The music on the "Song of Roland" is made with an analog computer - the 1975 Roland System 100 - and defines a term: Analog Computer Music (ACM). Compositionally, the computer music we investigate here is algorithmic - it is made by or with machine intelligence. Computer music is not human music, but music made by calculating, perhaps even thinking machines. Today these machines are digital. It is easy to forget that the digital age was preceded by the analog age when all computation was analog. The abacus would be such a thing, as was Babbage’s Difference Engine, but the first analog computer was the Antikythera mechanism, dating back at least to 100 BCE.
Roland`s System 100 was released in 1975 as a series of modules that, if put together as a whole system, provide 2 synthesizers, one with a keyboard for humans to play, a sequencer and a mixer with speakers. What makes this interesting is that only two modules, poetically named 102 & 104, make a complete analog computer that is capable of producing music on its own. These form quite literally the brain (module 104) and the voice (module 102) and all interaction takes place between the two since there is no keyboard for any human to play. Humans have to change knobs on the brain, the sequencer module 104, to effect any change in the order of sonic events that the System 100 emits. In traditional music the communication between brain and voice, between the sequencer and the synthesizer, is one-directional. The brain instructs the voice and the voice obeys. Perhaps not intended by the engineers at Roland, it is possible to re-wire the two modules so that a feedback loop occurs. Changes in the brain trigger changes in the voice, but now, changes in the voice will trigger changes in the brain. The brain is aware of the voice, as the voice is also aware of the brain. As a system, one could postulate that this qualifies as a self-aware entity. Let`s not get too absorbed in this. What we can define with certainty is that module 102 & 104, connected to each other, produce system-music. We merely steer the interaction between brain and voice. We do not determine each aspect of the music, constraining ourselves to observe and modulate the state of the modules, the state of the system as a whole. This is Analog Computer Music (ACM).