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

What's this noise?

Astrophotography
  • Please log in to reply
10 replies to this topic

#1 belliott4488

belliott4488

    Messenger

  • *****
  • topic starter
  • Posts: 403
  • Joined: 05 Oct 2020
  • Loc: MD, US

Posted 09 December 2021 - 08:37 AM

Here is an extreme stretch to emphasize some noisy color bands across an image of M31. Is this a recognizable form of noise? What's the best way to try to remove or reduce it?
 
M31_noisy_med.jpg
 
Thanks,
Bruce
  • lee14 likes this

#2 sharkmelley

sharkmelley

    Fly Me to the Moon

  • *****
  • Posts: 5,233
  • Joined: 19 Feb 2013
  • Loc: UK

Posted 09 December 2021 - 09:20 AM

Which camera are you using?



#3 lee14

lee14

    Apollo

  • *****
  • Posts: 1,450
  • Joined: 19 Dec 2009
  • Loc: CNY

Posted 09 December 2021 - 09:39 AM

The issue needs to be solved at a more fundamental level, but this is the result of some quick tweaking of your image in PS.

 

Lee

Attached Thumbnails

  • post-344153-0-61494600-1639056991_thumb fix.jpg


#4 belliott4488

belliott4488

    Messenger

  • *****
  • topic starter
  • Posts: 403
  • Joined: 05 Oct 2020
  • Loc: MD, US

Posted 09 December 2021 - 09:40 AM

Which camera are you using?


It's the Nikon D5000 in my signature. It was on an 8" f/3.9 Newt with a GSO coma corrector (which I think I still need to get better adjusted, but that's another matter). This image is from 70x120s lights, 30 flats, 34 darks, and 30 bias frames. I was also auto-guiding, not that that likely contributed to this noise signature.



#5 belliott4488

belliott4488

    Messenger

  • *****
  • topic starter
  • Posts: 403
  • Joined: 05 Oct 2020
  • Loc: MD, US

Posted 09 December 2021 - 09:42 AM

The issue needs to be solved at a more fundamental level, but this is the result of some quick tweaking of your image in PS.

 

Lee

Thanks, Lee - that's encouraging!

 

I've progressed to the point where I think my greatest weaknesses are now in the processing stage. Did you do anything other than careful stretching, maybe of individual color channels?



#6 lee14

lee14

    Apollo

  • *****
  • Posts: 1,450
  • Joined: 19 Dec 2009
  • Loc: CNY

Posted 09 December 2021 - 09:47 AM

Since the overall color is good, I just desaturated the reddened streaks. I darkened the sky background just a tad, and applied the basic sharpening tool.

 

Lee



#7 sharkmelley

sharkmelley

    Fly Me to the Moon

  • *****
  • Posts: 5,233
  • Joined: 19 Feb 2013
  • Loc: UK

Posted 09 December 2021 - 10:01 AM

It's the Nikon D5000 in my signature. It was on an 8" f/3.9 Newt with a GSO coma corrector (which I think I still need to get better adjusted, but that's another matter). This image is from 70x120s lights, 30 flats, 34 darks, and 30 bias frames. I was also auto-guiding, not that that likely contributed to this noise signature.

It's possible you are seeing the Nikon coloured concentric ring problem caused by the raw data compression.  Given the D5000 is only a 12bit camera and given the compression table it uses, the D5000 is affected worse than most Nikons.

 

The only way to know for sure is to convert your raw files to DNG and use the ring removal tool before feeding the resulting files into your stacking program:

https://www.cloudyni...ngs/?p=10867894

 

Alternatively, upload a raw light and a raw flat to a file sharing site and I can easily check them.

 

Mark



#8 belliott4488

belliott4488

    Messenger

  • *****
  • topic starter
  • Posts: 403
  • Joined: 05 Oct 2020
  • Loc: MD, US

Posted 09 December 2021 - 11:06 AM

It's possible you are seeing the Nikon coloured concentric ring problem caused by the raw data compression.  Given the D5000 is only a 12bit camera and given the compression table it uses, the D5000 is affected worse than most Nikons.

