There is a Point at Which Methods Devour Themselves || Pedro Oliveira

The piece places the listener at the center of the question 'what makes a voice human?', as a rendition of Elomar Figueira Mello's 1989 song 'Balada do Filho Pródigo' unfolds in the room.

In this rendition, the voice of Greek singer Aimilia Varanaki – who learned the song by ear and without any knowledge of Brazilian Portuguese – is constantly de- and reconstructed through resonant filters and spectral synthesis, blurring the lines between recording, reproduction, re-synthesis, and reverberation, while cutting across the textures of language, prosody, articulation, and the materiality of sound.

This work is part of a long-term research project on the so-called 'dialect recognition software', a proprietary technology that utilizes similar techniques as those employed in the sound piece. This software has been in use since 2017 exclusively by the German Office for Migration and Refugees in cases of undocumented asylum seekers, as supporting evidence towards deportation. The error margin of this software is reported to be around fifteen to twenty percent.

Commissioned by the Max-Planck Institute for Empirical Aesthetics (Frankfurt, DE)