The project stages the editorial content as a series of scenographies conceived as spaces of representation. Dioramas, stage objects, and constructed environments condense the texts’ themes into recognizable visuals, positioned in an intermediate space between construction and reality.

The images are generated through a structured prompt system that frames each scene as a miniature set, drawing key elements from the texts and shaping them through lighting, camera, and rendering choices. This designed approach makes it possible to produce images that differ from one another while remaining part of the same visual universe.

To enhance interactivity, depth data extracted from the generated images is used to simulate a simplified three-dimensional space. A real-time shader assigns depth to each pixel, while the cursor acts as a moving light source, revealing nearby elements and letting distant ones dissolve into fog. Color blending completes the system, giving each scene a distinct chromatic identity within a shared visual language.

Depth, light, and fog guide the viewer’s gaze, rewarding exploration while withholding a complete view. The act of revealing, partial and contingent, mirrors the speculative logic of the essays, where meaning emerges through inference and atmosphere rather than fixed, total description.