I got the files. Again, the overall file structure is the same for both the working and non-working files. (You called the non-working file "corrupt", but I don't think it is.)
The two commands executed by Geosetter are very different, and the resulting XMP is also very different. But simply changing the XMP content shouldn't cause problems for any reasonable XMP reader. It is still valid XMP, although it may contain some irrelevant/unnecessary information such as image dimensions and bit depth, etc. You can try this command to see if it is just the XMP content that is the problem:
exiftool -tagsfromfile IMG_20211024_131137_noncorrupt.dng -xmp IMG_20211024_131137_corrupt.dngIf the resulting file works in your DAM, then it was definitely the XMP causing the problem. If so, you really need to go back to your DAM people because it is valid XMP. (Just for fun, I also ran the "corrupt" XMP through both the RDF validator
https://www.w3.org/RDF/Validator/ and the PDF/A XMP validator
https://www.pdflib.com/pdf-knowledge-base/xmp/free-xmp-validator/ and neither showed any problems.)
But if it doesn't work, the other difference is that the "noncorrupt" file also contains IPTC. If this happens, run this command on the new "corrupt" file from the last command:
exiftool -tagsfromfile IMG_20211024_131137_noncorrupt.dng -iptc:all IMG_20211024_131137_corrupt.dngIf this now works with your DAM but the previous version didn't, then the difference is the IPTC. Again, the IPTC is technically fine. But due to tests you have run earlier, I suspect the the XMP is the source of the difficulty.
- Phil