Often in programming the handling of UTF-8 data is mis-managed. This happens in new applications which are mis-configured and in old applications which started with latin1 table character sets. The result of these errors are UTF-8 sequences where an extended character belongs.
API Skeletons has developed a tool which analyzes an entire MySQL database on a per-byte level and corrects UTF-8 sequences, even multiple encoded, into proper UTF-8 characters. The output of this tool is a properly encoded database. We can correct your data.
Each invalid UTF-8 character may be double or triple encoded. Each case is unique. If you have a database you need to correct please Contact API Skeletons