To view the list of presenters: CIRMMT Student Symposium - List of Presenters
General Assembly and Student Symposium 2015 overview
Oral Presentations (9:30-11:15am, 2:05-2:25pm & 3:30-4:15pm)/
Présentations orales
(09:30-11:15, 14:05-14:25 et 15:30-16:15)
9:30 - Cedric Camier: Does reverberation affect auditory motion perception?
We report three experiments measuring the upper limits, defined as auditory velocity thresholds beyond which listeners are no longer able to perceptually resolve a smooth circular trajectory in various reverberate conditions. These thresholds were measured for white noise, band-limited white noise and band-limited white noise mixed with a pure tone, in different reverberation conditions: acoustically dry room, two simulated source-image-based reverberations and natural reverberation with different configurations of loudspeaker arrays. Experiment 1 took place in a dry room and thresholds were measured with and without a reverberation simulation of an actual reverberant room. In Experiment 2, various simulated reverberation parameters were tested in the same dry room, and two different loudspeaker configurations were tested in a reverberant room. Experiment 3 investigated the effect of audio source type in simulated reverberation condition and for high velocities. No significant effects were observed among reverberation conditions, suggesting that the upper limit is robust against reverberation
9:55 - Caroline Shung: Defining Mechanical Voice Dose in Singers
Vocal disorders are among the most common communication disorders across the lifespan with an estimated 9% of individuals having voice problems at some point in time. Professional singers in particular are at a high risk for developing phonotraumatic vocal fold lesions due to excessive muscle tension and friction during voice production. Recent studies have shown that some tissue mobilization therapies can improve the healing response to acute phonotrauma as compared to complete voice rest, highlighting the complex role of mechanical excitation in both tissue damage and healing. This study uses physiological, aerodynamic and acoustic measures of singing voice to define a metric representative of the accumulated mechanical stress due to phonation in singers. This metric termed mechanical voice dose can be used to predict the physical and functional effects of mechanical excitation associated with different singing exercises.
10:20 - Kai Siedenburg, Na Eun Kim, Shrinkhala Dawadi, Stephen McAdams: Sounds unheard of: Familiarity and source identification in memory for musical timbre
Previous research has considered memory for timbre to be based on an echoic type of information persistence that retains sensory traces. Although the importance of semantic components has been discussed, their impact on memory, as well as the role that stimulus familiarity plays in memorization has not yet been studied experimentally. These questions are relevant not only for theories of memory, but also for contemporary musical practice, where the creation of novel timbres is a rich field of artistic exploration.
We compared memory fidelity for familiar acoustic, and unfamiliar, digitally synthesized tones that were not readily identifiable (as observed in a previous experiment, N=15, which also confirmed reduced familiarity). A sequence of three tones was followed by a delay and a probe sound. Participants were asked to quickly indicate whether the probe had been part of the sequence. Experiment 1 (N=30) compared musicians’ memory for acoustic and synthetic sounds over two delay periods (2 vs. 6s); further the position of old probes was varied (1,2,3), as was the perceptual similarity (as obtained by previous ratings, N=24) of new probes and the sequence (high, low). All factors affected memorization performance significantly. Moreover, listeners’ performance improved for the synthetic sounds but not for the acoustic sounds, exemplifying online familiarization. Experiment 2 (N=24) compared silence with an articulatory suppression and a visual distractor task with a 6-s delay. Musicians demonstrated reduced performance for acoustic and synthetic sounds in both suppression conditions, underlining memory’s reliance on attention (visual distractor), as well as suggesting a (perhaps subvocal) rehearsal-based mechanism even for non-identifiable tones (articulatory suppression). Effects from Exp. 1 regarding familiarity and familiarization were reproduced. The data acquisition for nonmusicians is ongoing.
In sum, this is the first study to exemplify the importance of stimulus familiarity and active memory trace maintenance in the cognitive processing of musical timbre.
