Hi Benjamin,
thats a tough question, which usually lead to long discussions in other threads. But I try to answer this one with my own thoughts, although I'm not sure that others may agree. But what I'm sure of is, that there are many others that don't agree
From what I have learned, I know that the SNR in a single frame increases with exposure time. So when the signal doubles, the noise increases only by the root of 2. That means, the longer you expose your sub, the better the SNR will be. But obviously, there are some limitations like light pollution, seeing or guiding performance. I am not going scientific again, because that may be a battle I loose. I'm no expert on that. But for the basics, I prefer this video:
https://www.youtube.com/watch?v=1PUTWfWgD0g However, where I live, the conditions are ok-ish. So my goal is to maximize the SNR in my subs without overexposing. Light pollution is not a big deal for me in my opinion. So The limit of the exposure time is the point, where there are some pixels starting to get overexposed. You can see this in NINAs Statistics panel. There is the MAX field, where you can see the max brightness a pixel can have. Behind this number in brackets is the number of pixels, that are already overexposed in your image. I usually keep this in the range of under a thousand pixels. This might be ok if you have 10 to 20 million pixels in total. But it depends on the target. For a bright object (i.e. globular clusters), this might be too much. For really faint things, this number might be even a bit higher. So you can decide for yourself, where the useful data should be found in the dynamic range of your sensor. That is how I usually deal with those things. For narrowband, the guiding performance of your setup might limit the max exposure time. If the stars get too blurry while exposing 1800 seconds, it might not be worth it to maximize the SNR.
I guess thats not exactly the scientific way you are looking for. But I doubt, there is a generic rule you can use without taking some other things into account. For me, it's just a mix of experience, trial and error and confidence in the experience of other photographers (looking at the image details of their images).
The Optimal Exposure Calculator brings up too short exposure times for my taste as well. I assume, the developers are referring to the known presentation of Dr. Robin Glover, which seems to split the astro photographers community in half. He seems to be much cleverer than me, but there are some things that are at least questionable. (Without any further comment on this)
So I hope this might be of some interest to you. If not, please just ignore me...
CS
Christian
