Noise Scraper

Noise Scraper is a collaboration between Oscar Cass-Darweish and Hannah Sawtell that explores shared interests in the creative uses of open source browser-based tools that enable dynamic relationships between image and sound through audience-activated installations and performances. These configurations enable playful engagement with digital images, their surfaces and intersections through reconfiguring Hannah’s previously developed Useful Tools synthesiser app to produce sound from the surface of inputted images. The adapted app underwent its first live testing, taking place across Vivid Projects and BRIG Cafe through a series of performances in an evening event in December 2025.

Through different projects, time with Vivid Lab has focussed both Hannah and Oscar’s work with work around questions of access, aesthetics of digital representation and real-time data processing in distinct ways that converge in this collaboration.

For Noise Scraper Hannah considers the labour that exists within the digital image. Who made it, and why, and who it is for. In the age of AI, the digital image exists as a site for extraction. As the product of human labour images and sound have an agency outside of simple and quick analysis, they have a foundation in interactions and exchange that is not just about currency, but collaboration and sharing.

As an action formed by mutual dialogues, Oscar has experimented with the coding for the Useful Tools synthesiser to sample visual data from images, video and live feeds to trigger sound. Using the low-level computer vision process of tracing contours to assign sounds to clusters of pixels in images, where colours change sharply. Contours are usually used to start isolating shapes which can then undergo more complex recognition processes to track objects and individuals in images and video. Here, paired with sounds in repetitive loops, their activation through performance invites deliberation on the aesthetic modes by which machines make sense of visual information.