Skills, Leadership, and Experience of Lead Researcher: Ichiro Fujinaga

Fujinaga is the team leader and head of the Content axis. He was one of first developers of OMR technology (Fujinaga 1988, 1996) and has published more papers than anyone else on the subject. In large part due to his research, OMR technology is now accurate enough to convert digitized images of scores into symbolic format, facilitating the work of the Searching and Analysis axis. Since commercial OMR software can only deal with modern music notation, for the past 8 years he has concentrated on extending OMR to early music notation from the 14th to the 17th centuries (e.g., Ouyang et al. 2009; Pugin et al. 2007a, 2007b, 2007c; Vigliensoni et al. 2011). Fujinaga is also one of the lead developers of the open-source document analysis software Gamera (based on his OMR software) (MacMillan et al. 2002a, 2002b, 2001; Pugin et al. 2008).

Fujinaga directs the Distributed Digital Music Archives & Libraries (DDMAL) research lab, consisting of over a dozen graduate and undergraduate students. He has extensive experience managing large teams of researchers. He was Acting Director of the Centre for Interdisciplinary Research in Music Media and Technology (CIRMMT) from 2003–4, managing several large grants involving its over 30 faculty members and 100 graduate students across three Québec institutions. As the coordinator of the Steering Committee of the International Society of Music Information Retrieval, he has overseen the successful growth of a series of conferences on music information retrieval since 2001. He has been involved in many large joint research programs. A select listing includes an Intel Grant at John’s Hopkins University with over 20 researchers ($1.7M; 1999–2001); an NSF Information Technology Research Grant, involving 12 Co-PIs ($1.5M; 2002–6); an NSF Digital Libraries Initiative Grant with 8 other researchers ($0.5M; 1999–2002); a Community-University Research Alliances Grant from SSHRC with 8 other Co-PIs ($1M; 2005–10); a Major Collaborative Research Initiatives (MCRI) Grant from SSHRC with 16 other Co-PIs ($2.7M; 2007–13); and another MCRI with 41 other Co-PIs ($2.3M; 2009–15). As a PI, he has received nearly 5 million dollars in funding from SSHRC, CFI, and the FRQSC since arriving at McGill University in 2002.

Mechanisms to Coordinate the Team’s Work

Fujinaga and Cumming coordinate the research and research dissemination of the Content and Searching & Analysis axes, respectively. Meetings and workshops are organized by the project manager. MIRAI members and their students meet weekly to present work in progress, establish new research projects, maintain our websites and social media presence, discuss additional funding opportunities, and recruit new members and students. With teleconferencing, we communicate regularly with collaborating members who reside outside of Montréal; for example, Bain, Helsen, and Lacoste, Co-PIs on the Cantus Ultimus project. With the support of the FRQSC Emerging Team Grant, we hosted the following events that brought us face-to-face with our international collaborators and disseminated our work to a wider audience:

  • 7 July 2015: SIMSSA Workshop at the Medieval and Renaissance Music Conference; presentations by Fujinaga, Cumming, Bain, Hankinson, Helsen, Lacoste, and Pugin.
  • 27 June 2015: SIMSSA Workshop at the joint meeting of the International Association of Music Libraries, Archives and Documentation Centres (IAML) and International Musicological Society (IMS); presentations by Fujinaga, Cumming, Hankinson, and Pugin.
  • 7 November 2014: Workshop luncheon at the joint meeting of the American Musicological Society (AMS) and the Society for Music Theory (SMT); presentations by Fujinaga, Cumming, and Bain.
  • 6 November 2014: Working group and discussion of SIMSSA at the joint meeting of the AMS/SMT.
  • 29 September 2014: CIRRMT Workshop (McGill) highlighting recent developments on the SIMSSA, ELVIS, Cantus Ultimus, and Optical Neume Recognition projects; presentations by Fujinaga, Cumming, Bain, Hankinson, and Wild.

