When we talk about Auto-Tune, we’re talking about two different things. There’s the intended use, which is to subtly correct pitch problems (and not just with vocalists; it’s extremely useful for horns and strings.) The ubiquity of pitch correction in the studio should be no great mystery; it’s a tremendous time-saver.
But usually when we talk about Auto-Tune, we’re talking about the “Cher Effect,” the sound you get when you set the Retune Speed setting to zero. The Cher Effect is used so often in pop music because it’s richly expressive of our emotional experience of the world: technology-saturated, alienated, unreal. My experience with Auto-Tune as a musician has felt like stepping out the door of a spaceship to explore a whole new sonic planet. Auto-Tune turns the voice into a keyboard synth, and we are only just beginning to understand its creative possibilities. (Warning: explicit lyrics throughout.)
Music is a transmission medium for emotions. A confident and definite performance comes across. When you have a singer do take after take after take in search of technical perfection, you often end up with the sound of a bored and annoyed singer. Bored and annoyed singers are a drag to listen to, no matter how accurate their pitch is. Auto-Tune makes it impossible to sing wrong notes, so you can always use first takes, when the performance is freshest. You can also take tossed-off improvisation and make it sound studio-perfect. Auto-Tune inspires fearlessness, which inspires playfulness, which produces good feeling for everyone in the room. What more could you ask from a music tool?
Auto-Tune isn’t just a source of pleasure. It can also evoke dread, as it does at 2:15 below.
For my tastes, the most musical uses of Auto-Tune come from contrasts. The extreme perfection works best when balanced by roughness and rawness elsewhere. In “20 Dollar” by M.I.A., her wordless Cher-effect melismas are balanced by loosely pitched uncorrected singing on the choruses and unpitched rapping in the verses. Also, M.I.A. layers distortion and reverb on top of the melismas to harshen them and remove their bubblegum quality.
When you give Auto-Tune an ambiguous or microtonal pitch, you get the characteristic warble between adjacent scale tones. The warble has a delightful set of qualities of its own. It introduces new rhythms into previously rhythmless sustained notes. If you add a little digital delay, the warble locks satisfyingly into the beat of the song. A quick fillip to a neighboring chord tone that would normally pass unnoticed by singer and listener alike suddenly takes on dramatic musical significance when exaggerated by Auto-Tune. Cultures that favor melismatic vocal techniques naturally find this effect to be fascinating. For example, here’s an Algerian tune called “Lkit li nebghih” by Cheba Djenet. (Hat tip to Jace Clayton for this example.)
By flattening the vertical pitch aspect of a singer’s voice, Auto-Tune draws out the horizontal qualities, the vibrato (as opposed to tremolo), the nasalness vs throatiness, the overtones and partials.
Auto-Tuned speech has remarkable qualities of its own, as rappers discovered years ago. Human speech is strongly tonal to begin with. When you automatically tune it to the closest piano-key pitches, you can more easily hear the melodies that were already present.
By quantizing and digitizing information, you make it easier to memorize and replicate it. I find myself humming phrases from the Gregory Brothers’ videos the way I hum Andrew Lloyd Webber. Digitized sound information is easier to memorize, store and copy. The subtle nuances of the Double Rainbow guy’s speech with all the pitches on a continuous spectrum are difficult to remember and imitate, but once he’s Auto-Tuned, it becomes effortless.
It’s easy to make jokes about talentless singers who can become famous using Auto-Tune. Spend some time in the studio, however, and you discover quickly that to sound good through Auto-Tune, the singer has to be good to begin with. The software can’t add emotion, tone, rhythm, or charisma. We shouldn’t have been at all surprised to discover that T-Pain sings like an angel without Auto-Tune.
Even more than T-Pain, Kanye West has come to be the avatar of Auto-Tune in hip-hop. My friend Greg Brown has some thoughts about that.
I’ve been listening to 808s and Heartbreak and Twisted Fantasy. I’m really enjoying them. Far more than I thought I would. I think Auto-Tune here is somehow protective for Kanye when he is expressing emotion in a genre where that is not really smiled on. I haven’t quite put my finger on it, but I think the dehumanizing of the human voice is somehow a foil for the expression of inner turmoil. It’s haunting… The hard part for me to wrap my head around is the fact that Auto-Tune is a filter, a dehumanizer, and it manages to make Kanye both closer and more human.
Maybe Auto-Tune heightens emotion by making the melody totally unambiguous. It gives the sung notes an organ-like clarity and distinctness, and slight pitch nuances get exaggerated into stairsteps and warbles. The filter changes the voice’s upper partials in odd ways that add to the pathos. Also, once we’ve come to expect the filtering, removing it can be a dramatic effect. Kanye’s raw singing voice is so comically bad that when you hear it unfiltered, it’s startling. In this day and age, hearing such a major pop star sing terribly is more remarkable than hearing him sing perfectly.
In his essay “Understanding Kanye: Sweet, Sweet Robot Fantasy, Baby“, Mike Barthel describes Ye as turning himself (figuratively) into a robot.
It wasn’t the raw emotion of humans, but the synthesis of emotional impulses and mechanical restraint, a computer’s inauthentic attempts at automatic expression which nevertheless sprung from a real human need to communicate.
Here we have it, the perfect encapsulation of what it’s like to be a feeling human being in a hypertechnological, hypercapitalist society. Auto-Tune gives that indefinable feeling a literal voice. No wonder it’s so popular.