The aim of the project was to develop an automatic soundtrack system which reacts to a video clip. The system should extract relevant information from video footage edited by the group and, in real time, generate a musical composition that reinforces the emotional content of the sequence.
The main emotional stimulus in the video was identified as the development of colour. Therefore the system was designed to modify the soundtrack according to the intensity of different colours, their position in the video container and overall brightness/darkness balance.
The soundtrack is generative composition based around a chord sequence in C major which varies in sympathy with the video content through a combination of stochastic processes, fixed relationships and randomisation.
The system itself can be interpreted as having three intimately linked sub-systems:
Video analysis and data extraction utilising Max’s video extension, Jitter.
Translation of extracted data into usable information which has some sort of musical meaning.
Various methods of generating music in Max/MSP by synthesis/sampling and audio processing.
Project by 3rd year Audio Technology Students at GCU:
Steven Myles
Matthew Murphy
David Maguire
Andrew Dick
