Skip to content Skip to navigation


You are here: Home » Content » 4 - Methodology


Recently Viewed

This feature requires Javascript to be enabled.

4 - Methodology

Module by: José Menezes. E-mail the author

4 Methodology

For this research two studies were conducted: Study 1 focused on performance processes in free improvisation and Study 2 on performance ideologies and backgrounds of the improvisers. Separate accounts of these studies are given. This choice of methodology takes inspiration from previous research on standard jazz performance (Schögler, 2003 and 1999; Gibson, 2006; Dorffman, 2005 and 2008). The present study makes application of the methodological approach employed in these studies to free improvisation.

4.1 STUDY 1

Performance Characteristics of Free Improvisation

In this study an improvised performance by expert musicians was analysed using automatic feature extraction from audio, focusing on the "best moments" of performance. Each one of the musicians was separately questioned about which moments they considered to be the best. By overlapping these moments a series of six segments was found. These segments represent the moments unanimously considered by the musicians as “best moments”. The musical features of these segments were extracted using Mirtoolbox.

The portfolio of data presented here comprises the audio recording of the complete performance with sound separation of microphones (each microphone in a separate channel); the video recording of the performance; the quantitative data extracted by Mirtoolbox.

4.1.1 Participants

For this study three male musicians aged from 37 to 51 were recruited, all of them regular practitioners of free improvisation.

The criteria for approaching and inviting musicians for the project were not primarily guided by instrumental proficiency or command of any specific musical language (classic, jazz or other). The processes of interaction in free improvisation can function meaningfully even between improvisers with different levels of instrumental skill. (Stenström, 2009: 309). My choice of a trio setting was guided by the communicational possibilities of this formation, which enables dialogue of each improviser with two different musicians and narrative lines. The choice of musicians to take part in this study was oriented primarily by their attitude towards free improvisation, their life-long dedication, their seriousness and commitment to the genre. With this in mind I contacted PC, an active saxophone and flute player, a professional musician dedicated almost exclusively to free improvisation, with several CDs released under his own name and working in the area of composition for theatre, cinema and animation. As I gave him a very general explanation of my study, he willingly accepted to participate. It was left to him to choose the other elements of the trio. He chose BP, a clarinet player and MM, a cellist, both of whom PC regularly plays with and who are equally committed to the genre. I contacted BP and MM in order to give them a brief explanation of the project and to ask for their collaboration and permission, which was immediately granted. It was agreed with PC that, if the recording quality was of an acceptable level and the musical outcome pleased the musicians, the master recording of the concert would be granted to PC for future edition and CD release. In order to keep the naturalistic surroundings of the project a token entry charge of 5 Euros was fixed. Moreover, it was agreed that the total amount coming from the entries would be handed to the musicians at the end of the concert. None of the musicians participating in the project received any other financial incentive to take part.

4.1.2 Materials

A 34 minute improvised performance by the trio was audio and video recorded. As we are dealing with totally improvised music the performance had no prescribed score. No pre-composed materials had been prepared and there was no rehearsal material for this performance. Besides alto sax PC doubled on soprano sax and flute. BP played soprano and alto clarinets

4.1.3 Procedures

My first concern was towards creating an environment for data collection that was as naturalistic as possible. I decided on organizing a concert in a Lisbon venue, a small cultural centre ( where regular art exhibitions, conferences, film festivals and concerts regularly take place. Art rock and free improvisation have a regular presence in their small concert room. This way I tried to avoid bringing to the moment of performance any interfering elements that could arise from playing freely improvised music in front of an audience not acquainted with the genre. Moreover, the room, seating an audience of 40 people from a total of 80, offers video recording facilities of semi-professional standard that would be of practical use for the project. When contacting the venue’s management, I informed them about my reasons for organizing of the concert. The project got instant permission and support.

Considering the relatively short performance (34 minutes) and considering that people could go in and out of the concert room, no intermission was planned. Moreover, it would interrupt the flow of improvisation and would bring to the performance disruptive elements of a social character such as the audience reaction, expressed opinions about “How are you enjoying this?”, the players perception of critics or fellow musicians in the audience; all elements that could alter the creative flow of performers.

In the days following the performance, and prior to the personal interviews to be included in Study 2, the video recording of the concert in mpeg format was sent to each player, asking them to identify in the video timeline the “best moments” in the concert. No indication was given about what “best moments” could mean. From overlapping the musicians’ individual answers, a series of consensual “best moments” was found. Those moments would be the focus of further qualitative and computational analysis.

