Hyeon Min Kim (aka Una Lee)

"ALTERNATIVE” STORYTELLING : spoken voice and narrative in sound-based works

2019

Supervisors:

This is a portfolio-based PhD thesis within the academic framework of Sonic Arts. It illustrates how sound-based artistic practice can be employed to convey fictional narratives within a symbiotic exchange with each other. It consists of a portfolio as the main body of submission and an accompanying commentary elaborating the ideas and thoughts embedded in the works. The portfolio consists of 9 sound-based works, which are enhanced by other various artistic media and deliver stories written by myself: some of these stories were invented in order to foreground the technological concept implemented within each respective work, while in others the technological means was employed based on the scenario of the work. As a collection of works which sit amongst multiple artistic disciplines such as creative writing, theatre-making, filmmaking and performance art, this thesis will prospectively inform sound-based arts practice about the potential of integrating and utilising fictional narratives and concurrently communicate alternative ways to deploy sound to narrative-based disciplines. 

Pablo Sanz

Environmental Composition Approaches. A Portfolio of Works

2018

The portfolio features ten major projects grouped into four clusters attending to their primary modes of listening. The first cluster corresponds to extensive fieldwork in the central Brazilian Amazon and its primary outcome, the immersive durational ambisonic project series entangled (xirminja nahpy benry). The second group corresponds to m·e·s·h [earthworks], a long-term body of work featuring pieces for intimate headphone listening which are manifested in releases, broadcasts and public events in various contexts, and as dedicated gallery-based exhibitions. The portfolio includes twenty pieces from this ongoing project series. The third cluster features four site-determined temporary sound works in architectural and public spaces created in response to their specific conditions and situations. The four projects within this cluster are: !!! (volumen), La Casa Encendida, Madrid (2014); kazum®, Rhizzo: De Supermarkt, Amsterdam (2014); PS4, PS2 Gallery, Belfast (2016); and untitled (sharawadji), Fundación March, Madrid (2016-17). The fourth group features sonic experiences in a live performance event context, termed ‘immersive concerts’. The projects within this group are: [in]audible, spectrum (ståtic limbø), out of earshot, and twilight concerts. They have been performed at several events internationally.

The ecological thinking which permeates this body of work goes beyond the widespread ‘green’ frame of understanding, challenging notions of ecology and nature which take the environment as something exterior to the human and culture. Distinctions and hierarchies between animate/inanimate, human/ non-human, subject/object, natural/artificial, and knowing/experiencing are blurred in explicit and subtle ways in the projects. This practice proposes to move away from an anthropocentric understanding of the world, de-centring the human in favour of a concern for the non-human and encouraging truly ecological awareness. This ecological thought goes beyond strictly biotic understandings of vitality and the non-human, giving attention as well to the agencies of materials, energies (electricity, electromagnetism, light, sound), weather, architecture, and technology. Recognising vitality beyond the human makes it also possible to think sounds and works of art as phenomenological entities that in some strong sense display “something like agency and something like affect” (Morton 2017).

The modes of listening fostered in this portfolio emphasise intensified, sustained forms of attention, and enactive, non-prescriptive, participatory, sensuous, and direct felt experience. The works operate as catalysts for shifts in perception, able to function as transformative tools, as agents for change and personal investigation, cultivating alternative forms of being and thinking. Listening becomes a political act, intended to resist dominant tendencies in contemporary societies, moving away from anthropocentric, representational, narrative, utilitarian, and visual-centric paradigms.

The doctoral submission consists of a written account (thesis) alongside a standalone web application and original project materials stored on a USB flash drive. The thesis combines an ethos and critical framework with a description and reflection on process. The text sets out a context for the appreciation and assessment of the works, drawing from diverse practices and theoretical arguments, and mobilising ideas and speculations on the nature of perception and reality itself, art, music, ecology, listening, phonography and electroacoustic sound composition. The thesis engages with relevant literature, situating the practice within contemporary cultural ecologies (societal-political-philosophical), citing perceived controversies and concerns regarding prevailing listening practices, towards the outline of a personal artistic position. The description of process is understood expansively to embrace writing that corresponds to creative and technical strategies (involving fieldwork, studio work, and curatorial choices), commissioning and production contexts, and impressions on how the listeners-participants are engaged.

This research contributes the perspective of an artist working within a fluid community of practice, proposing alternatives to canonical assumptions and representations of creative practice engaged with environmental sound, and of environmentally-concerned and ecological art. Furthermore, the PhD submission offers a detailed and personal perspective on the contemporary condition of doing work and seeking its dissemination and viability while pursuing an ethical attitude which informs the works and related activities.

Robin Renwick

Topologies for Network Music

2017

Supervisor: Pedro Rebelo 

This thesis presents a body of theoretical research, combined with a portfolio of art works, residing within the field of network music. Network music is a musical practice in which conceptual, technological, ideological, and/or philosophical concepts of the network are included in the design, composition, production, and/or performance process.  This thesis contains critical analysis of three historical examples of network music, as well as critical reflection of three artistic responses. The analysis investigates how contemporary technologies allow increasingly complex perceptual and technological understandings of network concepts, ideologies, strategies, and topologies to be explored within network music. The thesis also contains description, critical reflection and analysis of one original artwork, which investigates a theme that emerged during the research process. The original work investigates how a performance topology, adapted from the field machine learning, alters the perception and interpretation of the network, as well as the impact it has on the involved musical agents and the performance process.

The works included with this thesis constitute the portfolio and they are the creation of the author.

Bryan Whitelaw

Franz Liszt's Piano Sonata in B Minor: Context, Analysis and Hermeneutics

2017

Supervisors: Aidan Thomson & Yo Tomita

Franz Liszt has all too often been discarded as the virtuosic showman, despite the fact that his orchestral works have often gained great praise and attracted scholarly engagement. However, one also finds striking development of formal design and tonal harmony in many of the works for his principal composition medium, the piano. This thesis seeks to explore the genesis of Liszt’s magnum opus for the instrument, the Sonata in B Minor, to uncover the time-frame in which the work was composed and from where its inspiration was derived.

Considering the historical analyses placed upon the work that deal with structural design, as it pertains to the paradigm of Classical sonata form, I wish to utilise a practical application of James A. Hepokoski and Warren Darcy’s Sonata Theory in order to survey the Sonata in B Minor’s landscape in a new light. In particular, I offer new perspectives on the validity of the double-function form, insight into the rhetorical layout of a rotational discourse, and propose a nuanced analysis befitting of this striking work.

Finally, I shall explore the implications of a hermeneutical reading of the Sonata to engage with the discursive meanings contained within the work as a whole. The composer’s literary influences and the evidence uncovered within the body of this thesis help to support a possible narrative that may have been a significant influence upon Liszt’s composition of this great work.

Diogo Alvim

Music Through Architecture - Contributions to an Expanded Practice in Composition

John D’Arcy

Portfolio of Compositions

2016

Supervisors: Michael Alcorn, Sinead Morrissey

This thesis provides a commentary on a portfolio of original compositions produced through a series of practice-as-research projects. These compositions explore the mediation of text in aural and situated experiences. The mediations of text (adaptations, translations and montages) are approached from the perspective of an artist working in the broad field of sound art producing performances, installations, locative-media and radiophonic work. This practice embraces interrelated disciplines primarily concerned with the production and reception of work in aural and situated modes of delivery, often composed for specific contexts in Northern Ireland. This commentary explains the motivation behind the research, makes explicit creative decision-making processes and contextualises the work within historical and contemporary artistic research and practice.

