Aire v.3 is an audiovisual album generated with machine learning tools that analyze patterns of the pollution from different cities around the world: Mexico City, Bogotá and Sao Paulo. The data comes from the predictive system of the Resource Watch, designed by the World Resource Institute and the NASA. Interspecifics has created a series of sound and visual compositions showing the pollution dynamics in the air of each of these three cities.
The first stage of this project revolves around an audiovisual experience in 10 minute segments where cityscapes made of elevations from 3D maps and satellite images are shown alongside generative sound compositions.
The pollutants are analyzed in a saturation scale, with each one assigned its own sound identity that responds to the flow of data by the means of a modulation or control signal. The most relevant and particular patterns detected by the machine learning tool developed by Interspecifics–, have the function of activating and silencing sound events as they occur, establishing the structure of the entire composition.
This project was developed with the support of the Air Quality Department of the World Resource Institute in Mexico, and with advice from Dr. Beatriz Cárdenas.
released April 16, 2021
Code, music and visualizations by Interspecifics
(Emmanuel Anguiano, Leslie Garcia, Paloma López and Felipe Rebolledo).