This way points of qualitative change have been identified, points where high intensity of communicative interaction is required (Schögler, 1999: 81). In order to understand how “best moments” differ from other points in performance, some musical features were chosen and quantitatively analyzed using Music Information Retrieval (MIR) techniques. MIR is the interdisciplinary science of retrieving information from music and allows a musical document to be described by a set of features that are directly computed from its content (Orio, 2006: 2). Selected excerpts of the audio and video recording were analysed through this method. From this combination of data analysis methods I intended to provide a cross-examination between the subjective impressions reported by the players regarding interaction and communication and the objective changes in audio signal reported by MIR analysis.

  • 4.1.4 Apparatus

Audio was recorded by a professional technician with his own professional hardware. This ensured both the quality of equipment and the quality of audio recording. The audio data of the performance was recorded into a Mac laptop using Apple Logic Pro 8 software via a Motu 828mkII audio interface and Focusrite OctoPre 24-bit/96 kHz ADAT Card.

Table 1 – microphones used

Table 1
# Instrument Microphone Type
1 Flute Studio Projects C4 cardioid
2 Alto andSoprano saxes Shure SM57 cardioid
3 Cello AKG 414 TLII direct input from cabinet hypercardioid
4 Soprano clarinet Shure SM57 cardioid
5 Alto clarinet Shure SM57 cardioid

Microphone placement was a major technical concern, on which depended not only a good sound quality but more importantly the best possible separation between instruments. Aware of the fact that, with acoustic instruments playing close one another, a complete separation between instruments was impossible, cardioid microphones were chosen and special attention was given to microphone placement.

Microphone 1 was placed close to the embouchure hole of the flute.

Microphone 2 was used to capture both alto and soprano saxophones. When capturing the alto, it was placed close to the saxophone bell. When the soprano saxophone was being used, this microphone was placed between the player’s hands in a central position in relation to the horn. The cello was captured by microphone 3, placed near one of the f-holes and also by direct input from the amplifier used by the player for his own monitoring. Both clarinets were captured by mic 4, placed in the central part of the horn and mic 5 pointing at the instrument’s bell. A problem was detected during the greater part of the performance: the body movements of the musicians and consequent placement of instruments in and out of the capture range of the microphones constituted a problem during the whole performance and created fluctuations of quality and separation in the recording. This problem was especially noticed on the woodwind instruments.

The performance was video-recorded with a tripod-mounted high-definition Sony HDSr5 camcorder. The camera was place in a fixed central position in front of the stage. In order to achieve a smaller file size, and since the high definition recording was not essential for this project, the recording was converted to a mpeg file with the following specifications:

Size: 2.41Gb

Duration: 35’ 43’’Video: NTSC DV, 29,97 fps, with a resolution of 720x480Audio: 48000Hz, 16 bit, stereo

  • 4.1.5 Data Analysis

Musical features from the performance were extracted using Mirtoolbox 1.2.3 (June 2009 version). This Music Information Retrieval (M.I.R.) is a MatLab toolbox developed by Olivier Lartillot, Petri Toiviainen and Tuomas Eerola at the Department of Music of the University of Jyväskylä in Finland. It is conceived in the context of the “Tuning the Brain for Music” project financed by the European Union (FP6-NEST). It is free, open source software that can be downloaded from the developer’s webpage ( offers an integrated set of functions dedicated to the extraction of musical features from .wav and .au files. Its design is based on a modular framework whose building blocks form a basic vocabulary which can be freely articulated in new original ways. Before data extraction all the .wav files analysed in this study have been normalised, meaning that the amplitude of all audio files was increased to the maximum level without the introduction of any distortion.

Among the questions concerning the differences between “normal” moments in performance and those considered the “best” by the musicians, some become prominent: Are “best moments” louder or quieter then the rest of performance? Do they comprise a greater or lesser density of notes? How does the timbral quality of the group vary in those moments? In search for answers to these questions four musical features were considered relevant and are analysed in Study 1. They encompass different areas: Intensity, rhythm activity and spectral analysis. These features are:

