossible to say which version has better stars. The background or nebulosity shouldn't be affected by star reduction.

If your stars are blown out in their linear form, then that's an acquisition issue. If they're blown out later on, that's a processing issue.

I like to use star reduction for my broadband images, but removing stars is the wrong approach imo. Dimming the stars is what really works magic. The dimmer they become, the more the background starts to show.
  • psandelle and dan_hm like this

#10 dan_hm

dan_hm

    Viking 1

  • *****
  • Posts: 596
  • Joined: 31 Dec 2012
  • Loc: New Jersey

Posted 14 August 2020 - 10:10 AM

PI users check out Adam Block's "Star De-Emphasis" method:

 

https://adamblockstu...cles/star_demph

 

I used it on my last image of the Cygnus Loop and I think it did a much better job than Morphological Transformation. The process of trying to match the noise values was a pain but I found it didn't need to be that accurate for the effect to work.


  • psandelle and Bretw01 like this

#11 schmeah

schmeah

    Fly Me to the Moon

  • *****
  • Posts: 5,925
  • Joined: 26 Jul 2005
  • Loc: Morristown, NJ

Posted 14 August 2020 - 11:06 AM

Over the years, I have grown to like a dense background starfield, and avoid star reduction if at all possible. Also the diamond shape stars and tinker toy artifacts that I see in many images with excessively reduced stars is unpleasant. But I don’t think many are of the same opinion, based on lack of comments whenever I post an image with a “natural” starfield.

 

Derek


  • psandelle likes this

#12 pyrasanth

pyrasanth

    Gemini

  • *****
  • Posts: 3,114
  • Joined: 08 Jan 2016

Posted 14 August 2020 - 11:27 AM

I emphasise my original point "it is what it is"- you have a lot of stars in your image that is because there are a lot of stars to be captured! What you do with them after in the image is very much your choice but I personally would never try & deliberately remove them. I have edited a very bright star that has potentially ruined a nice image- I just clone it with a fainter star of similar colour.


Edited by pyrasanth, 14 August 2020 - 11:30 AM.

  • psandelle likes this

#13 ngc1535

ngc1535

    Vendor - Caelum Observatory

  • -----
  • Vendors
  • Posts: 54
  • Joined: 18 Sep 2008

Posted 14 August 2020 - 06:18 PM

PI users check out Adam Block's "Star De-Emphasis" method:

 

https://adamblockstu...cles/star_demph

 

I used it on my last image of the Cygnus Loop and I think it did a much better job than Morphological Transformation. The process of trying to match the noise values was a pain but I found it didn't need to be that accurate for the effect to work.

 

Please see my update for a simplified method (and do subscribe so you will be "in the know!").

 

https://www.youtube....dSDoJfjM4&t=35s

 

Thanks,

Adam


  • psandelle likes this

#14 schmeah

schmeah

    Fly Me to the Moon

  • *****
  • Posts: 5,925
  • Joined: 26 Jul 2005
  • Loc: Morristown, NJ

Posted 14 August 2020 - 06:23 PM

I emphasise my original point "it is what it is"- you have a lot of stars in your image that is because there are a lot of stars to be captured! What you do with them after in the image is very much your choice but I personally would never try & deliberately remove them. I have edited a very bright star that has potentially ruined a nice image- I just clone it with a fainter star of similar colour.

Cloning a star as you describe -is- deliberately removing a star ... then just replacing it with a different star. Just as objectionable if indeed you find removing stars objectionable. I think I would draw the line between stretching, sharpening, minimizing data and replacing it altogether. 
 

Derek



#15 klaussius

klaussius

    Viking 1

  • *****
  • Posts: 706
  • Joined: 02 Apr 2019
  • Loc: Buenos Aires

Posted 14 August 2020 - 06:43 PM

The area around M20 is really interesting exactly because it is full of stars.

 

To me, the second image is more natural and interesting. Yes, it is soft. It shouldn't be that soft. But I'm talking only about the "amount" of stars.

 

Sometimes, stars do detract from the image, when the cover the target. I like stars in my image, but I've been in the position of wishing to reduce the number or prominence of stars a bit.

 

I find filters are a great way to achieve that. HaRGB or UHC-RGB, mostly because filters do dim stars in a natural and unintrusive way.

 

You can get a similar effect with post, but the natural quality you can get with a filtered luminance layer cannot be beat IMHO.

