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

Horizontal banding in image

  • Please log in to reply
17 replies to this topic

#1 Adrian3245

Adrian3245

    Lift Off

  • -----
  • topic starter
  • Posts: 6
  • Joined: 28 Nov 2023
  • Loc: Australia

Posted 01 December 2023 - 08:57 PM

Hi everyone,

 

Been imaging for close to a year and have recently come across horizontal banding issues in my latest image.

 

I suspect the banding may be caused by the temperature of the sensor and my darks not being correctly matched but I’m not too sure.
 

The temperature was particularly high that night with it being about 30 degrees Celsius at the start of the night and it ending at about 24 degrees Celsius at the end of the night (start of summer here and it can get pretty hot).

 

I did try and create a dark library to try and see if i could match the exif temperature data but that did not help (Also realised that the actual sensor temperature would differ from the recorded body temperature and would make it very hard to match up - if not impossible)

Finished image 
 

Light.jpeg

 

Stacked and stretched dark master 

 

Dark.jpeg
 

My setup

  • Canon EOS R7 (unmodified) 
  • Askar 65PHQ
  • Skywatcher Star Adventure GTI
  • ASIAIR Plus + dew heater

Image details

  • 648 * 25 second light frames (about 4.5 hours of integration)
  • 25 * 25 second dark frames 
  • 30 * 1/8 second flat frames
  • ISO 1600

Any help is appreciated 

- my first time imaging during hotter nights



#2 vidrazor

vidrazor

    Fly Me to the Moon

  • *****
  • Posts: 5,240
  • Joined: 31 Oct 2017
  • Loc: North Bergen, NJ

Posted 01 December 2023 - 09:33 PM

That's why I always shoot darks right after a light session, to have darks at the same sensor temp.

 

30°C is not all that warm for a camera, I've shot in similar temps with my mirrorless and DSLRs and have not experienced that.

 

So you shot those darks at home and got the same stripes? Something's up  with your camera, especially since you mentioned you've been shooting with it and never saw that before.

 

Are you able to borrow another camera and see if you get the same effect under the same temperature conditions?



#3 Adrian3245

Adrian3245

    Lift Off

  • -----
  • topic starter
  • Posts: 6
  • Joined: 28 Nov 2023
  • Loc: Australia

Posted 01 December 2023 - 10:20 PM

I do try an and take the darks right after the lights, unfortunately sometimes it doesn’t pan out and I have to try and get it as close as possible. 

 

That is good that the temperature should not be the issue but is a bit concerning if it’s my camera that is acting up.

 

I should be able to borrow another camera and run it under the same conditions.

 

I also checked through some of my older dark frames and my oldest dark frames to see if I can identify the issue in them.

I found a set that was taken under slightly cooler conditions (about 2 - 3 degrees cooler), also happened to be one of my oldest sets. There is striping but it is negligible, at least in comparison to the dark master I posted.

It appears that the striping is worsening over time though - will need an hour or so to be 100% sure of that though 



#4 BQ Octantis

BQ Octantis

    Cosmos

  • *****
  • Posts: 8,552
  • Joined: 29 Apr 2017
  • Loc: Northern Virginia, USA

Posted 01 December 2023 - 11:36 PM

G'day Adrian,

 

Welcome to Cloudy Nights!

 

Your banding is quite benign. The EXIF sensor temperature is generally pretty good for doing darks library matching, but in high temperatures and high Bortle, the unique gradients from the exact conditions are harder to match. These can even be hard to match at the end of a session if the ambient temperature (and EXIF temperatures) have changed dramatically. So it helps to know how to manage banding in post.

 

If you're not dithering, it helps remove walking noise cause by the finer contrast structures from the mismatch. However, yours instead exhibits fairly large-structure bands—along with opposite gradients in the red and the blue. If you use Siril you can eliminate both, first with a degree 1 polynomial subtraction background extraction, followed by a banding reduction (default settings, but with Protect highlights checked).

 

Lastly, your cores are blown out in both stars and galaxies. I highly recommend a GHT stretch over a linear + gamma cctf stretch. GHT allows you to stretch the shadows while preserving the highlights.

