From Visual Developer Magazine #53, January/February 1999


Track Ripping



A strong open standard for music binaries will allow a Net-based music market to develop, and allow niche music to make money.

The CEO of a small software company did something interesting a year or so ago. He bought a pair of 9GB hard drives (back when they were still expensive and a little exotic) and then stayed up all night for a week transferring tracks from his music CDs to hard disk—a process called track ripping. What he has now is the ultimate jukebox, with a couple hundred hours of music randomly selectable from disk and playable through a clever little $10 program called WinAmp. (

WinAmp plays MPEG Audio Layer files, or MP3 files as they're usually called. They store sound at a density of about a megabyte per minute, depending on the sampling rate. A cheapo 4.3 GB hard drive can thus store over sixty hours of music—and you'll never have to juggle a jewel case. This is a very good thing. A strong open standard for music binaries will allow a Net-based music market to develop, and allow niche music to make money. This is important for guys like me, who listen to things like obscure Shaker hymns, Percy Grainger's march music, and one-hit wonders from the sixties like the Peppermint Trolley Company. You don't find that stuff in record stores. Pearl Jam is what moves volume, and rack space is at a premium.

I have a lot invested in vinyl (I hate to think how much) but I would buy it again as clean, crackle-free AAD tracks—assuming I could buy it by the track. Alas, this is the record labels' worst nightmare, and I was stunned at the slathering rage the labels are directing at the entire MP3 concept. Not only won't the labels sell music this way, they won't sign any band who has ever released a single to the Net as an MP3.

Why? The current pop CD business model assumes that CDs will sell for $14.99 and no less. With rare exceptions, pop CDs contain one or at most two good songs. The rest is trash, composed by bands who aren't composers just to fill the disc out so people will pay fifteen bucks for one "hit." This is one reason music sales have slowed recently. Boomers have replaced as much of their vinyl as they can, and now most people wait for a "greatest hits" CD before buying current bands. In fairness to the labels, it's tough to make money selling a physical CD for less than $15. If some labels started selling quality tracks for a buck apiece on the Net, store traffic would tank, and a lot of other labels would go under. In fairness to consumers, current CDs represent a different kind of "track ripping" in that all but one or two tracks on a disc are ripoffs.

Right now it's a brittle standoff. Sooner or later, somebody is going to break ranks and try it, and the entire music industry will change. The vanguard will be cult music that isn't getting shelf space now. Eventually, larger bands who don't need to take orders from the labels will begin selling singles on the Web, and then it's all over. The old triumvirate of record label, radio station, and music store will fold in on itself. The bands will create coop labels with other bands to spread admin and promo costs around, and the online stores will likely be owned either by the coops or by radio stations, which will naturally favor music that they can sell. The biggest bands will make less money, but the smallest will make a lot more, and won't have to embarrass themselves by padding CDs with hasty low-quality tracks.

There are opportunities here for programmers, who like music, can create utilities to manipulate music, and even make music, as composers, performers, or both. (I've always been surprised at how many developers have their own bands—and Coriolis CEO Keith Weiskamp is a concert-quality pianist with degrees in both computer science and music.) We're on the edge of an amazing rearrangement of an entire industry. I can't wait for the fireworks to start.