I don't know if encoding information is embedded in the IGOR files or
not.  Currently the stock parser just leaves everything it reads in in
bytes.  For compatibility, the igorpy module attempts to convert those
byte strings to Unicode, but it just assumes that the encoding used in
the file matches the locale or default encoding used by your system.
This could be a portability issue.  Until commit

  commit fe7006e3
  Author: W. Trevor King <wking@tremily.us>
  Date:   Sat Jul 21 07:50:09 2012 -0400

    Replace igor.igorpy parsing with translations from igor.packed.load.

The igorpy parser used sys.getfilesystemencoding() to guess the
encoding, but that encoding is actually used to encode file names, not
file contents.  locale.getpreferredencoding is a better guess, but
it's still just a guess.