[Rhythmbox-devel] ID3 Tag reading behavior

[Sergio Bobillier <sergio bobillier gmail com>]

Hello

A few months ago I made a bug report about Rhythmbox on Ubuntu bug tracker here: https://bugs.launchpad.net/ubuntu/+source/rhythmbox/+bug/703697 time passed and I received no answer the bug is still unrecognized and unassigned. The bug report claimed that Rhythmbox was failing to recognize Unicode Japanase character from ID3 tags, the bug report was inaccurate in fact the problem is not Rhythmbox failing to recognize Unicode characters it is just a peculiar behavior among media players.

I have some Japanese music which off course had Japanese characters on the ID3 tag information of the Mp3 Files but Rhytmbox always displayed them wrong, some characters where represented like boxes with little numbers in them (like unrecognized characters). It didn't matter how many times I tagged and re-tagged the Mp3s even from within Rhytmbox the characters kept displaying wrong in Rhythmbox.

I tried to open those Mp3 files in all the media players I had, Winamp, Windows Media Player, VLC Player (Linux and Windows), iTunes (on Machintosh), Amarok, Songbird and Cowbell (is not a media player but an ID3 tagger) and in all of them the characters where displayed correctly the only other player that shared the same awkward behavior was Banshee. I couldn't understand why these two player had the same issue I guessed that maybe they were sharing the same ID3 Tag library.

I found by accident the cause of all this. When trying to re tag the songs (like for the 100th time) in an attempt to make Rhythmbox read the tags correctly I tried to totally remove all the tags using Winamp i removed the ID3 v1 tag and then I removed the ID3 v2 tag and then I saved the changes to the file and surprisingly the wrong displayed characters emerged in Winamp, I opened the ID3 Tag window once again in Winamp and it showed another ID3 v2 tag there. So, the file had two ID3 v2 tag headers on it (O.o I don't know why or how they got there).

I removed that other ID3 v2 tag and re-tagged the files, then opened them in Rhytmbox and at last it displayed the correct characters :D YAY!

So, my guess here is that Rhytmbox and Banshee read the first ID3 Tag header that they find on the file while most modern media players read the last one (or maybe read borth and keep the last one they read).

I just wanted developers to know this particular behavior just... in case. I don't know if you are going to make any changes but anyways I just wanted to share this information with you.

Sincerely

Sergio