An Abstract Control Space for Communication of Sensory Expressive by yurtgc548

VIEWS: 3 PAGES: 18

									An Abstract Control Space for
 Communication of Sensory
Expressive Intentions in Music
        Performance
   Canazza, S., De Poli, G., Roda, A.,
              Vidolin, A.

       Presented by Aaron Yang
Path of Expressive Music
         Composer
             
           Score
             
         Performer
              
       Acoustic Signal
             
          Listener
       Conceptualizing Expressive
              Intentions
   Categorical
   Dimensional




                       PerformanceWorm (Goebl 2002)
                               Loudness and tempo
      Understanding Musical
Performance (Camurri et al., 2001)
 Layer 1 – Physical Signals: audio data
 Layer 2 – Low-level features and statistical
  parameters
 Layer 3 – Mid level features and maps
 Layer 4 – Concepts and structures
                    The Test
   Mozart’s K 622 Concerto for Clarinet in A major
   Performed:
       Light
       Heavy
       Soft
       Hard
       Bright
       Dark
       Natural
             The Analysis
 Double Factor Analysis
 Matrix of replies and factor score
 Multi-Dimensional Scaling (MDS)
 Cluster Analysis
 Space of Musical Expression


                 Robinson’s
                 agreement
                 index
               Experiment 1
 Performer plays the piece with the six
  adjectives in mind
 Performances are recorded
 Listeners rate the performances on a scale
  based on the adjectives
 2 groups-
     12 post graduate from a music conservatory
     12 with no musical training
Matrix of Euclidian distances
Multidimensional Scaling of
        adjectives
Factor Loadings
Factor Plot
Cluster Analysis
Experiment 2 – New adjectives
    nero (black),             dolce (sweet),
    greve (oppressive),       aereo (airy),
    grave (serious),          lieve(gentle),
    tetro(dismal),            spumeggiante
    massiccio (massive),       (effervescent),
    rigido (rigid),           vaporoso (vaporous),
    soffice(mellow),          fresco (fresh),
    tenero (tender),          brusco (abrupt),
                               netto (sharp).
Experiment 2
Cluster Analysis
    Human Computer Interfaces
 Graphic panel – user learns how to use it
 Multimodal – through movements and
  non-verbal communication. The interface
  determines what the human intention is.
      Expressive Human-Computer
               Interaction
   Energy (x axis)
   Kinetics (y axis)
   Ktempo (z axis)
Once upon the time

								
To top