TAGS IN DIGITAL MEDIA FORMATS

 

---------------------------------------------------------

CONTENTS:

Metadata Tags in Digital Media Formats

        What's a Metadata tag?

        Why metadata tags are crucial

Types of Metadata tags

Tagging Tools

Next Time...

---------------------------------------------------------

 

 

Recently, a user on the Audiotron (a digital media device) "help" forum posted the following query:

 

        I have a question I am hoping someone can help me out with. 

        All of my Grateful Dead songs in iTunes have the Artist as

        "Grateful Dead".  But, in the AudioTron, the songs show up

        under "Grateful Dead" and "Gratefull Dead".  A search in

        iTunes for "Gratefull" shows nothing.  Does anyone have

        any idea what may be going on? 

 

This question comes up at least twice a week on that alias alone.  Users edit the artists' names over and over, but the changes don't show up in their devices.  They blame their device or media player, but in reality the problem resides in their mp3 metadata tags. I thought a general primer on the cause of this and related problems would be useful to reviewers on this list.  This week I'll discuss what tags are and how they work, and next week I'll discuss some tagging hassles, their causes, and solutions.

 

WHAT'S A METADATA TAG?

 

Metadata Tags (or more simply, “Tags”) are packets of information that are included in an mp3 or wma file along with the audio.  Metadata (“data about data”) includes artist name, song title, album title, genre, etc… about the music in the file.  Different file formats use different tagging schemes, but the basic procedure when a CD is ripped is always the same: the ripping tool creates a file “container” and puts metadata into the tags.  Then it transcodes the music into either mp3 or wma format and adds the music to the file.  

 

Why tags are crucial

Tags make it possible to organize and use media libraries. When a user scans their hard drive for new music files using a media player or device, the player finds the files, reads the tags, and constructs a media library index.  After that, instead of a flat list of thousands of files, users can then easily play music from a particular genre, artist, or album. The same is true in a device - the Audiotron, for example, reaches out across the network to find music files, scans them, and builds a local index file.  My Nomad Zen does the same on its local hard drive as media is moved onto it.  Tags make all of this possible.

 

So even though most users have never heard of them, tags are very important - so important that most enthusiast users spend many hours "cleaning up" the tags in their media libraries. But tags are only as good as the data they contain, and are also only useful if everyone creates and reads them in the same way - and often neither of these conditions is met.  But before I talk about that, let's talk about the major tags standards in more detail.

 

TYPES OF TAGS

 

MP3 Tags - ID3v1 and ID3v2

The inventors of early digital media formats didn't imagine users would have enough music to need tags.  MP2 and later MP3 files initially didn't include tags, but as users' media collections grew, an informal - and very limited - tagging standard called ID3v1 (ID3 version 1) emerged.  Because it was limited in size (ID3v1 used fixed-length "padding space" at the end of an mp3 file) it included only very basic information, and limited the fields to 30 characters.  Initially, ID3v1 didn't even include track numbers, so albums would be played in alphabetical or random order.  An illustration of an id3v1 tag is shown below in Figure 1. 

 

Figure 1 – an ID3v1 footer

 

Subsequently, some folks defined ID3v2, which allowed longer names, custom genres, and fixed some other nagging problems.  An ID3v2 tag is illustrated below in Figure 2. 

 

Figure 2 - An id3v2 Tag Header

 

The most recent ID3 version is ID3v2.3 (for more information, see www.id3.org), and further versions have been proposed.

 

A typical mp3 file includes *both* id3v1 and id3v2 tags.  Thus, the file starts with an ID3v2, then the music data, and then ends with the id3v1. See http://www.caulton.org/mp3_tags.jpg for the structure of a file shown in one of Microsoft’s internal testing tools - first the ID3v2 tag, then the music data, and finally the ID3v1 tag.

 

Figure 3 - An mp3 file – first id3v2, then the song, then id3v1 tag

 

Almost all digital media players and devices that support MP3 audio also support some or all parts of the ID3 tagging scheme.

