Skip to main content

I was listening to Morrissey sing Dear God, Please Help Me, on his recent album Ringleader of the Tormentors, when I had one of those moments of revelation that sometimes happen in pop music. As that light, vulnerable tenor floated above the song's gliding beat, I realized that I was hearing something that has been banished from whole sectors of the recording industry: a man singing out of tune.

He's not far off the pitch, most of the time, though enough to notice if you have an ear for that sort of thing. One descending line sags a little each time he sings it, especially at the words "track me down," when he slides a full tone flat at the trailing end of the phrase. He goes sharp, too, pushing a bit too high each time he reaches the last and highest-pitched words of the final refrain, "but the heart feels free." At this point in the music, Ennio Morricone's orchestral arrangement has reached its full magnificence, all the tension Morrissey sang about in the opening lines has dissipated, and you're left with the very tactile symbolism of a voice straining upwards as the heart feels its freedom.

Dear God, Please Help Me is one of the most beautiful songs of the year (also one of the most mischievous, since it portrays a sexual tryst as an encounter with divine mercy), and its flaws are part of its beauty. It might not have turned out nearly as well if Moz were up to date, and used Auto-Tune like everybody else.

Auto-Tune, in case you haven't heard, is a top-selling software program that corrects pitch for musicians who can't always do it themselves. It became headline news three years ago, when country singer Allison Moorer attached stickers on her album Miss Fortune that read: "Absolutely no vocal tuning or pitch-correction [technology]was used in the making of this record."

A lot of the discussion since has focused on whether you're cheating if you use Auto-Tune, or programs like it. But at a time when people use samplers and old vinyl to make hit records without singing or playing any instrument themselves, it seems a bit quaint to say that Fatboy Slim is playing by the rules, but Paris Hilton isn't.

What matters more is what auto-tuning software is doing to musical styles and to our experience and expectations as listeners. It isn't just a tool, but an instrument that's reshaping the character of popular music, song by song.

Auto-Tune was introduced in 1997 and has become standard equipment in most recording studios. Run anyone's voice or instrument through it (one at a time -- it can't tune a chord), and the program adjusts the pitch of each note to match the nearest comparable value in a preset scale. It's quick and easy to use, and if you listen to popular music, you're hearing its effects every day.

"It's on about 100 per cent of records that you hear on the radio," singer Neko Case said recently on PBS's The Tavis Smiley Show, exaggerating just a wee bit. She was trying to explain to Smiley why he had so often gone to concerts and found that the live thing (as he put it) "doesn't sound anything like the record that you fell in love with." Case told me last year that she asked the engineers at Toronto's Iguana Recording Studios, where she and Darryl Neudorf mixed her latest album, which singers coming through there still don't use the program to tweak their vocals. "They said, 'Just you and Nelly Furtado. Everyone else does,' " she said.

Bob Dylan, Neil Young or Tom Waits probably don't have much use for auto-tuning, because a ragged delivery is integral to their style. To hear one of these performers (whose careers were all well-established before Auto-Tune came along) singing perfectly in tune would be like seeing Dylan with a facelift.

Auto-tuning is audibly rampant in the R&B scene, and among young punk and emo bands. Green Day's hit single Wake Me Up When September Ends is almost a love song to pitch-correction technology. Singer Billie Joe Armstrong delivers this pop-punk ballad about insecurity and pain in an unnaturally steady stream of pitch-perfect vocal sound. It might be an organ playing, except that it's still obviously a human voice, backed by heavily distorted guitars that are also dead in tune from start to finish.

Green Day presents itself as a punk band, and punk has traditionally gone for a messed-up, do-it-yourself aesthetic that scorns overt displays of skill or polish. In that context, the perfectionism of Wake Me Up When September Ends feels like the last nail in Joe Strummer's coffin. Punk used to be about losing control and not really wanting it back; Green Day makes the genre seem as tidy as a golf green. You might say that Green Day was always headed in that direction ("They're not punk, they're plonk," said the Sex Pistols' John Lydon earlier this year), but auto-tuning may have boosted them to the summit of neatness.

American Idiot, Green Day's latest album, has sold almost seven million discs (including legal downloads) and topped the charts on both sides of the Atlantic. The more they and other auto-tuning musicians succeed, the more people become habituated to their kind of pitch perfection. Dead-centre pitch is becoming the new norm. As it does so, a lot of popular music's expressive capacities may wither away.

Sarah Vaughan had a famously good ear, yet she purposely sang away from the centre of the pitch when the music seemed to need it. "You'll dry all my tears," she sings in Lover Man, her voice bobbing below pitch between each of the first words before landing deliberately flat on "tears." She's back in tune by the very end of the phrase, but along the way she has given the line a sexy, needy fluidity that suits the song's dreamy construction of the perfect lover.

Glenn Gould, writing about Barbra Streisand in her prime, said that her singing presented "a seemingly limitless array of available options," one of which was her constant push-pull relationship with the centre of the pitch. You could say the same thing about old-time blues singers such as Leadbelly, whose forceful rise above the "right" pitch on the phrase "your house catch afire!" is as meaningful as the bent "blue" notes of the White Stripes' recording of Little Bird. Much of the expressive power of supposedly reductive forms such as blues and rock comes from all the shade and colour lurking between the "correct" pitches, and it works on you even if you can't consciously detect the deviations.

Antares Audio Technology, which developed and sells Auto-Tune, says that its software doesn't have to strip out expressive details. Only when you put it on the fastest retune setting does it suppress all pitch variations ("the now-infamous 'Cher effect,' " to quote the Auto-Tune manual). That includes vibrato, which from a computer's perspective is just a regularized way of singing out of tune.

"An appropriately selected slow setting can leave expressive gestures intact," says the manual, citing vibrato and portamento (gliding between pitches). You can even use the program to beef up the expressive gestures, adding vibrato or changing its speed or amplitude, "to allow the creation of much more convincing vibratos." Going all the way, you can strip out every distinguishing nuance, add a completely synthetic vibrato, or rebuild from the bottom using the Physical Modeling Vocal Designer, "a radical new vocal tool" whose capacity for virtual redesign of throat, vocal cords, mouth and lips "allows subtle vocal modifications or vocal tract models well beyond limits of physical human anatomy."

Very interesting, Dr. Frankenstein; but most Auto-Tune users are looking for a shortcut, not ways to increase their labours in the studio. They want a gizmo to clean up the one or two takes they've already got, to avoid having to record a dozen more takes and maybe cut and paste between them. Noting the total absence of vibrato in the eerily well-tempered Wake Me Up When September Ends, I'm betting that Green Day ordered something close to the full Cher and called it a day.

And it doesn't stop there. Auto-tuning programs can be used in live performance (including karaoke), with no discernible sound delay. You need never hear how your favourite singer really sounds, unless they forget to switch the thing on, or stray into a situation in which it's not convenient to do so. Michael Bublé is a mediocre song interpreter, but he has a nice sound and on his recordings appears to have a flawless ear. But when he sang his single Home on last spring's Juno Awards broadcast, he was off pitch from beginning to end, and not in a good way (check out the video on To judge from several other live-performance tapes, that's not unusual for him.

Maybe his fans don't notice. Maybe they do, but accept his shortcomings as part of the immediacy of live performance. I kind of like the fact that he let it all hang out at the Junos. The mask of technology slipped away, and the musician had to show what he could do, and what he couldn't.

"Sometimes on the record I hit notes a little funny," says Neko Case of her recent Fox Confessor Brings the Flood, "but it's kind of like, as my friend Brian Connolly would say, 'Humans were here.' " If you want perfection, try mathematics.

Interact with The Globe