Experience, Accomplishments, and Complementarity of Regular Team Members

Julie Cumming

Cumming manages the Searching & Analysis axis. She is a leading figure in Renaissance music and digital musicology (2014, 2012). Her book (1999) is the definitive study of the motet from 1400–75. Since coming to McGill in 1992, she has been collaborating with Schubert. Together they have produced many collaboratively written papers (e.g., Cumming & Schubert 2015a, 2015b, 2015c; Schubert & Cumming 2012) and conference presentations, organized three conferences, and jointly supervised dozens of graduate theses. Cumming has been part of several large teams of researchers. She is the leader of one of the two primary research axes of the SSHRC Partnership Grant SIMSSA, lead by Fujinaga, and a Co-PI of the SSHRC Partnership Grant “Early Modern Conversions” ($2.5M; 2013–18), together with Sinclair. She has been a Co-PI on a SSHRC MCRI Grant entitled “Making Publics” ($2.5M; 2005–10), in which she developed an expertise on early music printing (Cumming 2010, 2012), and the PI of the Digging into Data Challenge Grant, “Electronic Locator of Vertical Interval Successions (ELVIS),” with 9 Co-PIs, including Fujinaga and Schubert, as well as 17 students. The ELVIS project broke new ground in computerized music analysis. She has worked closely with Fujinaga in the development of OMR for Renaissance music prints and manuscripts, and for plainchant. Cumming currently supervises 2 postdoctoral researchers, 8 doctoral students, and 2 master’s students. Several of her students work in Fujinaga’s DDMAL lab.

Peter Schubert

As an expert on counterpoint pedagogy, Schubert and will manage (with Ethier) the counterpoint analysis project within the Searching & Analysis axis. He has published two textbooks on counterpoint—one on 16thcentury counterpoint (1999) and one on 18th-century counterpoint with Christoph Neidhöfer (2006)—two chapters on historical music pedagogy, and many articles on contrapuntal structures in the music of Willaert, Lassus, and Palestrina. He has collaborated frequently with Cumming on papers in which they show how knowledge of improvised counterpoint changes the way we think of compositional technique in the Renaissance. He is a member of CIRMMT, having collaborated on two experimental projects, one on the perception of Renaissance variation techniques with McAdams, and one on historical tuning systems with Fujinaga and Wild (Devaney et al. 2013, 2010). He has worked on computer modeling of counterpoint as a Co-PI on the SSHRC Partnership Grant SIMSSA, lead by Fujinaga. Schubert also brings to MIRAI extensive experience as a performer. Currently, he conducts three vocal ensembles: the Schulich School Singers, Les Chanteurs d’Orphée, and VivaVoce (which released a CD on the Naxos label in 2007).

Audrey Laplante

Laplante’s research concerns music information-seeking behaviour and music information retrieval systems (Desrochers et al. 2013; Laplante 2015, 2014a, 2014b, 2014c, 2013, 2010b; Laplante et al. 2011). Laplante will co-lead (with Chiasson-Taylor) the user interface project, assisting both axes in their development of user-centred interfaces. She is a Co-PI on the SSHRC Partnership Grant SIMSSA, lead by Fujinaga; and the PI on the SSHRC Insight Development Grants “Chercheurs 2.0: Étude des usages et des facteurs d’adoption des médias sociaux par les chercheurs canadiens en contexte de collaboration interinstitutionnelle” ($74,028; 2015–17) and “L’indexation de musique à grande échelle : exploration du potentiel de nouvelles méthodes d’indexation pour accroître la visibilité de la chanson francophone sur le Web” ($69,735; 2012–14) (Kessler et al. 2014a, 2014b). Her prior work focused on the music information-seeking behaviour of young adults (2012, 2011, 2010a). By examining how young adults discover new music and interact with current music information sources, she was able to identify behaviours that were not well supported by current systems and to provide specific recommendations for the design of better music information systems. Prior to joining the Université de Montréal, Laplante worked as a music librarian at Concordia University, which allowed her to work closely with music scholars, acquiring a good understanding of their information needs.

Cory McKay

McKay’s expertise in multimodal music analysis and processing will be put to use in the Searching and Analysis axis. His previous work as project leader for the jMIR project has provided him with extensive experience producing software that can collect, analyze, organize and process diverse kinds of musical information, both in combination and separately. McKay’s contributions to MIRAI will focus on researching techniques and developing software for extracting metadata from musical scores and audio recordings, and making this information accessible and useful to music researchers. During his doctoral studies, McKay worked in the DDMAL lab and Fujinaga was his supervisor. Since his graduation in 2010, McKay and Fujinaga have continued to collaborate, publishing over 24 peer-reviewed papers (Fiebrink et al. 2005; Fujinaga et al. 2008; McEnnis et al. 2006a, 2006b, 2005; McKay et al. 2010a, 2010b, 2009, 2012, 2009a, 2009b, 2008, 2007, 2006a, 2006a, 2006b, 2006c, 2005a, 2005b, 2005c, 2004; Thompson et al. 2009; Vigliensoni et al. 2010). McKay is a Co-PI on the SSHRC Partnership Grant SIMSSA, lead by Fujinaga.

