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

Light leakage taking long exposure darks

  • Please log in to reply
10 replies to this topic

#1 dwassem

dwassem

    Lift Off

  • -----
  • topic starter
  • Posts: 13
  • Joined: 06 Jan 2021

Posted 21 September 2021 - 11:06 AM

Hi all:

 

I am trying to take my first dark frames.

The problem is there seems to be significant light leakage around the edges, and I am wondering if this is a problem.

 

I took the camera of the scope, put the lens cap on, taped in the lens cap with layers of foil and tape, wrapped the whole thing in blankets, and put these in a dark room. Therefore, I wonder if the camera emits its own light somehow, e.g. the on/off led light bleeding in.

 

Does anyone have experience with this?

 

Thanks!

 

Camera: asi1600mm

10m exposure

 

 

Attached Thumbnails

  • 20210921_112743.jpg


#2 Rudi Bjoern

Rudi Bjoern

    Vostok 1

  • *****
  • Posts: 147
  • Joined: 28 Jan 2014
  • Loc: Danmark

Posted 21 September 2021 - 11:20 AM

What gain and temperature?
Looks like amp glow.

#3 KungFood

KungFood

    Mariner 2

  • -----
  • Posts: 249
  • Joined: 18 Jul 2014
  • Loc: Dallas, Tx

Posted 21 September 2021 - 11:20 AM

What temp are you cooling the camera to, and are you leaving room in the blanket cocoon for the exit vents from the cooler?


  • Oort Cloud likes this

#4 idclimber

idclimber

    Surveyor 1

  • *****
  • Posts: 1,854
  • Joined: 08 Apr 2016
  • Loc: McCall Idaho

Posted 21 September 2021 - 11:43 AM

The only thing that I can tell from your image is that was taken as it was displayed on the iPad. The image does not look correct to me because it is missing the signature amp glow that the 1600mm has. It is also a bit brighter than a dark is typically displayed. 

 

How the image looks is completely dependent on the software you are using and if it is presented in a linear format or if it is stretched. A dark frame should look completely black if displayed in the orriginal linear form. This is also the case for a light frame with the exception of a few stars that are close to being clipped. Displaying images in the linear form is not particularly useful as you really can't see much. As such most capture software will apply a temporary stretch. 

 

It is only when the software stretches that data that the image either looks grey in the case of a dark or details in the deep sky object begin to appear. 

 

The main problem with stretching the images is that minuscule differences in brightness between adjacent pixels is greatly magnified. If you acutely use a tool to see what ADU value between 0 and about 65,000 (16bit display mode) is on each pixel you will likely find the brightest pixel is about 625 while the darkest is about 600. Those actual measurements are based on gain, exposure and the noise of your sensor. 

 

Each software that I have used to analyze images is a bit diffrerent. But you should be able to get information about average, min, max ADU values on any image. 

 

If you upload an originall FIT file of one of the darks, I will be happy to open it and compare it to my darks for the 1600mm. 


  • jdupton likes this

#5 OldManSky

OldManSky

    Fly Me to the Moon

  • *****
  • Posts: 5,438
  • Joined: 03 Jan 2019
  • Loc: Valley Center, CA USA

Posted 21 September 2021 - 11:54 AM

Yes, the 1600 does produce "amp glow," which means it produces its own light.  It's not leakage.

I'm kind of having a hard time with the 10-minute exposure, though...trying to figure out why you would do 10-minute exposures...?

 

With that camera, that's probably not a good idea.  



#6 maximalz

maximalz

    Sputnik

  • -----
  • Posts: 30
  • Joined: 12 Nov 2019
  • Loc: France

Posted 21 September 2021 - 01:53 PM

Hey, like OldManSky said, this camera can take 10 mins exposures, but you really should be asking yourself WHY would you do 10 mins exposures. That topic is very complex and vast, but I highly suggest Cuiv the Lazy Geek's video on gain/offset/exposure time to get a better grasp on this. There are plenty of research videos on the topic of shorter vs longer subexposures that have proven that very long exposures are rarely worth it (unless you know why you're doing it).

 

Unless you're planning on doing extremely faint and unusual targets, chasing exposure time is a common beginner's trap. A better approach is trying to balance dynamic range and read noise your camera produces. The more gain, the less DR but the less read noise too. Take into consideration the fact that the shorter your subexposures, the less risk your subs get ruined by any variable out of your control (wind, planes, extreme satellites trails, ...). It's all a balancing act smile.gif Unless you're really experienced and know WHY you're doing any exposure over 300s (5mins), I highly suggest you simply don't.

