When I teach swing, I like to play examples of the same piece of music with and without swing for ease of comparison. My favorite comparison is between “Dance of the Sugar Plum Fairy” from the Nutcracker Suite and “Sugar Rum Cherry” by Duke Ellington. This isn’t an exact comparison, though, because Ellington does more than change the time feel; he also changes the instrumentation and structure. I wanted to find an example where the same music repeated identically with and without swing. The problem is that so far as I can tell, no such piece of music exists. But then I realized that it would be easy to make this piece of music myself, by warping something out in Ableton Live and applying different groove settings.
I decided to use Bach’s Prelude No. 1 from the Well-Tempered Clavier as performed by Glenn Gould, because it’s a slow, steady stream of straight sixteenth notes. I added the drum intro from “Soul Love” by David Bowie, which is also laid back straight sixteenths. With piano notes and drum hits on every sixteenth note subdivision, the effect of swinging those sixteenth notes would be maximally obvious. My original thought was to alternate straight with swung sections. As I was trying different swing feels from the Ableton Groove Pool, I realized that it would be better to cycle through several different feels, rather than alternating straight and swing. After some experimentation, here’s what I came up with.
At the beginning of the track, you’re hearing Glenn Gould and the Bowie beat with their original straight time feel. Over the next thirty seconds, the swing ratio gradually increases until it is a very wide 71%. Then it gradually decreases until by the 1:00 mark, the feel is completely straight. The swing increases again until it peaks at about 1:30, and then straightens back out by the end of the track.
The swing percentage system was the invention of Roger Linn, who first used it in the Linn LM-1. Linn’s system has become the most common way to specify swing settings in drum machines and DAWs. To understand what the percentages mean, you need to know some terminology. A measure of 4/4 time is divided into four beats, each of which is a quarter note long. Each of those quarter notes is divided into two eighth notes, which are the basic pulse unit in most current American groove genres. Each eighth note pulse unit is subdivided into two sixteenth notes. The first sixteenth note in each pair is called the onbeat subdivision, and the second one in the pair is called the offbeat subdivision. To make your drum machine swing, all you need to do is make the onbeat subdivisions longer, and make the offbeat subdivisions correspondingly shorter. The result will sound extremely fake, because there is a lot more to swing than just stretching and shrinking your subdivisions uniformly. But because drum machine swing is so simplified, it’s a good pedagogical starting point.
Anyway, the swing percentage represents the onbeat subdivision’s length relative to the overall pulse unit.
- In straight time, the onbeat and offbeat subdivisions are the same length, 50% of the pulse unit each, so Roger Linn calls that “50% swing.”
- At 57% swing, the onbeat subdivision is four sevenths of the overall pulse unit, and the offbeat subdivision is three sevenths. That is close to even, so this is a light and subtle swing feel.
- At 60% swing, the onbeat subdivision is three fifths of the pulse unit, and the onbeat subdivision is two fifths. Now the swing feel is more pronounced and obvious.
- At 67% swing, the onbeat subdivision is two thirds of the pulse unit, and the offbeat subdivision is one third. This two-to-one ratio is also known as triplet swing or 12/8 shuffle, and it sounds more like blues or country than like funk or hip-hop.
Most groove-based American music of the past hundred years uses between 50% and 67% swing. That said, you are certainly allowed to use wider swing if you want; jazz drummers sometimes get up to 75% or beyond at slow tempos.
While Roger Linn’s percentage system is widely used, it’s not the only convention out there. Some DAWs use percentages with a different scale. Digital Performer uses 50% for straight time like Linn does, but it defines triplet swing as 100% rather than 67%. (This is also the convention we used for the Groove Pizza.) FL Studio uses 0% for straight time and 100% for triplets, which is the system I find appealingly intuitive, even if the numbers no longer directly describe the length of the onbeats and offbeats.
Musicologists who study swing usually talk about it in terms of onbeat-to-offbeat ratios (also called beat-upbeat ratios). In these terms, straight time is 1:1, moderately heavy swing is 1.5:1, and triplet swing is 2:1. Milton Mermikides prefers to describe time feels in terms of rhythmic cents, 1/100th of a pulse unit. This enables you to use the same system for swing as you do for phrasing ahead of or behind the beat: if John Coltrane is playing straight time 10 cents behind the beat and Elvin Jones’ swing feel uses 60-cent-long onbeats, then Coltrane and Jones will line up on all the offbeats. This enables Coltrane to play with a different time feel from his rhythm section while still locking in with their swing.
One big difference between the way humans swing and the way computers do it is that humans don’t maintain a consistent swing ratio. Fernando Benadon found that not only do jazz musicians change their swing feel from one solo to the next, they vary it from one beat to the next. So when you say that Miles Davis tends to use a swing ratio of 1.2:1, that’s only vaguely approximating a whole universe of nuanced complexity. Benadon observes that jazz soloists swing more widely than bassists and drummers, and that they usually swing more widely at phrase endings. There is so much more microstructure waiting to be discovered through close study of jazz musicians’ timing. The concept of a swing ratio is a blunt analytical instrument at best, but at least it’s a place to start.