Add audio description to a short video with ViddyScribe

If a tutorial says “click here” while its cursor moves silently, the soundtrack is missing an instruction. Audio description supplies important visual information for someone who cannot see it. The job is to make the action understandable without requiring the listener to reconstruct it from hints.[4]

ViddyScribe can draft descriptions, generate narration and assemble a described video. This walkthrough follows its September 2026 documentation for the Enterprise workspace; we have not tested generation firsthand. Start with a finished video lasting a minute or two that you are entitled to upload.[1]

Before you begin

You need a ViddyScribe Enterprise account. The official pricing page lists Starter at $100 a month, including 50 minutes, with additional minutes at $2 each. A seven-day trial is offered by request, with watermarked outputs restricted to non-commercial use. Check the charge before starting.[3]

Use a computer or tablet for editing; the editor does not support phones.[2]

The free individual plan serves people with low vision or other accessibility needs and gives uploads Extended AD. This guide uses Enterprise’s Standard AD option and review controls.[3]

1. Decide where the extra words will fit

Watch the source video once and note information the soundtrack leaves out: a control’s name, an important action, a slide’s conclusion or text that appears only on screen. W3C’s guidance prioritises information needed to understand the video; it does not require a running inventory of every visible object.[4]

2. Generate a draft you can review

1. Select New Generation in Library, upload your video, then Continue.
2. Choose AD Video for a finished described MP4, or AD Audio if you need audio files. Below the output cards, check Review script before rendering, or Review script before creating audio.
3. Select Standard or Extended AD, choose the language and voice, and listen to the voice sample. Add any useful custom instructions.
4. Select Generate. With review enabled, the service prepares the script and waits before making the final audio or video.

For a software demonstration, an illustrative custom instruction is: Describe the control selected and the result that appears. Use the visible interface labels. Keep descriptions brief enough to fit the chosen timing. This is a starting instruction, not an example of output we generated.

3. Edit for the listener

When the draft is ready in Library, open Review & render, or Review script for audio. Play the video while reading the AD Script tab. Select a description to change its wording or timing; add or delete descriptions where needed. Edit with AI can shorten or reword a passage. Put a mispronounced name in the Pronunciation dictionary.[2]

Use present-tense descriptions of visible actions and include essential screen text. Avoid guessing intentions or adding an interpretation the image does not establish. W3C recommends placing narration with or just before the relevant visual information, rather than making the listener wait until after the action.[4]

For an illustrative tutorial showing a person choosing a folder, “She confidently organises her files” adds an unsupported judgement and omits the useful action. If the screen actually shows it, “She selects the Invoices folder, then clicks Move” gives the listener something concrete to follow. Check the words against your own clip.

Preview edited descriptions in the free robotic voice to judge wording and timing. To hear the selected narration voice, use the voice menu’s Use this voice control. The originally selected voice is included; a different voice can add a charge. Save version, then Render, and choose the required output.[2]

4. Export the version people can use

Open Export or download the completed files from Library. Keep the script as well as the final output so the wording and timing remain available for later correction. ViddyScribe also offers captions and a transcript; these are separate downloadable assets.[2]

For a self-contained described version, choose AD Video, an MP4. For a compatible player’s additional audio track, use Standard AD’s Mixed audio, which combines the original soundtrack and descriptions. The AD-only file contains narration alone and is intended for mixing. Extended AD changes the runtime, so it needs its own video.[3]

If the draft contains too few descriptions, Standard AD may be running out of pauses; try Extended AD. For an extra audio track, use exactly the same video edit so timings match.[1]

If narration competes with dialogue, shorten it or adjust its timing. Confirm that essential descriptions remain audible and arrive with or before the action.[4]

Play the exported result from beginning to end with the picture out of view. Can you follow the actions, identify the controls and understand the final state? Then check the picture for accuracy. A successful finish is a usable described clip whose instructions survive being heard on their own.

Report an error or suggest a correction