 

When I pick a target, I usually think whether I'll want the stars or not before I start. If I'm not into the stars in that region, I'll use filters. Otherwise, I'll go unfiltered. It's not something you can fix in post, you have to plan for that, but it's what I prefer to do.


  • psandelle and RedLionNJ like this

#16 Peter in Reno

Peter in Reno

    Voyager 1

  • *****
  • Posts: 11,248
  • Joined: 15 Jul 2008
  • Loc: Reno, NV

Posted 14 August 2020 - 08:50 PM

Why does the second image look excessively soft? Do the original and calibrated subs look soft or did you use excessive noise reduction which can easily make the image look pretty soft?

 

Personally I never mess with stars no matter how few or many stars are on the images. If the stars are many, leave them alone because there are supposed to be many stars.

 

I'm not sure what your goal is? Accurate representation or faking the images to please you? If it's the latter, you'll never win an APOD if that's what you're aiming for. 

 

If the original and calibrated subs look soft, you need to go back and inspect your equipment and find out why the stars look out of focused. If you are using standard Moonlite focuser, the drawtube could be slipping due to Crayford design and you may need to adjust tension of the focuser.

 

Peter 


  • psandelle and RedLionNJ like this

#17 dan_hm

dan_hm

    Viking 1

  • *****
  • Posts: 596
  • Joined: 31 Dec 2012
  • Loc: New Jersey

Posted 14 August 2020 - 10:01 PM

To actually comment on the OP's images, I don't see how the second one is to any degree "starless." The stars actually look more obnoxious in it than in the one on the left. That, and it looks like it's been excessively noise-reduced, or as if it was out of focus.

 

On the topic of star reduction/removal in general, I find it critical to image quality for my widefield images taken at high gain with the ASI1600. The stars saturate excessively and I personally find the image to improve appreciably with star reduction. I have never done a starless image, but I think they're done well a lot of the time. Starless APODs have been done. All you're doing is taking out a bunch of tiny dots that sit between Earth and the object you're really trying to capture. 


  • psandelle and RedLionNJ like this

#18 Peter in Reno

Peter in Reno

    Voyager 1

  • *****
  • Posts: 11,248
  • Joined: 15 Jul 2008
  • Loc: Reno, NV

Posted 14 August 2020 - 11:13 PM

To actually comment on the OP's images, I don't see how the second one is to any degree "starless." The stars actually look more obnoxious in it than in the one on the left. That, and it looks like it's been excessively noise-reduced, or as if it was out of focus.

 

On the topic of star reduction/removal in general, I find it critical to image quality for my widefield images taken at high gain with the ASI1600. The stars saturate excessively and I personally find the image to improve appreciably with star reduction. I have never done a starless image, but I think they're done well a lot of the time. Starless APODs have been done. All you're doing is taking out a bunch of tiny dots that sit between Earth and the object you're really trying to capture. 

That's true about starless images awarded for APOD but that does not happen very often. Good point though.

 

For wide field images, you could drastically reduce the gain to near or at zero gain to avoid saturated stars. It would probably require longer sub-exposure times but that should not be too difficult.

 

Peter 


  • psandelle likes this

#19 dan_hm

dan_hm

    Viking 1

  • *****
  • Posts: 596
  • Joined: 31 Dec 2012
  • Loc: New Jersey

Posted 15 August 2020 - 12:15 AM

That's true about starless images awarded for APOD but that does not happen very often. Good point though.

 

For wide field images, you could drastically reduce the gain to near or at zero gain to avoid saturated stars. It would probably require longer sub-exposure times but that should not be too difficult.

 

Peter 

For broadband yes. With narrowband it seems hard to justify going lower than unity gain though. 



#20 Peter in Reno

Peter in Reno

    Voyager 1

  • *****
  • Posts: 11,248
  • Joined: 15 Jul 2008
  • Loc: Reno, NV

Posted 15 August 2020 - 12:47 AM

For broadband yes. With narrowband it seems hard to justify going lower than unity gain though. 

I agree with you but the OP's original images appear to be broadband.

 

Peter 


Edited by Peter in Reno, 15 August 2020 - 01:29 AM.

  • psandelle likes this

#21 bobzeq25

bobzeq25

    ISS

  • *****
  • Posts: 21,108
  • Joined: 27 Oct 2014

Posted 15 August 2020 - 01:54 AM

This is about 4 hours of LRGB- ha.   

 

But the field seems like far too many stars..    

 

Which version do you find more pleasing?

 

edit,

it is almost as if the morphological transformation reduced some of the light background. 

