In the summer of 2004, I devoted an unquantifiable amount of time I'll never get back to rating each and every song in my iTunes library. It was often haphazard and included an expected amount of sleep-deprived misjudgment.
The following February, I created a plot of the amount of songs with each rating (1 to 5) joined by a faux-curve. To my surprise, it wasn't too far from a normalized bell. Just recently, it occurred to me to re-sample that data, and here it is:
I was curious about the possible shift to the right, so I decided to plot a normalized version, too:
It led me to draw a few conclusions. First, I've slowly adjusted my original ratings of some of the original batch, often "upping" their value. Second, I've either been less "harsh" as I rate new additions, or I've been more picky about what I choose to add.
What I find even more interesting is my Artist listening history for the last three years, thanks to last.fm and lastgraph. Through tremendously complex stacked-line graphs, I can clearly see blips of which bands I was into when, and make my guesses of why: album release, friend's recommendation, or deeper emotional reason. File this one under geek.
June 24, 2008
i + Tunes
Subscribe to:
Post Comments (Atom)
2 comments:
Great post! Have you had any worries that if you lose your iTunes library you will lose all this work. I ran into a case where I reformated my HD and it caused me to go through the painstaking process of figuring out a way to make sure ratings and play counts carried over.
well, the ratings themselves aren't as volatile as the play counts. I was able to transfer the ratings data in May of 2006 to my new laptop without issue using the xml version of the iTunes library auto-generated for consistency and running it though a find-replace script. The binary file keeps the play counts, however.
in the future, as long as I don't switch to a PC (as the binary library formats differ), I should be always able to rely on my time machine backup.
Post a Comment