Device, system and method for generating an accompaniment of input music data

A method for generating a music database for use in a device and method for automatically generating music, the method comprising:

• receiving live music data in real-time via a microphone;
• segmenting the music data into music patterns;
• extracting a plurality of characteristics of a music pattern as metadata of the music pattern, the characteristics including at least chords, energy per bar, density of notes, lowest pitch, and mean pitch of a music pattern;
• selecting transitions in the received live music data, to (i) use the transitions as music patterns and (ii) extract a plurality of characteristics of a music pattern as metadata of the transitions;
• and
• storing the music patterns and associated metadata in the music database,
• wherein the music database is configured to output real-time audio that is automatically generated based on the plurality of characteristics of the music patterns and the plurality of characteristics of the transitions associated with the live music data.

A music database comprising circuitry configured to:

• store music patterns including segments of music segmented from live music data received in real-time via a microphone;
• store metadata including a plurality of characteristics of a music pattern, the characteristics including at least chords, energy per bar, density of notes, lowest pitch, and mean pitch of a music pattern;
• store metadata of transitions of music patterns, the transitions are selected from the received live music data and the transitions are used as music patterns, and the metadata includes a plurality of characteristics that are extracted from the music patterns;
• and
• output real-time audio that is automatically generated based on the plurality of characteristics of the music patterns and the plurality of characteristics of the transitions associated with the live music data.