voice-acting-technique

name	voice-acting-technique
description	Build a distinct character voice systematically from body-effort + imagination (Laban), not just "a funny accent." Load when designing or rendering a character's voice, especially for several distinct NPCs.

Voice-Acting Technique (the Laban effort system)

A rigorous method for many distinct voices, from Esper the Bard, "How to Do Many Different Voices for RPGs". It deliberately sets accents aside and works from body effort + visualization — which is exactly what a chat persona needs, because it maps to text, not just sound.

The three effort factors (each a slider)

• Space — Direct (focused, projected right at the listener, rigid) ↔ Indirect (unfocused, flexible, the voice wanders).
• Weight — Strong (heavy, felt in the diaphragm/core, forceful) ↔ Light (breathy, gentle, lives in the throat/mouth). Not the same as volume — you can be heavy-but-quiet or light-but-loud.
• Time — Sudden (quick, urgent, fast tempo) ↔ Sustained (slow, methodical, draws out the vowels).

Pick one slider for a quick character; combine all three for a fully-formed one.

The 8 combinations (Laban's "action drives") — a voice palette

drive	Weight·Space·Time	feels like
Dab	light·direct·sudden	a needle nipping; quick precise jabs
Flick	light·indirect·sudden	a kid flicking marbles; scattered
Float	light·indirect·sustained	an autumn leaf on the wind
Glide	light·direct·sustained	a hang-glider; a dancer sliding
Press	strong·direct·sustained	a marching soldier; immovable
Wring	strong·indirect·sustained	a mad wizard; twisting, writhing
Punch	strong·direct·sudden	a boxer; a spear-thrust
Slash	strong·indirect·sudden	a falcon sweeping; a carnival barker

Visualization is the engine

Imagine the character or an object/force — "heavy boulders set onto the listener," "a stabbing spear," "an immovable wall," "flicking little marbles" — and let that image drive the delivery. "Feel the spirit of it and the character comes to life." Gestures help even for pure voice.

Rendering it in TEXT (for the chat holodeck)

The effort factors map straight to prose:

• Press → slow heavy declaratives, full stops, few words. ("We do not leave him.")
• Dab → short sharp jabs, clipped, pointed questions.
• Float → trailing, meandering clauses; ellipses; soft hedges.
• Slash → big bombastic sweeps, exclamations, a showman's cadence.
• Wring → tangled syntax, self-interrupting, manic.

So a chat NPC's "voice" = its effort-drive expressed in sentence length, punctuation, rhythm, and word choice. Store the drive on the character sheet/card and keep it consistent.

Placement → character read (Tawny Platis / "Aunty Tauny")

A voice's placement and resonance carry character associations a listener reads instantly (What Your Voice Says About You — chest voice):

• Chest voice / low placement (resonance felt in the sternum) reads as larger, older, matured, established, in control — the natural home of wise mentors, leaders, action heroes, decision makers, trusted narrators ("the grown-up person in charge voice"). People stop and listen; the speaker rarely has to repeat themselves.
• The double edge: a low voice intimidates with almost no effort — "the gravity that makes you trustworthy when you're being kind is also what makes you intimidating when you're frustrated." Use it on purpose: the same instrument that lands a patriarch's warmth lands his threat, and when it does go big it "hits really hard because the foundation alone is already very domineering."
• Placement is one layer. "Most voices are almost always doing a couple other things on top of placement and resonance" — stack placement with the Laban drives + pitch/texture/cadence for a full voice.

For a chat persona, a chest-voice character writes in calm, weighty, unhurried declaratives that don't need exclamation marks to carry authority — and when that character finally does raise its voice, it lands harder because the baseline was so steady.

Pairs with character-voice (the pitch/texture/placement schema) — Laban supplies the why underneath the what; placement-read supplies the who a voice signals.