The caption channel is two bytes a frame.

CEA-608 rides two bytes per field on line 21. Field 1 goes past once a frame and carries services CC1 and CC2, so one caption service gets two bytes every frame and nothing more — 59.94 B/s at 29.97. A basic character is one byte, so a caption is typed onto the screen at two characters per frame: a full two-row, 32-column pop-on is 64 characters plus 20 bytes of doubled control codes, which is 42 frames — 1.4 seconds — of wire before it can be displayed.

The engine is free and needs no account. Paste the SRT or the WebVTT unmodified. The wire is a queue, not a budget: it carries one caption at a time, so a caption whose transmission has not finished when its cue time arrives is simply late — and the wire is still busy when the next one wants to start, so the lateness accumulates. The first caption in a file is the one that usually fails, because it has to start crossing the wire before the programme starts. And a second service halves the first: adding Spanish on CC2 takes a byte a frame out of English on CC1, and nothing in the English file says so. Finally readability and deliverability are different tests — a reader tops out around 21 characters a second and the wire types at 59.94, so a file nobody could read too fast spends at most 35% of the channel on text. What fills the channel is control codes, extra rows and second services, none of which a subtitle checker measures. Every lane ships a worked example that costs nothing. Writing the report needs credits.