ID3 Tagging

When it was first released, the basic MP3 file format contained simply audio and information such as bitrate and sample rate which was needed to render the file. However, the nature of MP3 files allowed people to add information which would be ignored by normal MP3 players and which could be used to embed metadata (additional information about the file such as its name, the artist and the genre of the music). The way in which this information was stored became known as ID3 Tagging (ID3 is the string used to identify the tags within the file). ID3-enabled players could then display the information, if present, while playing the file.

Newer audio formats, such as Real Audio and Microsoft's WMA have followed this example and their files include support for metadata in the basic specification.

IMPORTANT NOTE (Updated): Recent developments in the way Microsoft's Media Player intentionally 'mangles'" and corrupts industry standard ID3 V2 tags has become an ongoing battle between Windows Media Player, Apple's Quicktime, and RealMedia's Real Player. Windows Media Player is the real culprit here. I'm sure many of you have encountered WMP's infamous, "Cannot Play Media" message, after reading and corrupting the ID3 Tags from other encoders or players. Here's the quick fix:

1. Obtain and install Apple's iTunes Program.

2. Add the folder of mp3's or the mp3 file Windows Media Player has corrupted.

3. Highlight the imported mp3 file(s) in the iTunes file window.

4. Right click on the highlighted file(s) and select "Convert ID3 Tags . . . "

5. Select ID3 Tag Version 2.4

6. Let ITunes update the ID3 tag(s) of the corrupted file(s)

7. Voila!! The mp3 is ready to be played in any Media Player again!!

How I Tag My Archives . . .

I've attemped to standardize on the following ID3 Tagging Standards throughout our Archives:

The Tagging Process--Explained.

Let me give you an example of the processes we undertake in tagging the shows in our collection and perhaps you can get some ideas for tagging your own collection--or send us your favorite tagging secrets and we'll post them here:

Every new file you receive is suspect. Treat every new recording that way. Take nothing for granted--not the name, not the series, not the dates. . . nothing. If you do identify a reliably tagged source, then by all means make a duplicate of the folder and recordings you keep from that source and use them as a reference until you complete your own tagging effort. Be a good neighbor of that source as well. If you identify misnamed tags or data from that source, let them know what you corrected and your provenances. That keeps the feed-back loop both useful and instructive.

LISTEN to the file. Yeah, we know, this may seem obvious, but you wouldn't believe the number of collectors who too often bypass this seemingly essential step.

LOG as much information about the recording as you HEAR, including sponsors, performers, technical personnel, a particular 'provenance' to date the episode in time (e.g., the mention of that day's date, or day of the week, or special event or occasion associated with the news that day). Even something as innocuous as an advertisement. Example: You're listening to an episode dated 1949, but you hear a commercial for "the brand new 1951 Ford". What's wrong with this picture? The file was dated wrong by whomever tagged it previously.

Do NOT refer to someone else's log until you've logged the episode yourself from beginning to end. Why? We often tend to accept another authority's log as accurate beyond question. All too often, others' logs are just as prone to error as your own might be. Comparing your log to someone else's is the very last step you want to take in this process.

Use every tool available to you, to nail down any episodes that don't seem obvious, or don't contain cues to provide the essential information you need to completely identify it. There are a world of resources available to you for this purpose. Among them, books on either collecting or logging Golden Age Radio recordings. I have several of these listed on Publications Page. Other excellent sources are the internet, Radio Historical Societies, and media museums--both brick and mortar and electronic.

Here's a Tip: One of our subscribers, Andrew Steinberg is an absolute stickler for tagging accuracy and takes great pains to catalog the errors he finds among mp3 recordings in circulation. He's compiled all of them on his website here: http://www.geocities.com/nightkey5

Having logged as much information as you can HEAR, you can begin to attempt to identify the show/episode at hand. The more episodes you have, the easier it will be to nail down day/date sequences in the abscence of any other obvious cues within the recording itself. Once you've exhausted all other resources available to you, you can begin to match them up with known logs from respected archivists you've grown to trust over time.

Once you've assembled all the information you need to completely identify the episode or series at hand you can begin to properly and methodically tag the episodes at hand. I'll give you an example of the steps I take, step-by-step and you can compare my methods to your own:

Step #1: We run every series through our version of Otter (individually, since Otter seems incapable of compiling any more episodes than it has in it's default database. There are over 200,000 in our own), to see whether Otter has all of the episodes at issue in it's database. This is by no means an essential step, as Otter is just as prone to error (and both buggy and crash prone) as any other log source, but it's a another 'sanity check' for an entire series--especially the longer running series with hundreds or thousands of episodes. This is your classic 'garbage in, garbage out' situation. If Otter's database entries bear no resemblance whatsoever to the episode information you've logged (and it often doesn't, since Otter's datasources are highly suspect), it's time to consider scrapping the log information in Otter and loading your own. Learn to trust your own hard work and due diligence.

