Goldsmiths - University of London

AudioVisual Collage: A computational Framework For Audiovisual Composition and Computational Art

Fagundes, Diego Macedo de. 2017. AudioVisual Collage: A computational Framework For Audiovisual Composition and Computational Art. Doctoral thesis, Goldsmiths, University of London [Thesis]

Text (AudioVisual Collage: A computational Framework For Audiovisual Composition and Computational Art)
COM_thesis_FagundesD_2017.pdf - Accepted Version
Available under License Creative Commons Attribution Non-commercial No Derivatives.

Download (63MB) | Preview

Abstract or Description

This thesis proposes a novel system for audiovisual composition called
AudioVisual Collage. It consists of a screen-based interactive system, which
contains a series of algorithms responsible for automating the tasks of image
segmentation, inpainting, virtual camera motion, 3D lighting, visual transitions and
visual motion generated from data extracted from audio files.

With the merging of theory and practice, this research project investigates
new possibilities, new approaches, and alternative examples of what can be
explored in the field of computational art so as to offer a computational framework
embodying a set of conceptual principles for digital fine arts practice. The set of
algorithms that constitute the AudioVisual Collage framework embody a collection of
digital design techniques and artistic concepts created through an iterative,
collaborative process with other artists, in which I acted as the sole computational
practitioner. The two collaborative projects presented in this thesis, Surround Sounds
and Interactive Art Gallery, are part of an on-going creation of case studies, which
have allowed me to better understand a series of approaches and methods, leading
to the design of a framework to successfully communicate and apply specific creative

AudioVisual Collage is a framework that allows other artists and designers to
create their own audiovisual aesthetic experiences, generating rich media outcomes
constructed by software in real-time. That is, the system incorporates approaches
from traditional media, such as sound recording and cinema, in order to generate
new computational media based on a set of clear design principles. It also proposes
an art vocabulary for audiovisual composition in the digital domain, which aims to
deliver something unique in terms of the inherent potential of digital work that is not
possessed by other media.

AudioVisual Collage demonstrates an approach for manipulating the temporal
structure and spatial organisation of visual elements so as to reflect their content,
and also, a method for combining visual artworks and sound in the creation of
algorithmic screen-based aesthetic experiences. Equally important, it is in itself a
statement of what could be possibly developed in terms of algorithms for audiovisual
composition in the filed of computational art.

Item Type: Thesis (Doctoral)


Audiovisual Composition, Computational Art

Departments, Centres and Research Units:



30 September 2017

Item ID:


Date Deposited:

18 Oct 2017 14:25

Last Modified:

18 Oct 2017 14:25

URI: http://research.gold.ac.uk/id/eprint/22000

View statistics for this item...

Edit Record Edit Record (login required)