Miguel Cerdeira Marreiros Negrão

Parameter Field Spatialization: The development of a technique and software library for immersive spatial audio

2016

Supervisor: Michael Alcorn

This thesis describes parameter field spatialization, a novel technique for creating and controlling spatial sound patterns formed along a surface, when working with sound syn- thesis and signal processing in a computer-music environment. Its main purpose is the creation of spatially-dynamic immersive sources in electroacoustic composition. This technique generates multiple decorrelated signals from a given sound process definition, which when spatialized at different locations can create a single enveloping auditory event with large width and heigth. By modulating parameters of the sound process dif- ferently for each signal it is possible to create a spatial surface pattern. The modulation signals are generated based on a mathematical model which assumes a surface encom- passing all the loudspeakers and describes an abstract pattern in this surface through a mathematical function of time and surface coordinates, called parameter field. This research investigates whether parameter field spatialization can successfully create and precisely control spatial surface patterns, and how these patterns can be made into a compositional parameter in computer music.

The technique was implemented in ImmLib, a software library for the SuperCollider audio-synthesis environment. Several specific examples of the combination of para- meter fields with sound processes were investigated from a perceptual point of view, in listening sessions using two loudspeaker systems, one spherical and the other a vertical rectangular grid. A group of composers was invited to use the software for their own work, producing three pieces presented in public, which were analysed regarding the use of the technique. From this practical work findings relating to the most effective strategies regarding the use of parameter fields were outlined.

Eduardo Patricio

Spatial referentiality and openness: A portfolio of environmental sound compositions.

2016

Supervisor: Pedro Rebelo, Paul Stapleton

Through a creative portfolio and an analytical and critical commentary, this research investigates the use of spatial references in the composition of semi-open environmental sound works. The portfolio explores a number of strategies to make use of spatial references as formal compositional components to enable more intuitive performance/reading experiences. The pieces present a number of electronically mediated scenarios in varied formats; concert, installation and mobile application. Counting on the intuitive way one tries to constantly identify surrounding spaces, each piece uses physical (performance/presentation spaces) and representational devices (illustrations, maps, video projections, spatialised sound etc.) to articulate and delimitate semi-open artistic experiences. Such ambiguous scenarios are enabled by both the unpredictability of elements of each work and the dependence on the subjective interpretations of the agents involved in the process. The creative processes presented here in a descriptive, analytical and critical manner attempt to make an artistic contribution and provide documental material for future reflection about related practices. 

Tullis Rennie

Composition and Identity - A portfolio of context-based sound works following interdisciplinary ethnographic methods 

2016 

Supervisor: Simon Mawhinney 

This research explores matters of social, cultural and personal identity through composition. The portfolio for this PhD is made up of seven works: Carioca Sound Stories, Selarón: A Great Madness, Manifest, The Metropolis Sounds, Muscle Memory, Everybody (Wants To Be The DJ), and Walls On Walls. They are presented variously as fixed media compositions, a multi-channel sound installation, and a participative community project.

My practice-based research has developed a musical language that seeks a threshold between ethnographic-cultural documentation and composed sound arts practice. This commentary engages with how such an approach might affect attitudes towards a number of broader issues concerning sound composition. Key themes include: the act of listening; the practice of field recording; notions of soundscape composition; ethnography in sound; music and identity; sampling; participative approaches to sound art. 

Many of the portfolio works aim to serve their subject matter reciprocally; interacting with the particular sites, sources, interviewees and participants discussed through the works. As such, these compositions aim to provide cultural insights for myself, for others directly involved, and for the wider community of interest relevant to each piece.

This research presents how an ethnographic approach to listening, sound, music, and ‘the field’ has affected my compositional outlook and the representation of multiple identities within these works.

Emily Diane Robertson

‘Tones out of nowhere’: Situating the Design and Development of Graphic Notation for Network Performance

2016

Supervisors Pedro Rebelo, Franziska Schroeder

This thesis presents results from an ethnographic interview-based study of the practices and processes of composers who create graphic notation for networks performance.  Through these interviews, as well as literature reviews and case studies, this thesis examines the effects of factors in the network of music environment on the design and development of graphic notation and the ways in which composers frame, play with, and respond to the network in music performance.  This thesis also includes reflections on a portfolio of network music and graphic notation experiments conducted as part of personal artistic practice and a key component of this research process.

Koichi Samuels

Enabling Creativity: a study of inclusive music technology and practices at The Drake Music Project Northern Ireland

2016

Supervisors: Franziska Schroeder, Fiona Magowan

The Drake Music Project Northern Ireland (DMNI) is a charity with the aim of enabling disabled people to compose and perform their own music independently through the use of music technology. Thus, DMNI is a charity that works at the intersection of music, disability and technology. This research contributes to raising further awareness on the issue of inclusion and disability, and at the same time presents an example of a charity working on practice-based and technical solutions to transcending both material and social disabling barriers to music making. Interviews, observations and professional perspectives on DMNI techniques and inclusive music practices were gathered through a sixteen-month ethnographic study of the charity between 2013 - 2015.

In this thesis I explore the ways in which people produce exclusions and barriers to inclusion whilst using computer-based music technology. In addition, I argue that a music technology device’s potential to be used in accessible ways, or to be inaccessible to certain users is not determined by its design. Through practices of adaptation, or by creating assemblies of devices, even interfaces that are not matched to the specific requirements of a certain user can provide access to music making. I argue that a relational understanding of “independence” serves to reveal a layer of activity beneath simply the physical ability to perform musical actions unaided, and recognises that independence also exists in the choices and opinions of the individual. I argue that the practices of resistance to various barriers and constraints to music making at DMNI are highly improvisatory and creative. Moreover, looking at the practices of music making, and the design and adaptation of devices I discuss throughout this thesis, I argue that DMNI provides a space and platform for disabled musicians to exercise acts of resistance against individual, social and material barriers.

Michael Dzjaparidze

A Portfolio of Original Compositions using a Physically Inspired Sound Synthesis Approach

2015

Supervisors: Paul Wilson, Maarten van Walstijn

This thesis accompanies a portfolio of compositions and, in addition, discusses a number of compositional approaches which make use of physical modelling and physically inspired sound synthesis methods for the creation of computer generated music. To this end, a software library has been developed for the purpose of the real-time simulation of systems of inter-connected 1D and 2D objects to be assembled by the user. Although the developed software has proven to be indispensable for producing the creative results, the emphasis of assessing the research should be on the contribution of the portfolio to the field of musical invention, as the goal was not to add any novel technical knowledge to the  field of physical modelling. Instead, my aim was to explore in depth the creative possibilities of technical research carried out by others and to show that it can be utilised in a form which aids my own creative practice. From a creative perspective, it builds upon concepts and ideas formulated earlier by composers Jean-Claude Risset and Denis Smalley, centered around the interpretation of timbre and sound as constructs which actively inform compositional decision-making and structuring processes. This involves the creation of harmony out of timbre and playing with the source-cause perception of the listener through the transformation of timbre over time. Lastly, the thesis offers a discussion of gesture and texture as they commonly appear in electroacoustic music and tries to justify my own personal preference for focusing on the development of texture over time as a means to create musical form and function.

