Closed Captions vs Subtitles vs SDH: The Differences Explained

These four terms get used interchangeably in delivery specs and creative briefs, and the confusion is expensive. Deliver translated subtitles where a broadcaster asked for closed captions and the file comes back rejected, because it is missing the non-speech audio information a deaf viewer needs.

Two separate questions are hiding inside these labels. What does the text contain? That is captions versus subtitles. Can the viewer turn it off? That is open versus closed. Keep those axes apart and the terminology stops being slippery.

Closed Captions vs Subtitles vs SDH vs Open Captions#

What Are Closed Captions?#

Closed captions are a same-language text representation of everything audible in a program, written for viewers who are deaf or hard of hearing. That means dialogue, but also speaker identification, music cues, and non-speech sounds such as [door slams] or [tires screech].

“Closed” refers to the delivery mechanism: the text travels alongside the video rather than inside the picture, so the viewer decides whether to display it. On US television that carriage is an EIA-608 or CTA-708 caption stream embedded in the video signal. In file-based workflows it is a sidecar such as SCC or MCC.

Closed captions are the format regulators care about. FCC rules under 47 CFR Part 79 require captions on nearly all US television programming and hold them to four quality standards — accuracy, synchronicity, completeness, and placement. Our FCC closed captioning requirements guide covers what each standard means in practice.

One practical constraint worth remembering: EIA-608 captions are capped at 32 characters per row across a maximum of four rows. If your source text was written for a subtitle format, it will need to be reflowed before it fits.

What Are Subtitles?#

Subtitles are a localization deliverable. They assume the viewer can hear the audio perfectly well but does not understand the language being spoken, so they carry translated dialogue and nothing else. A subtitle track will not tell you that a phone is ringing off screen, because a hearing viewer already knows.

That difference in scope drives everything downstream. Subtitle formats allow more characters per line — Netflix and most streaming specs permit up to 42 characters across two lines for Latin-script languages — and they support proportional fonts, italics, and positioning that 608 captions cannot express. Whether that styling survives delivery depends entirely on the format you hand over — on YouTube, for example, only a TTML upload carries italics, bold, and color into the player.

Subtitles are also where reading speed becomes the dominant QC concern. Netflix caps adult content at 20 characters per second and children's content at 17. Reading rate violations are one of the most common reasons a subtitle file gets bounced back. See our guide on translating subtitles for broadcast for the full workflow.

What Is SDH (Subtitles for the Deaf and Hard of Hearing)?#

SDH is the bridge between the two. It carries caption content — speaker IDs, sound effects, the full audio picture — but delivers it in a subtitle file format rather than a 608/708 caption stream.

SDH exists because most modern platforms never had 608/708 support to begin with. Blu-ray, streaming services, and mobile players handle timed text as IMSC, TTML, or bitmap image tracks. SDH lets a deaf viewer get caption-grade information on hardware that has no broadcast caption decoder in it.

Because SDH lives in a subtitle format, it inherits subtitle styling flexibility: proportional fonts, colour, and precise positioning, with the same 42-character-per-line ceiling most streaming specs impose. It can also be translated, which a closed caption track normally is not.

What Are Open Captions?#

Open captions are burned permanently into the video picture. There is no separate track and no toggle — the text is pixels, indistinguishable from the rest of the frame.

This is a delivery choice, not a content type. You can burn in captions or subtitles; what makes them “open” is only that the viewer cannot switch them off.

Open captions dominate social video, where autoplay is muted by default and no caption toggle exists in the feed. They are also common for in-venue playback, digital signage, and any platform that strips sidecar files on upload. The trade-off is permanence: fixing a typo means re-rendering the video, and you cannot offer a second language without a second render. For vertical social formats specifically, see captioning for mobile 9x16 video.

Which File Formats Carry Each Type?#

SRT deserves a specific warning. It is the most portable timed-text format in existence and carries almost no information — no positioning, no styling, no speaker metadata. It is excellent for moving text between tools and rarely acceptable as a final broadcast deliverable. Closed Caption Creator supports the full range of broadcast and streaming formats, and converts between them without a round trip through SRT.

Which One Does Your Delivery Spec Need?#

• US television broadcast — closed captions, delivered as SCC or MCC, or embedded as 608/708. Required by the FCC.
• Streaming platform (Netflix, Prime, Disney+) — SDH in the source language, plus translated subtitles per territory. Check the platform's timed text style guide for reading speed and line limits.
• Web video on your own site — WebVTT captions. If you are a US state or local government entity, the DOJ's ADA Title II rule sets WCAG 2.1 Level AA as the binding standard, with compliance dates now falling in April 2027 for larger entities and April 2028 for smaller ones.
• EU market — the European Accessibility Act has applied since June 2025. Our EAA video compliance guide covers the scope.
• Social and short-form — open captions burned into the picture, sized for a 9x16 safe area.

