denktank 2010
Post on 11-Sep-2014
455 views
DESCRIPTION
Presentatie onderzoek Olmo Cornelis - 17 dec 2010 - IPEM GentTRANSCRIPT
![Page 1: Denktank 2010](https://reader037.vdocuments.net/reader037/viewer/2022102813/5411e20b8d7f72c9738b4659/html5/thumbnails/1.jpg)
Theoretic and artistic research studying opportunities of symbiosis
of Western and non-Western musical idiom
(2008-2014)
Olmo Cornelis
![Page 2: Denktank 2010](https://reader037.vdocuments.net/reader037/viewer/2022102813/5411e20b8d7f72c9738b4659/html5/thumbnails/2.jpg)
OutlineIntroduction• Background
Problem
Definition
Methodology
Case studies
Conclusion
![Page 3: Denktank 2010](https://reader037.vdocuments.net/reader037/viewer/2022102813/5411e20b8d7f72c9738b4659/html5/thumbnails/3.jpg)
Background
Olmo Cornelis- Musicology @ Ghent University
- Composition @ University College Ghent
- PhD @ University College Ghent / Ghent University
Previous work: DEKKMMA
“Digitization of the Ethnomusicological Sound Archive of the Royal Museum for Central Africa”
![Page 4: Denktank 2010](https://reader037.vdocuments.net/reader037/viewer/2022102813/5411e20b8d7f72c9738b4659/html5/thumbnails/4.jpg)
DEKKMMA
• Digitization of the sound archive:- Audio: 50.000 sound recordings – 3.000 hours of
music - 33.000 items digitized.- Meta-data : 35.000 items digitized- Provide contextual data
• Development of database and website
• Exploration of MIR-techniques.
http://music.africamuseum.be/
![Page 6: Denktank 2010](https://reader037.vdocuments.net/reader037/viewer/2022102813/5411e20b8d7f72c9738b4659/html5/thumbnails/6.jpg)
Deliver detailed descriptions of music from another culture- Different from Western musical idiom- (till now) Western concepts were used
- Correct descriptions are needed for:- Online collections- Research- Respect cultural heritage
Actual problem I
![Page 7: Denktank 2010](https://reader037.vdocuments.net/reader037/viewer/2022102813/5411e20b8d7f72c9738b4659/html5/thumbnails/7.jpg)
How to transfer this knowledge into an artistic production?
How to transform musical elements into another culture?
How to enrich a compositional language with specific ethnic musical characteristics?
How to create a personal compositional idiom with implicit influences?
Actual problem II
![Page 8: Denktank 2010](https://reader037.vdocuments.net/reader037/viewer/2022102813/5411e20b8d7f72c9738b4659/html5/thumbnails/8.jpg)
Work PathStudy of problem: research question
Definition (definiëring en afbakening)
Theoretical analysis / set up framework (methodology)
Cooperation musicologist and computer engineer
Testing existing applications
Build platform for ‘objective’ analysis
Case studies
Result (diverse output)
Listen and encounter
Define research question
Set up frame work (methodology)
Experimental phase
Composition
Experiment
Evaluation
Composition
Concert
![Page 9: Denktank 2010](https://reader037.vdocuments.net/reader037/viewer/2022102813/5411e20b8d7f72c9738b4659/html5/thumbnails/9.jpg)
MethodEthnic music?
Three categories:– Ethnic music (in strict sense)
From cultures without written tradition
– Non-Western classical music– Religious and court music from cultures with a written – tradition outside Europe
– Folk music– From written culture, but not part of a classical
tradition
![Page 10: Denktank 2010](https://reader037.vdocuments.net/reader037/viewer/2022102813/5411e20b8d7f72c9738b4659/html5/thumbnails/10.jpg)
Historical tendencies
- Initially- Aversion towards music
- Music connected with dance and rituals
- After WO2- Collecting systematically- Describing semantically
- 21th Century- Digitization (conservation, accessibility)- Computational ethnomusicology
![Page 11: Denktank 2010](https://reader037.vdocuments.net/reader037/viewer/2022102813/5411e20b8d7f72c9738b4659/html5/thumbnails/11.jpg)
Objective measurement:
- Not using Western musicological assumptions
- Approach actual characteristics
Music Information RetrievalSignal based research
Computer assisted analysis
Method
![Page 12: Denktank 2010](https://reader037.vdocuments.net/reader037/viewer/2022102813/5411e20b8d7f72c9738b4659/html5/thumbnails/12.jpg)
Music Information Retrieval
Nowadays:
- audio is a digital object
- huge online collections
- personal consumption
=> Music discovery!