Fionnuala Fagan-Thiébot

The Sound of Memory: An artistic exploration of personal and cultural narratives in post-conflict communities.

2015

Supervisors: Paul Stapleton, Martin Dowling

Through a creative portfolio and reflective writing, this thesis explores how divergent sound-art practices can represent oral history narratives found in post-conflict societies such as Northern Ireland and Bosnia. Recently collected interview materials have been transcribed, edited and transformed through the application of a newly devised artistic methodology, which draws on verbatim theatre. This re-telling of personal narratives takes place through intertwining the practices of verbatim song-writing, sound design, live performance and installation art. The primary aim of this research is to increase access to the often forgotten knowledge found in individual experiences, thus augmenting more generalised historical narratives of marginalised communities.      

Peter Stitt

Ambisonics and Higher-Order Ambisonics for Off-Centre Listeners: Evaluation of Perceived and Predicted Image Direction

2015

Supervisors: Maarten van Walstijn, Stephanie Bertet

Higher Order Ambisonics is a spatial audio technique that aims to recreate a sound image over as large a listening area as possible. Only limited investigation has taken place into localisation with Ambisonics and Higher Order Ambisonics at off-centre listening positions. This thesis presents the results of three psychoacoustic localisation experiments investigating off-centre localisation of first and third order Ambisonics under different conditions: in acoustically damped studio conditions, investigating transient versus ongoing, non-transient stimuli for the third order system, and for a large array system with increased arrival time delay between loudspeakers. A detailed analysis of the results of each experiment is carried out to determine the robustness of the tested ambisonic systems and whether the variation is principally due to differences between listeners or through the variation of individual listeners. Comparisons are made between the results of the three experiments to determine the influence of changing the stimulus or increasing the arrival time delay between loudspeakers, where the relative gains of the loudspeakers was found to be perceptually more important than increased time differences between them. The significance and usefulness of these results can be increased by comparison with models for prediction of human localisation, where a robust model would afford fast evaluation of ambisonic systems and allow system optimisation for off-centre positions. Therefore, evaluation is performed for two binaural models on their ability to predict the results of the psychoacoustic experiments. This includes a model that has previously been used for prediction of off-centre listening positions for higher order systems and WFS. A modified version of the Lindemann model, which includes a precedence effect inhibition, is also evaluated, having previously been applied to two channel lead-lag experiments. Finally, the energy vector model, which is prevalent in the Ambisonics community, is extended to include elements of the precedence effect. The binaural models, the standard energy vector, previously untested for off-centre positions, and the precedence-extended model are evaluated by comparison to the perceptual results. The robustness of each of the binaural and vector models is discussed in the context of their use as predictors of localisation at off-centre listening positions. The predictions of the precedence-extended energy vector are found to exhibit the lowest deviation from the perceptual results.

Eduard Zatriqi

Portfolio of compositions

2015

Lawrence Zazzo

"Not local beauties": Handel's bilingual oratorio performances, 1732-1744

2015

Supervisors: Sarah McCleave

Patricia Alessandrini

Composition as interpretation through performative electronics

2014

Supervisors: Pedro Rebelo, Maarten Van Walstijn

The Introduction presents the notion of composition as an act of interpretation: all of the works of the portfolio are based on existing repertoire, which is thus seen to be 'interpreted' through Computer-Assisted Composition (CAC) processes, entailing the combination, modification, analysis, and transcription of recorded materials of the 'interpreted' work.

Chapter 1 documents the use of real-time physical modelling synthesis in the Max/MSP programming environment. In Funeral Sentences for ensemble and electronics, a 'string meta-instrument' is created through the combination of violin, cello, guitar, percussion - the latter two often bowed to better blend with the others - and physically-modelled strings. menus morceaux par un auto moi réunis for guitar and electronics combines real-time physically-modelled strings with physically-inspired stochastic models (PhiSM) of percussion instruments. In the electroacoustic series Nani, the noise content - such as breaths and consonants - of an a cappella recording is used to excite percussion models.

All of the works documented in Chapter 2 engage with visual manifestations of sound in addition to instruments and speakers, including transducing sound through objects and interactive video projection: in Adagio sans quatuor (2010), suspended plates are bent by motors to tune their resonant frequencies, while three-dimensional (3D) physical modelling simulates bending motion for one of the two instruments; Adagio pour l'absence (2010) maps data from a game environment to interactive networked performance including real-time score generation and 3D physical modelling synthesis; Mismoded (2011) features real-time concatenation of audio and video; and Gurre-Klänge (2012) combines video projected on resonating metal plates and other surfaces with interactive performance involving improvisatory feedback systems.

The Conclusion considers the compositions in the portfolio in terms of the provenance of sound in relation to visual and aural perception, and introduces future projects which continue in the direction of reinforcing visual correlates and the physical presence of sound.

Robert Casey

Embodied Music: A Portfolio of Original Composition

2014

Supervisors: Professor Michael Alcorn, Professor Pedro Rebelo

Niall Fredrick Coghlan

Physiological Correlates of Emotion as Interaction Channels with Artistic Applications: Artwork and Experiments

2014

Supervisors: Paul Stapleton, Ben Knapp

Links between music and emotional state have been posited for many years, dating back to Aristotle, and periodically taken up by musicians, musicologists, philosophers and most recently neuroscientists. Mounting evidence suggests that music is used by many as a form of mood regulation and there is a long history of music as therapy, usually realised as performance or group work, with more recent studies showing links between music listening and positive healthcare outcomes.

In addition to interest in the relationship between music and emotion, there has been a recent upsurge in research into emotion as a channel for technological and creative interaction with music, multimedia artworks, computers and new digital instruments. It is exploring this aspect of emotion-based interaction that provides the motivation for this thesis, a desire to utilise emotion (and the physiological characteristics associated with affect) as an interaction channel. In order to achieve this, further exploration of the associations between emotion, physiology and music is necessary.

To date studies in the field of music and emotion have faced a number of challenges, not least in the definition of and delineation between emotions, moods, feelings and other affective states. Other challenges include limited sample sizes in experimental work, a past focus on the western classical tradition and the difficulties involved in ecological assessment of emotion.

As context for my own work, this thesis initially presents an overview of the dominant theories relating to emotions, and in particular their interaction with music and the mechanisms and audio features by which music may induce emotion. An overview and critique of artworks utilising signifiers of affective state is also presented before detailing my own artistic contributions as co-creator of artworks utilising physiological correlates of emotion. An overview of a large-scale public experiment (Emotion in Motion) designed to explore and deepen knowledge of some these issues is presented, along with an analysis of the data collected. This analysis found relationships between specific musical features such as tempo, mode and dynamic range and participant experiences of factors relating to emotions such as valence, arousal and engagement.

The concluding chapter summarises the findings and examines potential directions for future research as well as applications of emotionally and physiologically aware technologies, including affective computing, healthcare and interactive artworks.

Gerard Gormley

Microsound: A Portfolio of Original Compositions

2014

Supervisors: Paul Wilson, Pedro Rebelo

This portfolio of six electroacoustic compositions and accompanying thesis present an investigation of microsound through a variety of lenses.

Microsound is defined as both sound particles consisting of extremely short transients, and also sounds containing low amplitude levels. In this research I investigate how considerations of microsound can influence and shape compositional decisions, and show how this influence can be traced within my own compositional approach. The six compositions that comprise the portfolio explore different aspects of dynamic and temporal microsound, as well as microsound in relation to concepts of noise. The compositions are: Surface (2010), a two-channel piece for tape that examines microsound in relation to soundscape and listening and recording spaces; Testure (2011), composed for a combination of two and five channels, investigates microsound via an exploration of masked

audio contained in instrumental sound sources; Bunker (2012) and Odessa (2012), interconnected compositions composed, respectively, for eight and sixteen channels, examine short transient audio fragments in the context of noise and rhythm; Mochorel (2013), a two-channel piece, which proposes the idea of ‘noisescape’ as a merging of soundscape composition and noise; and finally ParkerSuite (2013), a work for GPS-enabled mobile /phones that was created in the context of an umbrella Project, Belfast Soundwalks, in which compositions are mapped to real-world environments using mobile app technologies. This last composition explores naturally occurring microsounds within the environment, e.g. birdsong.

The six compositions in the portfolio are linked thematically as well as through the application of different microsound compositional techniques. The thesis shows how each composition developed both as a standalone work and in relation to the larger portfolio, and demonstrates a strong connecting thread between the different works. The commentary further explores in-depth the concepts and techniques explored in the portfolio, and situates these ideas within an historical context. Historical influences range from Futurist ideas of noise to John Cage’s concept of ‘small sounds’; the compositional ideas and methods of lannis Xenakis, Karlheinz Stockhausen, Morton Feldman, Jakob Ullman and others; and musicological and theoretical concepts put forward by Curtis Roads, Dennis Smalley, Kim Cascone, Joanna Demers, R. Murray Schafer, and others. Through the different compositional investigations and associated discussions I conclude that microsound offers a rich avenue through which to explore contemporary ideas and methods in electroacoustic music, and show how these ideas can be effectively applied in a variety of ways.

Mark Rossi

Abstracted Materials: A Composition Portfolio

2014

Supervisors: Eric Lyon, Pedro Rebelo

This commentary will discuss the main concepts behind the electroacoustic music in the composition portfolio: Abstracted Materials.

Produced between the years of 2008 to 2013 at the Sonic Arts Research Centre, Queen’s University Belfast, these acousmatic works will be framed within various contexts, from the practical to the theoretical. Occasionally, the analysis will centre upon the strategies and methods applied to the formal aspects of the compositions and also consider ideas associated with listening as an integral aspect of sound art practice.

During the time I attended the Sonic Arts Research centre, the research encountered many shifts: from the psychoacoustic aspects of a sound, or sound making activity as a source of inspiration, to the intuitive use of materials that formed my personal associations with a musical process. This document attempts to weave a narrative from these experiences. My goal is to explain these musical concepts and to faithfully record the process of producing this portfolio of acousmatic works.

Lilian Lima Simones

The roles of gesture in piano teaching and learning

2014

Supervisors: Franziska Schroeder, Matthew Rodger

The experience of engaging with music through listening, teaching and learning would be impossible without a bodily interface, through which movement and music can be physically produced, experienced and understood. Physical gestures form a central part of the communication established between the teacher-student dyad in the communication of symbolic and functional musical knowledge. Factors such as gesture types (forms and meanings) and their specific outcomes in the teaching and learning processes have been consistently overlooked in the instrumental music pedagogical context. This thesis prioritises such undervalued topics, focusing its enquiry upon piano teachers’ hand gestures used to communicate with students during the teaching process and incorporates theoretical frameworks from disciplines such as: music-psychology, psycholinguistics, gesture studies, gesture-led educational re-search, imitation and observational motor-learning.

Three investigations were carried out to investigate the role of teachers’ gestures in piano teaching and learning. The first two combined qualitative and quantitative approaches – the results of which were used in establishing the first known categorisation of piano teachers’ gestures. Amongst the most intriguing findings were the relationship between teachers’ didactic intent and the forms of gesture they employed, and ‘gestural scaffolding’ (when teachers adapted particular gestural communicative channels to suit specific student skill levels). In the third investigation an experimental setting was used to observe and evaluate the role of teaching gestures in one-to-one instrumental tuition. Here different gestural teaching and learning conditions yielded multiple levels of learning effectiveness, implying a need for empirical understanding and establishment of gestural performance as a concept that can be applied to enhance learning across specific pedagogical contexts. As well as building a case for future investigations in this research area, this thesis opens a debate within studies of pedagogical practice in instrumental music teaching, whilst contributing more generally to discussions of how the body impacts upon music understanding.

Felipe Hickmann

Territories of Secrecy: Presence and Play in Networked Music Performance

2013

Supervisor: Pedro Rebelo

This thesis presents practical work and theoretical research spanning topics in music composition, performance and new media. It proposes a particular creative approach to performance practices mediated by computer networks, which addresses issues of presence and liveness through the application of game-derived systems to performance settings. The thesis includes a portfolio of eight compositions; these pieces were designed and performed with the goal of implementing strategies proposed during theoretical reflection, while also suggesting new ideas for further study. 

The theoretical component of this work investigates the notion that performance is affected in its attributes of presence and liveness whenever reproduction and mediation technologies come into play. Building on the views of media theorists and research in psychology, it identifies the roles played by agency and social cues in the perception of presence. After reviewing contemporary artistic practices in which participation is enabled by open and playful performance situations, the thesis proposes the use of game systems as a strategy for negotiating musical play in networked settings. 

The notion of secrecy provides a key conceptual reference that guides both theoretical enquiry and creative exploration. This investigation argues that every act of mediation entails an opportunity for selective concealment of ideas and actions, and therefore comprehends a wide creative potential. While music performance over computer networks may entail the loss of cues that are often taken as granted in ensemble play – such as shared pulse, breath and bodily communication – it may also suggest innovative modes of engagement, based on the regulation of information rather than its unimpeded disclosure. This premise is explored throughout the portfolio, which borrows ideas from contemporary social practices of secrecy and performativity.

Javier Jaimovich 

Emotion Recognition from Physiological Indicators for Musical Applications

2013

Supervisors: Ben Knapp, Maarten van Walstijn

This thesis investigates the emotional response of audiences to music via physiological indicators, with the goal of creating interfaces that use emotional state for music control, specifically in performance scenarios via electrodermal and cardiovascular measures. In the past two decades, multiple disciplines have shown interest in studying the relationship between music, emotion, and its physiological manifestation. However, despite the increasing attention, the actual mechanisms on how music modulates human emotion and how this correlates with physiological changes are still not well understood. Therefore, this topic provides interesting challenges to determine if musical emotions can be measured from audiences in ecological environments via physiological signals. In order to address this, there are several questions that need to be resolved; including how to measure physiological indicators of emotion in concert environments, what level of shared responses and variance can be expected from public audiences, and how to assess the induction of musical emotions on listeners. In order to answer these questions, the work in this thesis starts by measuring physiological indicators of emotion in music concerts, revealing high correlations between the physiology of performers and audience members, as well as associations between physiological changes and structural and acoustical features of the music. In order to assess felt emotion on listeners and how these are manifested via changes in physiology, a series of modular public listening experiments (Emotion in Motion) were implemented in Dublin and New York, collecting physiological data and self-report measures of emotion from over 4000 participants. Analysis of this database reveals a set of specific physiological indicators that show significant relationships with musically induced emotions. This thesis also contributes robust feature extraction tools for Electrodermal Activity and Heart Rate, and a methodology for synchronization of multimodal signals for musical performance research.