How to Create Captions and Subtitles#

Regardless of which type you are producing, the authoring work is the same up to the point of export: transcribe or import the text, time it to the audio, segment it into readable events, and check it against a style guide. What changes is the export target.

There are three ways the finished text reaches a viewer.

Sidecar file

A separate timed-text file travels alongside the video and the player renders it at playback. This is how closed captions, subtitles, and SDH all reach streaming platforms and web players. The advantage is flexibility: multiple languages, viewer-adjustable styling, and text you can fix without touching the video.

Embedded 608/708

Caption data is encoded inside the video essence itself, as an EIA-608 or CTA-708 stream. This is the broadcast path and the one live captioning uses. Our guide to embedding 608/708 captions walks through the process, and the embed and extract plugin handles it inside an existing edit workflow.

Burned into the picture

The text is rendered into the video frames during export, producing open captions. Once burned in, it cannot be removed or translated without a new render.

Professional caption and subtitle editors such as Closed Caption Creator handle all three from a single project, so the same timed text can ship as an SCC for broadcast, an IMSC for streaming, and a burned-in social cut without being rebuilt each time.

Click the screenshot to enlarge.

Screenshot showing Closed Caption Creator: a professional closed caption editor made for broadcast.

New to the process? Our complete guide to closed captioning video covers authoring end to end, and the Closed Captioning for Beginners course walks through it on screen.

Why the Distinction Matters#

Compliance

Regulators specify captions, not subtitles. An FCC complaint, an EAA obligation, or an ADA Title II requirement is not satisfied by a translated dialogue track. If a deaf viewer cannot tell that someone is knocking at the door, the deliverable has not done its job regardless of how clean the translation is.

QC and rejection risk

Each type is graded against different limits. A 608 caption file fails on a 33-character row that a subtitle file would accept without comment. A subtitle file fails on reading speed thresholds that captions are not measured against. Running the wrong style guide against a file produces false passes. Our caption and subtitle QC guide covers what to check for each.

Audience reach

Captions serve deaf and hard-of-hearing viewers first, but the audience is far broader: anyone watching in a noisy environment, on mute in a feed, or in a second language. Muted autoplay is the default on every major social platform, which is why open captions became standard there rather than optional.

Discoverability

Timed text is machine-readable. Search engines index caption and subtitle tracks, and the same text feeds transcripts, chapter markers, and translation workflows. A single caption pass can be repurposed into transcripts and translated subtitle deliverables without re-timing anything.

Frequently Asked Questions#

What is the difference between closed captions and subtitles?

Closed captions assume the viewer cannot hear the audio, so they include speaker identification and non-speech sounds alongside the dialogue. Subtitles assume the viewer can hear but does not understand the language, so they carry translated dialogue only. Captions are an accessibility deliverable; subtitles are a localization deliverable.

What does SDH mean?

SDH stands for Subtitles for the Deaf and Hard of Hearing. It carries the same information as a closed caption file, including speaker IDs and sound effects, but is delivered in a subtitle format such as IMSC, TTML, or a bitmap track rather than an EIA-608 or CTA-708 caption stream.

Are subtitles the same as open captions?

No. Open versus closed describes whether the viewer can turn the text off. Captions versus subtitles describes what the text contains. You can have open captions burned into a video, and you can have closed subtitles delivered as a toggleable sidecar file.

Do closed captions satisfy FCC requirements?

Only closed captions do. FCC rules under 47 CFR Part 79 require captions on US television programming and hold them to four quality standards: accuracy, synchronicity, completeness, and placement. Translated subtitles do not satisfy that obligation because they omit non-speech audio information.

How many characters can a closed caption line hold?

EIA-608 captions are limited to 32 characters per row across a maximum of four rows. Subtitle and SDH formats are more generous: Netflix and most streaming specs allow up to 42 characters per line across two lines for Latin-script languages.

Which format should I deliver: SRT, SCC, or IMSC?

Deliver SCC or MCC when the spec calls for broadcast closed captions, IMSC or TTML for streaming and ATSC 3.0 platforms, and WebVTT for HTML5 web players. SRT is the most portable format but carries no positioning or styling, so it is rarely accepted as a final broadcast deliverable.

Can I convert subtitles into closed captions?

Partly. Timing and dialogue convert cleanly, but a subtitle file has no speaker IDs or sound effect cues to carry over, and line lengths written for 42 characters will need reflowing to fit a 32-character caption row. The conversion is a starting point, not a finished caption deliverable.

---

Not sure which deliverable your spec is asking for? Talk to our team or start a free trial and export the same project to every format you need.

---

Resources#