SDH subtitles: what they are and how to write them
Seen SDH in a subtitle menu, or asked to deliver it? Here is what SDH subtitles contain, how they differ from closed captions, how to turn them on, and how to make a file.

Subtitles for the deaf and hard of hearing, shortened to SDH, are same-language subtitles written for viewers who cannot fully hear the soundtrack. Besides the spoken lines, an SDH track describes the sounds and music that matter and names any speaker the picture does not show, in a file built like ordinary subtitles. Closed captions serve the same viewers, but the term comes from US broadcast captions, which follow their own technical and display rules.
SDH at a glance
If someone asks you for SDH, they want more than the words people say. The table compares the three kinds of timed text you are likely to meet, using the Netflix Partner Help Center, MDN and the DCMP Captioning Key. Treat it as a general map, not a delivery spec.
| Plain subtitles | SDH | Closed captions | |
|---|---|---|---|
| Who it is for | Viewers who hear the audio but not the language | Viewers who cannot fully hear the audio | The same viewers as SDH |
| Dialogue | Yes, often translated | Yes, in the language of the audio | Yes, in the language of the audio |
| Speaker names | No | Yes, when the speaker is unclear | Yes, when the speaker is unclear |
| Sound effects | No | Yes, in square brackets | Yes, in brackets |
| Music | No | Mood labels, song titles, lyrics with ♪ | Music labels and lyrics with ♪ |
| Position | Centered, near the bottom or the upper edge | As subtitles, moved off on-screen text | Can sit under the speaker or near the edge |
| Format and delivery | Timed text file: SRT, VTT, TTML | Timed text file built like subtitles | US broadcast captions with their own display rules |
| On or off | Usually switchable, or burned in | Usually switchable, or burned in | Closed means the viewer can switch them off |
What SDH means
The name spells out the audience: people who are deaf or hard of hearing. Netflix describes SDH as same-language text that carries dialogue, music and sounds for viewers who cannot fully hear, built and styled like subtitles but matching closed captions in audience and content. In practice it is a normal subtitle file with extra lines for sounds and speakers.
UK broadcasting uses one word for both. Ofcom's guidelines for TV and on-demand providers just say subtitles, and note that subtitles for d/Deaf viewers hold more information than translation subtitles.
What goes into an SDH file
Beyond the dialogue, SDH adds speaker names, sound effects, music, and notes on tone. The guidelines handle each one differently, so every rule below names its source.
Speaker names
Netflix's English (USA) guide puts speaker IDs in lowercase square brackets, except for proper nouns, and only when the picture does not show who is talking. Two speakers in one subtitle each get a line that starts with a hyphen.
The DCMP Captioning Key puts a known name in parentheses on its own line, such as (Lena), and only after the audio or on-screen text has introduced it. Ofcom leaves the method open: colour, punctuation or position can mark a speaker, with a label when the source of speech is not obvious.
Sound effects
Netflix and DCMP both use lowercase brackets. Netflix asks you to describe what is heard, not what is seen, and to leave out sounds the picture already shows. DCMP names the source unless it is visible, and uses the -ing form for ongoing sounds, [water running], and the plain verb for sudden ones, [cup smashes]. Ofcom asks for specific labels, so [rain on the window] says more than [noise].
Music and lyrics
Netflix puts a music note and a space at each end of a lyric subtitle and writes lyrics word for word. It names a song in brackets, such as ["Late Train Home" playing], only when the title is confirmed, and otherwise uses a mood or genre label like [soft guitar music]. DCMP adds two notes after a song's last line and skips background music under 5 seconds. Ofcom accepts a # sign or a music note.
Tone and off-screen speech
When a face does not show the feeling, DCMP puts the manner of speech in brackets before the line, such as [nervously], and marks silence when lips move with no sound. Netflix asks labels to describe volume, speed and voice quality.
For off-screen voices, Netflix italicizes narration and speakers who are not in the scene, but not someone who is only out of shot, and it never italicizes the bracketed labels. DCMP moves the caption toward the side of the screen where the speaker is.
Here is a short SDH file in SRT that follows the Netflix conventions. Omar is off screen in cue 2, so he gets a label.
1
00:00:01,200 --> 00:00:03,000
[kettle whistling]
2
00:00:03,400 --> 00:00:06,100
[Omar] Can someone get that, please?
3
00:00:06,500 --> 00:00:09,200
-I'm on a call.
-[laughs] Fine, I'll get it.
4
00:00:10,000 --> 00:00:13,400
♪ Wait for me down by the station ♪
How SDH differs from closed captions and plain subtitles
SDH and closed captions
Netflix says SDH and closed captions serve the same viewers with the same kind of text. The differences are technical: closed captions allow fewer characters per row and use their own display and formatting methods. Netflix treats closed captions as an older US broadcast term and makes SDH, not CC, for its branded content.
