Voice Catalogue¶

The 12 OpenAI TTS voices, each a gender-tagged singleton in dsl/voices.py. Voice is split from the rig — you pick one when you cast a character, independent of the body. Import the ones you need:

from dsl.voices import ballad, onyx, nova, by_gender

A Voice carries its name + gender + casting description — not tone. Tone (how a character speaks) is the character's instructions. See Actors → Per-character voices for the full casting API.

Male (6)¶

Voice	Best for
`ash`	Bright, clear, youthful. Younger leads, optimistic protagonists, sharp professionals (doctors, lawyers), upbeat support.
`ballad`	Warm, melodic, measured — slight British lean. Grounded narration, middle-aged warmth, romantic leads, domestic-drama dads.
`echo`	Calm, smooth, neutral-pitch. Thoughtful detectives, professors, mature professionals — quiet authority.
`fable`	Warm British storyteller. Narrators, period-drama leads, literate villains, fairy-tale roles.
`onyx`	Deep, resonant, authoritative. Older men, villains, narrators with gravitas, commanders, imposing creatures.
`verse`	Versatile, expressive, theatrical. Sorcerers, knights, dramatic villains, performative delivery.

Female (5)¶

Voice	Best for
`coral`	Cheerful, warm, reassuring. Caring mothers, supportive friends, teachers — roles that put others at ease.
`marin`	Clear, professional, contemporary. Executives, journalists, doctors, modern leads.
`nova`	Energetic, friendly, higher-pitched. Enthusiastic support, younger women, warm friends.
`sage`	Calm, wise, mid-range. Matriarchs, mentors, healers — quiet depth.
`shimmer`	Warm, gentle, soft. Nurturing roles, romantic leads, dreamy / ethereal characters.

Neutral (1)¶

Voice	Best for
`alloy`	Balanced, even-tempered, gender-ambiguous (slight male lean). Stealth agents, narrators, AI characters — functional rather than gendered.

by_gender() includes neutral

by_gender("male") and by_gender("female") each return their own voices plus alloy, so a neutral timbre is always an available fallback.

Casting¶

Pick a voice at cast time — define once at the top, then reuse:

from dsl.actors import james, megan
from dsl.voices import echo, nova            # gender-matched voices

marcus = james("Marcus", voice=echo)         # male rig + male voice
nina   = megan("Nina",  voice=nova)
# … then in any shot:  marcus.say("…", action_name="talking")

…or inline for a one-off (voice by singleton or string name, plus a delivery note):

brian("The Stranger", voice="onyx", instructions="Cold. Clipped. Military.")

Defaults & gender matching¶

Every character also has a curated default_voice (used when you don't pass voice=), set in assets/character_descriptions.json. All 36 rigs are paired with a gender-matched default (0 mismatches) — so a male character never silently gets a female voice. This closed Issue #12. When casting a new pairing, match the character's gender to the voice's (or use by_gender).