Guest
The Hawthorne Effect from Llyr’s album 'Biome' on Mesh. Video by Xander Steenbrugge. ► Stream/Buy: Llyr enrols Generative Adversarial Networks (GANs) AI art expert Xander Steenbrugge to lend his creative coding to this audiovisual piece. It combines surging AI hallucinations with Llyr’s experimental sound design techno, created from a field recording of a swarm of insects. Depicting the delicate interplay between nature's serene environment and the potential destruction that human civilization can bring about when left unchecked, Llyr's “The Hawthorne Effect“ challenges us all on how we can coexist with nature. Released via Max Cooper’s Mesh Recordings - exploring the intersection between music, art and science - Llyr’s Biome project is an expansive audio documentation of natural beauty and human impact. ► Limited edition 12“ eco-vinyl available: Llyr: The rain had just stopped after a three day torrential downpour and we were sitting in the outdoor kitchen area in a longhouse on an organic farm on the outskirts of the rainforest. It was early evening and we were eating under the light of a simple lamp. A large insect flew up behind me, attracted to the light, and then another, and then another… Before we knew it we were in the middle of a swarm of insects. Before dashing for cover I quickly turned on my recorder, which was sat on its tripod, and recorded the inside of the swarm. I had just witnessed an event which wouldn’t have happened in that time and place if I hadn’t been there; sometimes our mere presence is enough to change things. This is reminiscent of the Hawthorne Effect from psychology, where studying a subject results in modifying that individual’s behaviour. (The study of workers at the Hawthorne Works, which gave the phenomenon its name, was even about the effect of lighting changes on activity.) In the context of Biome, the occurrence I experienced is a microcosm of the influence humankind has on the natural word. It is impossible for us to not have an impact, but it is up to us to moderate it and try to ensure a peaceful co-existence with nature. This recording of the high frequency beating of thousands of wings and the wing flaps of individuals who flew right up to (and bashed into) the microphones became the source for this song. A drumkit was created by layering individual insect sounds. Lush, harmonic strings were created by repeating many tiny grains of these recordings. A pulsatile alarm was made by sending the insect sounds through a Metasonix distortion unit feeding back into itself. As such, we hear the insects acting as individuals and also swarming as one. This the unsettling warning sound of an event that is wholly the result of an interaction between nature and humankind. Xander Steenbrugge: This video highlights the delicate interplay between nature's serene environment and the potential destruction that human civilization can bring about when left unchecked. To create a compelling narrative around this interplay, I used an advanced AI technique called GANs (Generative Adversarial Networks). The first step in this process is to create a training set of images the AI can learn from. For this, I curated a collection of thousands of images from the pristine rainforests of Borneo as well as various examples of human interference in natural environments (logging, forest fires, oil rigs, ...). The AI model is then tasked to learn the visual style and composition of those training images. Triggered by these conflicting visual elements, the AI model begins to hallucinate interesting (and often dark) visual impressions that vaguely resemble the themes present in the training data. This hallucinated imagery is then finally composited into a seamless visual storyline that perfectly matches the narrative of the track. Follow Llyr: Follow Xander Steenbrugge: Follow Mesh: Written, produced and mixed by Llyr (Gareth Williams) Video by Xander Steenbrugge Mastered by Chris McCormack at Blacklisted Mastering Ltd Additional mix support by Nicolas Bougaieff ℗ 2021 Mesh © 2021 Gareth Williams (Llyr) #Llyr #Biome #Mesh
