Automated musical transitions through rule-based synthesis using musical properties

We present a novel approach to automatically generating a smooth transition between two distinct compositions. Our system, Intelligent Transition Generator (ITG), generates a transition between two input compositions. ITG analyzes tablatures of the input compositions to extract relevant information, such as playing style, time signatures, chord similarities, and note durations to determine appropriate rules for creating the transition. The composition engine uses various algorithms to apply the rules based upon timing information. ITG typically generates 2-4 bars of random transitions that are smooth and musically harmonious. Our new method to automatically compose sections of music does not require guidance from a composer or a set of sample compositions and therefore, it provides an efficient approach, which could be used in the gaming industry for dynamic, event-based music. It also could be applied to other multimedia purposes such as generating transitions between tracks in media players.