Hi Steve:

Thanks for the link. I missed this article.

This ties in to my general complaint about the sorry state of the "auto-metadata" databases like
Gracenote/cddb (used by iTunes) and freedb. Aside from having inconsistent naming of composers,
conductors, soloists and vocalists, these databases are inconsistent about basic things such as
punctuation and the use of modern or Roman numerals in titling symphonic movements, and are
inconsistent in how titles in non-English languages are treated. The whole thing is a hellish
mish-mash.

iTunes is basically a "dumb robot" that just loads in rips from CD's or the equivilent, using
Gracenote to fill in the data fields. The reason all these problems arose is that Gracenote started
out as one of those low-quality "online group effort" things and then turned into a for-profit
company, and was all the more profitable by not hiring editors and metadata specialists to bring all
of this disparate information into standardized rules and structures. I can say from the experience
of loading all the Mercury Living Presence CD's into an iTunes library that I had to do some sort of
tag correction (ie metadata correction) on every single CD to make things like artist names,
composer names and the way sections of classical works were titled into a consistent and
easily-searchable/easily-sortable mass.

As to the issue addressed in the WSJ article, the "dumb robot" just breaks classical works into
"tracks" from the original CD. So you wouldn't pay $2.50 for a full performance of a symphony, you'd
have to find all the movements (hopefully they are named consistently -- but no sure bet -- so they
can be found together on a list generated by an iTunes store search) and buy them separately, then
put them into a playlist in order to play the symphony front-to-back in order. This whole system is
set up for buying single pop tunes. By the way, this system also hurts sales of full pop/rock
albums, so the record companies should be fighting this on all fronts.

When classical CD's first came out, I thought that movements should have been handled as
sub-chapters, which the higher-end players could separate if someone really wanted to skip the
second movement of a symphony, for example (and of course the record companies should have spent the
extra few dollars during the Red Book authoring sessions and put in CD Text based on some
agreed-upon naming/spelling/numbering structures -- a Red Book Stylebook -- and then these sloppy
auto-fill metadata problems wouldn't exist). Each work should have its own "main cut" or chapter on
a disc. So a typical CD might include two Rachmaninov piano concertos and then maybe a few short
pieces to fill out the 74-80 minutes (almost all classical CD's did not cheat the buyer as far as
fill out the time limits of the new medium, whether one thinks that's good or bad is a matter of
opinion, suffice to say the first thing I do in iTunes is make playlsits of the original LP sequence
or of individual works that I would want to sit through). The way that typical CD should have been
partitioned is that each work was a chapter and the movements would be subchapters. So the first
concert would be "track 1" and each movement would be 1.1, 1.2, etc. This is enabled in the Red Book
definitions, but seldom used, and thus most players were never capable of skipping to and among the
sub-chapters. But I'd bet most users just don't go skipping around with movements within a work.
Because the industry unimaginatively divided classical works into "cuts" just like on an LP, they
ended up with this handicap of each movement being an individual "track" in iTunes. I would bet,
given the minimal sales of most classical recording on iTunes, it would be impossible to go back and
make the "dumb robot" smarter at this point.

-- Tom Fine

----- Original Message -----
From: "Steve Ramm" <[log in to unmask]>
To: <[log in to unmask]>
Sent: Thursday, January 05, 2012 8:45 PM
Subject: [ARSCLIST] Interesting Article on iTunes and Classical music in yesterday'zs WSJ


> _http://tinyurl.com/6udz8kp_ (http://tinyurl.com/6udz8kp)
>
>
> Steve Ramm
>