Disclosed herein is a signal processing apparatus, including: extraction means for extracting a feature quantity indicative of characteristics of sounds represented by a sound signal, the sounds including different tones from within a 12-tone equal temperament and being arranged in an order of a musical scale based on a reference sound having a predetermined tone; and root decision means for receiving the feature quantity and deciding whether the reference sound is a root of a first chord of the arranged sounds, the decision being based on the feature quantity.