  1. 1) RMS energy (Root Mean Square Energy) is a feature in the area of musical dynamics; it was computed by the mirrms operator in Mirtoolbox and indicates the global energy of the audio signal.
  2. 2) The mireventdensity operator estimated the average frequency of events, i.e. the number of note onsets per second, a rhythm feature.
  3. 3) The high level analysis operator mirentropy in MIRToolbox returns the relative Shannon entropy, a value used in information theory which is “a measure of the amount of information the signal carries” (Shannon and Weaver, 1949 cited in Camarena-Ibarrola and Chavez, 2009: 5). The feature whose entropy is analyzed in this study is the composition of audio spectrum, by default in MIRToolbox. The relevance of spectral entropy as perceptual feature is stated by Camarena-Ibarrola and Chavez who build the construction of a robust audio-fingerprint model in this feature (2009).
  4. 4) Spectral centroid is a musical feature that represents the geometric centre of distribution of the audio spectrum. This feature is a good predictor of perceived brightness in sound especially when studying “bands and ensembles where there may be many notes of different timbres being played” (Schubert, Wolfe and Tarnopolsky, 2004: 656). It is computed by the mircentroid operator in Mirtoolbox.

The values found for these features in the so-called “best moments” were to be compared with the values found for the same features in the larger sections where these “best moments” occurred. Since saxophone and clarinet were recorded using two microphones I decided to consider for my study the mean value of data extracted from these microphones. Since the cello was recorded using direct input and a microphone, I opted to only consider for my study the data extracted from direct input. This way, and although a hyper-cardioid microphone was used to record the cello, I tried to avoid sound leakage.

  • 4.2 STUDY 2

Performers’ backgrounds and performance ideologies

In this study I expected to grasp the performers’ musical background, their perspectives about the experience of making music together and their views about what that particular concert meant to each one of them. This study includes interviews conducted with improvisers involved in Study 1. Data collected during these interviews was subsequently submitted to qualitative analysis. The portfolio of data for Study 2 includes the audio recorded interviews with the musicians and the transcription (in Portuguese) of those interviews.

4.2.1. Participants

The trio of male improvisers involved in Study 1: PC, playing saxophone and flute; BP, playing clarinets (soprano and alto) and MM, a cellist.The average age of interviewees is 46 years old and the average time of practice is 35.3 years. The average age they began playing music was 10.6 years.

4.2.2. Materials

The interview questions were structured around four different topics: a) questions 1 to 7 refer to identity issues and personal musical history; b) questions 8 to 12 regard performance issues; c) questions 13 to 20 focus on the particular performance recorded for study 1; d) the audience role in performance is addressed in questions 21 and 22.

The interview’s complete set of questions can be found in Appendix A.

4.2.3 Procedures

Subsequently to the performance each one of the musicians were individually interviewed. In each of these meetings, and given the semi-structured nature of the interview, the set of questions was the point of departure for an extended conversation about the topics. After this, I invited each musician to watch the video recording of the performance. At any moment the interviewee could stop the video and enlarge upon how he perceived that particular moment in the music. Special attention was given to the moments considered by the musician to be the “best”.

The interviews were audio recorded for subsequent transcription.

  • 4.2.4 Apparatus

Interviews were recorded with a Sony MDWalkman MZ-R70 mini-disk and ECM-MS907 Sony microphone.


Almeida e Costa J., S. and Sampaio e Melo A. (1999). Dicionário de Língua Portuguesa. Porto.

Bailey, D. (1992) Improvisation - Its Nature and Practice in Music, Da Capo Press, Inc.

Berliner, P. (1994). Thinking in Jazz - the Infinite Art of Improvisation. Chicago and London, University of Chicago Press.

Bloom , J.and Porter L. (2007). "Creativity in Jazz Improvisation." The multidisciplinary study of imagination. Available at [Accessed 30th October 2008]

Borgo, D. (2002). "Negotiating Freedom: Values and Practices in Contemporary Improvised Music." Black Music Research Journal22 (2): 165-188.

Borgo D. and J. Goguen. (2005). "Rivers of Consciouness: The Nonlinear Dynamics of Free Jazz." Jazz Research proceedings Yearbook25.

Boyle, P. (2002). “The Eternal Now: Music, Creativity, and Improvised Experience”. Available at [Accessed 1st October, 2009]

Burrows, J. B. (2004). "Musical Archetypes and Collective Consciousness: Cognitive Distribution and Free Improvisation." Critical Studies in Improvisation / Études critiques en improvisation 1(1).

Burrows, J. (2004a). Resonances: Exploring Improvisation and Its Implications for Music Education. Faculty of Education, Simon Fraser University. Doctor of Philosophy.

