Well if it's Unicode then what you are seeing is almost certainly the Byte Order Mark, which is prepended to most Unicode files. The BOM dictates whether the characters should be read in little-endian or big-endian format.