Sorting Song

Presented through the perspectives of two young narrators, Sorting Song shows where our designed world has become machine readable. This computer-generated film features photographic renders sourced from the SceneNet RGB-D indoor training dataset – a large-scale image library of 3D models, floor plans and objects – originally compiled by the Dyson Robotics Lab to develop machine vision for domestic robots.

Niquille’s work reveals how machines remove the boundaries and nuances provided by a human environment, and it explores the discrepancies between ‘photographic’ and ‘photorealistic’ in a digital context. In the film, a conversation between two young voices naively questions the nature of things and their distinctions, playing on the form of the nursery rhyme that teaches children to name objects and tell them apart.