 

The only way to know for sure is to convert your raw files to DNG and use the ring removal tool before feeding the resulting files into your stacking program:

https://www.cloudyni...ngs/?p=10867894

 

Alternatively, upload a raw light and a raw flat to a file sharing site and I can easily check them.

 

Mark

Ah, thanks! I had come across that thread in the past but hadn't paid enough attention to it, being unsure if it affected me. I'll be sure to read through the full thread when I can; I see already that there's a wealth of good information there, mostly from you. It looks like you did some really great work on this issue, so I'll probably be the latest person to add my thanks.

 

I'll upload my lights and calibration frames when I get home tonight, and if anyone feels like taking a look I'll be very grateful.

 

Note: A replacement for the old D5000 is definitely on my wish list, but I was thinking I would wait until the point when it had become the weak link in my imaging chain (other than me). Maybe I'm reaching that point (??).



#9 belliott4488

belliott4488

    Messenger

  • *****
  • topic starter
  • Posts: 403
  • Joined: 05 Oct 2020
  • Loc: MD, US

Posted 09 December 2021 - 11:15 PM

It's possible you are seeing the Nikon coloured concentric ring problem caused by the raw data compression.  Given the D5000 is only a 12bit camera and given the compression table it uses, the D5000 is affected worse than most Nikons.

 

The only way to know for sure is to convert your raw files to DNG and use the ring removal tool before feeding the resulting files into your stacking program:

https://www.cloudyni...ngs/?p=10867894

 

Alternatively, upload a raw light and a raw flat to a file sharing site and I can easily check them.

 

Mark

I went ahead and put all my image files as well as DSS log files and master files here in anticipation of asking for help processing later. If you feel inclined to grab sample light and flat frames, and to check them, I'd be grateful. I'll download your colored ring correction tool and see what luck I have with it in the mean time.


Edited by belliott4488, 09 December 2021 - 11:15 PM.


#10 sharkmelley

sharkmelley

    Fly Me to the Moon

  • *****
  • Posts: 5,233
  • Joined: 19 Feb 2013
  • Loc: UK

Posted 10 December 2021 - 01:57 AM

I went ahead and put all my image files as well as DSS log files and master files here in anticipation of asking for help processing later.

I downloaded your files and unfortunately I have some bad news.

 

As expected, the "ring artifacts" are embedded in the raw data of both the lights and the flats. This can easily be seen by dividing the red channel by the green channel and dividing the blue channel by the green channel as I have done here with one of your flats:

 

D5000_ChannelDivision.png

 

The bad news is that the rings are not caused by the data compression and so the tool I wrote will not remove them.  The rings look more similar to concentric rings I found in my Nikon Z6:

https://www.cloudyni...ting/?p=9641613

 

Since your camera is only 12bit, you won't be able to work around this by increasing the exposure to the point where the back-of-camera histogram reaches the halfway point because these rings are much more pronounced in 12bit data. 

 

Mark


Edited by sharkmelley, 10 December 2021 - 01:59 AM.


#11 belliott4488

belliott4488

    Messenger

  • *****
  • topic starter
  • Posts: 403
  • Joined: 05 Oct 2020
  • Loc: MD, US

Posted 10 December 2021 - 07:11 AM

I downloaded your files and unfortunately I have some bad news.

 

As expected, the "ring artifacts" are embedded in the raw data of both the lights and the flats. This can easily be seen by dividing the red channel by the green channel and dividing the blue channel by the green channel as I have done here with one of your flats:

 

attachicon.gifD5000_ChannelDivision.png

 

The bad news is that the rings are not caused by the data compression and so the tool I wrote will not remove them.  The rings look more similar to concentric rings I found in my Nikon Z6:

https://www.cloudyni...ting/?p=9641613

 

Since your camera is only 12bit, you won't be able to work around this by increasing the exposure to the point where the back-of-camera histogram reaches the halfway point because these rings are much more pronounced in 12bit data. 

 

Mark

Ah, that's too bad - but thanks very much for taking the time to look at this for me.

 

I might have to consider upgrading this camera sooner rather than later. It was only meant as a "starter" camera anyway, so I always knew that I'd either upgrade or give up on the hobby (which I have no intention of doing yet).

 

Thanks again for all your help - cheers!




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





Also tagged with one or more of these keywords: Astrophotography



Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics