Anatomy of a dropped project

Published
2014-02-11
Tagged

Sometimes knowing when to stop is as important as trying to keep going.

Stage 1: Identifying the Problem

Genres in iTunes don’t serve much of a purpose for me.

It’s not that you’re useless: it’s just that you don’t do anything of any use.

There’s three interlocking problems with genres, which are that:

  • They’re auto-defined, so a bunch of music drops into my iTunes library with pre-defined tags on it.
  • They’re kind of arbitrary: a track could be “rock”, or it could be “indie rock”, or you could call it “indie progressive rock”. The level of detail required differs from person to person, and even then, one person’s indie rock may be another’s acoustic rock or folk rock or some further division.
  • They’re purposeless: what information do I gain from knowing a track is “rock”? When I want to listen to music I don’t go “I want to listen to some rock”, I go “I want to listen to something aggressive/soothing/upbeat/whatever.”

The problem manifests itself in a couple of ways:

  1. Sometimes I’ll sit in front of iTunes trying to work out what I want to listen to, or I’ll start start paging through album covers looking for the album I want. This is the twenty-first century: there should be a better way of doing this!
  2. A number of albums never really get played, just because I’ve never thought “I’m in the mood to listen to this sort of music.”

Stage 2: Creating a solution

Most songs’ ID3 information has a couple of empty fields: I’m looking at Grouping, Composer and Comments in iTunes. Rather than dividing music into arbitrary idealogical categories1, I plan on putting tags in the Grouping field of the file. These tags allow me to define the mood of the music, including whether there are lyrics, whether the beat is regular, experimental, or absent, and whether the music is upbeat, soothing, aggressive, happy, etc.

I can then make a number of smart playlists based on what moods of music I want to hear. My “writing” playlist, for example, could be defined by:

1
Beat: regular
2
Lyrics: none
3
Mood: soothing

My “first thing in the morning” playlist might be:

1
Beat: any
2
Lyrics: yes
3
Mood: happy/upbeat

These next two stages are new and important. It’s something I think I’ve only become good at over the past three years:

Stage 3: Determine Effort-to-Payoff

At this stage it is very easy to get sucked into the project, immediately starting to add tags to songs that you listen to. Instead, I tag about ten songs. So far, so good, but I run into a problem: some songs are “energetic”, some are “sad”, some are “angry”…it feels like I have a couple of different categories here: mood and energy.

OK, so I split mood into “mood” and “energy”…do I retag everything?

Next, I find that some music doesn’t have a regular beat, but it’s not what I classify as “experimental”. Is this a new category? What should I call it? Do I want to condense the phrase “regular beat” for my music?

At this point I realise I’m circling a deep, dark rat-hole of productivity sink:

Image courtesy XKCD.

Stage 4: Refocus

Why am I doing this? I could be coding, or writing cool stuff, and instead I’m re-tagging music in iTunes. Right now I have about 6 600 songs in iTunes. If I re-tag ten a day it’ll take me two years to tag everything.

In reality, I only use a couple of playlists: my early-morning wake-up playlist, my writing playlist, perhaps one of upbeat but ultimately forgettable music…that sort of thing. I don’t spend the majority of my time in iTunes making new smart playlists leveraging the two years of (hypothetical) work I just put into tagging my music: so why bother?

Surely a better approach is:

  1. Set up a folder of “mood” playlists. When I listen to some music and I think it suits a particular mood, drop it into my playlists.
  2. (Optional) Set up a smart playlist of music not already in a mood playlist, sorted by least recently played. When I feel the urge to categorise music, I can sort as required.

Quick setup, and I don’t feel the urge to meticulously tag my music based on a set of unknown future criteria.

Which frees me up to

Step 5: Do something worth doing


  1. After all, what separates rock from pop from dance music, really? The instruments? The time signature? The themes addressed? This could be it’s own whole post, but let’s just agree right now that the distinctions between these genres are somewhat arbitrary and definitely quite complex.