# Video captions: styles that are readable and on brand

[Canonical HTML page](https://gradiently.design/guide/video-captions-design)

A lot of social video is watched with the sound off, which makes captions the script. Here is how to set them so they read at a glance and still look like you.

## The short version

- Good video captions are short, high contrast, consistently placed and kept clear of the platform's own buttons and text.
- On vertical video, captions usually sit in the lower middle of the frame, above the area where the app draws captions, names and buttons.
- Subtitle style guides from broadcasters and streaming services commonly cap a line at around 37 to 42 characters, with no more than two lines on screen.
- Open captions are burned into the video and always visible; closed captions are a separate track viewers can switch on or off.
- A brand caption style should change type, colour and weight, never readability.

Readable **video captions** follow four rules: one or two short lines at a time, a size that reads on a phone at arm's length, strong contrast against whatever is behind them, and a fixed position clear of the platform's own interface. On vertical video that position is usually the lower middle of the frame, not the very bottom, where apps place their buttons and the post's caption. Once those rules are met, typeface, colour and emphasis are where your brand comes in.

## Open captions and closed captions

**Open captions** are part of the picture: burned into the video, always visible, styled exactly as you designed them. **Closed captions** are a separate text track that the player displays and the viewer can switch off, styled by the player rather than by you. Social video mostly uses open captions, because they survive every repost and play the same in every app. Long form video on your own site or YouTube benefits from a closed caption track too, which search engines and screen readers can use.

| Feature | Open captions | Closed captions |
| --- | --- | --- |
| Where they live | In the video frames | A separate file or track |
| Viewer can turn off | No | Yes |
| Your typography | Fully yours | Set by the player |
| Best for | Reels, TikTok, stories, ads | YouTube, websites, long talks |
| Editable after export | No, re-export needed | Yes, edit the track |

Many brands do both: styled open captions for short social clips, a closed track for long form.

Captions are also an accessibility requirement, not only a style choice. The W3C's guidance on [captions for prerecorded video](https://www.w3.org/WAI/WCAG22/Understanding/captions-prerecorded.html) explains what they need to include, such as speaker changes and meaningful sounds.

## Where to put captions on video

On a 1080×1920 Reel, TikTok or Short, the app lays its own interface over the frame: the account name and post caption near the bottom, and like, comment and share buttons down the right edge. Captions placed in the bottom quarter disappear under them. Place yours in the lower middle, slightly left of centre if the line is long, and keep them in the same spot for the whole video so the eye learns where to look.

- Reel and TikTok 9:16: 1080 × 1920
- Feed video 4:5: 1080 × 1350
- Landscape 16:9: 1920 × 1080

Vertical video needs captions well above the bottom edge. In 16:9, the classic lower position works because the player's controls hide when the video plays.

On landscape video, the traditional position just above the bottom edge, centred, is still right. On feed video at 1080×1350, keep captions above the bottom tenth or so, where some apps overlay a mute button and progress bar. Exact interface positions differ by platform and change often; [social media safe zones](https://gradiently.design/guide/social-media-safe-zones) keeps the current map.

## Line length, size and timing

- **Two lines at most.** Three lines is a paragraph, and nobody reads paragraphs over moving pictures.
- **Around 37 to 42 characters per line.** That is the range broadcasters' and streaming services' subtitle guides commonly use. Social captions are often shorter: three to six words per screen reads fastest.
- **Break at natural phrases.** Split after a comma or before a new clause, not between an article and its noun.
- **Big enough for a phone.** On a 1080 px wide vertical video, caption text in the region of 48 to 64 px tends to read comfortably. Check it on an actual phone.
- **On screen long enough to read twice.** If you can't comfortably read a caption aloud before it vanishes, it is too fast or too long.

## Caption styles that stay readable

There are only a few reliable ways to make text read over moving pictures. What they share is a consistent contrast between the letters and whatever sits right behind them. Fancy styles that break that consistency look good in one frame and vanish in the next.

| Style | Looks like | Works best on |
| --- | --- | --- |
| Solid bar | White or dark text on an opaque rounded box | Busy footage, talking heads outdoors |
| Heavy weight, no box | Bold type in one ink colour | Calm backgrounds and designed motion |
| Active word highlight | The spoken word changes colour or weight | Short punchy clips, tutorials |
| Outline | Letters with a thin dark stroke | Mixed footage, when a box feels heavy |

Pick one and keep it for the whole series. Avoid soft glows and halos behind captions: they smear and still fail on bright frames.

Vertical frame where a caption crosses light and dark areas of the background unevenly

Without a plan for contrast, part of the line falls on bright colour and part on dark. It reads in pieces.

Vertical frame where the gradient calms behind the caption so every word reads clearly

The same line where the background makes room: calm colour behind the words, consistent contrast throughout.

That second frame is how Gradiently handles words over its moving backgrounds. A Mark knows where the text is: its calmest, best contrasting region moves behind the words and auto ink picks light or dark type line by line, without boxes, shadows or blur. For footage you shoot yourself, a solid bar is often the honest answer. [Readable text on photos](https://gradiently.design/guide/readable-text-on-photos) covers the options for busy images.

## Making captions on brand

Brand captions change the voice of the text, not its legibility. Use your body or heading typeface if it has a bold weight that reads small; otherwise pick a clear sans serif and save the brand face for titles. Choose one ink colour and one highlight colour from your palette and use them in every video.

- Caption ink: #ffffff
- Bar or outline: #111827
- Active word: #facc15
- Speaker two: #38bdf8

A complete caption palette: one ink, one dark support, one highlight, and a second colour only if two people speak.

1. **Write or check the transcript** Fix names, numbers and product terms first. Captions with typos are worse than none.
2. **Split into caption chunks** Three to six words per screen for social, up to two lines for longer talks. Break at natural pauses.
3. **Set one style** Typeface, weight, size, ink, highlight and position. Save it so every video matches.
4. **Place against the safe zone** Preview at full size with the platform's interface in mind, not just in the editor.
5. **Watch it muted** Play the video on a phone with the sound off. If the story still makes sense, the captions work.

With Pro, Gradiently's video editor has a caption track on its timeline: write captions one by one or import an SRT or WebVTT file, set the text size, ink and background colour, and the captions are burned into the exported video. The walkthrough is in [add captions to video](https://gradiently.design/guide/video-timeline-and-captions), and if you are starting from a recording, [talking head video](https://gradiently.design/guide/talking-head-video) covers framing and sound.

## FAQ

### Where should captions go on a vertical video?

In the lower middle of the frame, above the bottom area where apps place the account name, post caption and buttons, and away from the right edge where the action buttons sit.

### How many words should be on screen in video captions?

For social video, three to six words per screen reads fastest. For longer talks, up to two lines of around 37 to 42 characters each is the range subtitle guides commonly use.

### What is the difference between open and closed captions?

Open captions are burned into the video and always visible. Closed captions are a separate track that viewers can turn on or off, styled by the player.

### What font is best for video captions?

A clear sans serif with a bold weight and open letter shapes. Use your brand typeface if its bold weight reads well at small sizes on a phone.

### Do captions help people who watch with sound on?

Yes. They help in noisy places, for people who are deaf or hard of hearing, for viewers reading in a second language, and for anyone who misses a word.