Captions

Captions That Stay Readable on Any Screen

Captions are read at speed on a small screen. Two lines, one position, and a contrast treatment that works on any footage.

By the EditFlow Academy editorial team20 September 20267 min read

Laptop showing caption editing on a vertical video frame

Two lines, always

Caption reading happens in glances. Two lines of text can be absorbed while the eyes also watch the picture; three or more forces the viewer to choose between reading and watching, and they will choose reading.

Keep the line under about forty characters so it does not have to shrink on smaller phones. If the sentence is longer, split it into two timed captions instead of adding a third line.

Placement and safe margins

Place captions in the lower third but keep them clear of the interface area at the bottom of the screen, where platform controls appear. A practical margin is ten percent of the frame height from the bottom edge and eight percent from the sides.

Keep one position for the entire video. Captions that jump between top and bottom look like an accident, and viewers lose their reading rhythm while the text moves.

  • Two lines maximum, roughly forty characters per line
  • Lower third, lifted clear of interface controls
  • One fixed position for the whole video
  • Consistent font weight, size and alignment

Contrast on any footage

White text alone disappears on bright footage. The most reliable treatment is a semi-transparent dark panel behind the text, or a soft shadow combined with a subtle outline. Both survive dark and light backgrounds without needing to be adjusted shot by shot.

Test your treatment on the brightest and darkest frames in the edit before publishing. If it fails on one of them, adjust the treatment rather than changing the footage.

Avoid placing captions over faces, hands or product details. If the footage has nowhere clean, shift the composition slightly in the edit rather than moving the text.

Timing and accuracy

Caption timing should follow speech with a small lead. Cutting the caption exactly with the first syllable feels late on a phone; appearing a few frames early feels natural and helps viewers who are also watching the image.

Automatic transcription is a good starting point and a poor finishing point. Read the captions through once without the sound and correct names, technical terms and numbers, which is where automatic tools fail most often.

Key takeaways

  • Two lines, about forty characters, no exceptions for long sentences.
  • Lift captions clear of interface controls and keep one position.
  • Use a panel, shadow or outline so text survives bright and dark shots.
  • Read captions once with the sound off to catch transcription errors.
EditFlow Academy editorial team

EditFlow Academy editorial teamWritten by working editors who cut short-form video daily and reviewed against real client timelines before publication.

Should captions be burned in or uploaded separately?

Burned-in captions guarantee that the style you designed is what viewers see. Separate subtitle files are more flexible and accessible, and are worth adding as well.

What font size is safe?

Large enough that a viewer at arm's length can read two lines instantly. If you have to concentrate, the size is too small.

Do captions help people who can hear?

Yes. Many people watch with the sound off in public, and captions let them follow the video without switching it off.

Keep learning

Related guides

Editor working with headphones at a keyboard

Editing rhythm

Cutting for Pace: Rhythm in Short-Form Editing

How shot length, cut points and breathing room control attention in a vertical video.

Read the guide →

Editor reviewing footage on a laptop in a room lit by a desk lamp

Colour

A Simple Colour Pass That Fixes Flat Footage

One contrast adjustment, one white balance correction and a check on skin tones that works on automatic footage.

Read the guide →