 

Hope that helps.

 

Cheers,

 

BQ


Edited by BQ Octantis, 01 December 2023 - 11:45 PM.


#5 Adrian3245

Adrian3245

    Lift Off

  • -----
  • topic starter
  • Posts: 6
  • Joined: 28 Nov 2023
  • Loc: Australia

Posted 02 December 2023 - 02:02 AM

Hey BQ,

 

That’s good then that the banding is benign. 
That may have been the issue here then, with the temperature and possibly the light pollution (Bortle 5/ 6 where I am).
 

I haven’t dithered yet, it definitely wouldn’t hurt to do that though (will give it a go when I can image next, hoping tonight if the clouds stay away).

I also re-processed the image using the first degree polynomial subtraction and the the banding reduction as you suggested.

Also took a closer look at how I use GHS and noticed I was a bit too aggressive with certain stretches and did not use any highlight protection. 

Here are two variations of the processing I came out with (will process it a few more times as I think it will be very good practice)

 

light1.jpeg

 

Light2.jpeg

 

Not as many things are blown out compared to before and although the banding is still visible, it is an improvement over my original image.

 

Thank you for helping 

 

- Adrian


Edited by Adrian3245, 02 December 2023 - 02:15 AM.


#6 Adrian3245

Adrian3245

    Lift Off

  • -----
  • topic starter
  • Posts: 6
  • Joined: 28 Nov 2023
  • Loc: Australia

Posted 02 December 2023 - 03:09 AM

I have also finished looking at my dark frame stacks and it definitely does appear that the banding in the stacked frames has worsened over time 


  • vidrazor likes this

#7 vidrazor

vidrazor

    Fly Me to the Moon

  • *****
  • Posts: 5,240
  • Joined: 31 Oct 2017
  • Loc: North Bergen, NJ

Posted 02 December 2023 - 03:26 AM

This is not a perfect solution, but you can compensate for the banding in Photoshop/Affinity Photo by dropping in a layer with a solid color of the banding at the brightest banding luminosity, then change that layer transform to lighten. The make a copy of the original layer and merge them. Then use a levels adjustment layer to drop the density back down with the black point while also opening up the gamma (midpoint). Adjust to taste.

 

This is easier done if you separate your stars and apply it to a copy of the starless layer. Below are two examples of what I'm depicting. The first is the effect applied to the dark sample plate (which you don't apply it too, it's here just for example). The second is the effect applied to a starless layer, and then levels adjusted to darken the skies and open up the galaxy data. Then the stars are merged back in. How dark you want the skies is of course personal taste, I like mine dark :-)

Attached Thumbnails

  • lighten.jpg
  • chain.jpg

  • rcg1 and Adrian3245 like this

#8 BQ Octantis

BQ Octantis

    Cosmos

  • *****
  • Posts: 8,552
  • Joined: 29 Apr 2017
  • Loc: Northern Virginia, USA

Posted 02 December 2023 - 08:03 AM

Hey BQ,

 

That’s good then that the banding is benign. 
That may have been the issue here then, with the temperature and possibly the light pollution (Bortle 5/ 6 where I am).
 

I haven’t dithered yet, it definitely wouldn’t hurt to do that though (will give it a go when I can image next, hoping tonight if the clouds stay away).

I also re-processed the image using the first degree polynomial subtraction and the the banding reduction as you suggested.

Also took a closer look at how I use GHS and noticed I was a bit too aggressive with certain stretches and did not use any highlight protection. 

Here are two variations of the processing I came out with (will process it a few more times as I think it will be very good practice)

 

 

Not as many things are blown out compared to before and although the banding is still visible, it is an improvement over my original image.

 

Thank you for helping 

 

- Adrian

No worries at all, mate.

 

Dithering won't remove the large-structure bands, but temperature matching is very helpful. It also helps to let the camera do an hour of temperature soak at ambient to prevent too steep of a change. (Not much you can do if the night temperature changes drastically.)

