Humbug or Mazoo:  A hum-controlled mouth synthesizer

Money Laundering Scheme:  an electroacoustic composition

The Secret Lives of Objects

TTSTSVS:  Text To Speech To Singing Voice Synthesis

interactive

installation

musical

instrument

electroacoustic

composition

software

development

psychoacoustic

experiment

Abstract:  This paper introduces the Humbug, a new musical instrument that translates human humming into synthesized notes of the same pitch.  The motivation for designing an instrument that can be controlled entirely with the mouth is considered.  Both the electrical and mechanical aspects of the design are examined, as each had an influence on the other.  Instrument construction and software development are also covered in detail.  The paper concludes with practical notes on playing the Humbug, as well as suggestions for future development.

Abstract:  As its title suggests, this piece deals rather obtusely with the topic of money laundering.  The three main themes that carry throughout the composition are money, water, and movement.  There are five sections to the piece, each separated by a passing train and distinguishable by the type of water heard in the background.  Compositionally, the intent was to take the listener on a journey that ends where it started, paralleling the idea that laundered money is moved and manipulated but at the end of the day is still money.  The soundscape recordings were made in Belfast, Belvoir Park, and Bangor.

(Jan. 2009)

Abstract:  This paper describes a psychoacoustic experiment that investigates human preferences for ratios in both the auditory (tone intervals) and visual (rectangle proportions) modalities.  Moreover, the experiment seeks to answer the question: “If a test subject is presented with both audio and video stimuli simultaneously, would one modality be dominant in ratio preference selection?”  Test subjects were asked to adjust the proportions of audio-only stimuli, video-only stimuli, and then a combination of both audio and video stimuli.  The range of adjustment for all ratios in this experiment were limited to between 1.1:1 and 1.9:1 in order to eliminate the less-interesting 1:1 (unity) and 2:1 (octave) ratios.  The results indicate that in this range our audio senses alone prefer a major fifth (1.5:1) ratio, our visual senses alone prefer as close to a 1:1 ratio as possible, and our combined senses tend to compromise between the two at a major third (1.266:1).

(Nov. 2008)

(Mar. 2009)

Abstract:  This paper describes “The Secret Lives of Objects”, an interactive sound installation where participants can scan the barcode of any item and an animatronic toilet will then tell a random story about the item by using synthesized speech.  First, similar existing projects are documented.  The motivation behind the design is then discussed, along with how the design evolved from a horoscope reader to a storytelling machine.  Next, the techniques used in the generation of random story content are examined.  The hardware and software to run the installation are then laid out.  The effectiveness of the installation, particularly with regard to interaction design, is analyzed.  Finally, suggestions for future instantiations of this installation are considered.

(Jun. 2009)

Abstract: This paper describes the development of TTSTSVS, a new Mac OS X application that can quickly generate synthesized speech and singing.  This software takes a MIDI melody and text as input and intelligently aligns the phonemes of the text to the MIDI notes.  A brief background on phonetics, TTS (Text To Speech), and SVS (Singing Voice Synthesis) is given.  The software design process is covered in detail, paying particular attention to choice of TTS engine, usage of the “Repeat After Me” application, programming language selection, phonemic data parsing, MIDI data parsing, the phonetic pitch and timing reassignment algorithm, and finally GUI (Graphical User Interface) considerations.  The paper concludes with examples of typical usage and suggestions for future development.

(Jun. 2009)

Ratio Preferences in Auditory and Visual Modalities

iPhone application

Abstract:  This paper describes the development of Polyrhythm Hero, an iPhone game that explores the impact of haptic, audio, and visual modalities on the learning of complex rhythmic relationships.  First, the game is explained, along with the motivating factors behind its creation.  Next, a background in both polyrhythm pedagogy and similar mobile music applications is given to provide a context for this work.  The layout of the application is then covered, detailing the functionality of the settings mode, training mode, and gaming mode.  Next, the evolution of the technical design is covered, with particular attention to audio, visual, and haptic programming decisions.  Finally, the results of a quantitative and qualitative study on the game’s effectiveness as a rhythm-training tool are presented.  The paper concludes with a discussion of the implications of the experiment, along with suggestions for future work.

(Oct. 2009)

Polyrhythm Hero:  A multimodal rhythm training game for iPhone