Voices of the Realm: How Every NPC Gets a Unique Voice

In Embertold, NPCs don't just have dialogue — they have voices. Real, distinct, emotionally expressive AI-generated voices.

How Voice Assignment Works

When the AI Game Master introduces a new NPC, it also selects a voice that matches the character. A grizzled mercenary gets a deep, rough voice. A young elven scholar gets something lighter and more refined. A sinister warlock gets something that makes your skin crawl.

The voice is selected based on the NPC's description, personality, and role in the story. Once assigned, that voice stays consistent for the entire adventure — every time the NPC speaks, you hear the same voice.

Emotional Expression

Static voices would get boring fast. Embertold's voice system supports a range of emotional expressions:

• A frightened NPC's voice trembles
• An angry guard barks commands
• A whispering informant speaks in hushed tones
• A jolly tavern keeper booms with warmth

The AI Game Master marks each line of dialogue with the appropriate emotion, and the voice system adapts the delivery accordingly.

Voice Variety

Embertold maintains a diverse catalog of voices — male, female, young, old, across different tonal qualities. No two major NPCs in the same adventure should sound alike. The system tracks which voices have been assigned and avoids repetition.

Performance Details

Voice generation happens in the background. While you're reading the NPC's text dialogue, the audio is being generated. By the time you're ready to hear it, it's usually ready to play.

And thanks to the caching system, if the same NPC says the same line (across different sessions or players), the audio is served from cache — instant and free.

When Voices Play

NPC voice lines play automatically during gameplay when Immersion Mode is enabled. You'll see the text first, then hear the voice. If you prefer to read in silence, you can mute TTS in your audio settings without affecting other immersion features.

Your Character's Voice?

Currently, the player character doesn't have a generated voice — you provide the "voice" through your typed actions and dialogue. The world speaks to you; you speak to the world through text.

Voices transform NPCs from text on a screen into characters you remember. The gruff laugh of a dwarven ally. The cold whisper of a villain. The warm greeting of a friend. These are the sounds that make a story live.