Quote:
Originally Posted by dikman
Re- drones and balloons, fine for short exposures but for long(ish) exposures stability and a fixed location would seem to be a necessity. Trying to maintain constant, and long, accurate pinpoint tracking with a drone could be an interesting exercise. Not to mention a balloon, which is dependent on the vagaries of any wind (or atmospheric disturbances).
|
I had exactly the same thoughts when I first read the ChatGPT output - but then it occurred to me: NASA has already solved those issues with the SOFIA instrument (infrared observatory mounted in a Boeing 747 - now retired).
Sure - the average hacker / maker doesn't have access to the sort of resources of NASA, but I don't think it would be insurmountable to give it a try. Put a relatively short focal length / wide angle camera onto a gyroscopically stabilised mount (available on commercial hobby drones, for example), and shoot video with relatively short exposures (< 1 second say), rather than long exposure imagery. I think it would be possible to use stacking software to put together some reasonable images - e.g. night sky / landscape images, but taken from a different viewpoint to the more common ground-based pictures of the Milky Way above the horizon / tree-line.
As a thought-experiment into using AI for "lateral thinking", I thought it might prompt some new ideas for people looking for a new challenge - which it did, for me. Perhaps unsurprisingly - it turns out that it's not a totally original concept - a quick Google search for "drone-based astrophotography" shows people are already doing this:
https://www.youtube.com/watch?v=gvN5lnklDNs
https://fstoppers.com/aerial/drone-v...n-nikon-352660
(Note that in the second link, the subject used a drone to get his camera above the fog bank, so he was able to capture some nice images on a night which would have kept most amateur astrophotographers inside.)