Re: [ARSCLIST] More on cataloging

[David Lewis <davlew@xxxxxxxxxxxx>]

Karl Miller wrote:

As I continue researching alternatives to MARC, I came across an article
from the 7 August 2005 NY Times. I quote a portion of it below.

While I realize that the 1.9 Million catalog entries for recordings in OCLC
are not limited to individual song titles, and that the points of access are
limited in the commercial databases, when I read this quote below, it did
give me food for thought...perhaps comparing apples and oranges (pun
intended)...

 "If it comes to that, they'll find
that a lot has changed in the online music business since Apple
opened its wildly successful buck-a-song iTunes Music
Store in 2003. In that time, Apple's catalog has grown
from 200,000 songs to nearly 1.5 million, Apple has
sold half a billion songs and it has been joined by
similar stores run by Microsoft, Yahoo, Sony, Real
Networks, MusicMatch, Dell and even Wal-Mart."

My guess is that people are able to find what they want, and it took
the for-profit sector less than 2 years to create a database of 1.5
million records, when it has taken OCLC 40 years to create a database of
1.9 million catalog records for sound recordings. I wonder, what am I
missing in this comparison...

Karl,

The data is mostly licensed from AMG, Muze and Gracenote - Apple, Microsoft,
Yahoo, Sony, Real Networks, MusicMatch, Dell and Wal-Mart combine a cocktail
of all three services, and a few use a small in-house staff to address
individual issues as well. It took AMG fifteen years to amass the data we
now have, which amounts to about 500,000 pop CDs, 128,000 classical ones and
more than 50,000 DVDs. I can't speak for Muze and Gracenote, except that in
the case of Muze they bought the Music Sales database which was started in
the 1970s, and Gracenote was originally, and to some extent continues to be,
a database maintained through entries created, or shanghaied, from the
public at large.

> besides the fact that I would guess the various labels are supplying their
own information, 

NOT! They should, to all reputable services, to insure their own survival at
the very least. But they don't, and you have no idea how hard it is to get
some of them even to consider it. A few have wised up by now, and it has
helped them, I believe, to gain a slight advantage over others who don't. 

> information which is created
digitally (40 years ago I would wager all record companies
were probably using typewriters), hence a great deal of information in the
early years did not exist in any digital form, hence the time required to
enter that information, which would require more time to get information
in the database...

The formats of older years present a real challenge for data maintenance.
You can stick a CD in a drive and find out all kinds of things about it with
nary a stick of information. Doesn't work that way for an LP, cassette, open
reel tape or cylinder.

[Karl]
> yet the information created by the
companies these days could possibly be shared by OCLC/RLIN or whatever, a
notion which several have suggested is not viable...that the information
in the commerical databases is not subject to authority control...etc.
Further, iTunes is not describing an object...

I can't speak for Gracenote or Muze, but the AMG Free website is used as an
authority by libraries and music stores, particularly for birth/death dates,
issue numbers and that kind of data. The reason that commercial databases
"(are) not subject to authority control" is that there is no dialogue
between the commercial databases and the libraries. OCLC is prohibited by
their own guidelines from opening up such an avenue. But their database was
designed to catalogue books, not recordings. On the other hand, from a
proprietary standpoint an open structure like OCLC presents an immense
problem for a commercial data enterprise, which doesn't want it's product
all over the web without certain tagging and protections.

> However, I would assume people are able to find what they want.

If it's Britney Spears, yes, but if it's an analog recording of Lynn Harrell
playing a cello concerto only issued on LP, then probably not. If it's a
tune that Frank Sinatra recorded with Tommy Dorsey in 1940, then they might
have the problem of too many choices.


David N. Lewis