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

Video lecture on CMOS imaging and noise - from the UK Practical Astronomy Show

  • Please log in to reply
6 replies to this topic

#1 Bennevis

Bennevis

    Vostok 1

  • -----
  • topic starter
  • Posts: 126
  • Joined: 24 Dec 2016
  • Loc: United Kingdom

Posted 18 March 2019 - 01:09 PM

Just getting back into the astronomy thing - well I was at the practical astronomy astronomy show in the UK the other weekend, where I saw a talk by Dr Robin Glover, an expert on CMOS imaging.

 

The talk was about noise and controlling it vs. signal. It explained the maths behind "short subs vs. long subs" thing, which I found very interesting. It turns out long sub exposures and deep cooling may not be that important.

 

There were also graphs showing temperature vs. noise and an explanation of the maths behind stacking as well as the impact of sky glow and narrowband filters. It all made total sense to me.

 

Luckily it was posted it up on YouTube so here's a link: https://www.youtube....h?v=3RH93UvP358

 

Really useful information here and I'm interested to hear what you think.


  • numchuck and Jon Rista like this

#2 Jon Rista

Jon Rista

    ISS

  • *****
  • Posts: 22668
  • Joined: 10 Jan 2014
  • Loc: Colorado

Posted 18 March 2019 - 02:07 PM

Aside from a few small points, caveats I might make, I think his presentation was very good. Covered all the subject matter very well, and theoretically and mathematically he is dead on. A bit of a bummer that he wasn't able to cover the gain stuff, I think that would have been some useful info. 

 

I think there are some definite caveats to exposure length and dark current that should be accounted for. Mostly from a practical standpoint. While he is correct that lower read noise allows shorter exposures, it isn't quite just that simple. Total exposure time still matters, and if you swamp the read noise by the recommended factor (use his recommended exposure lengths), then you will want to get the same amount of total exposure time with any camera, regardless of what its read noise is. If you need 10 hours of exposure, using 3 second subs is going to be rather problematic...you'll need to stack 12,000 subs! From a practical standpoint you would want to use longer exposures than 3 seconds, even if 3 is recommended. That can sometimes get you into another practicality predicament, as if you have smaller pixels, your FWC may not be that large. FWC imposes an upper limit on how long you CAN expose. 

 

So there are definitely balancing factors and practicality factors to consider in the real world. That aside, great presentation!


  • David Ault likes this

#3 DuncanM

DuncanM

    Vanguard

  • -----
  • Posts: 2480
  • Joined: 03 Nov 2009
  • Loc: Arizona Sky Village or the rain forest

Posted 18 March 2019 - 05:38 PM

It also ignores download time, which can be a significant factor depending on the camera and camera/computer data link.


  • David Ault likes this

#4 Bennevis

Bennevis

    Vostok 1

  • -----
  • topic starter
  • Posts: 126
  • Joined: 24 Dec 2016
  • Loc: United Kingdom

Posted 19 March 2019 - 04:00 AM

That can sometimes get you into another practicality predicament, as if you have smaller pixels, your FWC may not be that large. FWC imposes an upper limit on how long you CAN expose. 

 

So there are definitely balancing factors and practicality factors to consider in the real world. That aside, great presentation!

I think he ran out of time is the problem Jon. The lecture was cut short - well it actually overran and when I went out there was a huge queue for the next guy Damian Peach's lecture. 45 minutes just isn't enough time for such a complex subject in my opinion. I sure hope they give him a slot at the next show which is longer! I agree on practicality. I mean my hard drive has limited space, but I suppose some imagers have multi-terabyte drives and so-on so it won't bother them. I saw somewhere online galaxies imaged with thousands of one second frames with a large Dobsonian reflector but I can't find the reference. That must have taken some CPU time to stack and process. I'm just using mine for EAA at the moment so it doesn't bother me.

 

But what does FWC stand for? I haven't seen the term before.



#5 billdan

billdan

    Mariner 2

  • -----
  • Posts: 213
  • Joined: 21 Feb 2012
  • Loc: SE Queensland

Posted 19 March 2019 - 04:13 AM

FWC means full well capacity, or the point where the pixel is fully saturated and can't store any more electrons

 

e.g. An ASI1600 has 20Ke full well capacity at gain zero (5 electrons per ADU) and the full well capacity drops to 4Ke at gain 139 or unity gain (1 electron per ADU).


Edited by billdan, 19 March 2019 - 04:16 AM.


#6 HughGilhespie

HughGilhespie

    Lift Off

  • -----
  • Posts: 6
  • Joined: 01 May 2012

Posted 19 March 2019 - 04:15 AM

FWC = Full Well Capacity.

 

I watched this video and it is really good! I do have the Sharpcap Pro software and I had no idea it was so versatile. I am now looking forward to having a good old play with it.

 

Regards, Hugh



#7 Bennevis

Bennevis

    Vostok 1

  • -----
  • topic starter
  • Posts: 126
  • Joined: 24 Dec 2016
  • Loc: United Kingdom

Posted 19 March 2019 - 10:51 AM

Oh I see yes of course! Gain is essential to any understanding of CMOS. He said he missed a few slides due to time limitations at the show, and I'm pretty sure he said something about gain. I will email and ask if he might consider sharing the missing slides. I bought a TEC camera at the show and some other bits and pieces to play with after the talk. I missed the Damian Peach one on planetary imaging but it's on the youtube channel too.




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