SubtitleMedic

Convert SRT to SCC

Turn SubRip subtitles into Scenarist SCC closed captions (CEA-608) for TV stations, YouTube and video editors.

Processed in your browserNo sign-up.srt

How to convert SRT to SCC

  1. Open your .srt file

    Drop the subtitle file into the box above, or click to browse. The sample under the box is a short film in SubRip, if you want to see the result first.

  2. Pick where the time code starts

    01:00:00;00 for a TV station or caption house, 00:00:00;00 for YouTube, editors and videos that start at 0:00.

  3. Check what changes, then download

    The list above the button says how many lines were broken to fit and whether any caption shows late. Save the file: episode.srt becomes episode.scc.

What SCC needs that SRT doesn’t have

An .srt file is a list of lines with a start and an end. An .scc file is the stream a TV decoder receives on line 21 of an NTSC picture: commands and characters, two bytes per video frame, with a time code saying when each burst is sent.

1
00:00:03,120 --> 00:00:05,480
Is this seat taken?

becomes

Scenarist_SCC V1.0

01:00:02;16	9420 9420 94ae 94ae 94f2 94f2 97a2 97a2 4973 20f4 68e9 7320 73e5 61f4 20f4 616b e56e bf80 942f 942f

01:00:05;14	942c 942c

The converter writes pop-on captions, the kind used for prerecorded shows: each caption is loaded off screen (9420, 94ae), placed on a row (94f2) and centred (97a2), then shown whole by End Of Caption (942f) on the frame the subtitle starts. At its end, Erase Displayed Memory (942c) clears the screen, unless the next caption follows straight away and replaces it.

How the captions are timed

Drop-frame time code at 29.97. NTSC video runs at 29.97 frames per second. Its time code skips the labels 00 and 01 at the start of most minutes, so an hour of labels is an hour of clock time. The semicolon before the frames (01:00:02;16) marks it. Each start and end is rounded to the nearest frame.

Loaded ahead of time. The characters of a caption are sent in the frames before it appears, so it pops on exactly on its start frame. A full two-row caption takes about 45 frames, a second and a half, to send.

Commands sent twice. Every command goes out two times in a row, the way broadcast captions are sent, so a decoder that misses a frame still gets it.

Parity on every byte. The top bit of each byte makes the number of ones odd. Decoders use it to throw away damaged bytes, and some reject a file without it.

Fitting subtitles into the caption grid

CEA-608 captions sit on a grid of 15 rows of 32 characters. Captions use the bottom rows, at most four of them.

Long rows are broken again. Subtitles are often written for 42 characters per line. A row over 32 is broken at the last space that fits, so a two-line subtitle may become three rows. Nothing is cut.

Very long subtitles are split. More than four rows become two captions in a row, sharing the subtitle’s time by how many characters each shows.

Centred at the bottom. Each row is centred using the indent and tab codes. {\an8} and other positions from the .srt are removed.

Characters the decoder can’t show. The CEA-608 table covers English and the main Western European letters. Curly quotes become straight ones, other accented letters lose their accent (What changes counts them), and characters with no equivalent are removed and counted.

Tips

Captions late? The tool says how many. A caption right after a very short one has no time to load; give the short line a few more frames or merge the two in the Subtitle Editor.

Wrong start? If the station wants 01:00:00;00 and the video starts at 0:00 in your editor, pick 01:00:00;00 here; the editor or the station maps it to the first frame.

Captions for the web too? Web players read WebVTT. Convert SRT to VTT writes it, and Convert SRT to TTML writes the XML captions streaming platforms ask for.

Check the file you made. Open the .scc in Convert SCC to SRT: it decodes the captions the way a TV would, so you can read back every line and time.

Frequently asked questions

Who asks for SCC files?

TV stations and cable networks in North America, caption houses, DVD authoring, and platforms that pass captions on to TV, such as Amazon and Apple. YouTube, Vimeo, Premiere Pro, Final Cut Pro, Avid and DaVinci Resolve also import SCC. If a delivery spec says “CEA-608” or “line 21 captions”, it means SCC.

Should the time code start at 00:00:00;00 or 01:00:00;00?

Use 01:00:00;00 for a TV station or caption house; broadcast masters start the programme an hour in, after bars and tone. Use 00:00:00;00 for YouTube, editors and any video file that starts at 0:00. The delivery spec usually says which.

Why are some of my lines broken differently?

A CEA-608 caption has a grid of 32 characters per row and 4 rows at most. A row longer than 32 characters is broken again at a space, and a subtitle of more than 4 rows becomes two captions that share its time. The tool counts both.

Why does a caption show a few frames late?

A caption is sent one word (two characters) per video frame and must arrive before it shows. When a long caption follows a very short one, there is not enough time to send it, so it appears as soon as it has arrived. The tool counts these; shortening the line or adding a gap before it fixes it.

Which characters can SCC show?

English letters, digits and punctuation, plus á é í ó ú ñ ç ü and the other accented letters of Spanish, French, Portuguese, German and Danish, and symbols such as ♪, ©, ® and ½. Other letters lose their accent (ă becomes a) or, when there is no plain letter left, are removed. Chinese, Japanese, Korean, Arabic and Vietnamese with tone marks need another format.

Are italics kept?

Yes. Italics and underline are sent as mid-row codes. Bold, colours and positions from the .srt are removed; every caption is centred at the bottom of the screen.

Are my files uploaded?

No. The captions are encoded by JavaScript in this browser tab, and the file never leaves your device.