 

WMA tags

WMA files use the Advanced Systems Format (ASF) file container, which includes a rich tagging scheme.  Literally hundreds of types of metadata and bitmaps are specified within the format.  See Figure 4 for an illustration of a WMA file in the MP9s Tag editor. Most WMA-enabled media players use code from the Microsoft Format SDK to read and write WMA tags, which simplifies development and reduces versioning problems (more on that in my next mail) that plague ID3 tags.  With the new ASF license and spec, it's easy for ISVs to create WMA reading and writing tools appear on any platform. For more details, see the ASF spec at http://www.microsoft.com/windows/windowsmedia/format/asfspec.aspx). All devices that support WMA also support the ASF/WMA tag standard.

 

Figure 4 – WMA tag in the MP9s Tag editor

 

 

TAGGING TOOLS

 

If tags were simple and the standards were carefully followed, tagging tools would be unnecessary.  Unfortunately, as we'll see in my next mail, things are often not that simple.  As a result, there are tag editing tools for every purpose.  I'll go into their use for solving various problems later, but for now wanted to at least mention them.

 

The Windows XP Format SDK

Many people don't realize that Windows XP includes the Windows Media Format SDK, which provides ID3 and WMA tag reading and writing functions to the Windows Shell and ISVs.  Thus, most media players and ID3 editors on Windows need not reinvent the wheel by researching and interpreting the ID3 and WMA specs and writing their own code.  As we'll see, Windows exposes this functionality to users in a number of useful places.

 

The Windows XP file Properties dialog

The Windows XP file properties dialog utilizes the Format SDK to read and write tags.  Users can simply select a file, click properties, and view/edit the tags in the file (Figure 5).  This also enables bulk-editing for multiple selections.  I use this all the time for quick-and-dirty changes to tags without booting up an editor or player while I'm navigating through my media library.

 

Figure 5 – id3 tag editing in the Windows XP Shell

 

 

Windows XP Shell "details" view

The shell uses the Format SDK to allow users to view and sort based on nearly every detail about their files right in the shell in the details view.  This view can be customized from a context menu from the column heading (Figure 6). Even more detail is accessible from the "more..." item on that menu (Figure 7).  I find this extremely useful for sorting and filtering through files based on format, bitrate, album, etc... right inside of the shell.

 

Figure 6 – Windows XP Shell “Details” View

 

 

Figure 7 – “Details” View Customize Dialog

 

 

Digital Media Players

Most digital media players include at least rudimentary tagging tools (many also built on the Windows Format SDK), and recently some have become quite powerful.  Windows Media Player 9 Series and MusicMatch both have nice ones that are useful for both bulk and individual file editing.  In MP9S, you access the through a context menu in the media library (Figure 8), and can do bulk or individual editing of over a hundred tag fields (Figure 9)

 

 

Figure 8 – Context Menu Access to the Tag Editor

 

 

Figure 9 – Bulk Tag editing in MP9s

 

 

 

Power User Editors

Sadly, these are often needed by most users for resolving complex tagging problems.  My personal favorite is Tag&Rename, which can edit WMA, ID3v1, and ID3v2 tags. There are many others.  Beyond the functions offered by media players, it also has very powerful tools for:

 

    * Creating folder hierarchies based on tags

    * automatically applying serial track numbers to files

    * Case Correction of tag content (Title, lower, or upper)

    * many other essential but geeky features.

 

I love these tools and use them often, but wish they were unnecessary.

 

Single-Purpose Editors

Just for fun, I can't resist mentioning one other class of tag tools, built by hobbyists.  One example is Otter (http://otterbarn.tripod.com/), a utility for organizing and tagging files from Old Time Radio shows. The program actually includes a database of the dates and show titles of every episode of hundreds of different radio shows from the 30's to the 60's.  Point it at a directory full of OTR files, and it parses the filenames to determine the title or date and classify the show.  Then it can fill in missing data, rename the files, and advise collectors of what shows they're missing that are known to be available.  This sort of tool is a tribute to the doggedness of some user communities and the importance of a strong specialized developer community.

 

NEXT TIME...

I'll discuss the many hassles and problems caused by missing tag data, inconsistent tag implementations, and the "informal" nature of the ID3 specification.  Then I'll show how to use the tools above to resolve many of the most common ones.