Not to minimize the importance of what you state, there is, perhaps,
another side to this discussion.
Continuing your example, Motown handed off LP album XYZ123 for
manufacturing. At least preserving the best copy of that master that
changed hands in the sense of the release for manufacturing is a good
thing. Much better than finding a scratched up LP at the Salvation Army!
I think at some point, when we're preserving commercial releases, we need
to be content with preserving the final vision that went to manufacturing.
The reason is that is what the entire artistic and production team agreed
to as the "final product."
If we want to trace back pieces, that is wonderful, but sometimes that
metadata is hazy. I don't think that diminishes the value of the final
product, although having this metadata makes the final product more
interesting, at least to some people.
Just my two cents.
For another two cents, one can start going crazy by finding variants on a
selection. For example, the single release of a song may be a different mix
from the album release. But, these are two separate "releases" -- changing
of hands from production to manufacturing.
At 11:17 PM 7/1/2003 -0700, Paul T. Jackson wrote:
>The problem with long-term anything is that back in the 60s people like
>Motown were making virtual performances via the mail. That is to say, they
>would lay down a track by a drummer somewhere in New York, put it with a
>track from somebody in Hollywood, and master those with back up music by
>Detroit Symphony musicians in Detroit. Those takes were put forth as a
>single performance. I don't know what their index looked like, or if they
>even had one. They did record some things on 35mm film as well as VHS
>tapes, because of all the tracks they were putting together. But if the
>index of who is who on the single track takes doesn't exist, what kind of
>preservation will be available to document what went on? Who was where,
>when, and how many of those single track takes were put with how many other
>Paul T. Jackson - Trescott Research
>Information Resources and Library Development
>[log in to unmask] <mailto:[log in to unmask]>