Certain characters incorrectly encoded with ISO-8859-1

[nxfi777]

I've noticed that sometimes, the ' character is parsed into ’ (which is the same in ISO-8859-1 single-byte encoding, rather than UTF-8). I cannot create an fully functional MRE, as this only occurs occasionally.

UPDATE: I have just found another case, where the σ (\sigma) in σ-algebras is encoded as σ-algebra