[charrua]

I use MySQL for the DB so I don't have MyMusic32.db and I have a big music library so the music DB dump from MySQL is about 60Mb.
I'll try to pinpoint the albums, artists and songs that trigger this error and let you know.
Thanks

Much appreciated.
You can supply a filter to qa, eg. "qa songs africa" (although it's the song after Africa that is causing the problem) which should speed things up. Partial filter matches also work. If you enable the logfile and logfile.verbose properties, the logfile information might help with any diagnosis/solution.
With Python2 I've been able to replace stdout and stderr with Unicode-savvy codecs, but have so far failed to get this to work in Python3 which explains why you're hitting this - I'll keep looking for a solution though...

You're completely right. Switching to Python 2.7.4 solves the problem.
The names of some japanese, chinese and russian albums that I have seem to be the problematic ones when using the script with the newer Python 3.1.1.
Let me know if you want the list of the names of the albums for testing purpose, but probably any non ascii chars will return the same error using the script with Python 3.x
Thanks again.