Closed Captioning — PoC
Technical prototype portfolio — demonstration environment
Pick one of 3 bundled sample audio clips below. A real speech-recognition model (Whisper, running locally) transcribes it with timing; caption-format rules — reading speed, on-screen duration, line length, overlaps — flag anything that wouldn't make a usable caption. Do not submit personal, confidential or client-owned information — this public demo only runs the bundled samples. Contact Chrisian to discuss a controlled evaluation. A reviewer edits or confirms each line, then exports a real .srt or .vtt file.
Each sample is built to exercise one specific QC check (or none, for the clean baseline).
Expected: No QC issues on any caption