The Validity of Autotune

or: New Technologies, Old Arguments

There's been a lot of fuss in the press over the past couple of years over the use of Autotune, and I just wanted to collect my personal thoughts on the matter.

For those of you who've been living under a rock and don't know what I'm talking about, here's a quick history (which sadly will be lacking links and is typed entirely from memory, on a phone keypad, due to my ISP, TalkTalk, having accidentally cancelled my account and not told me why despite having had over a week to call me back now, but I digress...), so you'll have to forgive me any errors.

Antares Autotune was originally developed based upon software designed for seismic analysis. It basically works to pitch-shift a sound wave, i.e. make it higher or lower pitched.

Pitch-shifting technology has been around for decades. Those voice-changers you can buy as novelty toys or to disguise your voice on the phone use the tech. Where Autotune differs is that a user is able to vary the amount that the sound is shifted, specifying which particular frequency (i.e. which note) they want that sound to make.

Better still, users can program in an entire sequence of notes, using a MIDI sequencer for example, and have the sound come out at the perfect pitch. Or, alternatively the software can analyse the note you're singing, and if it's a bit flat (slightly lower pitched) or sharp (slightly higher pitched), it can correct it automatically for you.

So, we all on the same page now? Good.

Now, I think it's instructive to consider Autotune in comparison to the electric guitar or the synthesiser keyboard. Each of these instruments began as a way to mimic an existing instrument. The earliest electric guitars were just acoustic guitars with a microphone mounted inside, making it easier to amplify the sound so a larger audience could listen. Then the pickup was invented. A pickup is essentially an electromagnet, mounted on the body of the guitar, just under the strings. As the steel strings of the guitar vibrate they induce tiny currents in the electromagnetic coil. These currents can then be amplified, giving an electrical signal that can be easily sent to a loudspeaker. A much more robust way to make a guitar louder.

The earliest uses of electric guitars attempted to sound as much like an acoustic guitar as possible, but steadily the sound evolved into the dizzying array of guitar sounds we have today. The story of the synthesizer runs almost perfectly parallel to this, a little later in history. Ditto drum machines. Ditto the use of computers in production.

In fact, historically, a whole range of technologies have been the product of the evolution of earlier, more limited technologies. Not just in the past hundred years either; the piano (or, to give it its full name, the pianoforte) was developed from earlier instruments like the harpsichord, itself an evolution from the harp.

I believe Autotune is no different. It's simply a new tool musicians can use. Originally it was used to try and make singers sound as close to an unaided singer as possible. Then Cher released Believe, in which a production accident made the use of Autotune unusually prominent. They spotted that it was a hook, left it in the final release, and had a massive worldwide hit as a result. That song went number one damn near everywhere, so of course other musicians sat up, paid attention, and stole the idea, pushing it further and further. Kanye released an album deliberately saturated with the sound, and it's become the sonic signature of T-Pain. So much so, in fact, that T-Pain recently teamed up with Autotune to release an Autotune iPhone app.

So, do I see Autotune as being bad for music? No, not at all. I find it amusing that so many of its critics level at it the exact same arguments that the musical establishment level at any new sound, be that Mozart or The Beatles. It's corrupting, it's leading people astray, it's a base influence beneath contempt, it's all that's wrong with modern music. Bad news, critics, the lesson from history couldn't be clearer or simpler: Adapt or die.

But as much as I'm keen to defend the use of Autotune, even I have a hard time with some of its applications. Specifically, when a singer is presented to us as being superhumanly talented, when in fact that's a lie. In recent weeks here in the UK the TV show X-Factor has been heavily criticised for being caught using Autotune, to which it has confessed. I understand that the show is for entertainment, but its core purpose is that of a talent show. It's a search for the best singer. How can you possibly judge who's the best if you're deliberately manipulating how they sound? How is it fair on the naturally talented singers that the show claims to be looking for if they have to compete with people who're deliberately and extensively augmented? I see it as no different to clean athletes being expected to compete against athletes on performance-enhancing drugs.

I think it comes down to Autotune still being a dirty word for some people, particularly members of the music press it seems. The way forward? Accept the tech is here to stay, stop making people afraid to admit use of it, and focus on the only thing that's really important:

Does it sound good?

Seej 500

Yorkshire, UK, 1/9/2010