Page 2 of 8

Re: Picking the correct exposure for Deep Sky

Posted: Sun Feb 16, 2020 3:43 pm
by admin
Hi,

Yes, you're welcome to translate this content and share it. I would be very grateful if you could make sure that your translated version includes a link to the original content and credits me as the author.

Thanks, Robin

Re: Picking the correct exposure for Deep Sky

Posted: Mon Feb 17, 2020 1:10 am
by Landhiker
Thank you very much Dr Robin!
Sure I will include a link to this thread and credit you as the author.
Thank you again!

Re: Picking the correct exposure for Deep Sky

Posted: Sun Mar 15, 2020 9:36 pm
by mwg3656
Hi Robin,

I have been a user for 2 years and I found your noise analysis (previous posts) helpful and the March 9, 2019 video. I would like to see more of the math behind the calculations from the table in the video.

I run the sensor analysis and divide Full Well Depth by Read Noise and usually pick the point where the sensor changes gain amplifiers. What I find when discussing with individuals is 1) too long of exposures with an outside source (airplane lights, satellites, tracking issues) causing them to unwanted artifacts or 2) too much gain (poor SNR, over exposed), because the ability to see an image in real time exciting.

Equipment:
RASA 11" f/2.2
ZWO 6200 color (Read noise 1.53)
Optolong L-enhanced filter (maybe medium band filter)
Typical settings gain 100 and exposure time 30 to 120sec depending on object

I usually do not image for fixed time, but until I see a "smooth" histogram in front of the major peak and low ripple in the end of the histogram. My judgement based on observing the histogram develop.

My question is calculating the optimum number (10% or 5% noise) frames to stack?

thanks,
Mark

Re: Picking the correct exposure for Deep Sky

Posted: Sun Mar 15, 2020 9:57 pm
by admin
Hi Mark,

I think you can still use the calculation results even though you don't image for a fixed time – the calculations are trying to optimise to give you the best possible output image in a certain amount of time (for instance one hour). The same settings are also going to be optimal for getting to a particular image quality in the shortest time possible. If you run the calculations either in SharpCap or in a spreadsheet calculator, you will notice that the recommended gain, exposure and offset values don't change if you change the total imaging time.

Hope that makes sense, Robin

Re: Picking the correct exposure for Deep Sky

Posted: Mon Mar 16, 2020 2:48 pm
by mwg3656
Hi Robin,

Thanks, I think I understand. I like to plan my astrophotography sessions and would be interested in spreadsheet to calculation the number of frames for each object before imaging. Where do I find the spreadsheet?

thanks in advance,
Mark

Re: Picking the correct exposure for Deep Sky

Posted: Mon Mar 16, 2020 6:40 pm
by admin
Hi Mark,

As far as I'm aware there isn't a spreadsheet readily available with these calculations in (unless somebody has followed the discussion above and put the calculations into their own spreadsheet). I was just commenting on the possibility of making that sort of spreadsheet if you felt inclined.

Cheers, Robin

Re: Picking the correct exposure for Deep Sky

Posted: Tue Mar 17, 2020 12:52 am
by oopfan
Mark,

You can check out my site here: https://snrcalc.now.sh/

The code is open source and available at GitHub (scroll down to the footer for a link). You can create a spreadsheet if you want from the formulae in calculation.service in the core/services folder.

It takes minutes to read the general information on the home page, followed by the detailed information at the Help link.

I used the site to help plan the effort behind these DSOs:
https://astrotuna.com/m81-bodes-galaxy-in-11-6-hours/
https://astrotuna.com/star-clusters-m35-ngc-2158/

All of my work is with a mono cam using LRGB filters. I use the site's calculator to get a feel for the amount of time needed to capture the DSO with the "L" filter. I shoot for SNR 15-20. To pick up the color I use another calculator at a different site that tells me how much RGB to capture. Generally speaking if I spend "X" time capturing "L" frames then I will spend "X" time capturing R+G+B frames. The RGB frames add to the total SNR but you get your biggest boost in SNR from capturing "L".

I don't have any feedback from people using OSC cams. If that is your camera then I can only recommend that you start somewhere. Start with a target SNR of 15 and then go out and capture those number of frames. Process it and decide if that meets your expectations. If not, capture more and add it to your inventory. You'll eventually get to a point where you are pleased with the result. The last important step is to plug the total time you invested in the DSO into the calculator to see what your total SNR is. If it says SNR 'X' then use 'X' as the target for your next DSO effort.

I don't LiveStack. I capture individual frames and process them in APP. If you do LiveStack then you can look into a technique that Minos developed for saving your LiveStack results every so often and then stack those stacks in APP. I think that is what Minos does.

Good luck,
Brian

Re: Picking the correct exposure for Deep Sky

Posted: Fri Oct 30, 2020 2:01 am
by anamik
Hello Dr. Glover,

Thank you for a lucid explanation. I may be reading this wrong but references to read noise in units of electrons and to the symbol as sigma_r lead to a units mismatch in the equation for stack noise.

Background sky noise ~ tQrp,sky ~ units of electrons

Read noise ~ sigma_r ~ units of electrons (mismatch in equation) or sqrt(electrons)?

Thanks,

Shail

Re: Picking the correct exposure for Deep Sky

Posted: Fri Oct 30, 2020 7:40 pm
by admin
Hi,

The background sky noise term is actually the variance associated with the shot noise. This has a value that is numerically equal to the number of electrons collected, but the units would strictly be electrons squared. This arises because the mean and variance of a possion distribution are numerically equal.

Robin

Re: Picking the correct exposure for Deep Sky

Posted: Sat Oct 31, 2020 2:25 pm
by anamik
Ah, got it. I should have figured it out from the mention of shot noise following Poisson statistics.

Thanks,

Shail