mb_detect_encoding when using other than UTF8 and Latin-1

Do you have a question? Post it now! No Registration Necessary.  Now with pictures!

I have a multi-language site where strings have to be converted to
UTF-8 for processing.
Charset detection using mb_detect_encoding() for UTF-8 and ISO-8859-1
is working great but not so with other charsets such as Cyrillic or
Japanese Characters. In these cases it returns always UTF-8, even when
I explicitly pass the right encoding in the encoding list parameter.

Does anyone know if there's any way to solve this or tell me where I
can find a subroutine to do universal character set detection? Thanks.

Site Timeline