What are sound descriptions in captions? SDH rules explained

Sound descriptions are the short notes inside captions that tell you what you cannot hear: a door slams, music plays, an audience laughs. Captions that include them are often called SDH, subtitles for the deaf and hard of hearing. This explainer covers the rules that TV and streaming captioners follow, how they differ, and which live captioning tools write sounds into the text of a conversation today.

What is a sound description in captions?

A sound description is a note in the caption text, usually in brackets, that names a sound other than speech, such as [door slams] or [applause]. The Captioning Key from the Described and Captioned Media Program (DCMP) defines sound effects as "sounds other than music, narration, or dialogue"¹. It says they are captioned "if it is necessary for the understanding and/or enjoyment of the media"¹.

Without them, a caption file only carries the words. A hearing viewer also hears the phone ring before a character answers it. A sound description gives that cue to a deaf viewer in the same moment.

What is the difference between SDH, closed captions and subtitles?

SDH subtitles are subtitles with extra information for deaf and hard of hearing viewers, while plain subtitles cover only spoken words. The captioning company Verbit says subtitles "are designed for hearing users", and that SDH adds "speaker tags, sound effects and other elements outside of speech"⁹. It describes closed captions as typically white text on a black background, which sets them apart from SDH⁹.

The words vary by country, and this is a vendor's explanation, not a standard. In the US, TV rules use the term closed captioning³. The sound rules in the next sections apply whatever you call the file.

What do US rules require for sound and music in captions?

US television captions must include music, sound effects and audience reaction to the greatest extent possible. The FCC rule at 47 CFR 79.1 says captioning "shall provide nonverbal information that is not observable, such as the identity of speakers", and covers "the existence of music" (with or without lyrics), sound effects and audience reaction³.

The same rule sets four quality standards: captions must be accurate, synchronous, complete and appropriately placed³. Synchronous means captions "shall begin to appear at the time that the corresponding speech or sounds begin"³. A [door slams] that appears two seconds late is a quality problem, not only a style problem.

This is a rule for television programming in the US. It does not tell a theatre, a lecture hall or a doctor's office what to do. For those, the Americans with Disabilities Act (ADA) asks for "effective communication", and ADA.gov lists open captioning, closed captioning and real-time captioning (CART) among the aids and services⁸.

How do the DCMP and Netflix rules compare?

The DCMP Captioning Key and Netflix's English style guide agree on brackets and lower case, and disagree on italics. The table sets the two side by side for the points that change what you see on screen.

Two rules from the DCMP matter for readers. Use specific words rather than vague ones, and avoid subjective words such as "beautiful" or "melodic" when describing music, so the note tells you the mood without giving an opinion¹. And caption background sound "only when they are essential to the plot"¹.

Why does the verb tense of a sound description matter?

Sound descriptions use the present tense because captions run in step with the sound. The Captioning Key says: "Never use the past tense when describing sounds"¹. It asks for the present participle for a sustained sound (a phone ringing) and the third person verb for an abrupt one (a door slams)¹.

For a deaf viewer this is practical. [phone rings] tells you the sound is happening now and you may want to look up. [phone rang] would suggest you already missed it.

Do live captioning tools describe sounds?

Some do, some do not, and the settings differ. The table lists what each tool's own page says. Sound labels in live tools are an extra layer on top of speech, and none of the pages read promises that every sound is caught.

Captio AI, a live captioning app for deaf and hard of hearing people, writes the sound inside the captions of a conversation. Its release notes show the bracket style the TV rules use, [doorbell rings], in the present tense⁷. Google's pages are the better choice when you use Android, and Apple's Sound Recognition is the better choice when you want an alert with no captions running⁴⁵⁶. The release notes do not give a full list of sounds or accuracy figures for Captio AI, so test the sounds that matter to you.

Where do sound descriptions help in TV, lectures and everyday life?

Sound descriptions matter most where a sound changes what the words mean. These are the settings the sources support:

• TV and media in the US: the FCC rule requires music, sound effects and audience reaction in captions to the greatest extent possible³. See TV and media.
• Streaming in English: Netflix's style guide asks for plot-pertinent sound effects, unless the picture makes them clear².
• Lectures and conferences: CART is one of the aids ADA.gov names for effective communication⁸. The sources read do not say whether CART captioners write laughter or applause. See lectures and conferences.
• Everyday conversations at home: Captio AI 1.1.5 shows a doorbell or an alarm in live captions next to the words⁷. See everyday conversations.
• Restaurants and group settings: Captio AI 1.1.5 lists laughter and music among the sounds it writes into live captions⁷. See restaurants and group settings.

What are the limits of sound descriptions?

A sound description is a short note chosen by a person or a program, so it can miss a sound or name the wrong one. TV rules ask for sounds "to the greatest extent possible" and the DCMP asks for judgement about which sounds matter¹³. Live tools add a second limit: they decide in real time, with no editor.

A caption is also not a safety alarm. For a smoke alarm or a fire, use a device made for deaf and hard of hearing people. For how Captio AI compares with Apple and Android alerts, read how to see a doorbell, alarm or crying baby in live captions. To see who is speaking as well, read how to see who is talking in live captions.

Start Live Captions when you want speech and sounds in one place. Captio AI is for in-person speech. It does not caption phone calls.