 

C& C welcome.

I usually reduce stars some in images of nebulae or galaxies to highlight the main target.  MT does darken the background, some use a star mask to avoid the effect.  I often like it.

 

Cloning a star as you describe -is- deliberately removing a star ... then just replacing it with a different star. Just as objectionable if indeed you find removing stars objectionable. I think I would draw the line between stretching, sharpening, minimizing data and replacing it altogether. 
 

Derek

Actually clone in a replacement star?  Oh, the horror.  <grin>

 

A case can be made that you're simply removing an unnatural artifact, a horribly bloated star that's just an artifact created by the camera.  Thereby making the image more natural.  That's my story and I'm sticking to it.  <grin>

 

One of my most popular images.  The whole starfield was reduced, also, this is a classic case of highlighting the main target.  Guylain Rochon (BackyardEOS, BackyardNikon) said it was one of his favorite California Nebula images, ever.

 

https://www.astrobin.com/285813/B/
 


Edited by bobzeq25, 15 August 2020 - 02:02 AM.

  • psandelle and pyrasanth like this

#22 schmeah

schmeah

    Fly Me to the Moon

  • *****
  • Posts: 5,925
  • Joined: 26 Jul 2005
  • Loc: Morristown, NJ

Posted 15 August 2020 - 08:48 AM

Actually clone in a replacement star?  Oh, the horror.  <grin>

 

A case can be made that you're simply removing an unnatural artifact, a horribly bloated star that's just an artifact created by the camera.  Thereby making the image more natural.  That's my story and I'm sticking to it.  <grin>

.

 

https://www.astrobin.com/285813/B/
 

I once took an image of the dumbbell nebula and I could never get the color right, so I clone stamped it out and replaced it with the ring nebula that I had taken earlier. Interestingly it looked much better in that starfield so I left it <grin>. Of course with pretty pictures we may do whatever we like and there is no right or wrong. But we all should remain true to whatever ideas we use to define our own personal integrity in this hobby. Adding something that is not there or completely replacing something that is (even if it does detract of the beauty of the image), unless it is truly artifact, somehow falsifies the image in my mind. In -my-mind. Just an opinion Bobzeq25. Not dogma.
 

Derek


  • psandelle, bobzeq25 and pyrasanth like this

#23 pyrasanth

pyrasanth

    Gemini

  • *****
  • Posts: 3,114
  • Joined: 08 Jan 2016

Posted 15 August 2020 - 11:47 AM

Cloning a star as you describe -is- deliberately removing a star ... then just replacing it with a different star. Just as objectionable if indeed you find removing stars objectionable. I think I would draw the line between stretching, sharpening, minimizing data and replacing it altogether. 
 

Derek

I hardly think cleaning up one star is tantamount to an astronomical travesty. Any manipulation of the data including the techniques you described is no worse than anything else that can be done and indeed since noise is inherent in signal where are lines to be drawn?- at the stars or the removal of the noise & light pollution. The observation you have made is pretty much pointless.


Edited by pyrasanth, 15 August 2020 - 11:51 AM.


#24 schmeah

schmeah

    Fly Me to the Moon

  • *****
  • Posts: 5,925
  • Joined: 26 Jul 2005
  • Loc: Morristown, NJ

Posted 15 August 2020 - 12:13 PM

I hardly think cleaning up one star is tantamount to an astronomical travesty. Any manipulation of the data including the techniques you described is no worse than anything else that can be done and indeed since noise is inherent in signal where are lines to be drawn?- at the stars or the removal of the noise & light pollution. The observation you have made is pretty much pointless.

Then why did you say to the OP regarding stars that you would "never deliberately try to remove them"? Just commenting on an inconsistency. I agree we all have to draw our own lines. And I never referred to cloning out a star as an "astronomical travesty". That implication was someone else's drama. 

 

Derek


Edited by schmeah, 15 August 2020 - 12:13 PM.


#25 bobzeq25

bobzeq25

    ISS

  • *****
  • Posts: 21,108
  • Joined: 27 Oct 2014

Posted 15 August 2020 - 12:26 PM

Then why did you say to the OP regarding stars that you would "never deliberately try to remove them"? Just commenting on an inconsistency. I agree we all have to draw our own lines. And I never referred to cloning out a star as an "astronomical travesty". That implication was someone else's drama. 

 

Derek

If you're referring to me, it was someone else's joke.  <smile>


Edited by bobzeq25, 15 August 2020 - 12:31 PM.

  • schmeah 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