A few years back, I wrote a post about how I use the star rating system in iTunes to help manage how I listen to music. I recently acquired a new MacBook, and because I made a few mistakes in the process of transferring music from my external drive to the MacBook, I lost some metadata, and that metadata included the star ratings I had accumulated over the last 3 or 4 years.

Some Context

I thought this was going to be a bit of a problem, but it turns out the way I use music has changed a bit and the star system isn't really as central to the way I do things anymore. I tend to think of ambient music these days as needing to be contextually relevant. And by contextually relevant I mean appropriate for the time of day, place, mood, weather, time of year, lighting, temperature, and whatever other circumstances all combine to make up the conditions present when I'm listening to music.

iTunes doesn't have a particularly good way to handle this. I use a lot of Smart playlists to approximate some of the functionality I'm after, but there isn't quite enough options to slice things the way I'd like. What I'm looking for is a way to mix quantitative behavior patters with things like my qualitative feelings on things. Here's an example:

I get up early in the morning, and around this time of year it's dark and the apartment is cold. It's also quiet because the hum of the air conditioner in the background isn't present. I'd like music appropriate for this context.

Using Notes

One way I do this is by using the notes field. I essentially use the space to "tag" tracks. A lot of the tags I use are descriptive in terms of what type of ambient music I'm listening to: drone, instrumental, etc. and then I'll add a few more words that describe the sound like "clicks" or "spacey" or "underwater."

And so I've also used the field to add in contextual information by using keywords like "morning" or "quiet" or "winter" or "autumn" and etc. And this works mostly well, except that the metadata isn't structured, and therefore hard to remember and be consistent about. There's a tradeoff between not being forced to use the exact right term and losing out on tracks I've labelled as "soft" when I make a playlist of "quiet" tracks.

Smart-er Playlists

So I'd like there to be a better system for structured metadata. Basically a "Genre" field that's totally suped up.

But what I'd really like to do is then take these smart playlists based on a combination of keywords and then have the system be smart enough to only play tracks noted as "quiet" and "morning" and "winter" that I've actually in the past listened to between 5am and 7am during the months of December, January, and February.

So what?

What I'm attempting to do with Ambient Music Blog Podcast is to do this work for you. I think of a context, and then I pick some tracks that to me feel like they fit. My longterm goal is to explore the ways in which the kind of music I talk about on this web site might be useful for improving experiences in the physical world. Like, for example, would it be easier to buy a car if the salesman's office was playing a really well-done and contextually appropriate playlist?

In the meantime, I'm hacking away at iTunes and trying to make it do something closer to what I'd like it to do.