10:45 - Marcello Giordano, Fredrik Gran, Marlon Schumacher: Haptic notifications for polytemporal/metric synchronization in ensemble performance
Our CIRMMT student award project leverages our previous results on the development of a tactile notifications system for musicians, and brings our technology in a real-world context for music creation and ensemble performance. We collaborated with the performers of the Quasar saxophone quartet (http://www.quasar4.com/), for the creation of a piece by composer Fredrik Gran which was performed in a public concert on March 12th 2015. Gran's composition explored structural and rhythmic relationships of a complexity difficult or impossible to achieve without a multi-user synchronization system. Using our haptic notification system and CLEF (the CIRMMT Live Electronics Framework), Quasar's rehearsals and concert were “computer-assisted” with four individual metronome click-tracks delivered to performers via vibrating actuators. Quasar's feedback on the device was extremely positive: they much preferred our system compared to auditory click-tracks, found it equally effective at synchronizing with a given tempo, and expressed interest to new collaborations involving haptic notification.
2:05 - Sarah Gates, Kit Soden and Moe Touizrar: Correspondence of orchestral and visual components in representing
basic emotional categories in dramatic scenes
The link between music and emotion in dramatic art, be it film, opera or any other related theatrical form, has long been recognized and theorized as an integral component in the success of a dramatic work. However, very little scientific research is geared toward understanding why and how the combination of music and image produces powerful emotional responses. Our project seeks to identify, categorize and demonstrate salient features of dramatic early 20th century music that represent basic emotional categories using an interdisciplinary set of research methods. We aim to identify and isolate the components of dramatic music that are able to clearly represent basic emotional categories toward the evelopment of a functional theory that describes the interaction of both visual and auditory components. Using stimuli from both opera and film noir, we collected a rich set of data from 120 participants that will make it possible for us to segment and coordinate musical scores with the collated affect descriptors for subsequent score analysis. Custom software was developed in collaboration with Bennett Smith at the Music Perception and Cognition Lab (MPCL) at McGill University that enables respondents to record real-time intensity, create event segmentations and apply overlapping, multi-level affect labels to audio and visual stimuli. By separating visual and audio components, we are able to control for, coordinate and compare the contribution of each component. In the coming year, a statistical model of real-time intensity coupled with a detailed emotional analysis will provide a new dimension for analysis by making it possible to coordinate affect descriptors and reported intensity over time with respect to both audio and video components, and subsequently to map empirically validated emotional regions directly onto the scores.
3:30 - Daniel Steele: The role of activity in urban soundscape evaluation
While the soundscape field continues to grow, there remain few opportunities to put its ideas into practice, evidenced by the lack of cities embracing soundscape approaches. A meta-analysis is presented providing converging evidence for the influence of urban activity (an observable behavior) on soundscape evaluation from the perspective of city users. Three studies focus on urban activities in various ways: 1) a series of interviews with urban planners, addressing the gaps between planners and soundscape researchers; 2) a study that varied (envisioned) activity while collecting evaluations of soundscape appropriateness; and 3) a study carried out using the experience sampling method where momentary, in-situ soundscape evaluations were collected at various points of the day, along with activity-at-hand, mood, and cognitive effort. Our findings on the role of activity point toward the importance of soundscape over physical measurements in urban design and provide a common link to achieve cross-disciplinary synthesis.
3:55 - Kelvin Chan: A comparison of piano performance evaluations done under audio-only, limited and full audiovisual conditions
The current study compared ratings of piano performances given under three rating conditions: Audio-only, Limited Audiovisual, and Full Audiovisual. 60 pianists with experience in performance and/or teaching were recruited to rate 10 recordings of stylistically diverse pieces, by four expressive criteria: Phrasing, Dynamics, Rubato, and Overall Quality. Results illustrate that ratings given under each rating condition are statistically different from those of the other two, with the Limited Audiovisual condition having the highest mean scores, while the Full Audiovisual received the lowest. In addition, Dynamics was found to be the only expressive criterion which received significantly lower scores than did Overall Quality. The most important finding of the study reinforces those of recent research, which show that the visual component of piano performances has a powerful influence on the perception of musical expression.