Step #2: Think ahead. Creating extremely verbose file names will prevent you from smoothly burning your collection to your media storage method of choice, be it CD's, DVD's, DAT Tape, or Hard Drive. Until the specification is enhanced you won't be able to burn either CD's or DVD's with overly lengthy file names. Place the longer name in the ID3-V2 tag within each recording. Use a short, but sufficiently succinct title for the file name.

Step #3: Find yourself a Tagging Utility you have faith in. The simplest and most readily available--and reliable--that I'm aware of is Apple iTunes' own tagging features. But there are over a hundred special purpose tagging utilities available, each with a different collection of features that may or may not appeal to you or suit your methods. We use Tag and Rename for our tagging, so that's the utility we'll use to describe the remaining steps.

Step #4: Convert all ID3-V1 tags to ID3-V2 tags. Why? You'll have more room to enter more accurate and complete information using the ID3-V2 tag. Even if you don't need the additional room NOW, you made need it later. It's a no-brainer. Just convert all of them first.

Step #5: If you've thought ahead entering or correcting the file names, you've already got three of the key elements of any internal tag as part of the file name: the date, the episode number, and the title or episode name. Here's what we do to every series we run through Tag and Rename:

a. We copy the entire file name to the internal 'Artist' field.

b. We remove anything in the file name before the date.

c. We use Tag and Rename's 'Remove Numbers from file Name' process, to remove the date fields from all of the file names (remember, they're still intact within the Artist field at this point). This leaves just the Episode Number and Title in the file name.

c. We use Tag and Rename's 'Extract' feature to extract the episode number into the 'Track' field of the ID3-V2 Tag. Now we're down to just the Episode Title in the file name.

d. We copy the Episode Title to the 'Title' Field.

e. Using Tag and Rename's 'Rename' feature, We 'rename' the file back to what it was before we extracted the Episode numbers and titles to the Track and Title fields.

f. Selecting as many of the files as we can at a time, we begin entering the 'common' information, such as Artists, Genre, Album Name, and Year. We then enter those three critical elements into every tag--the Year, especially, in case the file name ever gets mangled by some other process.

g. We review all of the changes we made to each field, then save the entire series.

h. We run the entire series back through both our Otter--to update it's database--and our own database, to record the presence of, updates to, and entry of any new episodes into the Master Database.

i. We prepare and print a revised directory printout of the series for our own future reference--or to act as a 'needs' or 'have' list for further episode searches.

Step #6: We burn the completely revised series to either CD or DVD, depending on the size and number of the episodes.

Why We Tag This Way.

Our rationale for adopting these standards throughout the archives are as follows:

• We recognize that these archives are not just audio recordings. They can be very visceral, emotion-provoking, historical and sociological records of a time gone by. To that end, we make every attempt to capture and accurately preserve as much as is known about a given recording.
  
• Since the folder in which the file resides is almost always the 'show' folder for a given series, we feel that the presence of the Show Name in the name of the file is both redundant and unnecessary.
  
• I feel very strongly that this is part of our stewardship of these archives, helping to accurately archive, preserve, and restore as many of these recordings as we can, for sharing and enjoying with even more generations to follow.
  
• The practical differences between the ID3-V1 tags and ID3-V2 tags are that the ID3-V1 tags are much more limiting, especially in the area of Genre. ID3-V1 tags were engineered with a built-in limitation of 255 genres that are internally referenced via a binary index encoded into the header of the .mp3. Almost none of the built-in genres are practical for application to Golden Age Radio archives.
  
• Whenever practical we code both ID3 versions into our archives. We recognize that not all current media players have adopted the decoding standards of the more expansive and extendable ID3-V2 Tags.
  
• The date field formatting is the absolute minimum essential field we try most to apply these standards to, when an accurate date is known. Those of us that collect and/or archive our collections via a database require that the date field be two position year, two position month, and two position day for efficient cataloging with a database.
  

We hope you find this information helpful and we welcome any suggestions you may have regarding more accurate or efficient tagging methods and techniques.
The following references can give you even more detail and rationale behind the ID3 Tag standards for both versions: