Music Library Cataloguing systems

We here at CKUT are working on a new library database. We have the skeleton of the database interface in place. However, it has raised a bunch of interesting and rather big questions about our cataloguing system. We use a basic code: format-genre code-first three letters of artist last name/band name. This leads to the issue namely that a lot of records have the same codes. So for example, the Beatles and the Beach Boys or every Johnson who ever made a Jazz record, which is a lot, all have the same code.
Was wondering what kind  of library cataloguing system is being used out there by other stations.

What is the purpose of this code? Is it for use by the system or by humans?

In our system, and most DBs, each DB record gets a unique ID number. Then there are a number of other metadata categorizations  for information and search purposes but not identification.

The DB System will need and create a unique ID for each record so you may as well use that. We  store physical copies using that number and an format code (since formats need to be stored separately) and this can be looked up to find specific album/records.

I agree with Magnus that it all depends on the current and
future purpose of your database.  I'd also suggest that one
goal should be to try and reduce duplicate work.

If part of the intent is to eventually digitize your library
for easy search and playback, then perhaps your best option
would be to use the database created by the music library of
a playout system.

As an example, Rivendell Radio Automation allows you to
import CD's (or any other audio) into its music library.
Upon import it creates an entry into its database that
includes a whole host of fields which are searchable.  They
include (but are not limited to)

Track title, Artist, year released, album, Composer, Record
Label, publisher, conductor, ISRC codes, scheduler/category
codes, user defined, and more.

If you ripped CD's into Rivendell and make sure the various
fields are filled out (for some CD's they auto populate if
the system can find the CD's info in the various online
databases) and use one of the many fields available as a "CD
library identifier"  code (think of this as the equivalent
of the Dewy Decimal System that we all learned to use in the
book library) you'd then have a searchable library which
would allow you to hear the audio directly (or put it on the
air), or check the CD identifier code to be able to find the
actual CD in your library.

The nice thing about this type of solution is you don't have
to build the database yourself, you'd just need to create
the system on how to put CD's on the shelf on your library
(or use an existing system such as the one your local public
library uses).

Yeah, your basic code is fundamentally flawed, I'm afraid, because it's symbol space is too small and filled with too many duplicates. Heck, from what I can see, it can't even handle two albums by the same artist, given that it isn't part of the code!

As other have said, I will echo: use a simple numerical code for your internal unique ID - preferably one which is assigned by your database itself, to avoid collisions even further. You can include other codes to help with searching or understanding as necessary. For example, we have a unique internal number, and a shelf ID which locates it in the library. These are separate to allow albums which haven't yet been placed on the shelf or which are moving on the shelf to be put in the database. I wrote this as a simple MySQL + PHP system 9 years ago, and that part has worked flawlessly.
Good advice from both Lorne and Magnus. We separate our physical library by genre so we use a letter and number system. I should also note that we have virtually no vinyl. We started the system before ripping and digital downloading became commonplace so now a physical CD number would have the format B1234D which indicates that it is Blues and we also have a digital copy. If we only have a digital copy it would be B0000D. Before cataloging digital only, you could just look for the highest numbers and find all the new stuff, now you actually have to sort by the add date or you will miss the digital only.
If we had it to do again I think we would just put a number on the CD and just sort by meta-tags. Oh, and log by the track instead of the albums, again using the meta-tags to sort into albums.

This discussion does have me thinking - there are a number
of library classification systems which include music.  Is
there a reason we keep trying to re-invent the wheel with
regards to putting ID's on CD's and stuff, instead of using
one of the existing standards?

As an example, I previously mentioned the Dewey Decimal
system which has been used for music, there is also the
Library of Congress system:

A quick google finds there are a few others out there.  This
of course would not eliminate benefits of having the
information in a searchable database, it would just
eliminate the need to invent your own system for identifying

If you get right down to it, each CD release has a unique # baked into the CD. But it is the exceptions which mess the system. In the case of CDs it is the homebrews which have the identifier of the CD manufacturer which mess the system (mostly).
Most of the other systems I know of are not "nimble" enough to adapt to vinyl, CDs, digital and whatever from so many sources. I would not say I am an authority on this topic, these are just my observations. I would be pleased to be proven wrong.

Good discussion.

>If you get right down to it, each CD release has a unique #
baked into the CD

This is the ISRC that you're talking about, and it isn't CD
based, it is track / recording based.  Each track has its
own unique ISRC, with the exception of course being the
independent releases which - independent artists can apply
for an ISRC for their recordings.  Some do and some don't.

Organizing a library by ISRC code would be quite difficult,
especially when CD's have multiple tracks.

>(mostly). Most of the other systems I know of are not
>"nimble" enough to adapt to vinyl, CDs, digital and
>whatever from so many sources.

When I go to my local public library their system has CD's,
DVD's, digital, and they used to have Cassette and Vinyl.
They even have music from local unsigned artists that all
get indexed using the same cataloging standards as
everything else available in the library.  It actually works
quite well, if I search their database to see if the library
has a particular selection I generally have no trouble
finding the content on the shelf.

My local library uses the library of congress system, but
there are a few systems out there in common use.

I realize that using a cataloging system that also is used
in a public library is likely overkill for a
campus/community radio station, but adopting an existing
system does have the benefit of not having to re-invent the

Leave a Reply

Your email address will not be published. Required fields are marked *