Audio Description

Audio description, sometimes called video description or described video, is a supplementary audio narration track that conveys visual information in video content. A trained describer narrates key visual elements, including physical actions, scene changes, on-screen text, facial expressions, and other visual details, during natural pauses in dialogue and narration. This additional audio track is mixed with the original soundtrack so that people who are blind or have low vision can fully understand the video content. Audio description focuses on what sighted viewers take for granted: the visual storytelling that happens between spoken words. In a film, a character's silent reaction, a significant object in the background, or a change of setting all carry meaning. Without audio description, someone who cannot see the screen misses this information entirely, leaving gaps in comprehension that range from minor confusion to complete loss of narrative. The practice has roots in theater, where live describers have narrated performances for blind audience members since the 1980s. It has since expanded to television, film, museum exhibits, and web video content. In the context of web accessibility, WCAG establishes specific requirements for audio description at different conformance levels.

Video is one of the most dominant content formats on the web. Educational platforms, news sites, marketing pages, social media, and corporate training all rely heavily on video. For the estimated 2.2 billion people worldwide with vision impairments, video content without audio description presents a significant barrier. Consider an instructional video that demonstrates a software interface. The narrator says "click the button in the upper right corner," but without audio description, a blind user doesn't know what button, what it looks like, or what happens visually after clicking it. The visual demonstration, which is the core value of the video, is lost. WCAG addresses audio description at two levels. Success Criterion 1.2.3 (Level A) requires either audio description or a full text alternative for prerecorded synchronized media. Success Criterion 1.2.5 (Level AA) requires audio description specifically. Since most accessibility policies target Level AA conformance, audio description is effectively required for prerecorded video content on covered websites. Beyond compliance, audio description is a matter of equal access. Videos that convey important information, whether educational content, product demonstrations, safety instructions, or entertainment, should be accessible to everyone in the audience.

Creating audio description involves several steps: **Script writing.** A describer reviews the video and writes a script that identifies visual information essential to understanding the content. The script must fit within natural pauses in dialogue and narration. Good description is concise, objective, and prioritizes the most important visual information. **Recording.** A voice actor or trained describer records the description script. The voice should be clear, neutral in tone, and distinguishable from the main audio track. **Mixing.** The description audio is mixed with the original soundtrack, typically on a separate audio track that users can enable or disable. On the web, this is often implemented as an alternative video file or a selectable audio track within the video player. **Implementation on the web** can take several forms: ```html <!-- Using the track element for description --> <video controls> <source src="presentation.mp4" type="video/mp4"> <track kind="descriptions" src="descriptions.vtt" srclang="en" label="Audio Descriptions"> </video> ``` Note that browser support for the `descriptions` track kind is limited. Many implementations use an alternative approach: providing a separate version of the video with audio description baked into the main audio track, or using a player that supports audio description toggling. **What to describe:** - Actions and movements that are essential to the narrative - Scene changes and locations - On-screen text, titles, and credits - Character appearances and expressions when relevant - Visual effects or demonstrations - Silent actions that advance the plot or convey information **What not to describe:** - Information already conveyed through dialogue or narration - Purely decorative visuals that don't affect comprehension - Every minor visual detail (prioritize what matters for understanding) **Extended audio description** (WCAG 1.2.7, Level AAA) pauses the video when natural pauses are too short for adequate description. This is appropriate for visually dense content like fast-paced action sequences or complex demonstrations. AI-generated audio description is an emerging field that may make the process more scalable, though human review remains important for accuracy and quality.