Gary (founder of this group forum) just encouraged us to post about our failed experiments. I was a bit hesitant to post about my failed fine-art attempts before, but I think Gary is right that maybe it will foster some good discussion and help others avoid pitfalls.
What I was trying to accomplish with the image above was a fine-art, high-contrast, black-and-white rendition of the Boston City Skyline and Charles River with the Orion constellation (and nebulae) coming out of the light pollution dome.
The reason it failed was the image quality is not good enough for a fine art print. I only achieved what I'd consider smartphone quality.
The saga ;) :
I had major obstacles in even achieving the above photo. Firstly, to get that perspective I was shooting from the top of a parking garage (took lots of scouting locations to find), and many nights the top of the parking garage was closed due to icy conditions. I just had to drive there to find this out, as there was no one to call. Secondly, in my planning, the photo was more interesting with the Orion constellation closer to the city skyline (different time of year) at 35mm on the ASI1600, but I realized in testing that the light pollution dome was too strong when orion was that low. This also changed my strategy with lens selection, and I ended up going much wider (14mm) at a different time of year. By going wider, I think I introduced serious bandpass issues with my 5nm Ha filter, as all the stars (even at center) were not round even stopped down to f/4.0. My initial plan was to also shoot the city in Ha, as my at home experiments showed I could get decent signal with long enough exposures. But again, I realized my tests at home were with different optics and the 14mm was just not delivering sufficient quality, so I ended up shooting the city parts (with mount off) with the Luminance filter. For me, this partly ruined the 'concept' of the photo which was "Boston in H-alpha".
So, what did I learn? The photo I was actually imagining is a mosaic image shot at perhaps 135mm f.l. I think this would give me the details I want for a print, and avoid the bandpass issues I was having with such a wide angle lens. The problems with the new plan: To get the FOV I want, this would be many panels. Given the difficulties with the location; I don't think the new plan is feasible. I still have this image stuck in my head, but I'm not sure I have the motivation to try again after sinking probably 40 hours in to this image, which did not meet my expectations.