Digital electronics are now commonly used to replace humans or other pieces of analog equipment in the process of music making. And from many points of view (cost, reliability, compression, stability) digital has been a great improvement. But from an aesthetic point of view, many people find the results disappointing, off-putting or worse. Analog vs. digital is one of those endless techno-religious wars in which values get confused with specs.
I think this is because for a long time, digital computers (broadly speaking) were not being used in the area of their creative competitive advantage, which is remembering and recombining, quickly creating interesting options for us to choose from and refine. (Which, now that I think of it, is probably something many analog fans would object to in the first place.) But now developers are focusing on this area and for some reason, the most interesting results all seem to be coming from Germany.
The latest example is Melodyne from Celemony in Munich. Melodyne 1.0 took an audio sound file and broke it into its constituent notes with a high degree of success if the audio was relatively clear and distinct. The notes then appeared as sound "blobs" which could be changed in length, pitch and tone and triggered individually:
Like a sampler, it allowed you to treat existing audio as if it were just another sound coming from your keyboard. But because it analyzed and represented the sound as individual notes in a performance, you could create variations on existing performances with an ease no sampler could match.
Last month, Celemony introduced Melodyne 2.0. What makes this version even more interesting is that, in theory, it can do the same thing for a multi-part performance, breaking it apart instrument by instrument, note by note, spreading the arrangement out in front of you for detailed manipulation. You can change a single note in a recorded guitar chord. Try out alternate versions of your favorite songs using the original audio. 1.0 allowed variations on a performance. 2.0 allows variation in a performance.
But I'm most excited about what happens when Melodyne doesn't work properly. When, instead of cleanly picking out the bassline from "Hit Me One More Time", it mistakenly combines some of the bass sound with some of the vocal and a bit of the drums to create some unimaginable Frankenstein of virtual "instrument" and melody that never existed in the original. What happens when you play that? Do you hear something that was latent in the original song? A molecular recombinant form with faintly recognizable DNA, like a distant relative of the original? Remixing will gain a whole new set of dimensions. That will be interesting.
(Celemony's video introduction of Melodyne 2.0 is interesting in itself. Strangely Kubrickian in tone. You almost expect to see a Hanso Foundation logo at the end.)
Comments