This thesis contributes a novel, cross-disciplinary framework to the field of sound studies. It examines how our inherent capacities as listeners are manifested in transitional urban environments, and the primary role of voice as a vehicle for perception in field recording and soundwalking practices. Using the conceptual triad of ‘node, counter-atmosphere and meshwork’ as its analytical device, this research considers the polyphonic physical, personal and social ecologies at play in our encounters within transitional spaces. By doing so, it highlights the importance of sound for countering their functionality and opening them up to a more engaged perception. In its theoretical scope, this conceptual triad draws on and re-contextualises existing terminologies from a variety of disciplines: urban planning and Kevin Lynch’s notion of the node; philosophy and Gernot Boehme’s theory on the atmosphere as well as Gaston Bachelard’s concept of seeping through; anthropology and Tim Ingold’s idea of the meshwork. Coined as a sonic perceptual ecology, this triad is a new analytical tool that is the immediate result of the practice developed as part of this research. Involving three consecutive stages, the work spans across intensive fieldwork, workshops, hybrid telematic soundwalks, radioart pieces, public events and performances engaging with different sites in London and elsewhere. This thesis presents a constellation of original outputs, essential to creating and understanding the novel conceptual framework of the sonic perceptual ecology. This is achieved by testing new methodologies, by analysing, in new terms and through the Sensing Cities interviews series, existing creative work and by developing a portfolio of practice that has been presented as part of commissions, conferences and curated events. Key to these activities is the proposition that we perceive not as authoritative presences but as organisms whose voice is, as Mikhail Bakhtin would suggest, a chain of human and non-human utterances.