Sarah Orr

Numerical Simulation of Coupled String Vibrations with Application to Physics-Based Sound Synthesis

2013

Supervisors: Maarten van Walstijn, Colin Cowan

The work in this thesis presents and discusses new techniques for modelling string instrument vibrations for the purpose of physics-based sound synthesis. The work focuses on coupled systems, where phenomena such as sympathetic vibrations should naturally occur. Finite difference (FD) methods are chosen for modelling string vibrations for their flexibility both in terms of local adjustments and possible extensions to non-linearities. The resonating body is represented using a modal formulation. Modelling the body in such a way has the advantage of scalability which can improve efficiency. This is shown to be possible without affecting the overall timbre of the sound. The models were formulated using idealised shapes such as the beam and plate, but the modal formulation is general for all linear systems. A technique for interfacing the FD model of the string to the modal formulation of the body is presented. In this way the advantages of both methods are exploited, improving the balance between accurate and efficiency. Initially, this coupling is formulated using only transverse motion but is then extended to include longitudinal motion. In simulations of a harp-like instrument where the strings are coupled to the body at an angle, results obtained with numerical experiments show that including longitudinal vibrations impacts the eigenmodes of the system and prove essential for accurately modelling sympathetic vibrations. Comparisons with previous studies validate these results. By applying the proposed method to model resembling a simplified piano, online changes to parameters such as soundboard density further demonstrate the proposed technique.

Dionysios Athinaios

A Portfolio of Original Compositions

2012

Andrés Cabrera

Control of Source Width in Multichannel Reproduction Through Sinusoidal Modelling

2012

Supervisors: Dr. Gary Kendall, Dr. Stéphanie Bertet

This document describes the development of a new sound processing technique to achieve increased perceived spatial  extent in multichannel reproduction through artificial decorrelation. This technique produces multiple decor related copies of a sound source, which when played back over a multichannel system produce an enhanced sense of space. Decorrelation is achieved through random modulation of the time-varying sinusoidal components of the original signal's spectrum, extracted using a modified version of the Loris sinusoidal modelling technique.

Modulation is coherent within critical bands of each output channel and can be applied to both frequency and amplitude to allow maximising effect while minimizing rtifacts. The amount of decorrelation between channels can be controlled through the coherency of the modulators, allowing for control of image spread and width. A comparative objective study was performed to determine the effects of the technique on IACC and models predicting spatial extent. Results were compared to decorrelation through random-phase filters. A cross-platform desktop application implementing the technique was developed.

Liam Gorry

Handel and his accompanied recitatives

2012

Supervisors: Sarah McCleave, Jan Smaczny

This dissertation provides the first comprehensive study ever undertaken of Handel’s accompanied recitatives. It traces the evolution of the medium within his vocal works from 1704 up until the composer’s final English language works in the 1750s. Using music manuscripts, wordbooks, the early printed editions and documentary evidence a detailed picture of Handel’s musical and dramatic choices relating to his accompanied recitatives comes into view.

The principle singers for whom Handel’s accompanied recitatives were originally written and the differences between accompanied recitatives in acted and un-acted works are discussed in detail, and possible influences on the young Handel, with regard to accompanied recitative, are also considered.

The connection between accompanied recitative and the supernatural in the composer’s vocal works is also discussed, and his use of accompanied recitative to distort closed forms and his reasons for doing so are explored. The problems relating to the terminology of eighteenth-century accompanied recitative are discussed at length, and a history of the medium is provided.

Giovanni de Sanctis 

In-System Parametric Calibration for Two-Microphone Wave Separation in Acoustic Waveguides 

2012

Supervisors: Maarten van Walstijn, Colin Cowan

The separation of waves in acoustic waveguides can be used to infer information from various acoustic systems. When applied to musical wind instruments, it allows novel techniques for in-depth investigation of the physics of the instrument, including the extraction of playing information in wind instruments, bore reconstruction and impedance measurement. Although a few approaches to this problem can be found in the literature, little attention has been given to the calibration of the system and above all to the assessment of its performance. The proposed methodology is based on a frequency-domain optimisation of a small set of parameters that best describe the system. The nonlinear optimisation problem that arises is formulated in order to exploit the considerable amount of a-priori knowledge given by the theory of propagation in ducts. The resulting cost surface is typically characterised by several local minima; however, the initial guess given by the nominal values for the parameters ensures the convergence to the desired solution. The main feature of the optimisation approach is that it can be applied in system and does not require any special calibration apparatus. As a consequence, it is also possible to track slow variations in the propagation due to changes in temperature and humidity of the medium during normal operation. A procedure for the assessment of a wave separation algorithm is also proposed; this shows that, under the same conditions, the optimisation approach improves the performance of the system. Finally, the proposed frequency-domain optimisation is also suitable for other applications such as in-air direction of arrival estimation. Preliminary results on these applications are also presented.

Andrew Dolphin

A portfolio of original compositions

2011

Supervisor: Paul Wilson

A collection of original works are presented that explore a broad spectrum of compositional possibilities.

Five open form non-linear works are introduced as sound toys. These are designed to be exploratory and playful compositional frameworks, or systems for composition, that are relevant to the fields of sound art and electroacoustic music. Compositional decisions are controlled or influenced by an end user (or player) in these playful open form works. Multidisciplinary compositional approaches and alternative mediums for the presentation and dissemination of electroacoustic artworks are investigated.

Five fixed media electroacoustic pieces are presented. Ideas of kinesis and kinetic energy as an external compositional agent, and multichannel spatial approaches are central to these works.

Cross-pollination of aspects of both aesthetic and technical concerns of each discipline are explored, and these fields have been allowed to feed each other.

Matt Green

In and Out of Context: Field Recording, Sound Installation and the Mobile Sound Walk

2011

Supervisor: Pedro Rebelo

This thesis chronicles the practical and theoretical research that has been undertaken in conjunction with, and in support of, the creation of a portfolio of original works. The portfolio comprises five works of sound art: two are sound installations, and three are what I call ‘mobile sound walks’. All five works are site-specific, meaning that they have been designed for a specific space, in response to a specific social and environmental context, and take into account the cultural, historic and political significance of the hosting site.

The first half of this thesis is dedicated to an in-depth discussion of the work of six artists with whom I align my own creative practice. These include the composers Pierre Schaeffer, Luc Ferrari and R. Murray Schafer, who are noted for their use of everyday sounds as compositional material; the sound installation artists Max Neuhaus and Bill Fontana, who create site-specific, environmental sound art works; and Janet Cardiff, a pioneer of mobile sound art.

In the second half of this thesis I provide a detailed analysis of my own work, with a particular focus on my mobile sound walks, which use location-aware technologies as a means to map sounds across a landscape. Such technologies have received little attention through existing sound art discourses. This thesis addresses this lack, by offering conceptual perspectives and methodological tools for understanding and producing what I call ‘locative soundscape composition’.

