A reader recently got in touch to ask about my Axioms of Classical Tagging article, giving me the strong impression that my article makes it all sound too hard to tag things correctly... and I think he may have been right on that score! So I thought I'd simplify that article here, reducing its long, discursive justification of things to a few simple principles. If you want to understand why things should be as I'm about to spell it out, read the full article. For a short-cut guide to tagging correctly, however... read on!
There are just three basic principles to grasp from the outset, namely:
- The basic 'unit' of organising a digital classical music library is: the composition, not the CD. It doesn't matter if a recording of one symphony is supplied on the same CD as another. They are two unique and distinct compositions, and need to be filed apart from each other accordingly. The physical CD is irrelevant when cataloguing your music.
- The extended composition name is a combination of ordinary composition name, plus the surname of the conductor (or other principal artist), plus the recording year. It doesn't have to be the conductor: in a flute concerto, the surname might be that of the flautist; in an opera, it might be that of the prima donna. But usually, the conductor's surname will act as the 'distinguishing recording artist', making this recording of a composition unique from any other.
- The physical model (i.e., the folders you store music files in) must follow the logical model (i.e., the stuff you put into your metadata tags), and vice versa. Specifically, the physical folder structure should be in the form: composer-full-name/genre/extended-composition-name. So you end up with something like Edward Elgar/Symphonic/Symphony No. 1 (Boult - 1968).
Once those principles are understood and accepted, the way you tag your music really follows as a mere direct consequence. Specifically:
- The extended composition name in the physical folder structure should be identical to the contents of the ALBUM tag. Thus, ALBUM="Symphony No. 1 (Boult - 1968)".
- The genre in the physical folder structure should be identical to the contents of the GENRE tag. Thus, GENRE="Symphonic".
- The composer full name in the physical folder structure should be identical to the contents of the ARTIST tag, Thus, ARTIST="Edward Elgar".
- The full name of the 'distinguishing recording artist' should go in the PERFORMER tag, but it's optional to do this. Thus, PERFORMER="Adrian Boult".
- The COMPOSER tag always matches the ARTIST tag, identically, though it's optional to have a COMPOSER tag at all. Thus COMPOSER="Edward Elgar".
- The TRACKNUMBER tag is a number starting at 1, incrementing in steps of 1 for each track that makes up a recording. A traditional symphony will therefore probably end up with TRACKNUMBER tags labelled 1, 2, 3 and 4. If there are more than 9 tracks to a recording, pad the digits with leading zeroes, so you get TRACKNUMBERS of 01, 02, 03...10, 11, 12 and so on.
- The TITLE tag contains (usually) the tempo indication for the matching movement (i.e., TITLE="Andante. Nobilemente e semplice - Allegro"). For non-orchestral works (such as operas, lieder, song cycles and so on, the identifying first line of text of the song or aria might be appropriate, such as TITLE="On a Poet's lips I slept").
- The COMMENT tag is used to store free form text describing the conductor, orchestra, choir and other soloists that might be performing the work, in a comma-separated list. The conductor always comes first, then the orchestra, then choirs, then vocal and instrumental soloists. Soloists have their function in brackets after their name (the function is either the name of the instrument they're playing, or perhaps the character they're playing, if it's an opera). Thus: 'Marissa Robles (harp)' or 'Peter Pears (Captain Vere)'.
There is no need for any more tags than this (though different programs might embed some extra ones of no significance for the present discussion), and all the album-wide tags have corresponding entries in the physical path-and-folder name where you store the actual FLAC files. Thus, COMPOSER and ARTIST tags match the first part of the physical folder name; GENRE matches the second part; ALBUM matches the third. Parts of the ALBUM also go on to determine what gets written into the PERFORMER and YEAR tags. That is, if the ALBUM ends "...(Boult - 1968)..." then 'Adrian Boult' will be the PERFORMER, and '1968' will be the YEAR.
Within a folder of music, too, the TRACKNUMER and TITLE tags will pretty much exactly match the full physical file name. That is, if I've tagged a file as being TRACKNUMBER='1', TITLE='Allegro con brio', then I will expect to see a physical file name of '01 - Allegro con brio.flac' in my file browser.
Putting it simply, if I see this in a tag viewer:
...then, I see ARTIST = Bedřich Smetana, ALBUM = The Brandeburgers in Bohemia (Noseda - 2008), GENRE = Orchestral, YEAR/DATE = 2008, and PERFORMER = Gianandrea Noseda. From that, I can expect to see a folder called /Bedřich Smetana/Orchestral/The Brandeburgers in Bohemia (Noseda - 2008). Within that folder, I expect to see a file called 02 - Act I Ballet.flac (since those are the Track Number and Title tags). Here's what I actually see in my file browser:
It is almost precisely as predicted, because my logical and physical data models are consistent with each other. The one mandates or implies the content of the other, in either direction: that way, I can browse my music collection from an operating system's file manager or a music player's library display in exactly equivalent ways.
And that's it, really: I'm not sure I can make it any simpler than that! But let me know if you can...