I tried the one from the sourceforge project referenced in foo_freedb's dialog box and a few other versions I found linked in some antique forum posts - no go.
All I get from freedb when Japanese chars are involved is ?????, regardless of the charset I select in the dropdown.

C.

The charset convertion library used by foo_freedb are named iconv.dll and you can get it from here.

Just download libiconv-win32 and extract iconv.dll from the bin directory. It's all that's needed. Place it in foobar2000 main directory (same as foobar2000.exe) and restart foobar2000 (foo_freedb.dll must obviously be in components directory). If you now click on the square with presumably "iso-latin-1" on it in the foo_freedb window you should get a drop-down menu.

I do get the dropdown - it's just that I get only "?????" in the results, which doesn't change at all if I change the charset - not even into different gibberish.

C.

For Japanese, try this in freedb settings window: http://freedbtest.dyndns.org/~cddb/cddb.cgi

Sorry to dig this thread up - I'd lost the bookmark. The link mixderax gave actually worked - thanks man, you're my hero!

Still, this begs a few questions:
1) Is there a list somewhere of such unofficial freedb servers? This one doesn't look very permanent to me.
2) Why aren't the databases of such servers merged with official freedb? (After all, it does support UTF-8)
3) Any reason why all entries consistingof a large number of actual '?' shouldn't be removed from freedb?
4) Last not least, is there something like a freedb proxy that can merge the results of multiple servers?

C.