Tag Standard Updated (Re: [vorbis] Quality & Tags)

[Glenn Maynard]

On Thu, Jan 31, 2002 at 01:01:23PM -0800, Jonathan Walther wrote:
> Glenn, could you list the information thats waiting for metadata?

I can list what I want for metadata:

Translations, transliterations (eg. romanization), and phonetic versions
of the same information.  (See Message-ID: <20020104085955.GC3686 at zewt.org>
for an example of this.)  It's important that different versions of a
single piece of information be grouped.

Every piece of information in tags should have an equivalent (preferably
identified by the same name) in the metadata format, so tags can be
generated from metadata.  Don't make me maintain two sets of data.
Having more specific data may be OK; for example, if there are separate
"title" ("December, 1963") and "popular-title" ("Oh What A Night"), they
can both go into "TITLE" in the tags.

One tricky thing about that is the "priority of information" problem
that's come up a few times.  Without some basic idea of how the metadata
will look like, and a resolution to the tag ordering question (are tags
a full ordering, a partial ordering or an unordered list?), it's hard
to define this problem well.

I'm not sure of the issues with actual streaming.  (For example, if a
radio station is streaming, and sends a new title when a song begins to
play, they'd probably want a way to tell players that the title no longer
applies when the song ends.)

Unsynched lyrics are obvious, I hope.  Synched lyrics has been a wishlist
item for a lot of people for a long time, too.  (I also don't know how this
is intended to work: how something in the metadata stream can refer to a
point in time in the audio stream.)

(Just to preempt any "id3v2 had lyrics, and nobody used them!" responses:
id3v2 was next to unusable.  The library was unstable, so everyone had
to implement the parsers themselves--which only a masochist would do--and
this led to incomplete support.  I'd have written a synched lyric editor
myself if id3lib had been stable.)

> Thinking about the goals put forward for the tagging standard, adding an
> ENCSETTINGS tag would clearly fulfill goal 1: Let the user know WHAT he
> is listening to.

I read #1 as "let the listener identify what he's listening to".  The
encoder settings used are just a property of the recording, just like
translations, lyrics, and so on.


-- 
Glenn Maynard

--- >8 ----
List archives:  http://www.xiph.org/archives/
Ogg project homepage: http://www.xiph.org/ogg/
To unsubscribe from this list, send a message to 'vorbis-request at xiph.org'
containing only the word 'unsubscribe' in the body.  No subject is needed.
Unsubscribe messages sent to the list will be ignored/filtered.