![Page 13: Denktank 2010](https://reader037.vdocuments.net/reader037/viewer/2022102813/5411e20b8d7f72c9738b4659/html5/thumbnails/13.jpg)
Music Information Retrieval
- MIR: describes content of music
- Many applications
- Search/retrieve/organize
- Three levels- Signal => physical- Symbolic => musical symbols- Semantic => metadata
![Page 14: Denktank 2010](https://reader037.vdocuments.net/reader037/viewer/2022102813/5411e20b8d7f72c9738b4659/html5/thumbnails/14.jpg)
![Page 15: Denktank 2010](https://reader037.vdocuments.net/reader037/viewer/2022102813/5411e20b8d7f72c9738b4659/html5/thumbnails/15.jpg)
Music Information Retrieval
Many applications:
- Identification, plagiarism detection, copyright monitoring, version detection, melody finder, alignment, recommendation, sounds like, genre detection, segmentation, instrumentation, organization by metadata, mood, artist…
![Page 16: Denktank 2010](https://reader037.vdocuments.net/reader037/viewer/2022102813/5411e20b8d7f72c9738b4659/html5/thumbnails/16.jpg)
– domain specific techniques are required to approach the particular characteristics and constraints of different ethnic music styles.
– Need for an interdisciplinary research field.– Automated approaches are needed because of large archives.– large variability of music, users, search intentions and
expectations.– Metadata, content-based description, musical concepts are
fundamentally different from Western standards.
(MIR) considerations
![Page 17: Denktank 2010](https://reader037.vdocuments.net/reader037/viewer/2022102813/5411e20b8d7f72c9738b4659/html5/thumbnails/17.jpg)
Work PathStudy of problem: research question
Definition (definiëring en afbakening)
Theoretical analysis / set up framework (methodology)
Cooperation musicologist and computer engineer
Testing existing applications
Build platform for ‘objective’ analysis
Case studies
Result
Listen and encounter
Define research question
Set up frame work (methodology)
Experimental phase
Composition
Experiment
Evaluation
Composition
Concert
![Page 18: Denktank 2010](https://reader037.vdocuments.net/reader037/viewer/2022102813/5411e20b8d7f72c9738b4659/html5/thumbnails/18.jpg)
Artistic research- Composition (imho):
- Innovative element- Content - Coherence- Aesthetic component
- Ethnic music: kaleidoscope of sounds- Symbioses: added value of two components
=> Develop a personal idiom with implicit influence
![Page 19: Denktank 2010](https://reader037.vdocuments.net/reader037/viewer/2022102813/5411e20b8d7f72c9738b4659/html5/thumbnails/19.jpg)
Musical elements
![Page 20: Denktank 2010](https://reader037.vdocuments.net/reader037/viewer/2022102813/5411e20b8d7f72c9738b4659/html5/thumbnails/20.jpg)
Musical elements
Text Timbre Pitch Time Dynamics
![Page 21: Denktank 2010](https://reader037.vdocuments.net/reader037/viewer/2022102813/5411e20b8d7f72c9738b4659/html5/thumbnails/21.jpg)
Musical elements
![Page 22: Denktank 2010](https://reader037.vdocuments.net/reader037/viewer/2022102813/5411e20b8d7f72c9738b4659/html5/thumbnails/22.jpg)
Musical elements