Butterfield, M. (2007). Interaction, Improvisation and Interplay in Jazz (Book Review), Department of Music, Franklin and Marshall College: 239-249.

Camarena-Ibarrola A. and Chavez E. (2009). Robust Audio-Fingerprinting With Spectral Entropy Signatures. Proceedings of CIARP14th Iberoamerican Conference on Pattern Recognition Guadalajara, Mexico.

Csikszentmihalyi, M. (1990) Flow: The Psychology of Optimal Experience. New York: Harper Collins.

DeVeaux, S. (1991). "Constructing the Jazz Tradition: Jazz Historiography." Black American Literature Forum25(3): 525-560

Dogantan-Dack, M. (2006). "The body behind music: precedents and prospects." Psychology of Music34(4): 449-464.

Dorffman, M. (2005). Groove! It's production, perception and meaning in Jazz. Music. Sheffield, University of Sheffield. Ma: 85.

Dorffman, M. (2006). Groove within a jazz rhythm section: A study of its production and control. Proceedings of the 9th International Conference on Music Perception & Cognition (ICMPC9), Bologna / Italy, The Society for Music Perception & Cognition (SMPC) and European Society for the Cognitive Sciences of Music (ESCOM).

Dorffman, M. (2008). Feeling the groove: shared time and its meanings for three jazz trios. Department of Music. Milton Keynes, UK, Open University. PhD 350.

Friberg, A. and Sundström A. (2002). "Swing Ratios and Ensemble Timing in Jazz Performance: Evidence for a Common Rhythmic Pattern." Music Perception19(3): 333–349.

Gabrielsson, A. (2003). "Music Performance Research at the Millennium." Psychology of Music31(3): 221-272

Gibson, W. (2006). "Playing in the field: participant observation and the investigation of intersubjective knowledge in jazz improvisation." Anthropology Matters Journal8(2).

Heffley, M. (2005). The Body as Generator, Analyst and Theorizer of Free Improvisation. Proceedings of the Leeds International Jazz Conference, Leeds.

Hyatt, Q. M. (2002). In praise of Michelle Cliff's Créolité. Raleigh, North Carolina State University. Master of Arts: 62

Horowitz, D. (1995). Representing Musical Knowledge in a Jazz Improvisation System. Proceedings of the International Workshop on Artificial Intelligence and Music, Montreal, Canada, Academic Press.

Iyer, V. (2002). "Embodied Mind, Situated Cognition, and Expressive Microtiming in African-American Music." Music Perception19(3): 387–414.

Iyer, V. (2004). "Improvisation,Temporality and Embodied Experience." Journal of Consciousness Studies11(3-4): 159–173.

Iyer, V. (2004a). Exploding the Narrative in Jazz Improvisation. Uptown Conversation: The New Jazz Studies. B. H. Robert G. O'Meally, Edwards, & Farah Jasmine Griffin. New York, Columbia University Press.

Johnson-Laird, P. N. (2002). "How Jazz Musicians Improvise." Music Perception19(3): 415–442.

Kaikko, H. (2008). Not Rehearsing but Performing Free Improvised Music. Proceedings of the "Studying Musical Performance” Symposium.

Kastin, D. (1996 - 2007). "Lee Konitz: Back to Basics." July 23th 2007, Available at [Accessed 30th October 2008]

Kuhl, O. (2004). "Phrase, gesture and temporality: a cognitive perspective on jazz improvisation." , Available at [Accessed September 29th 2004]

Kühl, O. (2006). The semiotic gesture. Proceedings of the Second International Conference on Music and Gesture, Manchester.

Larson, S. (2002). "Musical Forces, Melodic Expectation, and Jazz Melody." Music Perception19(3): 351–385.

Lewis, G. (1996). "Improvised Music after 1950: Afrological and Eurological Perspectives." Black Music Research Journal16(1): 91-119.

Liebman, D. (1991). A Chromatic Approach to Jazz Harmony and Melody. Rottenburg, Germany, Advance Music

Lindenberger U., Shu-Chen L., Gruber W. and Müller V. (2009). "Brains swinging in concert: cortical phase synchronization while playing guitar." BMC Neuroscience10(1): 10-22.

