CHAPTER

• 1 INTRODUCTION
• 1.1 Motivation
• 1.2 Broad Research Question
• 1.3 Previous Work
• 1.3.1 The Role of Timbre
• 1.4 More Refined Research Question
• 1.5 Contributions of this Dissertation

• 2 PREHISTORY OF MUSICAL ROBOTS
• 2.1 Introduction
• 2.2 Greek Antiquity and the Islamic Golden Age
• 2.2.1 Archimedes
• 2.2.2 Hero of Alexandria
• 2.2.3 Apollonius
• 2.2.4 Banu (Sons of) Musa
• 2.2.5 Ismail ibn al-Razzaz al-Jazari
• 2.3 17th to 19th Century Europe
• 2.3.1 Vaucanson
• 2.3.2 Keyboard Automata
• 2.3.3 Miscellaneous
• 2.3.3.1 Marreppe's Violinist
• 2.3.3.2 Friedrich Kaufmann
• 2.3.4 Manzetti
• 2.4 Conclusion

• 3 TIMBRE PRODUCTION
• 3.1 Introduction and Previous Work
• 3.2 Striking Mechanisms
• 3.2.1 Djembe Strokes
• 3.2.2.1 Bass
• 3.2.2.2 Tone
• 3.2.2.3 Slap
• 3.2.2 Factors Influencing Timbre
• 3.2.2.1 Hand Rigidity
• 3.2.2.2 Hand Morphology
• 3.2.3 Additional Considerations for Slap
• 3.2.3.1 Hand Size
• 3.2.3.2 Open Fingers
• 3.2.3.3 Sticks in Sabar
• 3.2.4 Timbral Evaluation
• 3.2.4.1 Methodology
• 3.2.4.2 Materials Used
• 3.2.4.3 Centroid
• 3.2.4.4 Comparisons
• 3.2.4.5 Discussion
• 3.2.5 Kiki's Hand
• 3.3 Arm
• 3.3.1 Three Segment Arm
• 3.3.1.1 Impact Angle
• 3.3.2 Inverse Kinematics
• 3.4 Future Work

• 4 REALTIME TRANSCRIPTION OF HAND-DRUM TIMBRES
• 4.1 Introduction
• 4.2 PreviousWork
• 4.3 Implementation
• 4.3.1 Onset Detection
• 4.3.2 Feature Selection
• 4.3.3 Realtime Classification
• 4.4 Evaluation
• 4.4.1 The Ecological Case
• 4.4.2 Loudness
• 4.4.3 Extended Techniques
• 4.4.4 Different Instruments
• 4.4.4.1 Cajon
• 4.4.4.2 Darbuka
• 4.4.4.3 Frame Drum
• 4.4.4.4 Bongos
• 4.4.5 More Subjects
• 4.5 Conclusion and Future Work
• 4.6 Acknowledgements
• 4.7 Corrigendum

• 5 TIMBRAL AUTOEVALUATION
• 5.1 Introduction
• 5.2 PreviousWork
• 5.3 Implementation
• 5.3.1 Control Space
• 5.3.2 Timbre Space
• 5.3.2.1 Features
• 5.3.2.2 Expected Distances
• 5.3.3 Map
• 5.3.4 Optimization
• 5.4 Evaluation
• 5.4.1 Computational
• 5.4.1.1 Trials
• 5.4.1.2 Trends
• 5.4.1.3 Unreachability
• 5.4.2 Perceptual
• 5.4.2.1 Methodology
• 5.4.2.2 Listening Styles
• 5.4.2.3 Controls
• 5.4.2.4 Other Features
• 5.5 Discussion

• 6 INTERACTIVE RHYTHM LEARNING FOR PERCUSSION ROBOTS
• 6.1 Previous Work
• 6.1.1 Algorithmic Composition
• 6.1.2 Interactive Computer Music
• 6.1.3 Musical Robots
• 6.1.4 Machine Learning
• 6.2 Rhythmic Models
• 6.2.1 Components of Rhythm
• 6.2.2 Representations of Rhythm
• 6.3 Implementation
• 6.3.1 Basic Approach
• 6.3.1.1 Prediction
• 6.3.1.2 Learning
• 6.3.2 Improvements
• 6.3.2.1 Timbre
• 6.3.2.2 Activation Function
• 6.3.2.3 Timing
• 6.3.2.4 Long-term Dependencies
• 6.4 Evaluation
• 6.4.1 Network Topology
• 6.4.2 Learning Specific Rhythms
• 6.4.2.1 Individual Rhythms
• 6.4.2.2 Corpus of Rhythms
• 6.4.2.3 Signal Rhythms
• 6.4.3 Learning Improvisation
• 6.4.3.1 Note Density
• 6.4.3.2 Meter
• 6.4.3.3 Syncopicity
• 6.5 Conclusion and Future Work

• 7 CONCLUSION
• 1.1 Extensibility
• 1.2 Future Work

• BIBLIOGRAPHY

APPENDIX

• A PSEUDOCODE FOR INVERSE KINEMATICS OF A THREE SEGMENT ARM

• B AUDIO BUFFERS DON’T DO WHAT YOU THINK THEY DO!