 

As for your darks library, here's what I've been working on for the past couple of days, and I think it's a nice balance of versatility. Attached is my current progress on my darks library (the layout was borrowed from another CN member who did that better than me!). Yellow is planned, green is done, the rest is not a priority for now. My plan is as follow :

  • use gain 76 on targets that have dark gas structures (such as the eagle nebula) : the high dynamic range should allow for more details in here
  • use gain 139 on pretty much everything else narrowband for now because that's the best balance of noise and DR most of the time
  • use gain 200 on anything too faint or if I need to take short subs for some reason
  • use mostly gain 76 on LRGB targets because most of these can really use higher dynamic range. Test things out for shorter subs with gain 139 and 200.

 

As for offset, I went and did Offset 60 because, while it's costing a little bit in dynamic range, i'm extra safe dodging dark pixels.

 

Someone with a much more practical approach than me please correct me if I'm wrong. I still need to test all of this out, but the theory should check out!

EDIT : Also attached is a 300s, -15°C, 139 gain and 60 offset dark sub from my library, with basic STF stretch in Pixinsight. The glowing edges is amp glow which you should notice in your light subs too.

Attached Thumbnails

  • darkslib.png
  • 300s139g60o-15d.jpg

Edited by maximalz, 21 September 2021 - 01:59 PM.

  • OldManSky likes this

#7 ChrisWhite

ChrisWhite

    Fly Me to the Moon

  • *****
  • Posts: 7,160
  • Joined: 28 Feb 2015
  • Loc: Northeast USA

Posted 21 September 2021 - 02:00 PM

Share a full resolution fits from the camera via Dropbox. Easy to tell you if it's good or bad, but the screen share doesn't help.

Also, nothing wrong with 10 minute exposures. Sometimes that's what you need.

#8 idclimber

idclimber

    Surveyor 1

  • *****
  • Posts: 1,854
  • Joined: 08 Apr 2016
  • Loc: McCall Idaho

Posted 21 September 2021 - 02:24 PM

With the 1600mm I routinely used 10 minute exposures when imaging with any of the narrow band filters. I still do with the 2600mm that replaced it. I also have very dark skies and a good mount which makes a huge difference. 

 

I look at how many pixels are clipped solid white, and back off if it necessary. Voyager displays this metric and makes it very easy to do a test image even before it is dark enough to actually image. 

 

On the 1600mm I used two gains/offsets. 139/20 and 200/40. Both at -20c. The lower gain was for any LRGB filters. The higher gain was on any narrow band target. 



#9 dwassem

dwassem

    Lift Off

  • -----
  • topic starter
  • Posts: 13
  • Joined: 06 Jan 2021

Posted 23 September 2021 - 10:02 PM

Thanks so much all for the quick and many replies!

 

What gain and temperature?
Looks like amp glow.

0 gain, target temperature -20 (but actual more like -18)

 

What temp are you cooling the camera to, and are you leaving room in the blanket cocoon for the exit vents from the cooler?

targeting -20, which it achieved outside, but inside it was more like -18. I left vents for cooling. At one point I thought that might be the issue so I tried to wrap it fully with ice. That warmed it up to -5 within the 10m exposure time, but at least I was able to confirm the light in the corners was still there

 

The only thing that I can tell from your image is that was taken as it was displayed on the iPad. The image does not look correct to me because it is missing the signature amp glow that the 1600mm has. It is also a bit brighter than a dark is typically displayed. 

 

How the image looks is completely dependent on the software you are using and if it is presented in a linear format or if it is stretched. A dark frame should look completely black if displayed in the orriginal linear form. This is also the case for a light frame with the exception of a few stars that are close to being clipped. Displaying images in the linear form is not particularly useful as you really can't see much. As such most capture software will apply a temporary stretch. 

 

It is only when the software stretches that data that the image either looks grey in the case of a dark or details in the deep sky object begin to appear. 

 

The main problem with stretching the images is that minuscule differences in brightness between adjacent pixels is greatly magnified. If you acutely use a tool to see what ADU value between 0 and about 65,000 (16bit display mode) is on each pixel you will likely find the brightest pixel is about 625 while the darkest is about 600. Those actual measurements are based on gain, exposure and the noise of your sensor. 

 

Each software that I have used to analyze images is a bit diffrerent. But you should be able to get information about average, min, max ADU values on any image. 

 

If you upload an originall FIT file of one of the darks, I will be happy to open it and compare it to my darks for the 1600mm. 

 

That might be it! I used the Zwo Asiair Pro which applies auto stretch.

I did not think of amp glow, as I believed this was usually coming in from the side from one specific direction rather than from all corners.

