MACHINE LISTENING – DATA AUDIT
You open a dataset, click on a file and hear a person cough three times. You click on another one, hear a voice counting from one to twenty. You notice an accent but you can’t place it. The next file is labelled ‘oooo.wav’, and the one after that ‘YAF_voice_sad.wav’ What are these recordings? Why were they made? What can they tell us about the datasets of which they are a part, and their uses? In this experimental workshop, participants will listen to datasets as sites of both power and possibility, critique and creativity. Moving between lecture, conversation and a series of listening tests, Machine Listening will consider how auditory datasets are compiled, stored and processed, and how different modes of listening might reveal a world of social relations embedded within. If possible, please bring laptops and headphones.