Can anyone give me a clear explanation on what the difference is between UTF-16 and ISO-8859-1 ? I know that UTF supports Unicode, and the ISO version is for devices or whatnot that don't support Unicode...but what exactly does that all mean ? From some reading around, it seems like ISO would be the most tolerant, so why isn't that the default rather then UTF-16 for most tag writing programs ? Are there specfic pros/cons to using each format ?

I have a music server jukebox that I'm interested in using, and I've found that it won't read the tags correctly in UTF-16 but does in ISO. Before I resave my entire collection in ISO, I would like to understand the implications, if any, in doing so.

Any help would be greatly appreciated.

UTF-16 writes two bytes (16 bits) per character while ISO-8859-1 writes only 1 byte (8 bits) allowing only 255 characters.

Ok, but what does that translate into from a user perspective ? Is it going to make a big different to winamp, my car's mp3 disk player, etc if I rewrite my tags in ISO instead of UTF-16 ?

When I use mp3tag, the default is UTF-16, how come ? Like I said, from what I can figure out, ISO seems to cut out the ability to use asian, cyrillic and other characters that would not be used in a Western alphabet. But ISO seems to be more compatible since many devices/players/whathaveyou have trouble reading Unicode. So wouldn't that make ISO the more logical choice ?

Any insight into this mystery would be great appreciated.

ISO-8859-1 (Often called "Latin 1") only supports Western European characters. That's includes the familiar A-Z alphabet, and a few characters with accents, graves and umlauts like ö

You should use UTF-16 unless you have a specific reason not to. It sounds like you have such a reason

It's as simple as that. There is a full international compatibility with UTF-16, and using codepages you can display the proper characters only from that particular codepage.

I agree with bensky and use UTF-16 (v2.3) for my entire collection because I have some devices that don't read v2.4 tags (UTF-8 encoded).

Btw., which one is more commonly used and supported by hardware player: a) UTF-16 LE or b) UTF-16 BE ?

Good question (I don't know the answer either), but I think both should be supported...

Interesting question.
So I tried it. Mp3Tag writes UTF-16 LE tags and they work OK in old Creative MuVo TX. Using a hex editor, I converted the TIT2 frame to UTF-16 BE (swapped the bytes and changed the BOM to FE FF) and the player displayed it correctly.

I guess that players will accept both "flavors" of UTF-16

Good. My guess is that it's actually a spec violation if a program only supports one flavor (when reading).

Like it or not, ISO-8859-1 is likely the most "supported." Unicode isn't 100% prevalent.

Get devices that support Unicode and dump those that don't.

I have my MP3s tagged in ID3v2.3 UTF-16. Besides working on most software players, they also work on my iAudio U3 and Nokia 6120 classic. This kind of "cross platform" makes me very happy.

Personally I think UTF-16 should be promoted over ISO, as this is the way to go to support world-wide characters in any systems.