Au Banquet des Visages

About

“Au Banquet des Visages was born out of both a great curiosity and an urgent need to write: I wanted to understand what human substance remains in a voice when all its characteristics (hesitations, intonation, melody, accents, speed, etc.) are gradually removed. The emergence of artificial intelligence in our lives just a few months ago has made this question a burning one, as it is rapidly and profoundly transforming our relationship with language and meaning, leading us to a kind of neutral flattening of both. This piece is also part of research I have been conducting for several years on the alteration of language, the dramaturgical role of electronic processing on the voice, and how an artificial voice can be constructed from deeply human material.


To build an artificial voice, I needed a human basis. So I chose to work with voice recordings voluntarily sent by strangers via WhatsApp, who were invited to reveal their desires and fears. These messages, anonymized and then transformed (through extreme stretching, transposition, and filtering), feed into a voice generation model that produces three states: broken sentences, fragmented language, and quasi-dissolved vocal material. Other electronic transformation processes, such as cross-synthesis (which combines very different materials, such as vocal consonants and small sounds of broken bones), then complete the electronic composition.

The pianist (playing on a Disklavier) is at the center of the setup. At first absorbed by the mass of the ensemble, it takes him a while to emerge, a bit like a voice trying to reform a language. Around him, timbres crumble, consonants wander, and syllabic fragments appear and disappear. In this process, the electronics function as a distorting mirror.

The piece, constructed like a small sound theater unfolding in successive layers, traverses three zones: a world that is still human, saturated with gestures; an in-between space; and then a stripped-down zone where the machine gradually takes over, while the pianist articulates only the remnants of language, eventually playing only the consonants.”



– Clara Olivares, january 2025