> you're essentially exploring a fractal overlaid on all possible audible frequencies, modulated by time. Another way to think of it is that every knob on the

> you're essentially exploring a fractal overlaid on all possible audible frequencies, modulated by time. Another way to think of it is that every knob on the synth gives you a new dimension and each patch is a single point in that n-dimensional space. Sound design means exploring that volume by changing coordinates. I look at synth usability (which is very similar to how a lot of game designers talk about fun) as in part a question of how that space is arranged. How much of that volume produces sounds users will like versus ones they won't? How continuous is it? If I like a sound at point X, how likely am I to like sounds near X, or will moving even a small distance radically change the sound? Can I intuit what neighboring regions will sound like based on the current sound? I think a big part of why subtractive synthesis has been so successful is that it works well in that model. Most parameter changes are continuous and intuitive. The space feels like it smoothly changes in useful and predictable ways as you navigate. FM synthesis is really cool, but it feels a lot more like a chaotic universe where there are tiny islands of brilliant sounds separated by empty volumes of noise, or indistinguishable FM bloops.