LMOS-AI

This project is another response to the meaning that is given to sound, in reference to Stuart Hall’s The Work of Representation. To represent something is to describe or depict it, and the essence of scoring is to be able to depict an emotion tethered to a scene or visual, to better guide the audience through the impression. This concept was being explored through a network of different StyleGAN2’s, (a generative adversarial network) specifically an Audio Reactive Latent Interpolations with StyleGAN. This neural network uses different found image datasets and they react to the frequencies of the input audio. I used the scores from my previous projects: 20-20000HZ, ORNIS, dr34m5, THE F1A2C1E and 花.

I aimed to create another scenario in which the audience is meant to create meaning from visuals and sounds that don’t sound cohesive or representative of what emotion the score is meant to portray. 

I used the emotions: BITTER, RELIEF. JOY, GRIEVE, & AVERSION

I didn’t mean to emphasize an aberrant sound in contrast to the visual, but rather intended confusion in whether it relays with the visual or isn’t meant to. It’s not meant to be easily read, as the scores are all representative of an auditory, grainy and textural consistency.