Posted on: May 3, 2006 in Technology, Music
The future of music: Big hooks up front
Technology and music have always had an interesting duality. Throughout history, it seems that most popular music is "designed" (perhaps unwittengly) to exploit the technology used to deliver and consume itself. What do I mean exactly?
My favorite example is that of Louie Armstrong’s horn. The early phonograph had a very narrow acoustic range, and had difficulty picking up certain frequencies. Louie’s trumpet, with it’s warbling pitch, and high treble, was able to pierce the audio haze of early recording devices. It wasn’t just that the trumpet was more popular an instrument then the flute for example; it was that the Trumpet was the perfect vehicle to exploit the medium it was to be delivered on. Other examples include Mark Katz’s hypothesis that violin vibrato was introduced to help early recording devices better pick up the sound, in a similar fashion to Louie’s trumpet.
Moving ahead in time, consider Pink Floyds album, “The Wall”. The Wall was designed to capitalize on the two, long-playing and uninterrupted sides of a piece of vinyl. Other attributes of the media are exploited as well, such as the high fidelity nature of most stereo system at the time, and even the dimensions of the album cover in order to display art work. Could “The Wall” have been as successful in today’s world of downloaded singles and iPod shuffle’s? Perhaps, but unlikely…
When I was at SXSW a few months ago, I found myself in an interesting discussion with Matt May (and others), that I kept on thinking about. We were talking about how current technologies were going to affect the type of music being produced.
Our best guess?
Most online stores or music websites offer you a 30 sec (or so) listen of the track before you purchase, so we figured that future music would be designed to exploit this. Who cares about song structure or melodic development! All that will matter in the future is to have the first 30 seconds of your tune “optimized” for a public bent on quick scanning. All of a sudden, leading off your track with that huge punchy chorus seems like a good idea. The future of music is “big hooks up front”.
May 5th, 2006 at 5:34 am
Interesting hypothesis, but I think that most companies that offer sound samples (LoudEye, AllMusic.com, MusicNet) don’t grab the first 30 seconds of the song. Many companies go 30 seconds into the song and then grab their sample from the 31-second mark to the 60-second mark. I know AllMusic has an algorithm that analyzes the entire song and grabs the 30 “Most Representative” seconds of the song (usually the chorus).
I agree that getting the hook into peoples’ ears faster is the best way to increase a song’s popularity, but the actual mechanics of placing that hook so that these different technologies can pick up that exact 30 seconds would be difficult.
Similarly, I remember hearing speculation that pop artists would start writing songs that would be easily translatable to a Ringtone, so that the hook of a Missy Elliott song would be instantly recognizable when somebody’s phone rang on the subway, and it would sneak into the public consciousness that way.
- Zac
May 5th, 2006 at 9:56 am
Great points Zac. And interesting example of the ringtones!
One other example I remember reading about (someplace) was how new artists cover a well known song, just so when people search for the name of the song on iTunes, the new cover is listed right there with the original. If you cover a high profile song, one can generate lots of listens (and presumably purchases), just by appearing in that list of search results. Here musicians seem to be exploiting the iTunes interface itself.
May 6th, 2006 at 6:53 pm
I agree with Zac, most 30 second samples aren’t the hook, but something else, often though, they are not from the chorus … but there are researchers that are working on chorus detectors (looks like AMG is using one of those). Still, I think the 30 second samples are mainly to help consumers ensure that they are buying the ‘right’ song. I don’t know of anyone who ever bought a song based upon a first listen to a 30 second sample - 30 seconds is just not enough to decide.
May 7th, 2006 at 2:05 pm
“Still, I think the 30 second samples are mainly to help consumers ensure that they are buying the ‘right’ song.”
Not necessarily. Perhaps the 30 sec sample is used as a teaser to entice people to download a free version of track. Generally, I agree with you (Paul), but I do think that the 30 second audio clip can also be used for music discovery. For example, I have listened to audio clips of new music on last.fm, and then have gone on to download whole albums based on that clip. I might be a freakish outlier, of course…
You both are quite right for pointing out that most clips will not come from the first 30 seconds though. There’s quite a bit of interesting research going on about how to create audio thumbnails and how to select the 30 most representative seconds of a particular track.
But given that 30 sec samples will be around for a while (I think they will because of bandwidth issues as well as the current state of music licensing), I still think that a track that sounds fantastic in a 30 sec clip might be quite different then a track that sounds great on a complete (or repeated) listen.
I remember reading an interesting (and perhaps related) story in Malcom Gladwells book “Blink”. In the 80’s, Pepsi launched a marketing campaign called the “Pepsi Challenge” where an individual was presented with a small sample of coke or Pepsi. The sweeter tasting Pepsi always came out ahead by a large margin, but when it came time to consume a whole can, most individuals preferred Coke. Gladwell explains that this is because the sweeter tasting Pepsi excelled in a small sample, but was too sweet when it came time to drink the whole can.
Is there a parallel to music here? I know I might be stretching, but I think there could be. If the 30 sec clip is “sweet” enough to convince someone to purchase, download, or at least take notice of the particular artist, that might be enough over time to skew popular music towards tracks that sound good during a short sound bite (as long as it also sounds good on a whole listen, of course).
I guess my point (I think it’s somewhere in here) is that music discovery through a 30 sec clip is something new. It introduces a new bias towards different artists who might not have excelled when music discovery took place primarily by listening to complete tracks over the radio, for example.
All right, I’m going to go drink a coke and listen to Coltrane. The whole thing…
May 7th, 2006 at 3:39 pm
I agree, the 30 second clip introduces bias towards different artists, and perhaps even more important, away from others. When the 33rpm vinyl came out it open the door for longer compositions. Pop/Rock artists of the 60s and 70s took advantage of this new opportunity and we saw the birth of album oriented rock. Sgt Pepper and The Who’s Tommy and the Dark Side of the Moon are excellent examples where songs seemlessly flow from one to another. Unfortunately our ‘advanced’ iPod technology just makes a mess of these albums. Try listening to Tommy on an iPod, every 3 minutes you are reminded how we’ve regressed as a two second audio gap interrupts the flow. Even some of the more advanced playlist formats such as XSPF don’t give control (as far as I can tell) over the spacing between songs.
The ipod generation with its 30 second clips, and the attitude of ‘no need to buy a whole album when you can get the single good song’ will no doubt send us back to a time similar to twhen the 45 was popular, where a song existed with very little surrounding context.