 

The next-level matching is trying to match the actual temperature curve of the lights. The ideal match would be with an alternating light-dark-light-dark… capture sequence. But if you plot out the temperatures of your darks, you might find you have a set that has roughly the same temperature profile (possibly in reverse):

 

post-273658-0-63421400-1691330548_thumb.

 

I find for my 600D, my darks converge in about 20 minutes, so for matching I select the 20-minute part of the curve that best matches the lights.

 

That said, I find that in high light pollution (Bortle 6/7) I still get better results when I capture at set the end of the session than pulling from my darks library.

 

Cheers,

 

BQ


  • Adrian3245 likes this

#9 DanMiller

DanMiller

    Gemini

  • -----
  • Posts: 3,297
  • Joined: 20 Jun 2022
  • Loc: Pennsylvania, USA

Posted 02 December 2023 - 08:59 AM

It's the Canon camera.  Canon cameras do that.  There is a banding script in Pixinsight that helps remove them.  Sirill also has a script to remove that banding which I actually think is better at removing it.

 

Dan


  • archiebald and Adrian3245 like this

#10 Adrian3245

Adrian3245

    Lift Off

  • -----
  • topic starter
  • Posts: 6
  • Joined: 28 Nov 2023
  • Loc: Australia

Posted 02 December 2023 - 08:58 PM

This is not a perfect solution, but you can compensate for the banding in Photoshop/Affinity Photo by dropping in a layer with a solid color of the banding at the brightest banding luminosity, then change that layer transform to lighten. The make a copy of the original layer and merge them. Then use a levels adjustment layer to drop the density back down with the black point while also opening up the gamma (midpoint). Adjust to taste.

 

This is easier done if you separate your stars and apply it to a copy of the starless layer. Below are two examples of what I'm depicting. The first is the effect applied to the dark sample plate (which you don't apply it too, it's here just for example). The second is the effect applied to a starless layer, and then levels adjusted to darken the skies and open up the galaxy data. Then the stars are merged back in. How dark you want the skies is of course personal taste, I like mine dark :-)

That looks like a pretty good solution, I don’t have Photoshop or Affinity but I do use GIMP occasionally and it should be possible to do in there. I will give it a try today when I have the chance.

I do personally like my skies dark as well (sometimes I will have it lighter though, depending on what was imaged)

 

Thank you for helping :)

 

- Adrian 

 

No worries at all, mate.

 

Dithering won't remove the large-structure bands, but temperature matching is very helpful. It also helps to let the camera do an hour of temperature soak at ambient to prevent too steep of a change. (Not much you can do if the night temperature changes drastically.)

 

The next-level matching is trying to match the actual temperature curve of the lights. The ideal match would be with an alternating light-dark-light-dark… capture sequence. But if you plot out the temperatures of your darks, you might find you have a set that has roughly the same temperature profile (possibly in reverse):

 

post-273658-0-63421400-1691330548_thumb.

 

I find for my 600D, my darks converge in about 20 minutes, so for matching I select the 20-minute part of the curve that best matches the lights.

 

That said, I find that in high light pollution (Bortle 6/7) I still get better results when I capture at set the end of the session than pulling from my darks library.

 

Cheers,

 

BQ

That sounds like a good idea, I usually temperature soak for about half an hour but upping it to an hour could help. 

 

I get what you mean with the temperature curves, it sounds like it would be very useful. I may be able to plot a few curves based off of some of my pre-existing image sets and ambient temperatures that I recorded at the time.

 

The light pollution part might be okay, in a Bortle 5/6 area but has some darker spots which can help.

Thank you,

 

- Adrian

 

 

It's the Canon camera.  Canon cameras do that.  There is a banding script in Pixinsight that helps remove them.  Sirill also has a script to remove that banding which I actually think is better at removing it.

 

Dan

I have heard that some canon cameras are bad for it especially at higher temperatures, didn’t think it may also be effecting the newer mirrorless ones. 
 

