Not my genre, but, on the topic of fish-eyes; I thought that astro-photography was actually one of the places that they were actually the more 'correct' tool for the job.
The fish-eye 'effect' isn't actually a distortion; they are rendering an enormously wide angle of 'curved' view onto a flat plane, with 'true' angular perspective.
Non-Fish-Eye, 'Rectaliniar' lenses, are the ones that are actually 'distorting' the image, squashing a curved plane of view onto a flat sensor plane.
For very small angles of view, from a long telephoto, the very small angle of curvature of the scene is so small that it 'tends' to flat, over the scale of projection, and is still reasonably close as focal length tends to the 'normal', Whilst as you enter the realms of UWA lenses, the angle of view is so large that the amount of 'distortion' by way of 'rectilinear correction', can be quite enormous.
Its like the maps of the world, squashing the orange-peel, you get the Mercator effect, where the at the poles, the angle of the earths curve is so great in relation to the perpendicular angle of view, you have to either segment it like a peeled orange, and leave 'gaps' between the flattened segments, or you have to enormously 'stretch' the ends of the segments to fill the gaps.
On paper, filling the frame, such a map of the earth looks more natural, but its actually enormously distorted, and measurements on the paper are not to scale depending whether you are measuring North-South or East West.
Fish-Eye images then, rendering the curved scene to a circle, and lacking the 'distortion' to squash the circle into a square, are actually a more natural and accurate rendering of the scene, and linear measurements on the image, are actually true-scale, lacking that squashed orange peel 'correction' to fit the circle in the square.
As said, it's not a genre I indulge in, but it is a 'quirk' of UWA and fish I have actually observed, displayed trying to use either, to work in confined spaces to get photo's of motorbikes, where things like handlebars or wing mirrors, that much closer to the camera, than the main subject, the side of the bike, a Fish-Eye often renders them much more naturally, where I have observed some quite 'unnatural' distortion of what should be straight lines, like a mirror stem, from a UWA, which isn't 'fishy' just plain weird! In some cases, not so much bowing as actually wobbling bending in more than one direction, as the rectilinear correction hammers the orange peel flat at the edges, to make the circle, square!
I suppose that with Astro-photography, the camera-to-subject distance is so enormous, that something a few million miles closer to the camera at the edge of the frame, than in the middle, that the perspective doesn't give such a marked 'wobble', but I would still have thought it would have bee discernible.
Is the more accurate 'uncorrected' perspective provided by Fish-Eyes then, actually not an advantage in Astro-Photo, or only in scientific astro-photo, where images are used for plotting ad mapping and will be measured, where for observational Astro-Photo it's merely the aesthetic merit that matters?
