Over the years, I have done a fair amount of tinkering with smart playlists in iTunes in a never-ending quest to get that “ultimate” mix of great music out of my music library, so I was really excited last month when Apple added “Genius Playlists” to iTunes 8. While I had played around with the feature somewhat, I did not really give it a good workout until this past weekend. I took advantage of the eight hours I was going to be in the car driving up and back to a friend’s wedding, and loaded up my iPhone with 100% Genius Playlist-powered music to listen to over the entire trip.
What is a Genius Playlist?
Apple released iTunes 8 in September and Genius Playlists was one of the marquee additions to the update The concept is pretty neat Basically, you select one song out of your library, and then iTunes will pick songs from your music library that it deems “similar” to the selected song and automatically build a playlist of songs around it. Apple says that the algorithm to pick songs
is based not only on your individual profile and library, but also data about music preferences that it gets from the network of iTunes users. As a result, Apple claims that the accuracy / effectiveness of the Genius Playlist system will improve over time.
To get the Genius Playlists functionality working, you need to activate it in iTunes first by clicking “Turn on Genius” in the new Genius Sidebar. Once it is activated, iTunes indexes all of your music, and also gets your ok to send anonymous usage data about your music to the iTunes store. Apple has been very clear that all data sent to iTunes is anonymous, so your individual music preferences are not tied to you. After this indexing process happens, you are ready to go.
The process of creating a Genius Playlist is simple. You pick a song, click on the genius icon, decide how many songs to create in the playlist (25, 50, 75 or 100), and optionally click save if you want to permanently keep the playlist.
What Makes A Great Playlist?
If I had put together a wish list of the characteristics of the playlists that I wanted Genius Playlists to create for me, it would have been something like this:
- Mix in both “hidden gems” and well-known tracks. For example, if Genius was going to pick a Led Zeppelin song and always picked “Stairway to Heaven” and never anything else, I would have a serious problem with that.
- Select good music from my library that I have not listened to recently. I know I have great songs sitting in iTunes that I just have not heard in months if not longer. I want Genius to pick a handful of those into my playlist.
- Figure out songs that go great together that I would not have picked if I had manually created the playlist. I want iTunes to be smarter than I am about why certain music connects to other music in my library. That is the whole point, right!
When I was loading up my iPhone for the car trip, I decided that I was going to try it out with a number of different types of music. My ibrary probably is like yours in that there are several different genres or themes of music in it that would create one horrible playlist if I just hit shuffle on my whole library. One of my questions going into the test run was if Genius was smart enough to segregate the different genres.
How Did Genius Do?
I created five playlists of 100 songs each, and a couple of the playlists were just spot-on and hit on all of the above factors. To give you an example, I based one playlist off of The Band’s song “Atlantic City.” What types of songs did iTunes select? A smattering of Bruce Springsteen (he originally wrote “Atlantic City”), songs from Bob Dylan (The Band backed Dylan before they made it on their own), as well as artists within the same genre such as the Allman Brothers, the Rolling Stones, and Neil Young among others. One of the “hidden gems” it selected was “Incident on 57th Street” off of Bruce Springsteen’s The Wild, The Innocent & The E-Street Shuffle album — not a well-known song, but a great one. This playlist is actually now on a very regular rotation. Genius Playlists really nailed it.
In another example, I based a playlist on Pearl Jam’s version of “Love, Reign O’er Me”, a song which is originally by The Who. The Genius Playlist did exactly what I would have expected — it mixed both grunge artists with “Who-era” artists and songs. The playlist was a good mix of songs from Soundgarden, Led Zeppelin, Stone Temple Pilots, The Who, Smashing Pumpkins, and more. Another playlist was based on a Mighty Mighty Bosstones song, “Hope I Never Lose My Wallet,” and it did a great job picking ska bands and songs that just felt right alongside the Bosstones.
I did have one playlist that I thought was more wrong than right though. I based a playlist on “Same Old Song and Dance” by Aerosmith and somehow Genius Playlists made a connection to the “hair band” music in my library, and I ended up with way too many songs from the likes of Poison, Warrant, and the like. That playlist lasted about 3 songs before the skipped songs outweighed the ones I was actually listening to.
As a whole I would give Genius Playlists probably an eight out of ten. For the most part, it hit on what I was looking for, but most importantly to me, creating the playlists required basically zero work on my part which was HUGE! Moving forward it would be nice if Genius Playlists had a feature to give the song selection a thumbs up or thumbs down and then use the feedback to learn your musical tastes (Pandora has this feature). Also, at times it duplicated artists a little too much within the playlists, but I do believe Apple will be able to improve the algorithms once they analyze all that data that is getting sent to them. For me, the killer feature of Genius Playlists is the ease of creating good playlists in a matter of seconds. I expect to be using Genius Playlists quite often moving forward. You should definitely give them a try!