- Adrian


  • DanMiller likes this

#11 DanMiller

DanMiller

    Gemini

  • -----
  • Posts: 3,297
  • Joined: 20 Jun 2022
  • Loc: Pennsylvania, USA

Posted 02 December 2023 - 09:39 PM

That looks like a pretty good solution, I don’t have Photoshop or Affinity but I do use GIMP occasionally and it should be possible to do in there. I will give it a try today when I have the chance.

I do personally like my skies dark as well (sometimes I will have it lighter though, depending on what was imaged)

 

Thank you for helping smile.gif

 

- Adrian 

 

That sounds like a good idea, I usually temperature soak for about half an hour but upping it to an hour could help. 

 

I get what you mean with the temperature curves, it sounds like it would be very useful. I may be able to plot a few curves based off of some of my pre-existing image sets and ambient temperatures that I recorded at the time.

 

The light pollution part might be okay, in a Bortle 5/6 area but has some darker spots which can help.

Thank you,

 

- Adrian

 

 

I have heard that some canon cameras are bad for it especially at higher temperatures, didn’t think it may also be effecting the newer mirrorless ones. 
 

- Adrian

I don't know Adrian.  Mine is not a mirrorless one.  So, I don't know.  I found this post in DP review about a EOS R6 banding.  Not sure if this is your issue.  But, here is the link. https://www.dpreview.../thread/4543808

 

He is not an astro photropher.

 

Dan


Edited by DanMiller, 02 December 2023 - 09:40 PM.


#12 vidrazor

vidrazor

    Fly Me to the Moon

  • *****
  • Posts: 5,240
  • Joined: 31 Oct 2017
  • Loc: North Bergen, NJ

Posted 03 December 2023 - 01:10 AM

That looks like a pretty good solution, I don’t have Photoshop or Affinity but I do use GIMP occasionally and it should be possible to do in there. I will give it a try today when I have the chance.

I do personally like my skies dark as well (sometimes I will have it lighter though, depending on what was imaged)

 

I have heard that some canon cameras are bad for it especially at higher temperatures, didn’t think it may also be effecting the newer mirrorless ones.

You may want to consider getting Affinity Photo. It has a one time fee (no subscription), of normally $70 US, but is on sale for the Holidays right now for only $42US. I highly recommend it. Unlike GIMP, it has (among other things) layer transforms like Photoshop that make image data processing so much better and easier. it also has several astrophotography specific tools and can also work directly in 32 bit, and has 32 bit RAW development. Checkout the link above.

 

One of the program's developers also has a series of free astrophotographic macros (although he does request a donation if you can afford one, but it is not demanded) for automating all sorts of processes (check out his video). There are also general purpose macros for Affinity Photo as well, many which as also free. There's plenty of general as well as astrophotographic tutorials of Affinity Photo on YouTube, You can download a trail version of it to see for yourself, but really, it's worth it, especially now at it's sale price.

 

Yes, some modern-day Canons seem to also suffer from artifacts as some other cameras, and they're not necessarily heat related, so caveat emptor.


Edited by vidrazor, 03 December 2023 - 01:17 AM.


#13 Zambiadarkskies

Zambiadarkskies

    Surveyor 1

  • -----
  • Posts: 1,580
  • Joined: 17 Nov 2021
  • Loc: Zambia

Posted 03 December 2023 - 11:47 AM

I don't know Adrian.  Mine is not a mirrorless one.  So, I don't know.  I found this post in DP review about a EOS R6 banding.  Not sure if this is your issue.  But, here is the link. https://www.dpreview.../thread/4543808

 

He is not an astro photropher.

 

Dan

I haven't experienced any banding with the R6.  Amp glow - yes.  But not banding.  


  • DanMiller likes this

#14 MiraS

MiraS

    Sputnik

  • -----
  • Posts: 33
  • Joined: 25 Jun 2022
  • Loc: Germany

Posted 03 December 2023 - 12:38 PM

Usually Fitswork (there is also a english version available on the website) works really great with problems like this. Just open the image in Fitswork. Then click on processing -> background flatten -> lines to equal values. This works with either a light or dark. Sometimes you need to adjust the values a bit with the slider. I tried it with your compressed jpgs and it worked perfectly.



