Tutorials Accessibility

Building a Caption Style Guide

Intermediate · ~16 min

Overview

Once more than one person captions your content, or once you outsource it, the questions start repeating: how do we write numbers, do we caption swearing, how do we identify speakers, what happens with song lyrics. A style guide settles them once. Most of the individual answers matter less than having one, because inconsistency between episodes is more noticeable to viewers than any particular convention. This guide covers what to decide and offers defensible defaults.

What You Need

  • A sample of your existing captions, to see what you already do inconsistently
  • A decision about which broadcast or platform standard you are broadly following
  • Agreement from whoever actually produces captions, including any vendor
  • A document short enough that people will read it, one page is ideal
  • Examples rather than rules, because people copy examples
  • A review date, since platforms and conventions change

Steps

1

Start from an existing standard rather than inventing one

Broadcasters and platforms publish caption specifications, and adopting one as your base saves settling dozens of questions from scratch. Note which you are following and document only your deviations from it. This also means new captioners with broadcast experience already know most of your rules.

2

Settle the mechanical questions first

Characters per line, maximum lines on screen, minimum and maximum display duration, and reading speed target. These are the parameters most tools can enforce automatically, and they determine legibility more than any stylistic choice. Get them from your chosen standard rather than deriving them.

3

Decide speaker identification and non-speech conventions

How speakers are labelled. Name, role, format, and when labels appear versus when they are omitted. How non-speech sound is written and which sounds warrant it. Whether music is described, and how lyrics are handled. These are the decisions that most affect whether captions feel coherent across a series.

4

Write down the editorial calls

Verbatim or lightly cleaned. Whether filler words are retained. How profanity is handled. How numbers, times, and currency are written. How accents and dialect are represented, if at all. These have no universally correct answer and cause the most disagreement, which is exactly why they need documenting.

5

Include line-breaking guidance with examples

Line breaks are the most neglected caption decision and one of the most damaging when done badly, breaking mid-phrase measurably slows reading. Give the rule (break at clause boundaries, keep articles with their nouns) and, more usefully, three worked examples of good and bad breaks.

6

Keep it to a page and give it to whoever captions

A long style guide is not read and therefore not followed. One page of decisions with examples, handed to every captioner and vendor, achieves far more than a comprehensive document filed somewhere. Review it when platforms change requirements or when a recurring disagreement reveals a gap.

Pro Tips

  • Consistency beats correctness. Viewers notice changes between episodes more than any individual convention.
  • Write it as worked examples rather than rules. People copy examples and misinterpret rules.
  • Enforce the mechanical parameters in your tooling so they do not depend on anyone remembering.
  • Include a short do-not list, the errors you keep seeing, which is often the most-used section.
  • Give it to vendors before they start, not as feedback on their first delivery.

What You'll Learn

A style guide exists to stop the same questions being answered differently each time. Below: what it must cover, and why the specific answers matter less than you would expect.

What a Caption Style Guide Has to Cover

A usable guide settles four categories of decision.

Mechanical parameters: characters per line, maximum lines on screen, minimum and maximum duration, reading speed target, and positioning rules for avoiding on-screen text. These are enforceable in tooling and determine basic legibility.

Identification conventions: how speakers are labelled and when, how non-speech sound is written, how music and lyrics are handled, and how offscreen or unidentified speakers are represented.

Editorial policy: verbatim versus cleaned, treatment of filler words and false starts, profanity, numbers and units, dialect and accent, and how to handle inaudible or unclear speech.

Line breaking: where to break and where never to, ideally as worked examples. This is the most neglected and one of the most consequential.

A guide covering these four in one page is more useful than an exhaustive document covering edge cases nobody encounters.

Why Consistency Matters More Than the Specific Choices

Teams spend disproportionate energy debating individual conventions (whether to write out numbers, whether to caption filler words) and comparatively little on applying whichever answer they chose.

This is the wrong allocation. For most of these decisions, both options are defensible and viewers adapt to either within a few minutes. What viewers do not adapt to is the convention changing between episodes, or between segments of the same programme, which is what happens when several people caption without a shared reference.

Inconsistency reads as carelessness and creates real friction: a viewer who has learned that square brackets mean sound effects is confused when they start meaning speaker names.

The practical implication is to decide quickly, document, and move on. Adopt an existing standard as the default, spend your deliberation only on the handful of decisions specific to your content, and put the effort into making sure everyone actually works from the same document.

Where This Fits

This guide covers one specific part of captions and access. The wider picture, caption formats, reading speed, speaker identification, what automatic captioning still gets wrong, and a practical QA pass, is in Beyond Auto-Captions: Caption Quality, Styling, and Readability, which frames the discipline as a whole and links out to the detailed guides underneath it, including this one. If you are starting from scratch rather than solving a specific problem, read that first and come back here.

FAQ

Q: Do I need a caption style guide for a small team?
A: As soon as more than one person captions, or you outsource any of it. Below that it is still worth a short note to yourself, because your own conventions drift over months. The threshold is really about whether captions are produced by more than one memory.

Q: Should I write my own or adopt an existing standard?
A: Adopt an existing broadcaster or platform specification as your base and document only your deviations. It settles dozens of questions immediately and means captioners with broadcast experience already know most of your conventions, which reduces onboarding and vendor briefing considerably.

Q: What is the most commonly missed item in a caption style guide?
A: Line breaking. Breaking a line mid-phrase forces the reader to hold an incomplete thought while their eye travels, which measurably slows reading, and it is rarely covered. Give the rule and, more usefully, several worked examples of good and bad breaks.

Q: How long should the guide be?
A: One page. A long document is not read and therefore not followed, so comprehensiveness actively works against the purpose. Cover the mechanical parameters, identification conventions, editorial policy, and line breaking, using worked examples rather than prose rules.

Translate this page

Machine translation provided by Google Translate, on Google’s servers. We do not check these translations and they will get technical terms wrong. The English page is the authoritative one. Following a link sends this page’s address to Google. Your browser may also offer to translate this page itself, which keeps the request on your device.