Skip to content

Subtitles and audio

.srt and .vtt files next to the video are picked up by name: Title.en.srt, Title.eng.forced.srt, Title.en.sdh.srt, Title.pt-BR.vtt. The markers sdh, cc, hi and hoh mark captions; forced and foreign mark forced tracks. Everything is served as WebVTT, captions listed first and forced tracks last.

Text subtitles inside the container (SubRip, ASS/SSA, mov_text, WebVTT) are extracted on demand and cached under /transcodes/subcache. Bitmap subtitles (PGS, VobSub) are not offered. EMBEDDED_SUBS=off turns extraction off; sidecars still work. Extraction reads the whole file once, so on slow network storage turning it off is a reasonable choice. A brand-new file may show one track fewer on its very first view.

Captions are never on by default. Each person sets caption size (small, medium, large) and colour (white, yellow) under Account, and those apply on every device. The web player remembers the language you last picked in that browser.

A track menu appears when a file has two or more, labelled by the server (“English · 5.1 · AC3”). Picking a non-default track switches to a live transcode; see Playback and transcoding.