What audio description is
Audio description is a spoken account of what is visible: the image, the object, the scene or the action on screen. It is made mainly for blind visitors and visitors with low vision, though many other people use it too, such as visitors who are tired, who are looking elsewhere or who simply want more detail. It is different from a caption, which gives text for sound, and different from an ordinary label, which usually identifies and interprets an item without describing it.
A good description lets someone form a useful mental picture and understand why the item is in the exhibit. It does not need to be a complete inventory of every detail.
Where it is needed
Wherever the meaning of an exhibit depends on something seen, there is a need for a non-visual route. Common places include:
- Photographs, paintings, maps and diagrams on screens or walls.
- Objects in cases, especially when the visitor is meant to look closely.
- Video and animation, where visuals carry information that the soundtrack does not.
- Interactive screens, where a visitor should be told what is on screen and what can be done.
The last of these overlaps with screen readers, which handle interface text and controls, whereas audio description deals with imagery and content. Both are usually needed.
Writing the description
Start with what matters
Begin with a short identification: what kind of thing it is, roughly how large or small, and what it is made of or how it was made, if that is relevant. Then give the overall composition before the details. For a portrait, that might be who or what is shown, the pose, the setting and the general mood, followed by notable details such as clothing, objects or expression.
A reasonable order is:
- What it is, and its size or format.
- The main subject and overall arrangement.
- Important details, in a consistent order such as left to right or foreground to background.
- Color, texture and light, where they carry meaning.
- Anything the visitor needs to connect the description to the exhibit’s story.
Be objective, then interpret
Describe what can be seen before saying what it means. Say that a figure holds a raised hand and is turned away from the crowd, then let the interpretive label or the curator’s text explain the significance. Where an observation involves judgment, such as describing a face as sad, consider describing the features and then, if useful, the impression. Avoid telling visitors what to feel.
Use plain, concrete language
Short sentences and everyday words work best, which is the same advice given in plain language. Compare size with familiar things, but check that the comparison is accurate and widely understood. Explain unfamiliar terms briefly. Describe colors in simple terms, since many people who are blind have a good understanding of color from memory or concept, and the meaning of a color in a picture may matter.
Be careful with people
When describing people, include what is relevant to the item and what the original source or the subject would recognize. Avoid guessing at identity, ethnicity, gender or disability from appearance unless the exhibit text supports it or the information matters to the story. If the material is sensitive, the guidance on sensitive content applies to the spoken version too.
Keep length manageable
There is no single right length. Many teams offer a short description of a few sentences, with a longer one available on request for visitors who want more. A very long description can tire the listener, while one that is too thin may leave out the point of the image. Writing a short version first and a fuller one second is a helpful discipline.
Video and animation
For moving material, the description has to fit around the existing audio. Watch for natural pauses and write short descriptions of key visual events, such as who has entered, what has changed on screen, or what text appears. If there are no gaps, options include a separate version with extended pauses, or a version where the narration itself is written to carry the visual information. The latter is often the better approach when a piece is still in development, because building description into the script from the start produces a smoother result than adding it later. Teams that make short exhibit videos or motion graphics benefit from planning this early.
On-screen text should be read aloud if it is not otherwise spoken.
Delivering it
Audio description can be offered in several ways: through an audio guide, a handset, a phone linked by a code, a button on the exhibit, or built into the media player. Whichever it is, make it easy to find without sight, and do not require a visitor to navigate a complex visual menu to reach it. Think about sound too. A speaker playing description in an open gallery can disturb others, so private listening via headphones or a handset is often better, and sound bleed needs thought where speakers are used.
A transcript of the description supports deaf-blind visitors and people who prefer text, and it is a natural companion to the work on captions and transcripts.
Voice and recording
A clear, even voice that does not overact tends to suit description. Some teams use a different voice from the main narration so the listener can tell them apart. Record in a quiet space, read the script aloud beforehand to catch awkward phrasing, and check pronunciation of names and terms. Keep the script alongside the audio so it can be revised.
Testing and review
The most valuable step is to test with blind and low vision visitors, ideally before final production. Ask whether they could picture the item, whether anything was confusing or missing, and whether the length felt right. A short session with a few people often reveals more than any checklist. Update descriptions when displays change, and accept that descriptions are rarely perfect on the first try.
Published guidelines on describing visual content exist, and consulting them is sensible. Consider this article as a starting point and not a standard.