Next: Combining definitions Up: An introduction to procedural Previous: AI sound
"Relating to or comprising memory or knowledge concerned with how to manipulate symbols, concepts, and rules to accomplish a task or solve a problem"
The "problem" is producing sound that fits one or more constraints. Memory can exist in the form of stored knowledge in an expert system, weights in a neural network or Markov machine, genetic code in an adaptive system or just the last value of a register in an unstable recursive filter. Symbols and concepts are all about semantics, the meaning we give to the input, internal states and output of the system. The system doesn't really care about these, they are for our human benefit and may as well represent hatstands, badgers and windmills as far as the program cares. The important parts are really the input and output and that the mapping makes sense in terms that are understandable and useful. Input can represent a physical quantity like velocity or a game state like the approach of a level 4 boss. Output is either directly audio signals or control signals for an audio source (which may itself be procedural).