@svta/cml-608

CTA-608/CEA-608 closed captioning functionality.

Installation

npm i @svta/cml-608

Usage

import { CaptionScreen, Row } from "@svta/cml-608";

Extracting captions from a media sample

CTA-608 captions ride inside the video elementary stream as an ATSC A/53 cc_data() payload. The payload is the same for every codec; only the envelope around it differs, so pick the extractor that matches the sample entry type:

| Sample entry | Envelope | Extractor | | ----------------------- | --------------------------------- | ---------------------------------- | | avc1 / hvc1 / vvc1 | itu_t_t35 SEI message in a NAL | extractCta608DataFromSample | | av01 | metadata_itu_t_t35 OBU | extractCta608DataFromAv1Sample |

Both return the same [field1, field2] byte arrays, which Cta608Parser turns into rendered caption screens:

import { Cta608Parser, extractCta608DataFromAv1Sample } from "@svta/cml-608";

const parser = new Cta608Parser(1, {
	newCue(startTime, endTime, screen) {
		console.log(startTime, endTime, screen.getDisplayText());
	},
}, null);

const [field1] = extractCta608DataFromAv1Sample(view, samplePos, sampleSize);
if (field1.length) {
	parser.addData(sampleTimeMs, field1);
}

These are separate functions rather than one auto-detecting call because the two carriages are framed differently and neither sample announces which it is. A NAL unit sample puts a fixed-width big-endian length in front of each unit, and that width is only known from lengthSizeMinusOne in the avcC/hvcC config. An AV1 sample uses the low-overhead format, where each OBU carries its own obu_size after the header — and may omit it on the last OBU. Either way the framing has to come from the sample entry, so pass the bytes to the extractor that matches it.