What Granular Synthesis Actually Does
granular synthesis works by slicing an audio sample into tiny fragments — grains — typically between 1 and 100 milliseconds long, and then reassembling them in ways that may or may not resemble the original. you can stretch time without changing pitch. you can freeze a moment and let it shimmer. you can scatter the grains randomly and turn a voice into a texture.
what you cannot do is preserve the original sound. granular synthesis is inherently destructive. the moment you slice the sample, you've committed to a kind of violence against it. the question is whether what you build from the pieces is worth what you destroyed to get there.
The Ethics of the Grain
this sounds abstract, but it has practical implications. the relationship between source material and transformation in electronic music is genuinely complicated. when you granularize a vocal recording, you're taking something that was a performance — a specific person, in a specific moment, making a specific sound — and turning it into raw material. the humanity of the original gets distributed across thousands of tiny fragments.
some producers find this liberating. the voice becomes texture. it loses its specificity and gains a kind of universality. others find it uncomfortable. the voice is still there, somewhere, in the grains. you can hear it if you listen for it. the question of what survives transformation in music is one I keep returning to.
What Gets Lost
the thing that granular synthesis destroys most completely is continuity. a sound has a beginning, a middle, and an end. it has a trajectory. granular synthesis breaks that trajectory into pieces and reassembles them in a different order, or loops them, or scatters them. the original arc is gone.
sometimes that's exactly what you want. sometimes the arc was the problem — too predictable, too linear, too much like what the sound was supposed to be. the experimental music tradition of productive destruction has always understood that you sometimes have to break something to find out what it's made of.
Saving the Sound by Destroying It
the paradox of granular synthesis is that it can make a sound more itself by taking it apart. a voice granularized at the right grain size and density can sound more like the essence of a voice than the original recording — more present, more textured, more alive. you've destroyed the specific performance to reveal something more fundamental.
whether that's worth it depends on what you're making and why. there's no universal answer. but the question is worth asking before you reach for the granular plugin. what are you willing to lose? what do you hope to find?