Posters (11:20am-1:00pm) / Posters (11:20-13:00)
1. Mohammad Adeli and Deborah Egloff: Developing a physiologically-inspired vibrotactile interface
TBA
2. Rachel E. Bou-Serhal, Tiago H. Falk and Jeremie Voix: On the potential for artificial Bandwidth Extension of bone and tissue conducted speech: A mutual information study
To enhance the communication experience of workers equipped with hearing protection devices and radio communication in noisy environments, alternative methods of speech capture have been utilized. One such approach uses speech captured by a microphone in an occluded ear canal. Although high in signal-to-noise ratio, bone and tissue conducted speech has a limited bandwidth with a high frequency roll-off at 2~kHz. In this paper, the potential of using various bandwidth extension techniques is investigated by studying the mutual information between the signals of three uniquely placed microphones: inside an occluded ear, outside the ear and in front of the mouth. Using a Gaussian mixture model approach, the mutual information of the low and high-band frequency ranges of the three microphone signals at varied levels of signal-to-noise ratio is measured. Results show that a speech signal with extended bandwidth and high signal-to-noise ratio may be achieved using the available microphone signals.
3. Michael Di Perna, Michael El-Jiz and Ian Hattwick: Control Strategies for a Human-Conducted Quadcopter Ballet
How can a musician control the choreography of a quadcopter in real-time during music performance? Considering a musician utilizing a digital musical instrument in performance, we have access both to the raw data derived from their gestures as well as the musical output generated by this data. In this project we explore two related topics . First, the creation of more advanced control and modeling techniques to allow for the quadcopter to react in a desired way to both musical and gestural inputs. Second, which parameterized representations of human musical performance gestures are useful as input to quadcopter choreographic control? In addition, we will be addressing issues of appropriate gesture-music-quadcopter mappings as well as aesthetic considerations for the generation of quadcopter choreography.
4. Hubert Leveille: A Change is Gonna Come: Three Empirical Studies on the Evolution of Harmonic Syntax in Popular Music from the 1960s
The goal of this research is to investigate the evolution of musical tendencies of popular music in the 1960s through a large corpus study in order to identify any consistent changes in harmonic, tonal, and formal syntax. While writers have addressed the shifting musical paradigms of this period, noting that “[r]ock musicians no longer aspire[d] so much to be professionals and craftspeople” but “artists” (Covach, 2006, p. 38), it remains to be demonstrated that a correspondingly profound shift can be empirically located.
Three studies based on the Billboard DataSet (Burgoyne, et al., 2011; Burgoyne, 2011), a new corpus collecting 743 transcriptions of music popular in the United States between 1958 and 1991, will be presented. The first study looks at the incidence of multi-tonic songs throughout the decade, the second one focuses on the incidence of flat-side harmonies, and finally, the third study investigates the evolution of formal organization.
5. João Felipe Santos: Updating the SRMR blind reverberation time estimation method for variability reduction and use of multichannel information
Abstract: Reverberation time (RT) is an important parameter for room acoustics characterization, affecting perceptual sound quality and also speech intelligibility. Commonly, RT is estimated in an intrusive way, based on the room impulse response (RIR). In practice, however, measuring the RIR of an environment is sometimes not possible (e.g., for real-time processing), and the response is usually continuously changing. As such, blind estimation of RT based only on the recorded reverberant signals is of great interest. In this work, we have built upon a previously proposed RT estimation measure, the Speech to Reverberation Modulation Energy Ratio (SRMR). We updated the metric in order to reduce its variability and also to exploit spatial information by using multiple channels. Finally, we tested the metric with sounds recorded from solo musical instruments and ensembles, in order to evaluate its performance with music signals. Experimental results show the modified metrics outperforming state-of-the-art benchmark algorithms.
Keynote Address (2:30-3:30pm) / Conference Invitée (14:30-15:30)
2:30 - Gil Weinberg: Robotic musicianship at George Tech
The Robotic Musicianship Group at Georgia Tech aims to facilitate meaningful musical interactions between humans and machines, leading to novel musical experiences and outcomes. In our research we combine computational modeling approaches for music perception, interaction, and improvisation, with novel approaches for generating acoustic responses in physical and visual manners. The motivation for this work is based on the hypothesis that real-time collaboration between human and robotic players can capitalize on the combination of their unique strengths to produce new and compelling music. Our goal is to combine human qualities such musical expression and emotions with robotic traits such as powerful processing, the ability to perform sophisticated mathematical transformations, robust long-term memory, and the capacity to play accurately without practice. The talk will feature a number of computational approach for improvisation using genetic algorithms, hidden Markov models, and fractals which were implemented in 4 different robotic platforms: Haile, Shimon, Shimi and the robotic drumming prosthesis.
