Thanks, Ferrante. That is all correct. But let me briefly explain why I arrived at the values I have stated, which work for my location, but may have to be optimized for other locations.
First, I need to make a correction. I am not using 240 gain for LRGB, I am using unitary gain, i.e. 139 and 21 offset with the ASI1600MM-Pro, according to the tables that come with that camera.
I am using 240 gain for Ha and O3, 270 gain for S2. 120 s exposure for Ha and S2, 60 s for O3 in the city.
Obviously, all images should be recorded at 16bit depth.
Now to the reasons why.
I found the cloudy nights tables for which Wouter provided the link very helpful, but a bit too complicated. So I simplified them for practical purposes. The final result is nearly the same.
The cloudy nights tables recommend that the mean ADU should be 800 - 1100 ADUs above the dark ADUs, which in my case with the ASI1600MM are ~690 ADUs at 120s exposure time and -10 C sensor temperature. Given that I am imaging from the upper end of light pollution from the middle of a white zone, I aim at a mean ADU value of 1200 above the dark ADUs, so around 1900-2000. Read noise does not come into play under conditions of extreme light pollution, so one can safely neglect that, I think. It is extremely low for the ASI1600MM anyway.
The next consideration is that I want to avoid taking exposures longer than 120s. First, short exposures, as you point out, minimize guiding issues and result in fewer wasted frames. Second, where I live I have to contend with lots or airplane traffic. I have trails on approximately 10% of my 120 s exposures, so if I go longer, that % will only increase.
With these two constraints, my critical parameters are set. The remaining one, i.e. how to arrive at 2000 mean ADU with 120s exposure time max, is gain adjustment. That needs to be calibrated experimentally for the particular conditions and no specific gain can be given here, except for stating that the lower the better.
In other words, if you want to shoot longer LRGB exposures, you certainly can, I can just say that with all the noise coming from Bortle 8/9 light polluted skies, long exposure times are a waste. The dynamic range one wins by lowering the gain is more than offset by the light pollution, up to 240 gain in my case (the reduced dynamic range and increased noise becomes noticeable at 300 gain and gets rapidly worse above, so I definitely want to stay below that).
Aside from that, LRGB from a light-polluted city really gives acceptable results only on the brightest objects (like M42). I have almost given up on LRGB imaging from the city, I am almost exclusively using narrowband, i.e. Ha, O3 and S2. That still works, but even so, I definitely recommend filters of 5 nm or less bandwidth. The narrower the band, the less the effects of light pollution (obviously).
I hope this helps. Please let me know if the reasoning makes sense to you and if you see any room for improvement.
Cheers
Jo
PS: Check out the images I have posted under my profile. You can clearly see the progression in quality over time as I adjusted the values as stated above, especially when you compare the images of the Trifid with OSC and with narrowband. The noise in the OSC image is horrific, greatly mitigated by the narrowband imaging, even for the reflection nebula part.
And before I forget it, DEFINITELY get the cooled version!!!! Cooling the sensor is absolutely critical for reducing noise. Do not skimp on the $300 difference in price tag. I discovered how HUGE the difference is when I forgot to turn on the cooler one night last summer. The difference was between a good image and NO image, just noise and a wisp of haziness on top of it. Buying the uncooled ASI1600 only makes sense if you are living north of the Arctic circle and even there, global warming will render it useless within the next few years.
To illustrate the point: This is an image taken with a DSLR, though, from my area. ISO 400 30s, resulting mean ADUs 75! (on a scale from 0-255, so 8 bit). 500 exposures stacked and stretched to optimal range in PixInsight, but not processed further.
And this is probably the brightest LRGB object in the sky.