A noise phenomenology is proposed that considers noise as an information object that only has meaning when experienced by an actor. Environmental noise has a major impact on the health and well-being of citizens, particularly when the noise occurs at night. Noise level monitoring can fail to take into account the impact of noise on the lived experienced of citizens. A socio-technical multi-sensor architecture is used to record night time noise in a domestic urban environment in a longitudinal study in the UK. Results are triangulated with contemporaneous logs of occupants to isolate noise patterns that cause the most disturbance to sleep behaviour. The phenomenology identifies the need for a multidimensional interdisciplinary approach to developing sensor networks to acquire noise information for environmental monitoring.