| name | game-audio-implementation |
| description | Use when designing, recording, licensing, editing, integrating, mixing, localising, profiling, or testing game music, ambience, dialogue, UI, Foley, spatial audio, and adaptive audio systems. |
| metadata | {"portable":true,"compatible_with":["claude-code","codex"]} |
Game Audio Implementation
Create an audible feedback and storytelling system that remains legible, culturally permissioned, memory-conscious, localisable and useful even when the device is muted.
Prerequisites
Load the approved game-state map, narrative/cultural governance, rights policy, localisation plan and selected engine integration constraints.
Use When
- Building an audio bible, event map, buses, adaptive score, ambience, Foley, voice pipeline, subtitle timing, spatial mix, or audio performance budget.
Do Not Use When
- Audio is only a late asset-drop or living/restricted repertoire has not been cleared.
Required Inputs
Game states and event taxonomy, narrative/cultural boundaries, rights and performer agreements, device/speaker/headphone contexts, localisation plan, engine/middleware, memory/voice budgets and accessibility needs.
Workflow
- Define audio pillars, listener perspective, silence policy, event vocabulary and required non-audio equivalents.
- Map music, ambience, Foley, UI, dialogue and creature/character states to gameplay events and priorities.
- Commission or record only after repertoire, performers, consent, credits, territories and reuse rights are clear.
- Edit, name, loudness-check, loop, tag and version source/master/runtime files.
- Implement buses, snapshots, concurrency, variation, ducking, spatial ranges, occlusion and adaptive transitions.
- Integrate subtitles, pronunciation guides, localisation context and lip/gesture timing where required.
- Profile voices, decoding, streaming, memory, CPU and package size on target devices; test speakers, headphones, mute and interruption.
Quality Standards
- Pair every gameplay-critical sound with visual or haptic communication.
- Prevent repeated sounds from becoming fatiguing through variation and concurrency control.
- Preserve source, licence, performer, consent and cultural restriction metadata.
- Make mix states recover correctly across pause, focus loss, calls, backgrounding and route changes.
Anti-Patterns
- Music carrying required information. Fix: redundant communication.
- Sacred recording used as generic ambience. Fix: permission, contextual limits or original commissioned substitute.
- Unlimited simultaneous voices. Fix: priorities, caps and virtualisation.
- MP3/streaming/compression choice by habit. Fix: measure per asset class and engine.
- Voice recorded before final terminology. Fix: language and cultural gate first.
Outputs
Audio bible; rights ledger; event/bus map; recording and localisation pack; integrated banks/assets; mix and accessibility test; performance evidence and release stems.
References
Decision Rules
| Asset | Default treatment |
|---|
| Short frequent effect | Memory-resident only if measured and bounded |
| Long music/ambience | Evaluate streaming and transition cost |
| Critical dialogue | Subtitle, context note, pronunciation and fallback |
| Restricted or unclear repertoire | Do not record or ship |
Read Next
Use game-testing-polish for perceptual QA and mobile-game-release-liveops for privacy and download-pack policy.
