Parham Bahadoran

Iranian vocal tradition is based on a rich musical heritage and contains vocal styles distinct to the region, which differentiates the Iranian vocal tradition from the better-known Arabesque style of singing. This project aims to highlight the specific characteristics of traditional Iranian singing and in particular the ornamentation of Tahreer.

Bairong Han

A staple of the piano repertoire, Schubert's Impromptus No. 2 Op. 90 is considered to be one of Schubert's iconic clavier masterpieces. Current tools allow us to manually annotate beats from the frequency spectrum of a performance audio recording. By comparing the recordings of two renowned musicians—Horowitz and Zimmerman, an analysis of expressive performance tempo is undertaken, focussing on rubato and other variables of tempo/timing.

Jacob Harrison

The Kora is an iconic instrument in West African folk music, a 21-string 'bridge-harp' or 'harp lute' on which performers play expressive improvised melodic lines over bass ostinato grooves. Kora performances are often improvised and subject to interpretation, and so transcriptions and scores of particular pieces are rare or nonexistent. This study uses techniques from automatic harpsichord transcription to estimate the inharmonicity of Kora recordings, with a view to enable the developing of accurate note tracking algorithms.

Peter Harrison

Melodic discrimination tests have been used for many years for both individual differences research and experimental psychology research. This study presents and implements an automatic item generation algorithm for melodic discrimination tests, and models the difficulty of the resulting items using formal measures of melodic similarity and complexity.

Phelan Kane

This project provides an analysis and comparison of musical data drawn from cover versions by the same artist. The material selected for analysis stems from the Johnny Cash album ‘American IV: the Man Comes Around’. The album is a collection of ‘cover songs’ from the pop music canon and each song features heavy re-interpretations of the original. The main overarching aim of this project is to identify if there are any interesting patterns that arise from the data drawn from the cover version repertoire of one artist.

Angela MacArthur

This project analyses shot-for-shot remakes (i.e. remakes faithful to original plots and scripts) of films set about 20 years apart, in terms of emotional prosody, to determine useful comparisons over time in speech acted for the camera. It looks at American actors in English-spoken dialogue, across two genres (due to data availability) over two periods: 1937–1957 and 1974–2013.

Vanessa Pope

The inter-pulse intervals from two performances of ‘Not I’ by Samuel Beckett are used to examine actors’ use of silence throughout the 12-minute monologue and to identify significant silences in performance time. A 20 second window which includes a significant silence is then examined using word and syllable segmentation in Praat to compare the two performers’ approaches to silence.

Daniel Stoller

The mode of phonation is an expressive element of singing voices and can be described using the four categories normal, breathy, pressed and flow. On a dataset containing vowels sung by a female professional, I extract MFCCs as timbre descriptors and to train a neural network for automatic classification. The good performance demonstrates MFCCs are well suited for this task, in contrast to what was assumed in related literature.

Thomas Vassallo

Swing ratio, the long-short patterns found in consecutive eight notes is sometimes employed as an artistic choice by drummer. The demand for more human sounding sequenced drum patterns has inspired this research to attempt to extract swing ratio patterns in a selection of different musical genres.

Julian Zammit

There has been a growing trend of “sleep” or “workout”-related music since the boom of online music services. By looking into pieces tagged as such, we compare and contrast high- and low-intensity music. We look into finding defining characteristics for each tag and pick out distinct differences between the two.