I'm interested in getting the best possible star photographs with the S23 Ultra. It certainly excels in this area but I have some suggestions to Samsung for improvements.
My comments here pertain to getting super sharp star images with minimal "star trailing".
Astromode uses image-stacking of several automatically timed images over 4 min or more (I rarely use more than 4 mins as skies aren't dark enough here for 8mins). From the image info in the pict file, it looks like the timed images for the 3 cameras are:
20 sec. X1 main camera feq=23mm
15 sec. X3 wide camera feq=69mm
6 sec. X10 perisc camera feq=230mm
where feq are the equivalent 35mm film camera full frame focal lengths.
For sharp star images (without star trailing due to earth's rotation), the well known simple rule of thumb is the *minimum* exposure time for a camera lens with an feq value us:
t_exposure (sec) ~ 500/feq(mm)
so for the S23 Ultra these exposure times should be roughly:
22 sec. X1 main camera feq=23mm
7 sec. X3 wide camera feq=69mm
2 sec. X10 perisc camera feq=230mm
So the main camera is fine, the X3 camera (without extra zooming) is about 2 times longer so star training will be obvious, and the x10 is 3 times longer than optimum for star trailing. This is evident in the photos.
So I'm hoping (since these exposure timings can't be changed manually in astromode) that Samsung might change these fixed values in a future update😀
Of course one can use Pro mode or Expert Raw (not in Astromode) and set the exposure times manually for single image shots. However, the most impressive shots are the Astromode shots with image averaging and correction of of earth rotation (star movement) BETWEEN the timed stacked shots.
Finally, apart from star trailing issues, the focus setting must be accurately adjusted to get pin-point sharp images. This is tricky with the manual adjustment slider. I'm hoping Samsung improves this focus adjust feature to show the actual value in a dial visual (as in the recent improvements to the Samsung Photo Editor 3.3.10.8 interface). The 2 screenshots below show this. Also it would be excellent to include this focus value (e.g. .72) in the image metadata for accurate documentation.
Just found this ~ 2020 Nasa publication:
