Characters and encoding bullshit

So I've been trying to make a little C# application to parse a file name and its path to automatically tag my extensive library of music and have come upon TagLib to help me.

Now everything was going well until I fell upon some Japanese characters in my library. During parsing, it all appears to be fine, but when it executes the .Save() method, all characters become question marks (?). For example: "アールグレイ" will appear as such during debug, but on save, the actual tag will be "??????"

I have no idea if this is due to the library assuming the MP3 is in ID3v1 format, due to the .Save() method malfunctioning or something else. Has anyone used this library before and come upon a similar problem? Any idea how to fix this?

>>15
I'm glad you think UTF-16 is
1) A valid internal representation
2) A good to expose the programmer to when you're writing abstract string handling data structures
But you see, some of us don't live in the '90s and simply enjoy the benefit of byte arrays for which we don't give a flying fuck about encoding and more importantly: a sane default encoding so morons like FIOC programmers don't go around littering UTF-16 encoded text everywhere.
Take Haskell for example. UTF-8 everything. No need to re-encode shit unless some retard like you sends UTF-16 over the network because you forgot to waste CPU cycles going from one variable length encoding to another. Then of course there's the byte ordering thing to take care of, but I assume that FIOC will automatically use a healthy mixture of different endianness, you know, because sane and consistent internal representation doesn't matter right?