臺灣博碩士論文加值系統

本論文提出一種單音調性音樂式之作曲演算法。此演算法將西洋音樂中協和音與調性中心的理論以基因演算法的函數實現至一個人性化且有邏輯的使用者介面上；並且，演算法所產生的旋律亦經由新一代的統計方法所測試。無論對於專業作曲者的靈感找尋或是一般大眾的娛樂效果，本系統所產生的有趣結果都有一定程度的實用性。此外，實驗結果和總結也對音樂學和聲學有所貢獻。本論文顯示出音高在旋律「悅耳性」的角色，也提供線索給該領域日後的研究。

This thesis presents a compositional algorithm for short monophonic tonal melody. It uses music theory knowledge of consonance and tonal center implemented in fitness function for genetic algorithm. The system was implemented in a user friendly intuitive graphic user interface. The generated melodies are tested by the novel statistical method. The system gives interesting results, which can be used by professional composer for
musical composition inspiration or by average user for fun and entertainment. The test results and summary can be also useful for musicology and auditory roughness. This thesis shows the role of pitch in melody “pleasantness” and gives many hints for future studies.

CONTENTS ................................................................................................................ 2
LIST OF FIGURES .................................................................................................... 2
LIST OF TABLES ...................................................................................................... 2
1.1 MOTIVATION.............................................................................................. 3
1.2 BRIEF LITERATURE SURVEY AND RELATED RESEARCHES......... 4
1.3 CONTRIBUTIONS ....................................................................................... 5
1.4 ORGANIZATION OF THESIS.................................................................... 5
Chapter 2 BACKGROUND AND PRELIMINARY KNOWLEDGE ...................... 6
2.1 MATLAB MIDI TOOLBOX ........................................................................ 6
2.2 MELODY, TONAL CENTER AND TONAL MUSIC ................................ 8
2.3 CONSONANCE AND DISSONANCE....................................................... 11
Chapter 3 ALGORITHM DESIGN ......................................................................... 16
3.1 SYSTEM STRUCTURE ............................................................................. 16
3.2 GENETIC ALGORITHM .......................................................................... 18
3.3 FITNESS FUNCTIONS .............................................................................. 21
3.3.1 MEAN-VARIANCE FITNESS FUNCTION................................... 21
3.3.2 SCORING FITNESS FUNCTION................................................... 24
3.3.3 HALF TONE FITNESS FUNCTION .............................................. 28
3.4 DATA........................................................................................................... 31
3.4.1 TRAINING DATA............................................................................ 31
3.4.2.TESTING DATA .............................................................................. 33
3.5 TEST RESULTS.......................................................................................... 34
3.5.1 TESTING DATA RESULTS............................................................ 34
3.5.2 SUBJECT TEST RESULTS............................................................. 36
3.6 USER INTERFACE.................................................................................... 39
Chapter 4 CONCLUSIONS...................................................................................... 42
REFERENCES ......................................................................................................... 43
APPENDIX A Training Data ................................................................................ 45
APPENDIX B List of Midi Songs in Testing Data................................................ 47
APPENDIX C Test Results .................................................................................... 49
APPENDIX D GUI Documentation ...................................................................... 50

[1] Robert R., Machine Musicanship, The MIT Press, Cambridge, Massachusetts, 2001
[2] Pozzati G., Gen: A Lisp Music Environment, Computer Music Journal, Vol. 24, No.
3 (Autumn, 2000), pp. 42-47
[3] Manzolli J., Interactive, Evolutionary Textured Sound Composition. Proceedings of
the sixth Eurographics workshop on Multimedia 2001, Manchester, UK pp: 153 –
164, 2002
[4] Yu C., Computer Generated Music (URL:
http://brainop.media.mit.edu/online/net-music/net-instrument/Thesis.html )
[5] http://www.mmk.e-technik.tu-muenchen.de/persons/ter/top/roughness.html
[6] Helmholtz, H. Die Lehre von den Tonempfindungen als physiologische Grundlage
für die Theorie der Musik. Vieweg, Braunschweig,1863
[7] Kim, Y., Theories of musical hearing: 1863-1931: Helmholtz, Stumpf, Riemann and
Kurth in Historical Context, Columbia University, 2003
[8] Yonatan I. F., Igor O. V., Noh M.D., Garell P.C., Bakken H., Joseph C.A., Howard
A.M., Steinschneider M. Consonance and Dissonance of Musical Chords: Neural
Correlates in Auditory Cortex of Monkeys and Humans (URL:
http://jn.physiology.org/cgi/content/abstract/86/6/2761 )
[9] Schellenberg E.G., Trehub E.S., Frequency ratios and the perception of tone patterns,
Psychonomic Bulletin & Review, 1994, 1 (2), 191-201
44
[10] Ochinsky, V.V. Octave cycle and golden proportion (URL:
science.ncstu.ru/nii/cycles/otc/2001/ )
[11] Burns, E. M. "Intervals, Scales, and Tuning." In The Psychology of Music, second
edition, edited by Diana Deutsch, 215–64. New York: Academic Press, 1998.
[12] http://www.jyu.fi/musica/miditoolbox/
[13] http://rosemck1.tripod.com/jukebox-classical.html
[14] http://www.jyu.fi/hum/laitokset/musiikki/en/research/coe/materials/miditoolbox/
[15] Hochreiter R., Audible convergence for optimal base melody extension with
statistical genre-specific interval distance evaluation, Springer Berlin, Dec.1, 2005
(URL: http://homepage.univie.ac.at/ronald.hochreiter/pub/preprint/h-audconv-full.pdf )
[16] http://www.mididb.com/
[17] Krumhansl, C. L. Cognitive Foundations of Musical Pitch. New York: Oxford
University Press, (1990).
[18] Toiviainen, P., Krumhansl, C. L. Measuring and modeling real-time responses to
music: the dynamics of tonality induction. Perception 32, 6 (2003), 741-766.