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

Lets Clarify Some Terminology – Gradients

  • Please log in to reply
11 replies to this topic

#1 Salacious B Crumb

Salacious B Crumb

    Mariner 2

  • *****
  • topic starter
  • Posts: 262
  • Joined: 13 Apr 2017
  • Loc: Greater Atlanta Area, GA

Posted 09 February 2018 - 02:09 PM

There are a lot of different posts regarding this issue but rather than hijacking someone else' topic, here's a new one.

 

So now when I’ve figured out (fingers crossed) why my stars were sick, I would like to understand what’s going on with my background. What are the areas I have marked on my image and why aren't they dark as the sky (I believe) should be? Are they gradients? If yes, from where are they coming from? Light pollution? I image from an orange zone (20.3 SQM according to the iPhone app) with narrowband filters and most of my backgrounds look like this. My mean ADU’s have typically been low 400’s and up to 600. The part which I don’t understand is that when I look at the AstroBin I see great photos taken from more severely light polluted zones with less integration time, yet without a hint of background what my images show. So what takes? Is it all in post processing? Or are my calibration frames not doing their job? Are the ZWO filters affecting to this? Not enough separation from the left side of the histogram? What am I missing here?

 

 

Gradient.JPG

 

Disclaimer; So far I have only used the PI’s preprocessing script to process my images. And I’m still using the bias calibration frames instead of the flat darks. This just in case these facts make a difference.

 

 

Thanks,

 

- Sal



#2 bobzeq25

bobzeq25

    Voyager 1

  • *****
  • Posts: 13462
  • Joined: 27 Oct 2014

Posted 09 February 2018 - 03:01 PM

Here is absolutely the main thing you're missing.  Some of that (probably most or essentially all) is certainly real.  Look at this image:

 

https://www.astrobin...301686/?nc=user

 

The stuff on the lower left, with the dark lanes and bright areas, is very unlikely to be either light pollution or amp glow (one reason you want dark flats with the 1600).

 

Two things.  From the Earth, skies are never dark, the darkest parts of an image are often actual dust.  Why they stuck the Hubble out there. 

 

And, there's a lot of stuff out there.  <smile>

 

Those images with black backgrounds?  Likely to be someone who raised the black level too high in processing.  People do that for a variety of reasons.  Most imagers find a hint of gray more realistic.

 

Putting my Rosette next to yours, and rotating, I can clearly see hints of the same thing.   That's an old image, when I had much less experience, the black level on it is a touch too high.  Typical rookie mistake.  <grin>

 

https://www.astrobin.../full/240686/C/

 

Processing is everything.


Edited by bobzeq25, 09 February 2018 - 03:22 PM.


#3 Goofi

Goofi

    Cosmos

  • *****
  • Moderators
  • Posts: 8048
  • Joined: 03 May 2013
  • Loc: Coastal Southern California

Posted 09 February 2018 - 03:35 PM

To answer your questions

 

Space, no, yes, yes, no, yes, no, a lot.

wink.gif

 

Best advice I can share is to not compare your results to others and try to match their's. Let your data speak for itself.

 

If you really are worried about background levels, slew over to the constellation Leo and shoot some random area of the sky with the same sub lengths as you normally shoot and measure its background. Any patch of sky without background nebulosity will work.  Narrowband filters do not eliminate light pollution - they just minimize it. The wider your bandpass, the more light pollution gets through.

 

Gradients are changes in the background across the frame. The challenge with a nebula that fills the frame is determining what is true background and what is just faint nebulosity.  Here is a place where comparing to other images can prove fruitful. In this case, I find using a widefield view - much wider than my framing - is far more helpful. For your Rosette, I'd use something like this APOD.  You'll see right away that there is a lot of background nebulosity in the region, and it extends significantly in several directions. 

 

One thing I'll suggest is when you stretch in PI don't use the default STF stretch. CTRL-click the STF radioactive button and change the stretch from 25% down to 20%. See if that doesn't help some of your issues. You can always tweak using HT a little, but with how much the Rosette is filling the frame I think you'll be fine.



#4 Salacious B Crumb

Salacious B Crumb

    Mariner 2

  • *****
  • topic starter
  • Posts: 262
  • Joined: 13 Apr 2017
  • Loc: Greater Atlanta Area, GA

Posted 09 February 2018 - 03:38 PM

Thanks bobzeq25. Just to clarify, you are saying that most of the stuff showing should actually be there and should further be seen also in the images? I'm confused (which is nothing new)...

 

Here is a link to a gorgeous AstroBin Rosette and IOTD https://www.astrobin...?page=6&nc=user (I apologize if I broke some rules here). There is a clear difference between the darker background sky and the nebulosity. And that's what I thought Rosette should look like. Based on what I read now, my image above is not that far off, with post processing (the above is just a stacked image), adding contrast and what not I might be able to get something which actually resembles the reality up there.

 

 

- Sal


Edited by Salacious B Crumb, 09 February 2018 - 03:39 PM.


#5 Goofi

Goofi

    Cosmos

  • *****
  • Moderators
  • Posts: 8048
  • Joined: 03 May 2013
  • Loc: Coastal Southern California

Posted 09 February 2018 - 03:50 PM

Sal, you need to compare apples to apples ...

 

In the case of the image you linked to from David Dvali, his field of view is larger than yours.

A good exercise is to plate solve both and register one to the other so you can see.

A poor-man's way to see this is find a common feature in one and then locate it in the other.

 

What I'd suggest is in the lower right corner of your image you have two bright stars next to each other; find them in David's image (hint, his image is rotated differently than yours).  When you find them in David's image you'll see surrounding nebulosity .. which matches what you are showing.



