Dialogue Production in Long-Form Audio

Reading full narratives with a single voice makes character dialogue difficult to distinguish. Producing audio drama typically involves assigning specific voices to individual speakers, auditioning alternate line deliveries, and balancing background cues.

Here is the step-by-step workflow for assembling multi-character projects locally in yovoice.

Production Steps

Step 1: Script Ingestion and Sentence Segmentation

yovoice accepts plain text, Markdown, EPUB ebooks, and SRT subtitle files:

  1. Create a new 【Story Dubbing】 project in yovoice;
  2. Paste or import dialogue text; the editor splits lines into timeline segments based on punctuation;
  3. Use the [Character] Line prefix format to automatically map dialogue to specific actors.

Step 2: Character Casting and Voice Design

Assign distinct vocal timbres across roles:

  • Narrator: Select clean built-in voices from open-source models like Kokoro or Qwen3-TTS CustomVoice;
  • Lead Roles: Use zero-shot voice cloning with 5 to 15 seconds of clean reference audio (via open-source IndexTTS or VoxCPM2);
  • Supporting Roles: Use voice design prompts (e.g., “a calm, lower-register male voice with measured cadence”) to generate original voices directly.

Step 3: Emotional Parameters and Take Selection

Nuanced audio performance relies on comparing variations:

  • Emotional Adjustments: Set intensity levels for tension, sadness, or neutral delivery on specific lines;
  • Take Comparison (Takes): Generate 2 to 3 alternate takes for key lines and compare them directly on the waveform timeline;
  • Waveform Trimming: Adjust leading and trailing pauses to control scene rhythm.

Step 4: Loudness Compliance and Soundtrack Export

Deliver audio compliant with platform standards:

  • Target Loudness: Presets include -14 LUFS (modern streaming/podcasts), -16 LUFS (Apple Podcasts), and -18 LUFS (traditional broadcast), running automated peak-limiting checks;
  • Soundtrack Synthesis: Switch to the music panel to generate matching background themes with ACE-Step or Stable Audio 3, exporting finished mixes or individual stem tracks.

With your script and character assets ready, you can start building multi-character audio projects directly in the desktop app.