In analyzing a microphone signal, I imagine you could be interested in understanding if it has any structure that can be explored by other algorithms, or you may be interested in seperating the sound into perceptual components, such as voice and a guitar, or you may be trying to reduce the noise...