The algorithmic composition engine EKO, known in earlier development phases under the name Widerhall, is an autonomously operating, data-driven sound system that generates new sonic structures from existing audio recordings. At its core, the system does not aim to imitate musical styles or compositional rules. Instead, it translates processes of analysis, metadata, and algorithmic uncertainty into audible sound events. EKO should therefore not be understood as a “composer” in the classical sense, but rather as an instrument that interweaves analysis, archive, and algorithmic interpretation.
Fundamental to understanding the engine is the assumption that every original recording is more than a sound object. Each recording functions simultaneously as a source of data, a historical marker, and a trigger for a multitude of subsequent processes. EKO does not operate with abstract musical ideas, but with concrete physical, temporal, and spatial properties of sound. The system appropriates these properties in order to develop an autonomous sonic process of its own.
fritsch music 