What audio description is
Audio description is an additional narration track that explains important visual information in a video: actions, settings, on-screen text, gestures and changes of scene that are not clear from the dialogue and sound alone. It is designed mainly for blind and low-vision viewers, though some other people find it useful too, for example when watching without looking at the screen.
It is different from captions. Captions present speech and key sounds as text, mostly for people who are deaf or hard of hearing. Audio description does the opposite job: it turns the visuals into speech. A fully accessible video often needs both.
What gets described
A good description covers what a viewer would otherwise miss, and only that. Typical content includes:
- who has entered or left, and what they are doing
- the setting and any change of location
- facial expressions or gestures that carry meaning
- text shown on screen that is not read aloud, such as titles, names or slides
- visual jokes, charts or demonstrations that the dialogue assumes you can see
Describers generally aim to be objective and brief. They say what is visible rather than interpreting motives, and they avoid spoiling a reveal before it happens on screen. The aim is to let the viewer form their own understanding.
Standard versus extended description
Standard description fits into natural pauses in the original soundtrack. If a video has dense dialogue or continuous narration, there may not be room, and the description can end up too thin to be useful.
Extended description pauses the video so a longer explanation can play. This needs a player and format that support pausing, and it is more common in educational or technical material, where a diagram or demonstration needs real explanation. Which approach is possible often depends on the platform and the way the video is delivered.
How descriptions are delivered
There are a few common approaches, and availability varies by platform:
- A separate audio track. The describer’s voice is mixed with the original sound as an alternate track the viewer can select. Playback features of adaptive bitrate streaming can carry multiple audio tracks, though what a given player exposes differs.
- A second version of the video. Some creators publish a described version alongside the standard one, which works almost anywhere but doubles the files to manage.
- A text description. A written transcript that includes visual detail can help people using screen readers or braille displays, and it is sometimes the simplest option for short material.
Whichever method you pick, test it with the actual player your audience uses, since support for switching audio tracks is uneven.
Writing and recording descriptions
The usual workflow is to watch the video, find gaps in the dialogue and write short lines for them. A script is then voiced by a person or, increasingly, a synthetic voice. Human narration tends to carry tone better, while synthetic voices can lower cost and speed for some uses. Quality varies, and it is worth having a blind or low-vision person review the result if you can, since sighted creators often over-describe or miss what matters.
The description track should be balanced against the original audio so neither drowns the other, which is a normal part of audio mixing. Clear original sound also helps, because good audio quality makes narration easier to place.
Designing so less description is needed
Some videos are easy to describe and some are nearly impossible without heavy narration. A few habits reduce the gap:
- Say important things aloud. If a presenter says what they are pointing to or what a chart shows, the main content survives without sight.
- Read key on-screen text rather than only displaying it.
- Avoid relying on colour or gesture alone to convey meaning.
- Leave occasional pauses in dialogue where visuals carry the story.
This kind of planning is sometimes called making the soundtrack self-sufficient. It does not replace description for visually driven scenes, but it often shortens the script and improves the video for everyone, including people listening while doing something else.
Rules and expectations
Some organisations, broadcasters and public bodies have requirements for audio description in certain cases, and these differ by country and sector. This article does not cover specific legal duties; if you work for an organisation that may be covered, check the rules that apply to you. The wider set of expectations around captions, controls and description is outlined in video accessibility guidelines.
Offering description is not an all-or-nothing choice. Even partial coverage of the most important visuals, done carefully, can make a real difference for viewers who would otherwise be left guessing.