Pop-E: Musical Interpretation Model for
Polyphrase Ensemble

A polyphonic musical piece has to be rendered as it may exhibit independent expression of each voice, as well as considered synchronization. Pop-E, where ``attention transition'' is focused on in designing the model, generates expressive music performance as following;

1) giving each voice part an independent performance expression,
2) introducing a parameter related to attention transition, and
3) synchronizing onset timing between voice parts by using prolongational reduction and the rule of attention transition.

The above blue boxes shows a musical attention and the blue arrows means its movement. Parameters given to the system, for example, are; velocity of arpeggio’s start notes were stressed to 125%, apices were lengthened to 150%, the first notes given attention were lengthened to 200% in the accompaniment or to 150% in the melody, and so on.

Performance Examples：

• "Fantasie-Impromptu", the middle development (F. Chopin) (Jun., 2004)
  • MP3 (given constant parameters based on the method)
  • MP3 (just two notes of the previous performance are manually adjusted)
  • dead-pan (MP3)
• Minuette KV 1(1e) ( W. A. Mozart ) (Aug., 2005)
  • MP3 778 kb
  • WAV 8.5 Mb
• Rapsodie sur un theme de Paganini (Ravmaninoff ) (Aug., 2005)
  • MP3 664 kb
  • deadpan (MP3)
  • WAV 7.3 Mb
• Adagio from Swan Lake, ACT II (Tchaikovsky) (Aug., 2005)
  • MP3 976 kb
  • deadpan (MP3)
  • WAV 10.8 Mb