Page 1 of 1

Sky noise

Posted: Fri Apr 05, 2019 9:28 am
by festoon
Hi Robin,
I'm trying to recreate your optimum exposure time calculation in a script (as you use in the brain function).
Can I please ask - how do you calculate the sky noise from the sky electron rate?
The total sky electrons will be the total exposure time x sky electron rate
Do you assume the sky noise to be the square root total sky electrons or perhaps the total sky electrons/some value?

Re: Sky noise

Posted: Fri Apr 05, 2019 4:18 pm
by admin

The shot or Poisson noise on a number of electrons is always given by the square root of the number of electrons.

Cheers, Robin

Re: Sky noise

Posted: Fri Apr 05, 2019 7:01 pm
by festoon
Thanks Robin...
I'm trying to do the same calculation, but I'm getting very different answers...please see image below that calculates the optimum exposure for a given total stacking time, read noise, allowed increase of 10% in total noise, and sky electron rate.
IMG-3767_1.jpg (286.25 KiB) Viewed 1101 times
An example calculation based on this is shown below here
calc.PNG (26.32 KiB) Viewed 1101 times
Are you able to please let me know if this follows the same methodology as used in your calculations, or am I missing something?

Re: Sky noise

Posted: Sun Apr 07, 2019 7:28 pm
by festoon
Hi Robin,
By luck today I found your explantions here
Looking at this equation
equation.PNG (15.07 KiB) Viewed 1066 times
It seems there is one major difference between my equations above and your link above. The first term you have in this equation is the read noise of the final stack which is n x read noise^2. In my equations above I effectively used n^2 x read noise^2
Does the total read noise equal the square of the total read noise of the stack or n x read noise squared?

Re: Sky noise

Posted: Sun Apr 07, 2019 7:56 pm
by admin

Since the read noises uncorrelated from one frame to the next, you need to add it in quadrature when combining the read noise with other noise sources. The upshot of this is that if you were considering the read noise only, the total noise in the stack would be the square root of the number of frames stacked times the read noise of one frame.

Hope that makes sense, Robin