If the file is corrupted but readable, and you don't use block-level, it will back up the corrupted version in whole. The block-level we'd likely pick up the new damaged parts assuming the file is large enough to warrant a block level backup. Old versions will remain according to your retention policy. I haven't seen anyone report this particular issue, probably because old versions can be restored if there's an issue and SMART drive status should pick up the problem if you're monitoring. Is this something you've run into previously?

Thanks for your reply.

To the best of my knowledge, none of the files (mainly media) have suffered bit rot, but I have a new NAS since the prior one was old and one drive failed - thank you Raid 1! However, if you visit forums (e.g. Reddit), pretty much everyone will tell you to use BTRFS. But, based upon your explanation, CloudBerry would save me because I would notice that a new version was backed up, which would alert me to a problem (since I almost never modify these files).