Captions and subtitles
Turn high-quality local transcription into captions you can ship.
The caption workflow starts with local speech recognition and word timing, then turns reviewed cues into sidecar files or rendered video captions.
Vertical captioned video preview.
Workflow
Captions are a review surface, not just an export format.
Start with local STT
Run faster-whisper on your own machine, keep the original media in local storage, avoid a vendor per-minute transcription bill in the default path, and use transcript timing as the source of truth for captions.
Render the formats people actually use
Export subtitle files when the destination wants sidecars, or render a branded captioned clip when the destination is social video.
Captioning workflow questions
Can captions be generated from local transcription?
Yes. The caption workflow starts from local speech recognition and word timing, then turns reviewed cues into SRT, VTT, or rendered captions.
Does a caption workflow require a hosted captioning service?
No for the local path. Remote storage, hosted access, or cloud models are separate egress choices.
Transcription quality
Better captions start before the caption editor.
VoxNotch leans on high-quality local transcription, model choice, language hints, and reviewable timing rather than treating captions as a blind video effect. The meter is local compute, not hosted STT minutes by default.
Start with one short clip.
Transcribe it, inspect the cues, export a sidecar, then render the same moment as a captioned video.