Text to Speech and Accessibility: What It Helps and What It Does Not

Text to speech is often described as an accessibility tool, and for many people it genuinely is. A person who finds long passages hard to read, who has low vision, who is learning a language or who simply prefers to listen can benefit enormously from hearing text read aloud. It is worth being clear, though, about what it does and does not do, because adding a narration is not the same as making content accessible, and the difference matters.

This article sets out the real benefits, the genuine limits and the habits that make content usable by the widest range of people. It does not claim any standard or legal compliance, because that is a matter for the relevant guidelines and, where appropriate, a specialist.

Who benefits from spoken text

• People with dyslexia, for whom listening can be easier than reading.
• People with low vision, who may listen rather than read small text.
• Language learners, who can follow the text and the sound together.
• People who are busy, driving or doing something else with their hands.
• Anyone who finds a long page tiring and prefers to listen.

What narration does not do

A narrated video is not accessible to everyone merely because it has a voice. A deaf or hard of hearing viewer needs captions. A viewer who is blind needs the visual information described, not only read, which is called audio description. A person using a screen reader needs text that is structured, labelled and navigable. Narration serves some of these people and leaves others untouched, so it should be one part of a larger approach.

Captions first

If you can do only one thing, caption your videos. Captions serve deaf and hard of hearing viewers, people watching with the sound off and anyone in a noisy place, and they help non-native speakers follow along. They also make the content searchable. Good captions are accurate, correctly timed and include meaningful sounds where relevant, not only speech. The guide on subtitles on this site covers how to write ones that are readable.

Provide a transcript

A transcript is a plain text version of the content, and it serves more people than almost anything else. A screen reader can read it, a search engine can index it, a reader can skim it faster than they can watch the video and a person with a slow connection can use it when the video will not load. The transcript tool on this site retrieves existing captions from a video, which is a quick way to get started.

Describe what is on screen

For viewers who cannot see the picture, narration that refers to the screen is only useful if it also says what is there. A line like as you can see here tells a blind viewer nothing. Replace it with a description: the graph shows sales rising from March to June. This is good practice for everyone, since sighted viewers also benefit from being told what to look at.

Make the voice itself usable

• Choose a clear voice at a moderate pace.
• Avoid loud background music that competes with speech.
• Keep the audio level consistent through the video.
• Provide a way to pause and replay, which any standard player does.
• Do not rely on sound alone to convey essential information.

Accessibility standards and legal requirements vary by country and by sector. If you are producing content for an organisation with obligations in this area, consult the relevant guidelines or a specialist, rather than relying on a general article.

A practical order

If you are starting from nothing, do these in order: add captions, publish a transcript, describe what appears on screen in the narration and keep the audio clear. Each step helps real people, costs little and builds on the one before. Narration has a part to play in that, and it is one part, not the whole.