EKO is an algorithmic composition engine that engages in a speculative dialogue with the sonic memory of the world. Drawing from a vast database of recordings preserved in the Berlin Phonogramm Archive, EKO transforms historical sound documents into material for contemporary musical creation.
The archive comprises recordings spanning more than a century: from early wax cylinders made over one hundred years ago to recent field recordings captured in the present decade. These documents originate from a wide range of cultural, geographical, and historical contexts, forming an extraordinarily diverse corpus of sound. Within EKO, this material is not treated as static heritage, but as an active, living resource.
In a dedicated backend system, the archival recordings are analyzed, fragmented, and recontextualized. Based on these analyses, the composition engine generates new sounds and musical structures that emerge entirely from the archival material itself. Rather than reproducing or quoting the original recordings, EKO reconfigures their sonic characteristics—timbre, rhythm, texture, and spectral qualities—into new musical forms. In this process, historical sound is translated into a contemporary artistic language.
The engine consists of several digital instruments, specifically four synthesizers and three samplers, each designed to operate exclusively on material derived from the archive. No external sound sources are introduced; every tone, texture, and gesture can be traced back to the original recordings. EKO thus functions as a closed yet expansive system, in which the past continuously reshapes the present.
EKO positions itself at the intersection of archival practice, algorithmic composition, and artistic research. It explores questions of authorship, memory, and cultural transmission, proposing a form of music-making in which history is neither preserved nor erased, but actively reimagined. Through EKO, the archive becomes an instrument—one that resonates across time, cultures, and technologies.