Mendonça, Wallace W. and Beroggi G. (2002). Development of a Decision Logic to Support Group Improvisation: An Application to Emergency Response. Proceedings of the 35th International Conference on System Sciences, Hawaii

Mendonça, D. (2002a). "Decision Support for Improvisation: Prospects, Challenges and Opportunities." Available at[ Accessed 8th November, 2008]

Monson, I. (1996). Saying Something - Jazz improvisation and interaction. Chicago, London, The University of Chicago Press.

Montuori, A. (2003). "The complexity of improvisation and the improvisation of complexity: Social science, art and creativity." Human Relations565(2): 237-255.

Nettl, B. (1998). In the course of Performance - Studies in the world of musical improvisation. Chicago and London, The University of Chicago Press.

Nunn, T. (1998). Wisdom of the Impulse: On the nature of Musical Free Improvisation, published by the author, pdf version (IIMA) 2004.

Orio, N. (2006). "Music Retrieval: A Tutorial and Review." Foundations and Trends in Information Retrieval1(1): 1-90.

Pressing, J. (1987). Improvisation: Methods and Models. In Generative processes in music. J. Sloboda. Oxford, Oxford University Press.

Pressing, J. (1998). Psychological constraints on improvisational expertise and communication. In In the Course of Performance: Studies in the World of Musical Improvisation. Bruno Nettl with Melinda Russell. Chicago, University of Chicago Press: 47-67.

Ramalho G. and Ganascia J. (1994). Simulating Creativity in Jazz Performance. Proceedings of the Twelfth National Conference on Artificial Intelligence (AAAI'94), Seattle.

Reinholdsson, P. (1998). Making Music Together: an Interactionist Perspective on small-group Performance in Jazz. Department of Musicology. Uppsala, Sweden, Uppsala University. Doctor of Philosopy: 440.

Sansom, M. (2001). "Imaging Music: Abstract Expressionism and Free Improvisation." Leonardo Music Journal11(Not Necessarily "English Music": Britain's Second Golden Age): 29-34.

Sansom, M. J. (2005). Understanding Musical Meaning: Interpretative Phenomenological Analysis and Improvisation. Proceedings of the British Forum for Ethnomusicology Annual Conference – Music and Dance Performance: Cross-Cultural Approaches London, UK.

Sansom, M. (2007). "Improvisation and Identity: A Qualitative Study." Critical Studies in Improvisation / Études critiques en improvisation3(1).

Sawyer, K. (1992). "Improvisational Creativity: An Analysis of Jazz Performance." Creativity Research Journal5(3): 253-263.

Sawyer, K. (2002). "Improvisation and Narrative." Narrative Inquiry12(2): 319-349.

Sawyer, R. K. (2006). "Group creativity: musical performance and collaboration." Psychology of Music34(2): 148-165.

Schögler, B. (1998). "Music as a Tool in Communications Research." Nordic Journal of Music Therapy7(1): 40-49.

Schogler, B. (1999). "Studying temporal co-ordination in jazz duets." Musicae Scientiaespecial issue 1999-2000: 75-91.

Schubert, E., Wolfe, J. and Tarnopolsky, A. (2004) Spectral centroid and timbre in complex, multiple instrumental textures, in Proceedings of the International Conference on Music Perception and Cognition, Evanston, Illinois

Seddon, F. (2005). "Modes of Communication during jazz improvisation." British Journal of Music Education22(1): 47-61.

Stenström, H. (2009). Free Ensemble Improvisation. Faculty of Fine, Applied and Performing Arts. Gothenburg, University of Gothenburg. Doctor of Philosophy.

Weick, K. E. (1998). "Improvisation as a Mindset for Organizational Analysis." Organization Science9(5): 543-555.

Zagorski-Thomas, S. (2007). "The Study of Groove." Ethnomusicology Forum 16(2): 327 - 335.

Content actions

Download module as:

Add module to:

My Favorites (?)

'My Favorites' is a special kind of lens which you can use to bookmark modules and collections. 'My Favorites' can only be seen by you, and collections saved in 'My Favorites' can remember the last module you were on. You need an account to use 'My Favorites'.

| A lens I own (?)

Definition of a lens


A lens is a custom view of the content in the repository. You can think of it as a fancy kind of list that will let you see content through the eyes of organizations and people you trust.

What is in a lens?

Lens makers point to materials (modules and collections), creating a guide that includes their own comments and descriptive tags about the content.

Who can create a lens?

Any individual member, a community, or a respected organization.

What are tags? tag icon

Tags are descriptors added by lens makers to help label content, attaching a vocabulary that is meaningful in the context of the lens.

| External bookmarks