- Format: SDH ships as a subtitle-style timed text file. The W3C names WebVTT, SRT and TTML as caption formats, and Netflix's TTML rules cover subtitles and SDH alike.
- On or off: the W3C notes that viewers can show or hide closed captions, while open captions always stay on. Burned-in SDH becomes open captions.
- Placement: Netflix centers SDH near the bottom or the upper edge, clear of on-screen text. DCMP captions sit on the bottom two lines or under the speaker.
- Regional words: the FCC rule for US television, 47 CFR 79.1, says closed captioning. Ofcom in the UK says subtitles, and the W3C notes that some regions use subtitles for both.
Follow the name on the spec sheet. If it says SDH, send a subtitle file with the extra content. Our guide to subtitles vs captions covers the wider naming question.
SDH and plain subtitles
Plain subtitles carry speech. MDN's page on the track element says subtitles suit viewers who hear the sound but do not understand it, while captions add sound effects, music cues and who is speaking. Netflix adds that its subtitles are usually translations.
SDH also keeps more of the speech. Netflix's English SDH rules say not to paraphrase, and Ofcom asks for speech word for word as closely as possible. DCMP allows edited captions for educational media, but keeps quotes and song lyrics verbatim.
Take one moment at 00:00:21,000. The plain subtitle reads Wait for me! The SDH cue reads [car horn honking] on the first line and [Omar, shouting] Wait for me! on the second.
When you need SDH
- Streaming deliveries. Netflix expects the subtitle and SDH files it receives to follow its timed text style guides. Other buyers have their own specs, so read the delivery document first.
- Accessibility rules. The FCC rule for US television sets four caption quality standards: accuracy, synchronicity, completeness and placement. It also asks captions to identify speakers and note music and sound effects where possible, and Ofcom's UK guidelines ask for the same.
- Sound-off viewing. Ofcom notes that many people without hearing loss use subtitles, younger viewers in particular, and that subtitles help in noisy or public places. MDN lists muted playback as a case for captions.
For social video, the last point applies most often. A clip on mute loses the phone buzzing and the off-screen voice unless the captions carry them. See video accessibility and captions for more.
How to turn on SDH when you watch
Viewers meet SDH as a track in the player's subtitle menu. Netflix's help page says a title with SDH shows an SDH badge on its details page, and that some seasons or episodes lack it. In that case the subtitles may carry dialogue only, without sound or music context.
To pick a track on Netflix, start playing, tap or click the screen, or press up or down on a TV remote, then choose Audio & Subtitles. In the Apple TV app, the Subtitles or Languages button in the playback controls does the same job.
Apple's help page groups the two labels. If no SDH option appears on an iPhone, iPad or Apple TV, open Settings, then Accessibility, choose Subtitles and Captioning under Hearing, and turn on Closed Captions + SDH to use those tracks by default. Either way, a menu entry marked CC or SDH is the one that adds sound and music context.
How to write SDH: rules from each guideline
The guidelines share a goal but differ on details, so pick one per project. Our guides to subtitle timing and reading speed and subtitle formatting rules go deeper.
Netflix English (USA) Timed Text Style Guide
- Up to 42 characters per line and two lines per subtitle.
- Up to 20 characters per second for adults, 17 for children's programs.
- On screen from 20 frames at 24 frames per second to 7 seconds.
- Labels in lowercase square brackets, never in italics.
DCMP Captioning Key
- Mixed case, with capitals only for shouting.
- Two lines preferred, on screen for 40 frames to 6 seconds.
- For educational media, at most 130 to 160 words per minute by level, labels included.
Ofcom guidelines (UK)
- Word for word where possible, with light paraphrase for very fast speech.
- Placed low, but moved off the speaker's mouth or face.
- Lines split at natural breaks, with the BBC's guidelines cited for detail.
MDN documents two WebVTT tools: the voice span, <v Name>, and the line setting, where line:0 lifts a cue to the upper edge. The voice span is styling markup, so still type names in brackets for viewers.
WEBVTT
00:00:14.000 --> 00:00:16.500 line:0
[rain on the window]
00:00:17.000 --> 00:00:19.800
<v Lena>Close it, the floor is getting wet.</v>
How to make an SDH file with Vid Hub
Vid Hub writes captions from the speech in your video. It does not add speaker names or sound descriptions by itself, so you type the SDH layer.
- Open the caption editor at vid-hub.com and add your video. No account is needed to start. You get captions for up to 10 minutes of a video free, and paid plans caption up to 20 minutes of a video.
- Check every caption's text and timing against the audio, and fix what is wrong.
- Type speaker labels and sound descriptions into the captions, such as [Omar] or [phone buzzing], and adjust the timing so each label shows with its sound.
