I imported some images to my research-folder, and noticed some ‘remained’ blank and Scrivener was unable to interpret them as a valid image.
for example, I uploaded google.com-screenshot.jpg. I guess Scrivener uses the .com-screenshot.jpg to find the file extension to define Content-Type, instead of ‘only’ reading .jpg. (I noticed this, because everything passed the first dot is chopped in its filename and afterwards tried to re-import the same file, after renaming.)
I guess it is due to an misspelled PREG-pattern… an, maybe, even ‘better’ solution to this problem would be to interpret the Content-Type from the first bytes of each imported file.