Closed Caption Creator

Your delivery spec asks for 708 captions, and your file is SCC. That will not pass. SCC carries EIA-608 data only, as byte pairs in a 32 by 15 grid. It holds no CEA-708 data. For 708 requirements, use MCC, an embedded track, or a TTML-family format.

One EBU-STL text block holds 112 bytes of text. That is a hard ceiling, and accented letters use two of those bytes. The format stores accents as floating characters, so the accent byte comes first and the letter second. An A with an umlaut is C8h 41h.

Fixing a caption error after delivery costs far more than catching it before. Closed Caption Converter includes an Automatic QC filter that flags problems before broadcast.

A caption that covers a lower third or burned-in text breaks caption quality rules. Our new guide explains where a caption should sit and what survives export. Default to bottom center, inside the safe area.

Drop frame does not drop frames. It is a labelling rule for 29.97 fps video that skips two timecode numbers at the start of most minutes. The picture stays the same. Without it, non-drop timecode runs about 3.6 seconds short after one real hour.

Every caption spec wants speakers identified. They disagree on how. Netflix uses square brackets and keeps the label on the same line as the dialogue. The DCMP Captioning Key uses parentheses and gives the label its own line. UK practice writes the name in capitals with a colon.