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

Why does sky limited exposure time not depend on conversion noise?

CMOS DSLR Equipment
  • Please log in to reply
3 replies to this topic

#1 Astrogerdt

Astrogerdt

    Lift Off

  • -----
  • topic starter
  • Posts: 4
  • Joined: 26 Feb 2020

Posted 08 May 2021 - 08:15 AM

Hello all,

As far as I have read, it is generally considered reasonable to expose until you are background limited, so that the signal from the sky background swamps the readout noise to such an extent that even faint structures stand out from it. So far so good. I understand this quite well, and I get it right in the application.

 

But now I had a thought yesterday. If the signal should be much stronger than the readout noise, in order to lift weak details from it, shouldn't the signal also be much stronger than the conversion noise?

 

On my 200d, the readout noise at ISO 800 is 2.85 e-. According to the overview here, the conversion noise of the 200d is 2.6DN (I assume this is an average of all ISO values. Does anyone know more?), which corresponds to 20.8 e- at ISO 800. Wouldn't this conversion noise then actually have to be added to the readout noise, and exposed until the signal is significantly stronger than the sum of these two values? Or am I grossly wrong in my understanding?

 

If that were the case, then the exposure time until I am background limited would increase by a factor of roughly 10, which would drive up the exposure time considerably.

 

A small remark at the end: I know that you don't get exact values calculated this way, but only guide values. But that is not my point here. In practice, of course, the photographer still makes the picture.

I hope you can help me a bit with my understanding here.

 

 

Kind regards and CS
Gerrit



#2 deansjc

deansjc

    Mariner 2

  • -----
  • Posts: 224
  • Joined: 14 Jan 2018

Posted 08 May 2021 - 08:36 AM

CMOS Deep Sky Photography by Robon Glover.

I strongly suggest that, if you have not already, watch the above video available on YouTube.

I believe your questions directly relate.

John

#3 sharkmelley

sharkmelley

    Aurora

  • *****
  • Posts: 4,667
  • Joined: 19 Feb 2013
  • Loc: UK

Posted 08 May 2021 - 05:39 PM

Hello all,

As far as I have read, it is generally considered reasonable to expose until you are background limited, so that the signal from the sky background swamps the readout noise to such an extent that even faint structures stand out from it. So far so good. I understand this quite well, and I get it right in the application.

 

But now I had a thought yesterday. If the signal should be much stronger than the readout noise, in order to lift weak details from it, shouldn't the signal also be much stronger than the conversion noise?

There are two factors at work here. 

 

Firstly each individual exposure needs to swamp the readout noise, otherwise in your final stack your signal-to-noise ratio for the faint details will limited by the read noise.

 

Secondly you need lots of total integration time.  This is the only way to lift the faint signal above the noise in the final stack.  Of course the noise we are talking about is the sky background noise because the read noise has already been swamped in each exposure.

 

Mark


Edited by sharkmelley, 08 May 2021 - 05:40 PM.


#4 whwang

whwang

    Soyuz

  • *****
  • Posts: 3,589
  • Joined: 20 Mar 2013

Posted 09 May 2021 - 08:29 AM

Hi Gerrit,

I believe conversion noise should be included. The page you referred to is the DxO derived quantities. If you look at the input-referred read noise, the read noise there includes conversion noise.

Also, based on the DxO numbers you referred to, I can’t reach the conclusion of 20.8 e you mentioned. It’s way too high.


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: CMOS, DSLR, Equipment



Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics