v10/internal/design/spf/features/multi-language-audio.md

status, date, definition

status	date	definition
implemented	2026-05-25	implemented

Multi-language audio

SPF exposes HLS audio renditions with language metadata, chooses a default, and supports programmatic or user-driven switching during playback.

Decisions

• Preserve audio rendition identity and language in the presentation model.
• Separate user intent from the resolved active track.
• Use switchAudioTrack for resolution and the shared segment-loader/source-buffer path for midstream changes.
• Flush audio at the next-segment boundary when the selected rendition changes.
• Keep persistence of a user's language preference outside the playback engine.

Current sources

• packages/spf/src/playback/engines/hls/engine.ts
• packages/spf/src/playback/behaviors/track-switching.ts
• packages/spf/src/playback/behaviors/segment-loader.ts
• packages/spf/src/playback/actors/source-buffer.ts
• colocated tests

Codec-changing switches remain constrained by browser capability. Broader probing is tracked in capability probing.