- On a paid plan, download the captions as .srt or .vtt from the arrow beside Export. The .vtt file escapes &, < and > and leaves out empty captions.
- Need another format or fixed timing? Use the subtitle converter, SRT to VTT or the subtitle shifter.
Because of that escaping, a voice tag typed into a caption reaches the .vtt file as plain text, so add tags to the exported file if your spec needs them. To label an SRT or VTT file you already have, use the free subtitle editor.
The free export is a 480p MP4 with the captions burned in and a vid-hub.com mark. That is an open caption, and Ofcom encourages timed text files so assistive technology can read them. For an SDH delivery, send the .srt or .vtt, and see burned-in vs soft subtitles for the trade-off.
Common SDH mistakes
- Describing pictures. [he frowns] is visual. Netflix wants sounds, such as [scoffs].
- Naming every speaker. Netflix labels only speakers the picture does not show.
- Mixing brackets and parentheses for names. DCMP lists consistency as an element of quality captioning.
- Captioning every noise. DCMP keeps background sounds only when they matter to the plot.
- Vague labels. Ofcom asks for specific descriptions, so write [upbeat brass band], not [music].
- Guessing song titles. Netflix names a song only when the title is confirmed.
- Writing [speaking foreign language]. Netflix asks you to name it: [in Portuguese] for translated lines, [speaking Portuguese] for untranslated ones.
- Burning in when a file was requested. Viewers cannot switch open captions off.
Frequently asked questions
What is the SDH option in a subtitle menu?
It is a same-language track for deaf and hard of hearing viewers, with bracketed sounds, music notes and labels for unseen speakers. Choose it on mute or when the audio is hard to follow.
Are SDH and closed captions the same?
They serve the same viewers with nearly the same content. Netflix separates them by technical spec and style: SDH is built like a subtitle file, closed captions follow US broadcast methods.
Should I pick English or English SDH subtitles?
Pick English SDH when sounds, music or off-screen voices matter to you. Netflix notes that without SDH, a title's subtitles may carry dialogue only.
Why are some subtitles in square brackets?
Brackets mark text that is not speech, such as [phone buzzing]. Netflix and DCMP use them for sound effects, and Netflix also puts speaker names in them.
What file format should SDH be delivered in?
Follow the buyer's delivery spec. Netflix sets TTML rules for subtitles and SDH, and W3C WAI describes WebVTT as the most common caption format on the web.
Can I add SDH labels without paying?
Yes. The free subtitle editor adds labels to an SRT or VTT file you already have, and Vid Hub writes captions for up to 10 minutes of a video free. Downloading those captions as .srt or .vtt needs a paid plan.
Should SDH be burned into the video?
Usually not for a delivery, since viewers expect to switch closed tracks on and off. Burned-in SDH stays on for everyone, which can still suit social clips watched on mute.
How many characters per line does SDH allow?
Netflix's English (USA) guide allows 42 characters per line, two lines and 20 characters per second for adult content. DCMP sets no character count but prefers two lines.
Try it on your own video
Open the editor, add your video, and get captions you can edit, style and export.
Upload your videoNo account needed to start.
Sources
We read each page below on the date shown. Rules, prices and plans change, so check the page itself before you rely on it. A source we cite does not endorse Vid Hub.
- Netflix Partner Help Center: English (USA) Timed Text Style GuideRead 2026-11-07.
- Netflix Partner Help Center: Timed Text Style Guide: General RequirementsRead 2026-11-07.
- Netflix Partner Help Center: What is the difference between subtitles, SDH and CC?Read 2026-11-07.
- Netflix Partner Help Center: Do all subtitles need to adhere to the Netflix timed text style guides?Read 2026-11-07.
- Netflix Help Center: How to use subtitles, captions, or choose audio languageRead 2026-11-07.
- Apple Support: Change subtitles or languages for videos in the Apple TV appRead 2026-11-07.
- DCMP: Captioning Key, Elements of Quality CaptioningRead 2026-11-07.
- DCMP: Captioning Key, TextRead 2026-11-07.
- DCMP: Captioning Key, Presentation RateRead 2026-11-07.
- DCMP: Captioning Key, Sound Effects and MusicRead 2026-11-07.
- DCMP: Captioning Key, Speaker IdentificationRead 2026-11-07.
- DCMP: Captioning Key, Special ConsiderationsRead 2026-11-07.
- U.S. Government Publishing Office: 47 CFR 79.1, Closed captioning of televised video programming (FCC rule)Read 2026-11-07.
- Ofcom: Guidelines on Providing Television and On-Demand Access ServicesRead 2026-11-07.
- W3C Web Accessibility Initiative: Captions/SubtitlesRead 2026-11-07.
- MDN: <track>: The Embed Text Track elementRead 2026-11-07.
- MDN: Web Video Text Tracks Format (WebVTT)Read 2026-11-07.