 

I am not sure how to read the ADU values correctly but I tried in pixinsight. It seems to go from 800-900?

 

Appreciate your offer to have a look! Can you check it below?

 

https://1drv.ms/u/s!...1XOIIw?e=arqa1J

 

Yes, the 1600 does produce "amp glow," which means it produces its own light.  It's not leakage.

I'm kind of having a hard time with the 10-minute exposure, though...trying to figure out why you would do 10-minute exposures...?

 

With that camera, that's probably not a good idea.  

Thanks - looks like that was a rookie mistake!

 

Hey, like OldManSky said, this camera can take 10 mins exposures, but you really should be asking yourself WHY would you do 10 mins exposures. That topic is very complex and vast, but I highly suggest Cuiv the Lazy Geek's video on gain/offset/exposure time to get a better grasp on this. There are plenty of research videos on the topic of shorter vs longer subexposures that have proven that very long exposures are rarely worth it (unless you know why you're doing it).

 

Unless you're planning on doing extremely faint and unusual targets, chasing exposure time is a common beginner's trap. A better approach is trying to balance dynamic range and read noise your camera produces. The more gain, the less DR but the less read noise too. Take into consideration the fact that the shorter your subexposures, the less risk your subs get ruined by any variable out of your control (wind, planes, extreme satellites trails, ...). It's all a balancing act smile.gif Unless you're really experienced and know WHY you're doing any exposure over 300s (5mins), I highly suggest you simply don't.

 

As for your darks library, here's what I've been working on for the past couple of days, and I think it's a nice balance of versatility. Attached is my current progress on my darks library (the layout was borrowed from another CN member who did that better than me!). Yellow is planned, green is done, the rest is not a priority for now. My plan is as follow :

  • use gain 76 on targets that have dark gas structures (such as the eagle nebula) : the high dynamic range should allow for more details in here
  • use gain 139 on pretty much everything else narrowband for now because that's the best balance of noise and DR most of the time
  • use gain 200 on anything too faint or if I need to take short subs for some reason
  • use mostly gain 76 on LRGB targets because most of these can really use higher dynamic range. Test things out for shorter subs with gain 139 and 200.

 

As for offset, I went and did Offset 60 because, while it's costing a little bit in dynamic range, i'm extra safe dodging dark pixels.

 

Someone with a much more practical approach than me please correct me if I'm wrong. I still need to test all of this out, but the theory should check out!

EDIT : Also attached is a 300s, -15°C, 139 gain and 60 offset dark sub from my library, with basic STF stretch in Pixinsight. The glowing edges is amp glow which you should notice in your light subs too.

 

Thanks a lot. That gives me some peace of mind. Will also try and read up on exposure times.

I basically followed the advice in the book which was to "fill up" the histogram in a test shot to determine the right exposure.

 

Share a full resolution fits from the camera via Dropbox. Easy to tell you if it's good or bad, but the screen share doesn't help.

Also, nothing wrong with 10 minute exposures. Sometimes that's what you need.

Thank you - please see above for link

 

With the 1600mm I routinely used 10 minute exposures when imaging with any of the narrow band filters. I still do with the 2600mm that replaced it. I also have very dark skies and a good mount which makes a huge difference. 

 

I look at how many pixels are clipped solid white, and back off if it necessary. Voyager displays this metric and makes it very easy to do a test image even before it is dark enough to actually image. 

 

On the 1600mm I used two gains/offsets. 139/20 and 200/40. Both at -20c. The lower gain was for any LRGB filters. The higher gain was on any narrow band target. 

Thank you. I also looked at the histogram clip to decide on the 10m but I will experiment with other values as well as your recommendations for gain and offset



#10 idclimber

idclimber

    Surveyor 1

  • *****
  • Posts: 1,854
  • Joined: 08 Apr 2016
  • Loc: McCall Idaho

Posted 23 September 2021 - 11:30 PM

I downloaded the dark. It looks perfectly normal to me. ADU values are as you say so you are reading them correctly. The header showed it was taken at a gain of 0, offset of 50 and a temperature of -16.5. 



#11 dwassem

dwassem

    Lift Off

  • -----
  • topic starter
  • Posts: 13
  • Joined: 06 Jan 2021

Posted 23 September 2021 - 11:56 PM

I downloaded the dark. It looks perfectly normal to me. ADU values are as you say so you are reading them correctly. The header showed it was taken at a gain of 0, offset of 50 and a temperature of -16.5. 

Thank you. I appreciate the quick check. Good to know that I seem to be taken the darks correctly then.

 

At any rate, it looks like I need to start a bit of gain and reduce exposure time, so will take new darks accordingly.




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