Rebuilding iTunes

I have had a lifelong love/hate relationship with iTunes. I would like to tell you that it dates to when I was in the womb, but it’s not that tragic of a tale.

My music tastes aren’t simple. I have no one “thing” that I am into, really. My brother, in his brilliance, exposed me to the classics of the sixties and the seventies, and instilled in me a great love for all things classic rock, from the Who to Led Zep.

On my own, I found punk and fell hard for it; I still worship at the altar of the only band that matters, the Clash, and could get lost in the sounds of my high school and college days, with bands as diverse as the Squeeze and Jam and Cure, and as far flung as KLF and Bow Wow Wow and the one hit wonders of Modern English.

And over a dozen years ago, thanks to a small sidebar article about freelance Hellraiser and a little ditty called A Stroke of Genie-us, I found and embraced mash-ups, which can either be considered a genre of their own or prove that today, music defies genres.

Which is why, when we all switched from mix tapes to an amassed collection of MP3s, I found solace iTunes. Smart playlists made all the differences. With a little work and some Boolean logic, I could not only teach my iPod how to better play my music, I could teach it how to serve up music based not just on genre, but at intervals based on ratings. And, up until iTunes version 4.6, iTunes was actually smart enough to dynamically do this on the fly, on the device.

Best songs, rated five stars, would be played at least three days apart. Great songs, rated four songs, would be played no more than once a week. Middle of the road songs, no more than once every two weeks. The rest, rated at two stars, would be played no more than once every three weeks. And I could always, as needed, downgrade a song to 1-star, and it would drop out of the cue altogether, because I used 1-star as a holding pattern for deletion. Together, these formed what I called my base radio station, often just called @Art so that it showed up at the top of my options.

But I could refine that even more. I’ve long had a playlist called 91X that was @art and then any of a bunch of bands from my college days – U2 OR REM OR INXS OR Social Distortion OR Dead Milkmen and on and on. Applying the same Boolean logic kept the same play interval rules, and then sub-filtered to just the bands.

And over the years, as our inventories have grown, I have added specialty playlists for differing devices, but keeping them offset from each other. I have an older ipod that I use to use for long runs and long rides – that I loaded with long mashups. Stuff that was in the cue there wouldn’t show up in, say, the playlists on my phone. In the end, I could have songs set to lay only on one device at a time, but once play, could show up on another device.

There are a couple of things that really make all of this work. The first is having great Boolean rules. I write and rewrite and change and tinker with them, as needed. But that base radio station concept, described above, has been the anchor point for years and years. It’s more complicated – now I filter out classical and holiday music, for example, in the event some sneaks into my library by accident – but the basic idea hasn’t changed.

Everything is tagged. Genres as assigned, not by accident but by me. Ratings are assigned, but by accident but by me. And I lowball things by default, too. But using Boolean logic means having data to search; if I watch to make a playlist for music released in the 80’s, it gets easier if that stuff is in there (MediaMonkey is a big part of that, by the way).

Which is why I am where I am right now – a late night on a weekend, taking a break from rating songs, to blog a rating songs. I just added back in 85 volumes from Mash-Up Your Bootz Party, which works out to about 1600 songs when you throw in the extras, too. The bulk tags were already there (hard written to the files, via MediaMonkey), but there’s still the task of rating them. I could just bulk rate them something like 2-stars and be done with it, but there are pure-gold nuggets in there that need to surface. So, due diligence to find and bring them out – so the Boolean works and the integrated smartlists can do their things.

(Here‘s a much, much older piece I wrote about smart playlists, complete with pictures. Have questions? Just ask.)

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.