Caroline Traube

Traube will contribute to the Audio branch of the Searching & Analysis axis. She specializes in musical acoustics, specifically the analysis of musical performance parameters, such as timing variation, intonation, dynamics, and timbre (Bel et al. 2015; Bernays et al. 2014, 2013a, 2013b, 2012, 2011, 2010; Tillmann et al. 2011; Traube 2015, 2013). She is the founder of the Laboratoire de recherche sur le geste musicien (LRGM) at the Université de Montréal, which gathers performers, composers, musicologists, and scientists to tackle research questions relating to the gestural control and the verbal description of timbre and the analogies between the prosodic features of speech and the expressive parameters of instrumental performance. As part of MIRAI, she will develop tools to generate a complete a musical score with performance parameters that are not notated as well as information on the gestural control of instrumental sound.

Glen Ethier

Ethier will collaborate with Schubert on the counterpoint analysis project and with both Schubert and Cumming on their research into style change and the development of the tonal system. His prior work has focused on the perception of melodic accent and the development of an analytic methodology premised on the interaction and concordance of accent in counterpoint of the early 16th century. He has developed paradigms for defining and quantizing strengths for different types of melodic accent and applying these paradigms to re-notated versions of contrapuntal scores. With MIRAI, Ethier will create computer-based models to quantize the parameters for analysis over a much broader corpus of works than he was able to examine in his prior work, including early Renaissance counterpoint and 18th-century tonal counterpoint. Ethier completed his doctoral studies at McGill under Schubert.

Rachelle Chiassion-Taylor

Together with Laplante, Chiasson-Taylor will manage the user interface component of the MIRAI research program. She is currently Senior Music Archivist at Library and Archives Canada (LAC) and a Guest Professor of Musicology at the Université de Montréal Faculty of Music, where she teaches early music history and research methodology. At LAC, she has been closely involved in the development of digital capacity and processing procedures for the institution, which is designated as a Trusted Digital Repository, and has processed a number of born-digital music archives using a variety of non-proprietary software and complex preservation procedures. She also participates in a LAC study group on the archiving and retrieval of complex digital objects and has given papers on digital music processing within archival fonds. In addition to digital archives processing and web harvesting strategies, she brings to MIRAI considerable practical experience and knowledge of metadata standards in federal institutions internationally. Chiasson-Taylor has published collections of essays on early keyboard music, several journal articles on a variety of topics including digital music archives at LAC, the Canadian composer István Anhalt, and the involvement of musicians in espionage activities in Early Modern Europe (Smith & Taylor ed. 2013; Taylor & Knox ed. forthcoming; Taylor forthcoming, 2013, 2011, 2002, 2000). She is also an internationally recognized performer and has recorded three CDs of Renaissance keyboard music for the ATMA label. She brings archival science, historical musicology, and early music performance to the MIRAI team. Cumming was the adviser of Chiasson-Taylor’s doctoral degree in musicology at McGill.

Complementarity of Team Members’ Expertise and Quality of Their Collaborations

The success of our research program is ensured, in part, by the members of our team, due to the diversity of disciplines they represent and their rich assortment of skills. Fujinaga, McKay, and Traube are music technologists, Cumming is a historical musicologist, Schubert and Ethier are music theorists, Laplante is an information scientist, and Chiasson-Taylor is an archivist, musicologist, and performer. The members of the MIRAI team are at different stages of their academic careers: Schubert is a full professor; Cumming, Fujinaga, Laplante, and Traube are associate professors; McKay and Ethier are tenured faculty members at CÉGEPs; and Chiasson-Taylor is Senior Music Archivist for Library and Archives Canada. What brings us together is a shared commitment to creating a music library for the 21th century, a commitment amply demonstrated by our many collaboratively produced publications and conference presentations.

Many of MIRAI’s regular team members have a long history of successful research collaborations. The most mature partnership involves Cumming and Schubert, who have been working together since the early 1990s on style change, performance practice, and compositional process in Renaissance music. Their work has been supported by three SSHRC Standard Research grants (1995, 2001, 2009). Fujinaga has worked with Cumming on optical music recognition of early music since 2005. Since 2007, he has worked with Schubert on tuning in ensemble singing; a research funded by an FRQSC Recherche-Création Grant.