Geekfoolery

There are a lot of songs I like, but I just can’t remember what the hell it is, or who it is. This is partly because my music library used to consist of a ton of files that had descriptive file names like “lovinu.mp3″ or “rainsong.mp3″. Which is very helpful, right? I mean, all I to do is just have to narrow it down to any song that contains the word “love” or “rain” and I am there. And I had hundreds of these. It would have taken days to figure out what song is which, and then manually relabel the songs. Here are a couple tools that I have used to clean up my music library.

To deal with the first problem-song files that have incorrect or incomplete file names and metadata-there is a program out there will “listen” to your songs and fix the labels. This way when you look in iTunes (or whatever), you can see who and what it is. It’s really kind of mindblowing that this actually works, to be honest. The website that supports this tool is Musicbrainz.org, and they have a database of song information called AcousticFingerprints that they use to ID your music and clean the files. To use the database, you download one of the client programs from the site and run it on your machine against your library. I used it with iTunes on my Mac and it took care of about 90% of my mystery songs, and fixed a number of others that just had bad or incomplete tags.

Then I found is that I also had a lot of duplicate song files. No need to have two copies of the same thing. Apple’s iTunes has a “find duplicate” feature built in, which puts all your duplicate song files into one playlist. You can then decide if you want to delete them. Be careful, though… the feature is good, but not perfect. It may flag different versions of the same song as a duplicate if the filenames are similar.

If you don’t use iTunes, or are otherwise just trying to be difficult, there is a program called Noclone that also does tracks down duplicate files. It’s a $15 shareware fee, with a free trial. Two major differences from the iTunes function is that Noclone compare the files byte-by-byte, not just by filename. Second, it’s not limited to MP3s, it can find duplicates of any file. Nice if you’ve moved computers a couple times. It’s real easy to end up with multiple backup folders of stuff that’s just taking up space.

The next thing that you should clean up on your music library is the album art. It’s kind of superfluous, but with more music players displaying album art, it’s nice to have. iTunes 7 has the function built in and as long as the album art is in the iTunes Music Store, it should find it. If you don’t use iTunes. Media Monkey works pretty well, as does Tunesleeve. Generally I have found that all of these automated tools work well, but there are always going to be a handful of straggler files that you can’t find the art. You can go online and track down the album art yourself-Amazon is a good source (don’t forget to check the international Amazon sites if you have an import CD), as is Allmusic.com. Other options for the eclectic and hard-to-find: If it’s an independent artist, they may have their own website with images you can grab. If you have the actual CD in hand, you can scan the art yourself. Or, and purists will probably howl, I have been known just to go grab a random pic I thought was neat and make THAT the album art. My god, I am such a heretic.

I must bring up one final musical note (rim shot)… have you ever had that song in your head that you just can’t figure out what it is? All hail the Internet, this vexing scourge of the afflicted will soon be no more. Midomi.com, though still in beta, promises to figure out what song you’re thinking about just by having you hum or sing it to your computer. I say again: You hum the song, Midomi figures out what it is. It’s still in beta, but their database is designed to grow and become more accurate the more people use it. I tried it and get 1 out of 3 right (Strangers in the Night was the winner), which shows that either the concept is valid, or it’s just a commentary on how tone deaf I am.

Now all we need is the database that can figure what joke you’re trying tell based on how badly you’ve mangled the setup, the timing and the punchline.

Permalink | Trackback | del.icio.us Digg Reddit

Comments:

Comments RSS

Leave a comment