Dynamic Extent | A New Vocabulary | A New Language
Let's look at the language given to use by the elements described in the previous chapters. What can this language do? What can it express?
It can make really cool sounds (ref to examples, show the bubbles and moto examples). That's about it. You can do some basic sequencing using impulse, decay, pulse-divider, noise sources, et al, but your output will ultimately be either periodic or random. Interesting, but not powerful enough. (Pitch tracking and audio inputs)
The shortcomings that our language has are ultimately due to a distinction between sound and music. Sound has a continuous character---think of the vibrations hitting your eardrum. The vibrations never stop; they just keep on coming.
Think, on the other hand, of the music in your head. Things happen in that music. Music produces sound, but music speaks a different language, one that is fundametally discrete. As such, a computer language for music must reflect this discrete nature by describing music as discrete events occurring in time.
Beofre we describe the language that soundscrape creates, we need to create a bridge between the continuous and the discrete within the framework of unit generators.