Music has the power to reach listeners on a deep emotional level. I have been fascinated by these experiences since I was a little boy, four or five years old, trying to figure out the patterns in music that could elicit such intense psychological effects, and, as I later found out, even improve a person’s health. My quest led to academic studies in several different areas, from engineering to musical composition to psychology. I went on to found two companies, one of which is Re-Compose GmbH in Austria. Launched in 2008, it conducts research and development in machine learning and music/sound technology. It also engages in the commercial distribution of music production software such as I2C8 - Inspiring to Create, which was developed with the help of EU funding as part of the FET Open Innovation Launchpad scheme.
I2C8 is a piece of innovative, intelligent music production software that amateur and professional composers can use to generate and experiment with chords and harmony. It enables them to start from a simple musical idea and transform it into something that is as complicated as they wish, shifting the emphasis from formal music theory to a more experimental way of working.
The software works in a way unlike any other composition tool already on the market: it does not just ‘recommend’ chords or build sequences of chords, but generates a set of full chord sequences that satisfy a given structural pattern. To do so, it uses a machine learning model as the foundation for chord generation. However, I2C8 is aimed at composers and producers who want full control over their creative musical output and do not want robots to show them the way: the human user remains at the centre of the decision-making process. We want to bring about computer-supported creativity, not fully automated music generation.
In recent years, technological advances in music and Artificial Intelligence have led to a plethora of research results and methods with fantastic technical possibilities, but most have not been brought to a broader audience.. The origins of I2C8 lie in the FET project Lrn2Cre8, which led to new ways of generating coherent chord sequences based on machine learning and musical semiotics. Since then, we have worked to build a product that can make a real difference to people who use software to help them compose music. “Less is more” was one of our most important guidelines: our goal was to not overdevelop technological capability, but to satisfy a clear need.
EU funding allowed us to explore more innovative and experimental aspects of AI-supported music generation and state-of-the-art product development in the field of electronic music production, and the project has laid the foundations for further developments derived from the I2C8 core idea. On top of this, we have gathered invaluable knowledge of process optimization for our future work.