process

How the murmur is made

Voices uses public, country-specific trend and news feeds as raw material. It extracts short phrases, stores them by country and day, and generates speech variants for the fragments.

1. Collect public trend and news phrases from different national infospheres.
2. Normalize and rank the fragments so the loudest cultural signals rise to the surface.
3. Generate short speech clips in quiet, normal and intense variants.
4. Layer the clips in the browser with timing, volume and spatial drift.

The system is not designed to explain the news. It is designed to expose the pressure of repetition: how media fragments surround people before they notice them.

Generated audio is created from collected text. The installation does not need to know who you are to show the field you are living inside.