The premiere of Renga for White Noise was held at the Center for New Music and Audio Technologies (CNMAT) of the University of California, Berkeley on October 18, 2024. The telematic premiere was held at the Australasian Computer Music Conference (ACMC) on Futurity at the Australian Institute of Music, Sydney on October 22, 2024.
In the first version, two improvisers—one human, one AI—co-create a linked sequence of 30 and 20-second-long phrases respectively. Within this call and response, the human and AI perform their individual phrases separately. In the telematic version, three improvisers—two humans, one AI—between the US Virgin Islands (myself) and Australia (Kurt Mikolajczyk) perform individual phrases in a call and response sequence. The order of phrases proceeds from myself to Kurt to the AI, then back to myself, to Kurt, the AI, and so on. We alternate between 30 and 20-second-long phrases, a transmediation of the alternating 3 and 2-line verses in a renga sequence. The telematic version involves uploading and downloading files containing the data of musical gestures within a given phrase via a website in real-time, while sharing the audiovisual feed via Zoom, and while each improviser listens to each other then performs separately.
Two topics (timbre or rhythm) are explored throughout an algorithmic composition framework mapped to the principles of Japanese linked-verse poetry across impressiveness 1-4 and relatedness 1-4 measures, where an equal distribution between the two topics is reached. How impressive a given phrase is, and how closely or distantly it links to its preceding phrase, varies with respect to an ongoing memory bank of gestures. By learning the dial-based movements of the human's phrases on the Ableton Push 2, the AI produces gestures—on the same white noise instrument that we programmed for this work—including sounds that are impossible to be performed by human hands.
The work is intended to evoke the Zen experience of moment-to-moment awareness in renga and is inspired by Kawabata Yasunari’s transmediation of renga principles into his short stories and novels. The 5-7-5/7-7 mora count of renga verses is transmediated into the rhythm phrases. In timbre phrases, the filter-based processing of white noise is contextualized with respect to the white noise music of Jōji Yuasa from the 1960s alongside works by Merzbow. Being composed of the full spectrum of frequencies, as in Yuasa's work, white noise is likewise chosen here for manifesting the Buddhist principle of “many in one and one in many”.
Developed in collaboration with Kurt Mikolajczyk using Max and Javascript, various new aspects of this research-led interdisciplinary work will be advanced in 2025.