#6 Salacious B Crumb

Salacious B Crumb

    Mariner 2

  • *****
  • topic starter
  • Posts: 262
  • Joined: 13 Apr 2017
  • Loc: Greater Atlanta Area, GA

Posted 09 February 2018 - 06:24 PM

I know Goofi. But I like them apples...

 

 

- Sal



#7 bobzeq25

bobzeq25

    Voyager 1

  • *****
  • Posts: 13462
  • Joined: 27 Oct 2014

Posted 09 February 2018 - 06:38 PM

Thanks bobzeq25. Just to clarify, you are saying that most of the stuff showing should actually be there and should further be seen also in the images? I'm confused (which is nothing new)...

 

Here is a link to a gorgeous AstroBin Rosette and IOTD https://www.astrobin...?page=6&nc=user (I apologize if I broke some rules here). There is a clear difference between the darker background sky and the nebulosity. And that's what I thought Rosette should look like. Based on what I read now, my image above is not that far off, with post processing (the above is just a stacked image), adding contrast and what not I might be able to get something which actually resembles the reality up there.

 

 

- Sal

That's exactly what I'm saying.  Look at the image you cited side by side with yours, rotating one as necessary to match.  You'll see some of it.

 

Honestly, forget "reality".  Reality is gray, not color, so you artificially generate color.  Low contrast, so you increase it, particularly locally.  High dynamic range, so you compress it, maybe combine different exposures.  Often using masks to limit your adjustments to certain areas.  Here's an appropriate comment from the excellent book, Lessons from the Masters.

 

"When I first started color processing ten years ago, I was quite anxious.  I was certain that one day the "Color Police" would raid my dome and confiscate the equipment.  I played around with G2v balancing....  These days I just use my judgment."

 

What you see in even the best images is just someone's judgment.  People who claim to do "realistic" images often just mute the color.

 

Once you've stretched the image, reality flies out the window.  If you do LRGB imaging, the colors are inevitably altered in a non-linear way that's impossible to totally reverse.  Don't get me started about narrowband.  <grin>


Edited by bobzeq25, 09 February 2018 - 06:44 PM.


#8 Salacious B Crumb

Salacious B Crumb

    Mariner 2

  • *****
  • topic starter
  • Posts: 262
  • Joined: 13 Apr 2017
  • Loc: Greater Atlanta Area, GA

Posted 10 February 2018 - 10:48 AM

"What you see in even the best images is just someone's judgment.  People who claim to do "realistic" images often just mute the color."

 

I like this, thanks.

 

 

- Sal



#9 Jon Rista

Jon Rista

    ISS

  • *****
  • Posts: 21466
  • Joined: 10 Jan 2014
  • Loc: Colorado

Posted 10 February 2018 - 12:52 PM

Sal, you definitely have gradients in the image in your original post. While there is a lot of faint structure around Rosette, you have fairly evenly brightening corners, which is definitely not natural. That could be LP, it could also be diffuse internal reflections in your scope. It could be miscalibration from flats. Whatever the case, you definitely have gradients, and not all of the signal in those corners is "real and natural". 

 

I would first look into your flats, and make sure they are calibrating properly. This feels more like a slight miscalibration than anything else, although a very very close second it could also be some diffuse reflections (which can be difficult to trace down and resolve.) Usually, true LP shows up as a more linear gradient across the field, however if you have multiple sources of bright LP you can end up with more complex gradients. I don't think the radial nature of your particular gradients is normally caused by multiple LP sources...they would have had to be arrayed rather neatly around your field to create something radial like that, and would have had to have intense falloff.... seems unlikely. 

 

Anyway. I do indeed believe you have some gradients in there. If all else fails, then gradient reduction can help remove them, but it can be challenging, as Bob is right that there is also real structure in those areas as well. You need to separate the gradient from that structure. If you can fix the issue during acquisition...eliminate reflections and make sure flats are correcting properly...that is usually the better solution. If diffuse reflections ARE teh cause, then at the moment the only thing you really can do is gradient extraction, as resolving those kinds of issues usually requires reacquiring the lights and flats. 



#10 bobzeq25

bobzeq25

    Voyager 1

  • *****
  • Posts: 13462
  • Joined: 27 Oct 2014

Posted 10 February 2018 - 01:06 PM

Agree with Jon.  There is plenty of real structure there, but it's overlaid on some smaller gradients.

 

He's more precise than I am.  <grin>  Check out my sig.


Edited by bobzeq25, 10 February 2018 - 01:11 PM.


#11 Salacious B Crumb

Salacious B Crumb

    Mariner 2

  • *****
  • topic starter
  • Posts: 262
  • Joined: 13 Apr 2017
  • Loc: Greater Atlanta Area, GA

Posted 10 February 2018 - 05:28 PM

Thanks guys. My money is on Flats as I think my filing system has failed. That's why I just posted a reply to the Flat Darks post (below) as I think it's time to build new libraries and start to use the Flat Darks. I have to admit that I've been a little lazy with my Flats and plan now to start to take them more often. I have trees on the East side and our house on West so typically it takes me more than a few nights to collect 20+ hours from a single target. With the weather we have had, I think my Flats have "expired" before I had an opportunity to use them for all the different sessions.

 

https://www.cloudyni...0/#entry8392090

 

This said, I'm still not sure what to look for when trying to find Gradients. Can someone point me to the right direction or at least post an example so I know what to look for.

 

Again, I highly appreciate the support I get here.

 

 

- Sal

 

 

 



#12 Jon Rista

Jon Rista

    ISS

  • *****
  • Posts: 21466
  • Joined: 10 Jan 2014
  • Loc: Colorado

Posted 11 February 2018 - 04:19 AM

This may help:

 

https://jonrista.com...oundextraction/


  • keithlt, bobzeq25 and Salacious B Crumb like 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