Artificial reverberation is an unusually interesting signal processing
- ``Obvious'' methods based on physical modeling or input-output
modeling are too expensive
- We do not perceive the full complexity of reverberation
- What is important perceptually?
- How can we simulate only what is audible?