A spatial exploration of Bach’s mathematical precision in Prelude in C Major
For me, creative computing is about using digital tools to uncover hidden relationships and understand structures from fresh perspectives. By translating between sensory domains – sound, colour, motion, and space – these small experiments often feed into larger projects as inspiration and raw material. These three demos were made in 2024.
Wavelength visualisation
Music-color wheel mapping
Rearranging paintings into 3-dimensional colour space
In Find Find (2025), the walk cycles of the multi-legged creatures are built on a sine curve folded in half, like a repeating “m” pattern that drives each leg automatically. This saved the animators from animating each leg individually.
Clip of a crab walking
Clip of the starfish crawling on its tube feet
The octopus protagonist runs on a more complex rig that gives animators the choice between automated movement for swimming and manual control of individual limbs for grabbing and playing. The squids work on a similar system.
Clip of the octopus grabbing the cup
Clip of the squids swimming
The painted look is achieved by studying how stylus inputs (e.g., tilt, pressure, speed, etc.) affect brushwork, then reverse engineering those relationships through mathematical interpretations. Additional effects like ink drying and controlled randomness help make the result feel more organic. Overall, the computer paints each frame based on vector animation from After Effects and guided by computational instructions on painting style.
Demo of the computer painting the octopus
From animation to final render of the octopus
Some atmospheric effects – light rays, water waves, floating particles – are generated through simple customised interfaces directly as data for the computer to paint frame by frame later.
Generating water waves
Generating light rays
Generating floating particles
For the full story behind Find Find’s computational approach – why we chose it, the challenges we faced, and how we integrated it into our pipeline, see R&D of Find Find.
Before “AI” came to mean essentially ChatGPT, it was a far more diverse and philosophically rich subject in terms of how we frame different problems. I was hugely inspired during that period, so I did some small exploratory work – learning basics such as machine learning for recognising abstract images, numbers, and hand gestures.
This 2022 example combines the previously introduced computational painting method with Clipasso – a machine learning model that converts images into abstract sketches. It’s a fascinating area of research that delves into how we recognise things psychologically. Coincidentally, Picasso and Asian ink painting both explored the aesthetics of minimal lines with maximum effect.
A friend jokingly said the computer seemed “distracted” making the strokes in random order.
I prefer drawing storyboards on paper, so I developed a process that arranges and edits panels into an animatic automatically, to save time in production. Another potential use is in storyboard workshops for children and learners, letting them present their stories instantly without having to wrestle with the technicalities.
Demo of importing finished storyboards into After Effects as animatic
This is a 2024 example of me animating a ginkgo leaf with my hand like puppetry. What I’m really hoping for is an application of AI in art that isn’t so preoccupied with precision and polish, but instead allows organic quirks and spaces of imagination to seep through – making computer art a unique artform in its own right, rather than a replacement for human creativity or even reality itself.
Angela Wong's Portfolio 