Will Melodyne Revolutionize Sampling?

The fine folks at Celemony just released Melodyne, their polyphonic audio note editor. Yes, you heard me right: you can now edit audio clips the way you edit MIDI! Once only a distant figment of my imagination, this feature is now actually a reality.

What will this mean? Well, for one, it will be easier to correct an acoustical performance. If you wanted to alter a chord of a recorded piano performance for example, it will be easier than ever. You can just load up that recording in Melodyne and drag around individual notes in a chord to reassign their pitch and time position. Somehow, a magic algorithm doctor made the special voodoo needed to make it possible.

It’s mind boggling, really. See for yourself in this nifty video.

I really think this is going to be pretty revolutionary for sampling. I don’t personally sample other artists (even though I’d love to) because I don’t want to hire a lawyer for the clearances. (Also I’m a chicken.) But I do know that some artists like to grab a snippet and alter it so much that nobody will ever know where it came from. This will be a godsend for them, and not just to make the sample fit their track better… If one was in court, suing for sample usage, and the sample that was used had completely different notes and timing — it would be hard to prove that someone ‘stole’ the awesomeness from a prior track. And at the very least, it would make the case pretty interesting. 😉

Some advances you just know will mark a watershed moment in production technology. This might just be one of them.

On Producing Your Own Vocals

First of all, let me just say that it is never recommended.

The truth is, the producer hat and the vocalist hat are essentially at odds with each other. Just because you can do each one separately, doesn’t mean it will go smoothly when you try to do them together. The producer is the critical ear, the filter. Your producer brain wants to say no, do it better, that is not good enough. The producer brain is quite discerning. Then there is the performing brain — that is the part that must stay open. The performing brain must be supported, encouraged. There are no mistakes. The performance requires feelings, openness, vulnerabilty and safety.

Can you see how those things are nearly opposite? Stay too accepting, and you’ll never get that good take. Stay too critical, and you’ll never get that good take.

So, with this in mind, I am going to try to disclose some of my conclusions after sweating it out like crazy:

  1. Record first, edit later. Don’t think about editing on recording day, one hat at a time.
  2. Get ‘present’ – that stands to reason for any performance task. Be accepting. Too much desire for a perfect take almost guarantees a lack of results.
  3. Try taking on different characters. Who are you? Can you be someone else? This can help you get unstuck and give you some juice.
  4. Learn how to tell when you have enough good takes. Cover your bases, but don’t treat your voice like a sweatshop. You can always do pickups later.
  5. Pretend you are editing someone else’s voice. This is not you. Someone else hired you to do their hit song, and they happen to sound an awful lot like you.
  6. Learn to hear for microscopic details in a performance: breaths, little scoops, funny clicks or warbles. They will seem huge after 300 listens.
  7. Use autotune, don’t abuse it (unless you really want to.) It’s a great tool, which you can use to massage sneaky little words like “the” or “a” which people generally sing less accurately.
  8. Remember that a recorded performance is very different than a live performance. So don’t be shy about take looping.
  9. Don’t be afraid to screw up. In your home studio, time is free. Do it, and re-do it. Your morale may falter, but none of the time is wasted.
  10. Pat yourself on the back. Nobody else in the room to do it, and besides, no one is looking so you won’t look funny.