Rockbox mail archiveSubject: Re: Not reading genres
Re: Not reading genres
From: George Styles <george_at_ripnet.co.uk>
Date: Thu, 25 Mar 2004 12:38:35 -0000
I have actually written a utility to tag up MP3s with ID3 v1 called RipTag
Its freeware, and can use FreeDB to tag the songs (if you have the whole
----- Original Message -----
From: "Fred Maxwell" <rockbox_at_anti-spam.org>
To: "'Rockbox development'" <rockbox_at_cool.haxx.se>
Sent: Thursday, March 25, 2004 12:17 PM
Subject: RE: Not reading genres
> Jean Boullier wrote:
> > Second comment, I am using MusicMatch "for ages" now (I bought a
> > life-license long before Archos' birth) and I have never had any problem
> > with the core tags, whatever mp3 tool I have used with my files. If they
> > could read ID3v2 they could show the genre correctly, otherwise they
> > found the ID3v1 at the end and show what MusicMatch had put there.
> If you are using MusicMatch to rip CDs, then I strongly recommend that
> switch to Exact Audio Copy (EAC). MusicMatch rips "blindly" and will not
> detect CD reading errors and perform re-reads when errors occur. EAC
> Also, EAC seems to create tags fine. It is freeware and you can download
> at http://exactaudiocopy.de/.
> Is there information in ID3v2.x tags which is not present in ID3v1.1 tags
> that you absolutely need? If not, I recommend going with ID3v1.1 tags.
> They are compatible with everything whereas the v2.x tags are fraught with
> As Linus says, the specs are confusing and subject to interpretation.
> Sometimes MP3 files start with an ID3v2.x tag rather than an MP3 frame
> header. ID3v2 tags are of variable size and the MP3 header is not
> guaranteed to start immediately after the v2.x tag. Some tools apparently
> add extra null bytes between the header and the first MP3 frame. An MP3
> frame header starts with 11 sync bits (11111111b 111?????b) and the rest
> the header information follows. Needless to say, many files will contain
> things that look like a frame header. Although the ID3v2 spec includes a
> description of "syncsafe integers" and "unsynchronization bytes" which are
> designed to ensure that ID3v2 tags do not look like a frame header, their
> use is optional in most places. So there are many ID3v2 tags with byte
> sequences that look like frame headers.
> ID3v2.x tags are a mess and should be avoided. Hopefully, there will be
> unambiguous ID3v3 spec which specifically excludes the use of ID3v2.x tags
> in any MP3 file containing an ID3v3 tag.
> Fred Maxwell
Received on 2004-03-25