Two years ago, while recreating Ligeti's 'Poème Symphonique for 100 Metronomes' in software, I had to build my own audio/DSP-version of the Kuramoto model to be able to sync all the metronomes. I have written a post about the project, but looking back, I feel like my Kuramoto model software deserves a post of its own. A post describing the process and how I designed it as a dynamic multi-channel audio object in Max/MSP, able to sync 100+ oscillators on the fly.

Making neural networks smaller without harming their accuracy is difficult. Knowledge Distillation (KD) is a fairly new technique where larger models are used to train and improve the performance of smaller models. In this post, I walk through a recent research project where we examined if KD could enhance audio effects (VST) built with neural nets. And the short answer is yes, it can!

OSC is a high-res protocol (64-bit) for sending messages between musical devices, similar to MIDI (8-bit). Last year, I led some workshops on OSC focusing on its unique networking capabilities. In this post, I present my workshops as a written step-by-step guide, showing how to use OSC Timestamps and a technique called Forward Synchronization to effectively sync musical devices in Python and Pure Data.

Measuring video latency in networked music performances (NMP) is stressful. There is little time and resources at hand, and on-site conditions often differ alot from controlled research setups. In this post, I share a few practical methods and strategies for measuring video latency that guarantee precise and reliable data collection in the field, developed from years of professional experience.

By slightly adjusting the placement of your studio monitors, you can significantly improve the sound quality in your studio. In this post, I share an easy and practical way to find optimal placements for your studio monitors with just some basic tools and a little know-how.

Measuring audio latency in networked music performances (NMP) is stressful. There is little time and resources at hand, and on-site conditions often differ alot from controlled research setups. In this post, I share a few practical methods and strategies for measuring audio latency that guarantee precise and reliable data collection in the field, developed from years of professional experience.

I sometimes use the p5js libraries to make little audio-visual widgets and sonification examples in my posts. Recently, I decided to re-design and update my p5 and p5-sounds implementation to support newer versions of p5, React, Nextjs (v14.1), and TypeScript. In this post, I share all my code and detail my process.

Mixing live sound for networked music events, where several locations are connected together in real-time, is very different from mixing in an average concert setting. Having to manage different stages through an added network layer can be a real challenge, especially when it's showtime and the pressure is on. In this post, I share some useful steps and strategies that can help improve your networked audio engineering practice.

Thanks to modern browsers and open-source coding libraries, creating audio-visual art on the Web is easier than ever. In this post, I share some content from a workshop I recently held on ways to enhance data visualizations with model-based sonification in a browser using the p5js library. Interactive examples included!

While designing a new liveset, I wrote some Python code to help me map and navigate Ableton Live more efficiently using very few MIDI controls. Here is what I did and how.

My collegue and I teamed up with Telenor Research to see if it's possible to play music together over next generation 5G networks. Here are the preliminary results from our first audio latency tests on private and commercial 5G networks.