« Droidmaker: George Lucas and the Digital Revolution | Disk is the New Tape »

An Audio Engineering Arthur C. Clarke Moment

Arthur C. Clarke famously wrote that, “any sufficiently advanced technology is indistinguishable from magic.” Well, today I read about a new audio editing that performs magic: Melodyne Editor.

As readers of this blog know, I’m a guitarist and hobbyist recording engineer. I built out a ProTools HD recording studio in my converted garage in Portola Valley, CA back in the day, and my band Soul Patch recorded our two albums there, which we released on our label, Toothless Monkey Music, and numerous other albums were recorded there by my band-mate Nick Peters, who now runs his own label and studio (Bodydeep Music) out in Redwood City, CA.

Please pardon what follows – it is a bit of audio geekery, but anyone who is even superficially familiar with the capabilities of modern digital recording systems will likely be slack-jawed in disbelief when I explain what Melodyne Editor can do. (I have to give a tip of the hat to Thomas Dolby — I’ve been a long-time reader of his blog (and fan of his music), and it was his blog post that made me aware of this amazing tool.)

Anyway, all of this background is just to say that I know my way around digital audio and signal processing plugins. I’ve been a long-time fan and user of AutoTune (quite useful for cleaning up “almost right” vocal takes), which, amazingly, can put out-of-tune vocals back in tune. AutoTune is an example of pure technology magic, though some lament the effect it has had on musicianship and vocal performance.

Then, a couple years ago, I began experimenting with a new pitch processing audio editor called Melodyne. Not only did Melodyne offer the ability to correct out-of-tune instruments or vocals, but it broke the audio waveform down into discrete notes that could be slid around in pitch and time using a graphic editor. This took a step beyond AutoTune – not only could you tweak the pitch of a performance, you could actually move the notes around with your mouse in pitch and time. You could literally alter the melody and rhythm of a vocal or instrumental performance by dragging your mouse around. Pretty amazing, right?

Of course, as amazing as these pitch (and time) audio processing tools are, they have a big constraint: they only worked with monophonic material. You needed a track with a single singer on it, or an instrument that only plays one note at a time: this left out most parts performed by pianos, guitars, vocal choirs, a horn section or an entire symphony. Basically if any of the audio in which you wanted to fix pitch problems contained chords (more than one note played simultaneously) on the track, you were out of luck.

The newest version of Melodyne does something most in the audio world have considered impossible: it allows the editing of polyphonic material. You can literally reach inside a guitar track and retune an individual note within a chord. Or find an out-of-tune singer in a group of backup singers and fix just that singer’s out-of-tune note.

This, my friends, is magic. And one step closer to what I’ve long considered my ultimate fantasy audio engineering tool: software that could take a mono or stereo mixdown of a song, and break it out into a multitrack representation of each individual remix. This would allow anyone to take a favorite song, break it into its component parts and build a remix. There are many reasons why this is probably far more difficult than pitch shifting individual notes in harmonic material, but if anyone could pull this off, my bet is on the wizards at Celemony. Wow.

Note: check out this video on Celemony’s website (sorry no embed code) to see what some serious pros (like Herbie Hancock) think about Melodyne. If you watch long enough, you’ll find Living Color guitarist Vernon Reid refers to Arthur C. Clarke famous aphorism as well, which I didn’t discover until after I wrote this post!

April 9th, 2010     Categories: Uncategorized    
  • http://www.dealhorizon.com John Sharp

    This sounds very very cool, and is indeed non-trivial. Back in the day (1993 to be precise), I spent *weeks* in my studio retuning Asian music samples to Western (A=440Hz) standards for a two-volume CD-ROM set that I released of Asian music samples.

    For the curious: It was called "Heart of Asia" and a demo can still be heard @http://www.ilio.com/spectrasonics/heartofasia/ind

    Sounds like this little device could have condensed all of that heartache into mere hours. Thanks for blogging about this Ryan.

  • http://www.uebermeister.com Christoph Jaggi

    It is amazing to see how language and location are still major obstacles to information availability and common knowledge. Even in the age of the internet. Melodyne was presented more than a year ago (in Germany that is) and was covered online and in print.
    Here's a story from January, 15, 2009: http://www.heise.de/newsticker/meldung/Audiobearb

  • SJGooch

    If cepstrum processing (http://en.wikipedia.org/wiki/Cepstrum), is so cool and perfect, why is it that everyone can tell what it sounds like, and are now sick of listening to musicians and rappers through that particularly annoying type of cepstral distortion? It is a constant reminder that whomever you are hearing can’t sing! How long will it be until a wannabe musician can be retuned, and we can’t tell? (Sort of a “Turing test for the ears?”)