Kicked Out of the Hero's Party: But Now I Use My Software Whispering Skill to Rise in This World
README.md
Audio
The sound of Kicked Out of the Hero's Party: crossfaded music cues, looping ambience beds, one-shot sfx that duck the music, voiced story lines, and the SETTINGS volumes. Contract in design/spec.md. Everything is local to the peer and a no-op outside play; the voices spawn under a runtime AudioRig entity.
Music: Audio.music(cue?, fade?)
Three decks, equal-power crossfade (default 1.5 s); each track loops by crossfading into itself 5 s before its measured end.
| cue | mood |
|---|---|
title | sweeping bittersweet main theme |
dungeon | dark dragon-crypt drones |
tension | confrontation / humiliation underscore |
guild | cozy lute-and-fiddle guild/tavern theme |
whisper | ethereal harp + glassy synth (Software Whisper) |
battle | boss battle |
triumph | victory fanfare |
ending | emotional ending |
gentle | slice-of-life: acoustic guitar, warm piano, light flute (everyday calm) |
sad | slow solo piano + cello, lonely but not hopeless |
mystery | muted harp, celesta, pizzicato, investigation / suspicion |
night | soft e-piano and nylon guitar, moonlit and reflective |
Ambience: Audio.ambience(id?)
Three decks, 2 s crossfade, 4 s self-crossfade loop. nil stops.
| id | bed |
|---|---|
dungeon | stone hall: torches, echoing drips, wind through the exit arch |
crypt | sealed crypt, close and dry: dust and grit trickling, single drips, a cold draft |
guild | busy guild hall: crowd murmur, mugs, hearth, counter bell |
guild_night | guild after hours: hissing lanterns, a few hushed tables, quiet hearth |
forest | river forest by the Riverford bridge |
arena | duel arena crowd |
city | capital at dusk: bells, wind, distant crowd |
throne | vast throne hall |
night | summer night: cricket chorus, frogs, far owl, firefly shimmer |
market | midday market: merchants calling, coins, carts, awnings |
tavern | rowdy tavern: laughter, tankards, fireplace (no music) |
archive | hushed archive: candle flutter, page turns, quill, creaking shelves |
rain | steady rain on cobbles and roofs, gutters, one far thunder |
garden | morning palace garden: songbirds, fountain, bees |
Aliases: fireflies/crickets → night, lanterns → guild_night, library → archive, birds → garden, bustle → market, inn → tavern, storm → rain, drips → crypt.
Sfx: Audio.sfx(id, volume?)
blip advance click hover whisper_on whisper_off code_read code_edit code_patch glitch error coins level_up rank_up slam laugh door sword chapter. Per-id level, rate limit and pitch jitter; whisper/reward/drama sfx duck the music (attack 80 ms, release 0.7 s).
Voice
Audio.voice(id)→ boolean: the cue-sheet voices and per-character reaction barks inAudio.VOICE(seraphine_useless,bark_<cast>).Audio.voiceLine(lineId)→ boolean: the voice of a story line, when one exists (see the convention below). Stops any line still speaking, ducks music to 0.3 while it plays, and mutes the typing blips.Audio.voicePlaying()→ boolean,Audio.voiceRemaining()→ seconds left (0 when silent): auto mode waits on these.Audio.stopVoice()cuts the speaking line (advance / skip).Audio.hasVoiceLine(lineId)→ boolean.
Dropping in a voice actor's files
Every say / narration step has a stable line id <chapter>.<nnn> (for example ch1.004, prologue.012; the story writes them in game/story). Its voice is the soundClip named after the id with every . (and -, /, space) replaced by _, inside game/audio/voice/:
game/audio/voice/ch1_004.mp3 → imported as game/audio/voice/ch1_004.soundClip → plays on line "ch1.004"
game/audio/voice/prologue_012.wav → game/audio/voice/prologue_012.soundClip → "prologue.012"
Drop the audio files (mp3/wav/ogg) into game/audio/voice/; the importer makes the soundClips. No code change: the folder is listed by asset.list with literal selectors and re-listed (at most every 4 s) when a line has no voice yet, so files added while the game runs speak on their next line. Record dry, mono or stereo, peaking around −3 dBFS; lines play at a fixed gain on the voice bus. A line with no file simply plays nothing.
Settings
Audio.setMasterVolume(v), setMusicVolume(v), setSfxVolume(v), setVoiceVolume(v), setVolume(bus, v) with bus = master | music | sfx | voice (0..1), getVolumes() → { master, music, sfx, voice }. They drive the engine mixer channels (music, sfx, ambience follows sfx, voice), saved per player in runtime_data kohp/audio.
Audio.stopAll(fade?), Audio.debugState() (decks, gains, duck, settings) for tests.
Clips live beside it in music/, ambience/, sfx/ and voice/ (all generated with audio_gen; sfx/blip_tick is the single tick cut from sfx/blip). The original cues' levels were set from measured RMS; the 2026-11 additions (gentle sad mystery night, the eight new beds) are set by ear against them.