Throughout this thesis I draw upon a range of critical theories and method- ologies both from within, and from outside of, sound art studies, which help to shed new light on sound as it relates to, for example, the city, the politics of everyday life, the production of space and place, and the relationships between site-specific art, urban environments, and social actors.

Eoin Mullan

Physical Modelling Sound Synthesis by Digital Waveguide Extraction with Application to Computer Games and Virtual Environments

2011

Supervisor: Maarten van Walstijn

While the  first computer games synthesised all their sound effects, a desire for realism led to the widespread use of sample playback when technology matured enough to allow it. However, current research points to many advantages of procedural audio which is generated at run time from information on sound producing events using various synthesis techniques. A physical modelling branch of synthesis has emerged, primarily from research into musical instruments, and this has provided audio synthesis with an intuitive link to an environment's virtual physical parameters. Some physical modelling techniques, primarily modal synthesis, have been used to synthesise audio in real-time in interactive virtual environments.

This thesis presents a new technique called digital waveguide extraction which models the motion of a two dimensional membrane. The technique exploits the efficiency advantages of digital waveguides, which are commonly used to model one dimensional objects, by extracting harmonic subsets from the membrane's spectrum based on plane waves travelling in a single direction. The technique is shown to be mathematically equivalent to modal synthesis. A detailed method of implementation is provided, the output of which is shown to closely match that of modal synthesis in frequency content and amplitude envelope. A means of decreasing the computational complexity of the technique, while affecting minimal change to the synthesised sound, is proposed and justified. The saving can be applied to a varying degree which is useful in environments where objects can have different levels of perceptual importance and processing availability can vary due to factors unrelated to audio. Perceptual listening tests reveal that, without affecting any perceptual di erence to the sound produced, the computational complexity can be reduced by at least 72%, typically around 95% and under some circumstances beyond 99%.

Finally, the digital waveguide extraction technique has been linked to a purely geometrical analysis of a membrane. This provides an alternative understanding of the technique and indicates potential for a new wave-based approach to modelling non-ideally shaped objects and rooms.

Imogene Newland

The Piano and The Female Body: the Erotic, the Seductive and the Transgressive

2011

Supervisor: Pedro Rebelo

This thesis proposes an “erotics” of music performance based on a close analysis of the relation between the female body, the piano and the body of the spectator. This practice-led research consists of a written thesis and five original performance works using choreography, installation, film and photography.

As a pianist specialising in contemporary repertoire I have a particular interest in the overlap between gesture-based choreography and music performance. Through “embodied improvisation”, I explore physical expression in virtuosic repertoire using the analogy of sexual metaphor. Presenting the tactile relation between the female body and the piano as a concupiscent exchange of energies and desires, I propose a feminisation of the piano and the pianist's body based on the writings of Richard Leppert (1995 [1993]) and Susan McClary (2002 [1991]). Using choreographic theory I suggest “full” and “empty” body approaches to conceiving physical expression in music performance and how such approaches reflect on notions of gender ambiguity, abjection and the gaze (Butler: 2008 [1990]), (Claid: 2006), (Kristeva: 1982 [1980]) and (Mulvey: 2006 [1975]). The thesis concludes music performance as a seductive and transgressive act, supported by emerging theoretical findings and realised through original interdisciplinary practice-led research (Bataille: 2006 [1957]) and (Baudrillard 1990 [1979]).

Peter Bennett

The representation and control of time in tangible user interfaces (designing musical instruments for the manipulation of temporal media)

2010

Musical applications of tangible interfaces are successful because they solidify the ephemeral sounds and structure of music, presenting them to player as readily graspable bricks, pucks and tangible objects. This thesis proposes that a detailed study of how time may be used within tangible interfaces is required in order to allow a better understanding of how to design new systems with this capability.

Music and sounds are not existing physical objects; therefore representing them as objects provides the designer with a very wide selection of possibilities. Rather than following a prescriptive route of giving the designer a set of guidelines, this thesis takes the more open approach of creating a design-space and interaction model that can help synthesise new design ideas, helping to draw attention to lesser explored areas.

This thesis proposes and evaluates the temporal-MCRit interaction model; a conceptual model of how time can be used in tangible user interfaces, adapted from a standard TUI model. This new interaction model can be used to both analyse existing systems, and synthesise new ones. The aim in proposing the interaction model is to encourage the design of tangible interfaces that go beyond a simple spatial mapping of time, towards future designs that allow fluid, intuitive interaction with temporal digital media. The BeatBearing tangible rhythm sequencer is presented in this thesis as an instrument designed to explore the tTUI design space and develop the interaction model. Variations of the BeatBearing are compared to evaluate the ability of the temporal-MCRit in assisting the designer of new temporal-TUIs.

Vasileios Chatziioannou

Forward and inverse modelling of single-reed woodwind instruments with application to digital sound synthesis

2010

Supervisor: Maarten van Walstijn

Physical modelling of musical instruments aims to translate a set of physical model parameters into audio, by simulating the sound generation mechanism of the instrument(physics-based synthesis). The inverse process, physics-based analysis, addresses the extraction of the physical model parameters from the oscillations of the instrument.

Focusing on single-reed woodwind instruments, the main objective of this study is to formulate such an inverse modelling approach. Given a natural sound, measured under real playing conditions, physical model parameters are extracted, so that they can be used to resynthesise the original sound.

Under this inverse modelling scope, a physical model of the clarinet is created, with special focus on its non-linear reed-mouthpiece system. A two-dimensional distributed model is formulated, in order to study the mechanical behaviour of the reed, in the absence of acoustic feedback. This model is used to inform a lumped model formulation of the reed. Complex fluid dynamical phenomena are also incorporated in the lumped model, which is then coupled to a linear model of the instrument bore, to form a virtual representation of a clarinet.

Based on this lumped model formulation, the inverse modelling process consists of a two-step optimisation routine. Given the pressure and flow signals in the mouthpiece, the first step uses a parametrised relationship between the pressure and flow in the mouthpiece in order to estimate the parameters of a simplified model. Starting from this parameter set and simulating the oscillations of the instrument, the second optimisation step estimates the lumped model parameters, trying to match the original and the resynthesised sound. The accuracy of the process is tested based on a set of numerically synthesised data. Finally, the model is applied to various sets of measured data and the robustness of the parameter extraction process is discussed.

Brian Cullen

A Portfolio of Audiovisual Compositions for the ‘new media everyday’

2010

Supervisor: Pedro Rebelo

By examining three audiovisual compositions, A Natural Balance (2006), Pixel Parasites (2006) and Thrice Removed (2008), this thesis explores an approach that attempts to tackle problems associated with composing in the new media everyday. Each composition, which together I refer to as the ‘audiovisual portfolio’, addresses different aspects of the new media everyday. In a similar manner to television, film and videogames, the portfolio exploits computer graphics and sound manipulation techniques. However, there is an abundance of audiovisual content in the new media everyday. This thesis discusses the portfolio with respect to such content and covers issues relating to artistic communication in our creatively crowded society. It analyses the portfolio as both a critical response to, and an inescapable by-product of, the rich audiovisual experiences of our daily lives. In addition, it tackles the contradictory ways these experiences both excite and anaesthetise our senses.