#15 Phil Sherman

Phil Sherman

    Aurora

  • *****
  • Posts: 4,866
  • Joined: 07 Dec 2010
  • Loc: Cleveland, Ohio

Posted 03 December 2023 - 02:04 PM

Canon banding was so bad when Mike Unsold wrote ImagesPlus (now a free download from his web site) that he included a function that can remove horizontal and vertical banding. He applies an FFT transform to the image, drops the term that represents the banding, then reassembles the image from the remaining terms. The output image is identical to the input with the banding completely removed. I assume that other programs that remove banding are doing the same thing or have taken a different approach to removing the banding.


  • DanMiller likes this

#16 Adrian3245

Adrian3245

    Lift Off

  • -----
  • topic starter
  • Posts: 6
  • Joined: 28 Nov 2023
  • Loc: Australia

Posted 03 December 2023 - 09:21 PM

You may want to consider getting Affinity Photo. It has a one time fee (no subscription), of normally $70 US, but is on sale for the Holidays right now for only $42US. I highly recommend it. Unlike GIMP, it has (among other things) layer transforms like Photoshop that make image data processing so much better and easier. it also has several astrophotography specific tools and can also work directly in 32 bit, and has 32 bit RAW development. Checkout the link above.

 

One of the program's developers also has a series of free astrophotographic macros (although he does request a donation if you can afford one, but it is not demanded) for automating all sorts of processes (check out his video). There are also general purpose macros for Affinity Photo as well, many which as also free. There's plenty of general as well as astrophotographic tutorials of Affinity Photo on YouTube, You can download a trail version of it to see for yourself, but really, it's worth it, especially now at it's sale price.

 

Yes, some modern-day Canons seem to also suffer from artifacts as some other cameras, and they're not necessarily heat related, so caveat emptor.

It looks like Affinity Photo has a fair few features that would really be very helpful. I will give it a try today, seems like it will definitely be worth it.

 

I see what you mean, some very interesting artifacts with some of the listed cameras.

 

- Adrian

 

I don't know Adrian.  Mine is not a mirrorless one.  So, I don't know.  I found this post in DP review about a EOS R6 banding.  Not sure if this is your issue.  But, here is the link. https://www.dpreview.../thread/4543808

 

He is not an astro photropher.

 

Dan

It doesn’t appear to be the issue that I have been experiencing, I have experienced it before when recording at certain frame rates in artificial lighting though.

 

- Adrian

 

Usually Fitswork (there is also an english version available on the website) works really great with problems like this. Just open the image in Fitswork. Then click on processing -> background flatten -> lines to equal values. This works with either a light or dark. Sometimes you need to adjust the values a bit with the slider. I tried it with your compressed jpgs and it worked perfectly.

I gave Fitswork a quick try and it does appear to remove the banding extremely well (couldn’t see the banding which is really good)

 

- Adrian



#17 nsblifer

nsblifer

    Vostok 1

  • -----
  • Posts: 164
  • Joined: 27 Oct 2021

Posted 22 December 2023 - 06:54 AM

I can add to this. I ONLY see banding in my photos (with A73) if I shoot my darks on a separate night/and or before my imaging session without allowing enough time to acclimate the rig to outside temperature. Even when the temp change is negligible. This leads me to believe these mirrorless/dslr cams are much more sensitive to temperature change than we think.
  • Adrian3245 likes this

#18 archiebald

archiebald

    Messenger

  • -----
  • Posts: 438
  • Joined: 03 Feb 2009
  • Loc: Japan

Posted 22 December 2023 - 07:01 AM

It's the Canon camera.  Canon cameras do that.  There is a banding script in Pixinsight that helps remove them.  Sirill also has a script to remove that banding which I actually think is better at removing it.

 

Dan

Yes, unfortunately it is just a Canon thing, it's not a specific problem with the camera, they have been that way for as long as I can remember, even in low light terrestrial photography.

 

It's the main reason I finally shifted to a cooled astrocam.  General noise I could deal with but the banding was depressing.


  • DanMiller and Adrian3245 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