![Page 23: Denktank 2010](https://reader037.vdocuments.net/reader037/viewer/2022102813/5411e20b8d7f72c9738b4659/html5/thumbnails/23.jpg)
Musical elements
Axes:- Explicit <---------------------------------> Implicit- Effect <---------------------------------> Concept- Form <-----------------------------------> Musical- Music an Sich <-------------------> Functional- High level <---------------------> Low level MP- Physical <--------------------------------> Virtual- … <-------------------> …
![Page 24: Denktank 2010](https://reader037.vdocuments.net/reader037/viewer/2022102813/5411e20b8d7f72c9738b4659/html5/thumbnails/24.jpg)
Musical elements
![Page 25: Denktank 2010](https://reader037.vdocuments.net/reader037/viewer/2022102813/5411e20b8d7f72c9738b4659/html5/thumbnails/25.jpg)
Musical elements
![Page 26: Denktank 2010](https://reader037.vdocuments.net/reader037/viewer/2022102813/5411e20b8d7f72c9738b4659/html5/thumbnails/26.jpg)
Approach Artistic Research
1. Acquire knowledge and specificities of ethnic music
=> scientific study
2. Set up compositional experiments
=> in exhaustive range of possibilities
3. Create framework to work in
=> see previous slides
4. Create (with implicit ethnic influence)
=> in a personal style
![Page 27: Denktank 2010](https://reader037.vdocuments.net/reader037/viewer/2022102813/5411e20b8d7f72c9738b4659/html5/thumbnails/27.jpg)
Results
![Page 28: Denktank 2010](https://reader037.vdocuments.net/reader037/viewer/2022102813/5411e20b8d7f72c9738b4659/html5/thumbnails/28.jpg)
Framework TARSOS- Pre-processing:
- Audio signal improvement:- Noise reduction- Signal optimization
- Source separation (ONO2010)- Band pass filters
- Processing:- (existing) algorithms- Annotation optimization
- Post-processing:- Interpretation- Visualization- Export
- Data- Graphical
![Page 29: Denktank 2010](https://reader037.vdocuments.net/reader037/viewer/2022102813/5411e20b8d7f72c9738b4659/html5/thumbnails/29.jpg)
SIGNAL LEVEL• Test of available pitch trackers• Representation of output• Real-time analysis• Optimization results
SYMBOLIC LEVEL• Peak extraction• Peak analysis
– Correlation
OUTPUT• Sonification• Export
Framework TARSOS
![Page 30: Denktank 2010](https://reader037.vdocuments.net/reader037/viewer/2022102813/5411e20b8d7f72c9738b4659/html5/thumbnails/30.jpg)
Demo
![Page 31: Denktank 2010](https://reader037.vdocuments.net/reader037/viewer/2022102813/5411e20b8d7f72c9738b4659/html5/thumbnails/31.jpg)
Exploring African Tone Scales
Audio set:
• 901 audio files from the RMCA-archive
• Solo pieces played on:– Zither– Flute– Thumb piano – Musical bow
![Page 32: Denktank 2010](https://reader037.vdocuments.net/reader037/viewer/2022102813/5411e20b8d7f72c9738b4659/html5/thumbnails/32.jpg)
Exploring African Tone Scales
Remarkable finding 1:
the evolution of the increasing number of pitches in the scale through time.
![Page 33: Denktank 2010](https://reader037.vdocuments.net/reader037/viewer/2022102813/5411e20b8d7f72c9738b4659/html5/thumbnails/33.jpg)
Exploring African Tone Scales
Remarkable finding 2:
Increase of ‘regular’ intervals in recent years.
