Selected Software Projects

Rad Edit (2026)

 An audio editor based on the following concepts:
– Keyboard driven: a condensed, consistent, tracker-inspired grammar of keyboard shortcuts makes precise editing as flow-inducing as playing an instrument. No dials, no menus, no pop-ups, no hidden pages, just a waveform under your fingers.
– Fast, musical navigation and playback: you can immediately focus on beat-relevant chunks. The edited audio always plays in time and immediately responds to edits.
– Ultra-Destructive editing: The interaction (and the data) is almost stateless and abstraction free. Everything is just audio, or breakpoint curves, meaning its just you and the sound. No intermediaries. No getting lost in a huge project. Decisions are final, there is no second guessing yourself.
– Direct manipulation, WYSIWYG: your eyes and ears work together, the effects you are controlling are directly visible on the waveform or the spectrum. Rad Edit offers several brand new ways to directly, visually manipulate what’s crucial about the sound.
– Seamless interoperability between control signals and audio
– No presets: throw out those tedious lists of off-the-shelf one-size-fits-all presets. Rad Edits interaction model encourages commonly used transforms as gestures, and a temporary “palette” of reusable audio and transformations.

Mome (2026)

An exercise in the following question: if everyone can vibe-code a metronome app with any feature that they can think of in a matter of minutes, what could possibly be valuable about a paid app? Answer: when all around are headed toward more and worse, then the only thing is to offer less and better.

– Expansive, minimal interface
– Deeply considered click sound design controlled with an XY pad
– Rock solid timing
– Apple Watch-first design (always available on your wrist)

Ableton Meld (2024)

Meld is a bi-timbral soft-synth designed for immediate but deep sound design.
– Bitimbral, macro oscillators allow for layering of different synthesis techniques.
– Macro controls everywhere: a wide range of carefully curated DSP models all have well honed macro dials, tailored to be one big sweet spot. The macro philosophy provides consistent interaction patterns across the oscillator, filter and modulation sections.
– MPE-first: this is the first Ableton synth to be designed from the ground up to make mapping to polyphonic expression controllers easy and powerful.
– Inspired by Eurorack: we took the best ideas from the wild and wonderful world of modular synthesis to create something that is consistent with but offering radically different sounds from Ableton’s other synths.

Ableton Spectral Resonator (2021)

A reimagining of the vocoder using spectral sinusoidal synthesis. Transforms anything into a rich tonal instrument. Control with MIDI/MPE, and quantise every harmonic to a scale or a tuning system.

Own idea implemented for  Ableton by myself and team members. This video tutorial was created by myself for Ableton users.

Ableton Wavetable (2018)

Ableton’s canonical wavetable synth. I worked on the full stack, from the FFT-rendered wavetables for alias-free oscillators, to the iconic “donut” wavetable visualisation.

VibeAMP: Virtual Body Emulation Assisted Multidimensional Performance (2015)

An experiment into the potential of hand tracking devices to control multiple parameters of a synthesiser simultaneously.
Participants had to match a 6-DOF drum sound in time to a beat.
This experiment demonstrated that:
– hand trackers could considerably increase the throughput (bits per second) of information transfer from user to synth.
– time-constrained target matching (as found in musical interaction) does *not* follow Fitt’s law, it follows a different curve (Schmidt’s law).
– Virtual representations of the user’s hand in 3D space can aid memorisation of musical sequences of high-dimensional controls.

Tweakathon (2014)

An experiment that proved my assumption that using a hand tracker to change three synth parameters to match a target sound would follow Fitt’s law. According to this law, the rate of entropy reduction in an interactive search space remains  constant.

Sonic Zoom (2013)

An experiment that proposed an interface suited to the divergent stage of creativity. It demonstrated that users would naturally tend toward using the divergent interface early in the creative process, and later tend toward the convergent (classic 1D sliders) interface toward the end.

The Wablet (2010)

An experiment in scanned synthesis on a multi-touch screen. Users could interact with a mass-spring mesh which would be scanned along a drawable path and hence sonified. The app won a best paper award at DAFX conference and was featured in the Create Digital Music blog