The opera that builds a world from the places you describe

Before Róisín Daly entered From Dust, a facilitator asked her to describe places where she felt most and least herself, and an imagined place beyond ordinary reality. Her published account of the 2025 Holland Festival performance begins with that conversation. The scenery was about to acquire a personal history.[5]

Michel van der Aa’s opera gives each visitor a 24-minute journey through virtual reality, accompanied by the recorded voices and avatars of the six singers of Sjaella. It premiered in Rotterdam in December 2024. Its next announced installation run is at November Music in Den Bosch, from 20 to 29 November 2026.[1][6]

The unusual artistic problem is how to make something recognisably composed while allowing an audience member to change its substance. A place described before the headset goes on can become part of the world inside it. The questions also shape how that world behaves.

A description becomes a place

Before the premiere, Van der Aa told Kent Bye how the team turned generated images into environments using depth information, background removal and layers, like a diorama. Particles let the forms dissolve and reform. This instability suited imagined places that could never exactly match the pictures in someone’s head.[3]

Early descriptions such as forests or mountain ranges produced generic scenery, Van der Aa recalled. The team began encouraging more precise, imaginative answers. The conversation became part of the creative work.[3]

An illustrative distinction shows why: “a forest” leaves most decisions to the generator. “A forest where the trees are the height of houses and the path is covered in blue glass” supplies relationships that make a place particular. Neither description is a memory the system can verify. It is material for staging.

The composer told Dana Linssen that his team chose a locally run system and automatic deletion of visitors’ data after the experience. These are the production’s stated arrangements. They matter to the intimacy of an artwork that asks people to supply descriptions drawn from their lives.[4]

The music gives the journey its shape

Generative scenery sits inside an authored structure. In his May 2025 interview with Filmkrant’s Kees Driessen, Van der Aa described twelve possible routes, with some differences in the music visitors hear. One intake answer—whether someone would approach strangers—helps determine how close the singers may come. Personalisation reaches the distance between bodies.[2]

The vocal writing grew around Sjaella’s six voices and a loose artistic use of the Big Five personality traits, Van der Aa told Linssen. The framework supplied contrasting characters and musical colours. It was a way to develop the work, rather than a clinical assessment of its visitors.[4]

Recorded singers also needed to seem present. Van der Aa told Driessen that purely modelled figures looked too much like game characters, while volumetric recordings restricted interaction. The team developed a hybrid approach. Gaze mattered because a singer who follows a visitor’s movement can appear to be addressing that person.[2]

Daly’s account captures the complicated result. She described becoming readily absorbed, but also moments of self-consciousness about moving through a room in a headset, watched by unseen technicians. A close encounter with one of the singing figures startled her. The work’s intimacy could be awkward as well as involving.[5]

One person occupies the centre

That scale is built into the forthcoming November Music presentation: one person per time slot, with roughly an hour allowed for the intake and the 24-minute VR journey. The event asks groups to book separate slots for each participant. A shared outing produces separate encounters.[6]

Van der Aa described a composition with a common beginning and ending and a branching middle. He wrote those endpoints and a central destructive scene first, then connected them.[3]

This gives the generated places a dramatic purpose. Their interest depends on when they appear, which voice accompanies them and what the visitor has already been led to expect. Two people can arrive at the same ending having brought different worlds into the journey.

Report an error or suggest a correction