Florian Hollerweger

The Revolution is Hear! Sound Art, the Everyday and Aural Awareness

2010

Supervisor: Pedro Rebelo

This thesis presents a body of practical and theoretical work, which interprets sound art as a means of encouraging aural awareness in an everyday context. Through a methodological feedback loop of artistic practice and theoretical re- flection, strategies for the aestheticisation of everyday aural experience have been developed and situated within a wider context of contemporary aural cul- ture. The current state of this culture is critically examined. The widespread claim that we live in a deeply visualised culture is questioned. It is argued that contemporary Western society can also be characterised in terms of an increasing interest in auditory perception, which the development of sound art as an artistic discipline is one symptom of The technological mediation of listening is discussed and characterised in terms of a mobilisation and individ- uation of aural experience. An overview of different listening modes informs a discussion of six different perspectives on sound as a physical and perceptual phenomenon. Various listening practices, which have been proposed in the field of sound art, are presented.  It is argued that the development of sound art as an artistic discipline can be characterised in terms of an interest in the everyday as a source of sound material on one, and as an environment for aestheticised listening on the other hand. Sound art is proposed as a means of auralising the rhythms inherent to everyday life, and subtlety is identified as an aesthetic category for doing so. It is investigated how technological mediation can be applied to the aestheticisation of mobile and social listening experiences. The above issues have been addressed by means of artistic practice. The results of this process are presented as a portfolio of eight artworks, including sound installations, public interventions, site-specific electroacoustic pieces, graphical scores and mobile hardware projects.

Miguel Ortiz

Towards an Idiomatic Compositional Language for Biosignal Interfaces

2010

Supervisors: Ben Knapp & Michael Alcorn

This thesis presents a systematic approach to music composition with and for biosignal-driven interfaces. It gives a general view of the historical developments of biosignal-related art over the last forty four years, discussing the various ways in which artists have approached the creation of works that are driven by the electric physiological signals generated by the human body. It questions some of the established paradigms for physiological-level interaction between performers and musical systems, and proposes a model in which biosignals can be regarded as elements in the definition of novel musical instruments. It presents a set of technical hardware and software tools that allow for the easy and rapid deployment of biosignal interfaces as well as a conceptual framework for music creation based on physiological signals. It discusses a series of music compositions that where created before this framework was defined, outlining the shortcomings of these works and further presents a set of works that were created following the ideas presented in this thesis. It establishes a quantifiable difference between both sets of pieces, and how an idiomatic model can be applied to biosignal-driven instruments in order to take advantage of their inherently musical characteristics for musical creation. 

Alain Renaud

The Network as a Performance Space: Strategies and Applications

2009

Supervisor: Pedro Rebelo

Philip Noel George Strain

The Design and Evaluation of an Assistive Multimodal Interface

2009

Supervisors: Sile O'Modrain, Ray Yu, Graham McAllister

Henry Vega

A portfolio of original compositions

2009

Supervisor: Michael Alcorn

Tom Davis

The Ear of the Beholder: Ecology, Embodiment and Complexity in Sound Installation

2008

Supervisor: Pedro Rebelo

This thesis presents a body of practice and theory based work, based on a study of an application of models from Complexity Theory to music creation, with a focus on Artificial Life inspired techniques applied as musical processes. It identifies two important issues that are subsequently explored through the creation of sound installations. Firstly, issues pertaining to the employment of models from the field of Complexity Theory for the creation of music; a concern with transference of metaphor, a regrounding of referents in the specificity of music. Secondly, through a discussion of the employment of process in music it identifies a need to employ generative techniques that are more open in their nature, both on a structural and an interpretive level. It thus proposes a conception of process that is open in an ecological sense, in which a level of agency is prescribed to the technological agents that construct the installations, their environmental context and the human interactants within this context. It outlines a body of sound installations that exploit a concept of emergent musical structures as a process of timely perception, rather than fixed objects of empirical observation. In this conception, aesthetic meaning came forth not through the transitivity of the object but through the process of conjoined action between agents, the environment and the human interactant. This thesis formulates an understanding of sound installation as a situated and timely act that is best understood from an embodied, ecological perspective and develops a conception of engagement that is contingent on space, place and personal subjective experience. It outlines strategies that were developed through the process of creative practice for the creation and perception of complexity that exploit notions of embodiment within installation design.

Konrad Kowalczyk

Boundary and medium modelling using compact finite difference schemes in simulations of room acoustics for audio and architectural design applications

2008

Supervisor: Maarten van Walstijn

Simulation of acoustic spaces with the aim of developing virtual immersive applications and architectural design applications is one of the key areas in the field of audio signal processing. In this thesis, a complete method for simulating room acoustics using compact finite difference time domain (FDTD) schemes is presented.

A family of compact explicit and implicit schemes approximating the wave equation is analysed in terms of stability, accuracy, and computational efficiency. The most accurate and isotropic schemes based on a rectilinear nonstaggered grid are identified, and the optimally efficient explicit schemes are indicated.

Novel FDTD formulations of frequency-independent and frequency-dependent bound- aries of a locally reacting surface type are proposed, including a full treatment of corners and boundary edges. In particular, it is proposed to model generally frequency-dependent boundaries by local incorporation of a digital impedance filter (DIF), and the resulting formulae for compact explicit schemes are provided. In addition, a numerical boundary analysis (NBA) procedure is proposed as a technique for analytic evaluation of the numer- ical reflectance of the presented boundary models. The digital impedance filter model is also extended to model controllable surface diffusion based on the concept of phase grating diffusers.

Results obtained from numerical experiments and numerical boundary analysis confirm the high accuracy of the proposed boundary models, the reflectance of which is shown to closely approximate locally reacting surface theory for different angles of incidence and various impedances. Furthermore, the results indicate that boundary formulations based on the identified accurate and isotropic schemes are also very accurate in terms of numerical reflectance, and outperform directly related methods such as Yee’s scheme and the standard digital waveguide mesh. In addition, one particular scheme - referred to as the interpolated wideband scheme - is suggested as the best FDTD scheme for most audio applications.

Ravi Kuber

Developing an assistive haptic framework for improving non-visual access to the web

2008

Supervisor: Graham McAllistair

Haptic technologies have the potential to help the blind community overcome many of the challenges experienced when accessing the Web.  With limited design guidance available to web developers, haptic effects could be selected arbitrarily for use on a web page, with minimal consideration given as to how the sense to touch could assist a blind user.  Poor interface design is known to reduce the quality of the subjective browsing experience.  In this thesis, research has been conducted with the aim of developing effective spatial and navigational cues to address issues of accessibility on a web interface.  Using a structured participatory-based design approach, force-feedback cues have been developed to represent objects commonly found on a web page (e.g. images and hyperlinks).  The application of a modified version of the approach, has led to the design of tactile pin-based stimuli, which provides similar levels of structural and navigational support to the force-feedback cues.  Findings have informed a library of software, with accompanying guidelines for their application on a web page.  These are housed within a haptic framework.  This tool provides a vital reference for developers, allowing them to replicate effects on their own sites, and offers support during both the design and evaluation processes.  It is left to the discretion of the developer to include the mappings that are most appropriate to the context of the web-based task, and ensure that these cues are targeted to the needs of a broad range of blind individuals when using a tactile or force-feedback device.

