SDH stands for subtitles for the deaf and hard of hearing. An SDH track contains the dialogue plus the sounds a viewer cannot hear: who is speaking, music, and effects such as a door slamming or a phone buzzing. When a streaming menu offers both "English" and "English (SDH)", the second track is the one with those extra cues.
This guide explains how SDH differs from regular subtitles and closed captions, what an SDH cue looks like inside an SRT file, and how to translate SDH without losing the parts that make it useful.
What SDH subtitles include
A regular subtitle track assumes the viewer can hear the audio and only needs the words. SDH assumes they cannot. On top of the dialogue, an SDH track usually adds:
- Speaker identification when the speaker is off screen or not obvious, for example
[Maya]before her line. - Sound effects that matter to the story, such as
[door slams]or[phone buzzing]. - Music, either a description such as
[tense music]or the lyrics of a song that matters to the scene. - Manner of speaking when it changes the meaning, such as
[whispering]or[sarcastically].
Style rules differ between platforms. Netflix's English (USA) timed text style guide, for example, has a dedicated SDH section: speaker IDs and sound effects go in square brackets, in lowercase except for proper nouns, and a speaker ID sits on the same line as the dialogue it introduces.
SDH vs closed captions vs subtitles
| Feature | Subtitles | Closed captions (CC) | SDH |
|---|---|---|---|
| Written for | Viewers who can hear but need the words, often in another language | Viewers who are deaf or hard of hearing | Viewers who are deaf or hard of hearing |
| Sound effects and music | No | Yes | Yes |
| Speaker labels | Rarely | Yes | Yes |
| Language | Often a translation | Usually the language of the audio | Usually the language of the audio, and can be translated |
| Typical look | Subtitle styling at the bottom of the frame | Often white text on a black box, sometimes placed near the speaker | Subtitle styling at the bottom of the frame |
| Can be turned off | Usually | Yes | Usually |
Closed captions started as a television standard. The caption data travels with the broadcast signal and the viewer switches it on with a CC setting. SDH carries the same kind of information in an ordinary subtitle file, which is why Blu-ray discs, streaming services and online video platforms commonly offer SDH tracks.
Open captions are a fourth option: the text is burned into the video picture, so it is always visible and cannot be turned off or translated without re-exporting the video.
What an SDH cue looks like in an SRT file
SRT has no special field for SDH. The extra information is plain text inside the cue:
12
00:04:31,200 --> 00:04:33,900
[door creaks open]
13
00:04:34,100 --> 00:04:36,800
[Maya] Is anyone here?Because the descriptions are ordinary text, any tool that edits or translates SRT files treats them like dialogue. That keeps the file simple, but it also means the descriptions need the same review as the spoken lines.
When to deliver SDH
- A platform, broadcaster or client asks for an accessible subtitle track.
- Many viewers watch with the sound off, and the sound carries meaning: a joke built on a sound effect, an off-screen voice, a song that sets up a scene.
- You publish one subtitle file per language and want each file to work for viewers who cannot hear the audio.
Accessibility requirements differ by country and distributor, so confirm what your destination requires before delivery.
How to translate SDH subtitles
Translating SDH is mostly translating subtitles, with three extra checks:
- Translate the descriptions, not only the dialogue.
[door slams]should read[portazo]in a Spanish file, not stay in English. - Keep the conventions consistent. Use one style for speaker labels and sound descriptions in every cue. Character names normally stay as they are.
- Watch line length. Descriptions add characters to cues that may already be full. When a translated cue runs past the platform's line limit, shorten the wording instead of letting it overflow.
A regular subtitle file cannot become SDH through translation alone. The sound descriptions and speaker labels have to be written by someone who has listened to the audio. Add them in the source language first, then translate the finished SDH file.
Making and translating SDH files with Aivently
Aivently translates existing subtitle files, including SRT, WebVTT and ASS/SSA, into one or more languages. Bracketed descriptions and speaker labels are part of the cue text, so they are translated along with the dialogue while cue numbering and timing stay unchanged. The source and the translation appear side by side, and find and replace helps you keep labels consistent across the file.
Aivently's transcription turns audio or video into timed text you can edit. It does not identify speakers or describe sounds, so to build an SDH track from a transcript, add the speaker labels and sound descriptions in the editor before you export.
The file is measured before anything runs, so you see the exact credit total first: subtitle translation costs 1 credit per minute per target language, and transcription costs 1 credit per minute.
Translate an SDH subtitle file or transcribe audio or video to start a track you can label.