Remarkable finding 3:
Specific ‘irregular’ intervals (e.g. 160 or 370 cents)
Interval < 1960 1960-75 > 1975
min.2nd 1,46 1,87 2,20
maj.2nd 1,57 1,71 5,20
min.3rd 2,26 3,39 2,84
maj.3rd 1,25 1,28 2,78
4th/5th 2,55 2,56 5,31
sum 9,10 10,81 18,33
![Page 34: Denktank 2010](https://reader037.vdocuments.net/reader037/viewer/2022102813/5411e20b8d7f72c9738b4659/html5/thumbnails/34.jpg)
Pitch (melody)Language steers melos:
Ex.: Lisaka:
march
promise
poison
=> mono syllabic
=> anti-melismatic
=> verse: no identical melody
=> polyphony: parallel (no counterpoint)
Ngomi ngomi:
![Page 35: Denktank 2010](https://reader037.vdocuments.net/reader037/viewer/2022102813/5411e20b8d7f72c9738b4659/html5/thumbnails/35.jpg)
Rhythm and tempo
Preliminary research:
Experiment: Tempo extractionbeat / pulse annotationby human and by pc
Goal:- human tempo perception- set of audio with ground truth- test quality beat algorithms
![Page 36: Denktank 2010](https://reader037.vdocuments.net/reader037/viewer/2022102813/5411e20b8d7f72c9738b4659/html5/thumbnails/36.jpg)
![Page 37: Denktank 2010](https://reader037.vdocuments.net/reader037/viewer/2022102813/5411e20b8d7f72c9738b4659/html5/thumbnails/37.jpg)
![Page 38: Denktank 2010](https://reader037.vdocuments.net/reader037/viewer/2022102813/5411e20b8d7f72c9738b4659/html5/thumbnails/38.jpg)
Tempo
Set up:
- 70 audio fragments, 20 sec each
- 10 beat trackers
- 10 persons
- Task: tap the tempo
![Page 39: Denktank 2010](https://reader037.vdocuments.net/reader037/viewer/2022102813/5411e20b8d7f72c9738b4659/html5/thumbnails/39.jpg)
Temposong nr no.6 no.20 no.30 no.31 no.49 no.69 no.1
subject 1 129 65 190 72 92 46 53
subject 2 129 65 63 144 111 186 122
subject 3 129 65 63 95 46 121 107
subject 4 129 127 129 142 130 140 115
subject 5 129 x 69 144 178 238 121
subject 6 129 124 63 47 92 92 117
subject 7 129 132 63 144 92 93 221
subject 8 129 64 63 73 122 69 94
subject 9 129 130 125 96 44 92 65
median tempo 129 65 63 144 92 92 107-121
consistency
identic tempo 9 4 6 4 3 3 6
tempo octave 0 4 2 2 2 2 2
binary/ternary 0 0 1 3 0 3 0
other relation 0 0 0 0 4 1 1
meter 2 2 3 3 5 2 1
![Page 40: Denktank 2010](https://reader037.vdocuments.net/reader037/viewer/2022102813/5411e20b8d7f72c9738b4659/html5/thumbnails/40.jpg)
Microtiming
![Page 41: Denktank 2010](https://reader037.vdocuments.net/reader037/viewer/2022102813/5411e20b8d7f72c9738b4659/html5/thumbnails/41.jpg)
Rhythm
Western conceptually different:
Divisive, Z/L, symmetry
In 8: 3/3/2, 3/2/3, 2/3/3, 1/3/2/2, 1/2/1/2/2
Ewe drumming
![Page 42: Denktank 2010](https://reader037.vdocuments.net/reader037/viewer/2022102813/5411e20b8d7f72c9738b4659/html5/thumbnails/42.jpg)
EME- Electronic sound and
music controller, based on human movement and using custom made wireless motion sensors
- Hardware technology, software applications, music theory and concepts of Embodied Music Cognition
![Page 43: Denktank 2010](https://reader037.vdocuments.net/reader037/viewer/2022102813/5411e20b8d7f72c9738b4659/html5/thumbnails/43.jpg)
EMEApplications:demo, for musicians and non-musicians, a real-
time sound controlArtistic performance, i.c. Con Golese, a piece for
tape and voice, where an opera singer could control the tape composition through her gestures while performing. Bodily expressions - often gestures - ensued on the vocal part, feel natural and cause a closer blending of voice and tape. The possible real-time interaction of a performer with a tape composition creates new opportunities for electronic compositions.
![Page 44: Denktank 2010](https://reader037.vdocuments.net/reader037/viewer/2022102813/5411e20b8d7f72c9738b4659/html5/thumbnails/44.jpg)
![Page 45: Denktank 2010](https://reader037.vdocuments.net/reader037/viewer/2022102813/5411e20b8d7f72c9738b4659/html5/thumbnails/45.jpg)
![Page 46: Denktank 2010](https://reader037.vdocuments.net/reader037/viewer/2022102813/5411e20b8d7f72c9738b4659/html5/thumbnails/46.jpg)
Audio might not be sufficient to describe and/or retrieve ethnic music:
-> The idea of separating sound from the rest of its physical environment (movement, smell, taste, colour) may well be a weird “invention” of the West.
=> We cannot understand ethnic music correct without its social function and context!