Chui Chui Tan

An adaptive architecture to support web graphics exploration for visually impaired people

2008

Supervisor: Graham McAllistair

This thesis employs a user-centred approach to design and develop an extensible, adaptive system, the ACTIVE system to improve Web graphics accessibility for visually impaired people.  It is capable of adapting to its context of use such as graphics type, assistive technologies and user profiles in order to choose a suitable graphical multimodal application for the user.  Additionally, the system consists of a user model which contains information about a user in terms of their background, experience levels, and preferences.  By performing a series of experiments with visually impaired people, it reveals that people with similar profiles and experience levels prefer certain exploration conditions.  Consequently, a list of adaptation rules have been derived and applied in the system.  By using a feature-based approach, the ACTIVE system learns about the users from their previous interaction with applications and presents to them with their most preferable and appropriate interface.  The ACTIVE system was designed in accordance with usability and accessibility guidelines.  The system was evaluated with visually impaired people and the results reveal that it has improved the overall experience and satisfaction of people with sight loss when accessing graphics non-visually.  Furthermore, the adaptation accuracy level of the ACTIVE system increases with the degree of system use, where 96.67% of accuracy was achieved in the experiments.  This thesis shows the possibilities in developing a coherent, adaptive system by integrating various variables such as graphics types, assistive technologies and multimodalities.  It also demonstrates that adaptation can bring benefits to people with visual impairments in enhancing their graphics accessibility.

David Fee

Dereverberation of acoustic signals via adaptive filtering

2007

Supervisor: R Woods

This thesis investigates the blind dereverberation problem and proposes the use of predictive deconvolution as a tool for single channel dereverberation. Prediction deconvolution uses the principles of linear prediction to estimate a high order, minimum phase, inverse room impulse response directly from the reverberant audio signal’s linear predictive residual. Simulation results are present for speech and music sources reverberated with image-method generated room impulse responses. Due to the minimum phase property of predictive deconvolution and the nonminimum phase nature of rooms, higher-order statistics based adaptive filtering is also investigated. The subband kurtosis maximization adaptive filter is evaluated with speech and music sources. The room transfer function can be modelled as the product of a minimum phase and allpass component and predictive deconvolution, by reducing the effect of the minimum phase component, is shown to improve the stability and performance of the kurtosis maximization algorithm. Simulation results show up to 6dB reduction in reverberant error energy achieved with this combined approach. The use of linear predictive coding, to extract a decorrelated residual signal prior to dereverberation, is investigated with simulations demonstrating the effect of predictor order on performance. A sinusoidal-plus-residual model is proposed for audio signals with a large amount of harmonic content, such as the male singing voice, and shown to improve dereverberation performance. Finally listening test opinion scores are presented. These show that predictive deconvolution processed speech was judged to be closer to the anechoic source than the reverberant version by 65% to 70% of subjects; scores recorded for female speech processed with the combined predictive deconvolution and kurtosis maximization approach are higher than scores for kurtosis maximization alone.

Jason Geistweidt

A portfolio of original compositions

2007

Supervisor: Michael Alcorn

This dissertation focuses on a portfolio of original musical compositions created at the Sonic Arts Research Centre between February 2003 and June 2006.  The dissertation includes an analysis of the following works: Trans-mission; METROPOLIS ‘04; Terrestrial Variations; The orchestra of the noises of war ..; A letter from the trenches of Adrianopolis ...; Combine (after Rauschenberg); and Fuinneoga.  In his critique, the composer puts forward his own internal reasoning for the compositional act and points to his understanding of what are the salient features of the portfolio, discussing issues of function, form, and control.  In addition to the critiques, the portfolio includes scores, recordings, and software of the compositions discussed.

Martin Kuster

Inverse Methods in Room Acoustics with Under-Determined Data and Applications to Virtual Acoustics

2007

Supervisor: Maarten van Walstijn

With the advent of commercial surround sound systems there is a growing demand for a system that can convert existing mono and stereo recordings into a surround sound recording. A part of this problem is to generate the reverberation for the additional audio channels from the reverberation in the mono or stereo recording. In this thesis, it is investigated whether a room model can be constructed from one or two room impulse responses and in which a virtual surround sound recording can then be performed. The estimation of the room model parameters is based on the three well-known room acoustic models; the geometrical acoustic model with specular reflections, the eigenmode model and the diffuse field model. It is shown that the scope with the geometrical acoustic and the eigenmode model is limited but it is possible to obtain useful and consistent results for the room volume and the source-to-receiver distance from the diffuse field model. Based on these findings, the problem of generating multiple room impulse responses from one or two input room impulse response(s) is approached slightly differently. The very early part of the room impulse responses (the early reflections) is generated by a geometrical model with specular and diffuse reflections. The remainder of the room impulse responses are copies of the input room impulses obtained by convolution with a set of filters that control the coherence between them. The values for the coherence are given by expressions for the coherence between microphones with first order directivity in a diffuse field and these expressions are derived in the thesis. The results from objective and subjective tests indicate that this method works successfully.

Mevlut Evren Tekin

Intelligent Multi-Agent Systems in Polyphonic Score Following

2006

Supervisors: Yo Tomita, Christina Anagnostopoulou

Score following and automated accompaniment systems have been a much debated topic in the broader area of Music and Artificial Intelligence. While significant achievements have been made since 1984, these systems still lack some important abilities to provide a reliable accompaniment in certain situations, in particular their use by students and professional performers during learning, practice and rehearsal sessions.

This thesis describes a new intelligent score following and automated accompaniment system with a unique pipelined multi-agent system approach, which addresses the relatively unexplored area of providing reliable score following for performers at early stages of learning a new piece. Furthermore, the proposed system can provide automated accompaniment for such performances. The research focused on the development of an efficient algorithm which not only was capable of detecting ornaments, but also capable of identifying jumps, repeats and incorrectly-played notes with greater accuracy. Such problems were handled with little success by earlier score following and automated accompaniment systems since they were mostly designed to be used during concert perfor-mances. The proposed system also exhibited a flexible yet robust score-performance matching ability, which would allow the performer to try various styles when practising a new piece. The results of the experiments with some well-known piano pieces demonstrated that the system can achieve a good level of accuracy in practice situations as well as in concert performances.

Rachel Hosltead

Portfolio of original compositions

2005

Supervisor: Michael Alcorn

This Portfolio of Original Compositions highlights the development of my compositional approach during the years 2001 to 2004. Nine pieces of music are assessed:  Rain between the Showers for tape, For Two for oboe, clarinet and trumpet, Jatayu for flute/piccolo/alto flute and 8-channel tape, dissolving into light for clarinet, viola and piano, One.. many for solo soprano, flute, oboe, percussion, double bass, 4 drum kits, DJ, small vocal ensemble, choir, Balinese gamelan and electronics, Enchant for tape, Around the Corner:  Souvenirs of a day on Rathlin for high and low voice, kayagum and bass viol, Thar an bhfarraige gheal (Over bright sea) for seannós singer, Irish traditional fiddle, 3 violins, 2 violas, 2 cellos and tape and Music for Metropolis for amplified violin and double bass, electronics and tape.

Through this body of work, issues of style, technique, expression and form, and the roles of context, collaboration, intuition, compositional and pre-compositional processes are explored.