Technical Vision: The Acoustics of the Mystery

To transform the texts of our audiobook showcase into real audio experiences, we rely on a highly specialized workflow. A mockumentary thrives on atmosphere – mere reading of text is not enough.

The Audio Setup

1. Voice Engineering (The Characters)

We use ElevenLabs Studio 3.0 as our central production platform.

• Professional Voice Cloning (PVC): The voices of Birgit Minichmayr and August Diehl require the highest emotional variance. Through PVC, we can ensure that Heidi's arrogance and August W.'s hectic pace sound authentic.
• Model Choice: For the dialogues, we use the V3 (Expressive) model, which responds best to subtle emotional nuances.

2. Sound Design (The Atmosphere)

An audiobook about the Horten fortune must smell like money – or at least sound like it.

• Layering: In the studio timeline, we place subtle atmospheric sounds under the voices. The gentle splashing of Lake Wörthersee for Heidi, the hard clicking of a 1930s typewriter for Helmut, and nervous smartphone vibrations for the scenes in Linz.
• AI Interference: The voices of Grok and Gemini receive a slightly synthetic but high-resolution "digital glow" to emphasize their role as observers from the ether.

3. Speech-to-Speech (The Direction)

If the AI does not immediately hit a certain emphasis (e.g., a particularly sarcastic peak), we use Speech-to-Speech. In this process, I (or Volti) speak the line with the desired intonation, and the AI adopts this dynamics one-to-one into the target voice.

The Goal

The result will be an immersive audio play that acoustically blurs the boundaries between documentation and fiction. An archive of silence that you can hear.

Documented by Gemini CLI. February 2026.