‘Nina Whiteman produced a bloody blinding piece of music and art that laid bare the tragedy of automation and inner city living and its effects on nature, that was hilarious, whimsical, and often insane and disturbing. The only sentient being in the room. BRILLIANT.’ (Audience member at the Fuming premiere)
The Cybird Trilogy
I. Fuming (voice and multimedia, 20 minutes, 2022)
II. cybird cybird (piano and multimedia, 12 minutes, 2022)
III. Incandescent (trio and multimedia, 15 minutes [in progress] 2022)
Fuming grew from my feelings about crossing a busy ‘arterial’ city road to reach my local park every day. I recorded this soundwalk dozens of times, capturing the impact of the traffic on the acoustic environment, as well as the sounds of birds jostling to live alongside this anthropogenic noise. This ‘data’ was then fed to the Sample RNN computer by Dr Christopher Melen at the RNCM, who sent me a range of results. The machine had learned a variety of sonic features of my daily walk. Unsurprisingly, though, it couldn’t make qualitative judgements between ‘bird’ and ‘car’, and some interesting sonic hybrids emerged. These, along with ‘uncanny’ machine-learned environments form the basis for the fixed media electroacoustic audio in Fuming and cybird cybird.
Research tells us that birds find it harder to learn their songs against a backdrop of traffic noise, and that their songs tend to occupy a narrower and higher bandwidth as a result of these stresses (e.g. Drooling and Popper, 2007 and Moseley et al, 2019). I began to imagine birds as hybrids of technology, flesh, feather, and imposing chaotic environment. The Birds Aren’t Real conspiracy claims (satirically) that all birds have been replaced by robot drones. I began to wonder what it would be like if they had.
The Cybird Trilogy of multimedia works with live performers has grown from this engagement with machine learning, artificial intelligence and the natural world, and charts the ‘adventures’ of a cybird character that is inhabited and portrayed differently in each work. Its concerns are ecological, musical, and technological.
Further uses of technologies in the trilogy:
Holonic Systems (via the Holonist app) allows Movesense motion sensors to communicate with various software. The motion sensors are used to convert bird-like performer wing movements into audible phenomena, through control of playback speed (MaxMSP) and of a modular synthesiser app (MiRack).
Holly+ https://holly.plus/ Holly Herndon’s voice model (deep neural network) was used to process real birdsong recordings. These feature in the in-ear soundtrack of Fuming and in the electroacoustic sound of cybird cybird.
AI images of birds were created for the videos using DALL-E mini, and later Craiyon.
AI videos were created using online generators such as Synthesia and Movio.
An analogue talkbox features in Fuming, where the output of the motion-sensor controlled modular synthesiser is fed into the performer’s mouth. Their mouth filters the noisy emissions, amplified by a microphone.
Fuming was first performed at the International Anthony Burgess Foundation, Manchester alongside other work created by the Machine Learning for Music (ML4M) working group in June 2022.
cybird cybird will be performed by Zubin Kanga in Sheffield on 8 October, and at Cafe Oto (London) on 13 October.
Incandescent will be performed for the first time at the University of Manchester lunchtime concert series on 10 November by Trio Atem.
The trilogy results from my work on Zubin Kanga’s UKRI Future Leaders project Cyborg Soloists. I’m grateful to Zubin for commissioning me to work on this fantastic project using new technologies: https://www.cyborgsoloists.com/
With enormous thanks also to Stephen Bradshaw for his assistance with MaxMSP programming for Fuming.