![Page 47: Denktank 2010](https://reader037.vdocuments.net/reader037/viewer/2022102813/5411e20b8d7f72c9738b4659/html5/thumbnails/47.jpg)
Next- Composition
- Dag van het Onderzoek (March 2011)
- New proposals (COST, Tetra, FET Open)
- Pitch- Finalize application- Article methodology TARSOS (nearly finished)- Article application TARSOS (several options)
- Rhythm - Put documentation and experiment into article (advanced)- Set up framework for TARSOS
![Page 48: Denktank 2010](https://reader037.vdocuments.net/reader037/viewer/2022102813/5411e20b8d7f72c9738b4659/html5/thumbnails/48.jpg)
Dissemination
Publications
Compositions
Activities
Website
Doctoral Schools
![Page 49: Denktank 2010](https://reader037.vdocuments.net/reader037/viewer/2022102813/5411e20b8d7f72c9738b4659/html5/thumbnails/49.jpg)
Publications2010
- Cornelis O., Lesaffre M., Moelants D., Leman M., 'Access to ethnic music: advances and perspectives in content-based music information retrieval', Signal Processing, Elsevier, 2010 - Lidy T., Silla C.N., Cornelis O., Gouyon F., Rauber A. ,Kaestner C.A.A., Koerich A.L., 'Western vs. Ethnic Music: On the Suitability of State-of-the-art Music Retrieval Methods for Analyzing, Structuring and Accessing Ethnic Music Collections', Signal Processing, Elsevier, 2010 - Cornelis O., 'Het etnomusicologisch klankarchief van het KMMA, digitaliseren en beyond…', in Handboek Muzikaal Erfgoed (Ed. Schreurs E.), Resonant, 2010, in press
2009 - Cornelis O., Moelants D., Leman M., 'Global access to ethnic Music: the next big challenge?' Proceedings ISMIR 2009, Kobe, Japan. http://www.columbia.edu/~tb2332/fmir/Papers/Moelants-fmir.pdf - Moelants D., Cornelis O., Leman M., 'Exploring African tone scales', Proceedings ISMIR, (Kobe, Japan), 2009 - Lesaffre, M., Cornelis, O., Moelants, D., Leman, M., 'Integration of Music Information Retrieval techniques into the practice of ethnic music collections', Proceedings Unlocking Audio 2 (London), 2009
2008 - Antonopoulos I., Cornelis O., Moelants D., Leman M., Pikrakis A., 'Music Retrieval by Rhythmic Similarity applied on Greek and African Traditional Music', Proceedings, ISMIR 2007 - De Mey M., Cornelis O., Leman M., 'The IPEM_EME: a Wireless Music Controller for Real-time Music Interaction', Proceedings ICAD 2008, Paris - Cornelis O., De Mey M., Leman M., 'EME: a wireless music controller for Real-Time Music Interaction', Proceedings ARTECH 2008, Porto
![Page 50: Denktank 2010](https://reader037.vdocuments.net/reader037/viewer/2022102813/5411e20b8d7f72c9738b4659/html5/thumbnails/50.jpg)
Compositions
- Shaken- Con Golese- Nelumbo
- EME experiment:- ICAD- ARTECH- RESONANT- E-Dinges
![Page 51: Denktank 2010](https://reader037.vdocuments.net/reader037/viewer/2022102813/5411e20b8d7f72c9738b4659/html5/thumbnails/51.jpg)
Doctoral Schools
![Page 52: Denktank 2010](https://reader037.vdocuments.net/reader037/viewer/2022102813/5411e20b8d7f72c9738b4659/html5/thumbnails/52.jpg)
Website
http://webs.hogent.be/olmocornelis/index.html
http://tarsos.0110.be
![Page 53: Denktank 2010](https://reader037.vdocuments.net/reader037/viewer/2022102813/5411e20b8d7f72c9738b4659/html5/thumbnails/53.jpg)
30%
- Coaching students in Master
- Follow up English Master students
- Guide Potential PhD students
- New means:- Hercules- TETRA- COST- FET Open
![Page 54: Denktank 2010](https://reader037.vdocuments.net/reader037/viewer/2022102813/5411e20b8d7f72c9738b4659/html5/thumbnails/54.jpg)
http://tarsos.0110.be
http://music.africamuseum.be
http://webs.hogent.be/olmocornelis