Captions for Vertical Video That Stay Readable on Phones

Your captions looked fine in the editor, then you posted the clip and the like button, the username and the description text sat right on top of them. Captions for vertical video fail in a different way from widescreen captions, because the platform paints its own interface over the bottom third of your frame.

I run captioning for an archive that now publishes most clips in 9:16 as well as 16:9, and the rules we use for widescreen break almost immediately on a phone. Here is what we changed and why.

Why captions for vertical video need their own rules

A widescreen subtitle sits near the bottom edge, spans up to 42 characters per line and is read on a screen where nothing else competes for that space. On a phone held upright, three things change at once.

The frame is narrow. A 1080 by 1920 frame is only 1080 pixels wide, so a 42 character line either shrinks to an unreadable size or wraps into three or four lines.
The bottom is occupied. Every short form app overlays a caption field, account name, audio credit and progress bar across roughly the lowest 20 to 25 percent of the screen.
The right edge is occupied too. A column of like, comment, share and profile buttons runs down the right side, usually covering about 120 to 150 pixels of width.

If you place captions where widescreen habit tells you to, they disappear behind the app itself. That is one of the caption mistakes that make viewers stop watching, and on vertical clips it is the most common one we see.

The safe zone and where to put the text

We work to a single template that holds up across the big three short form platforms. In a 1080 by 1920 frame:

Area	Pixels from edge	Use for captions?
Top band	0 to about 250 from the top	No, search bar and tabs live here
Bottom band	0 to about 400 from the bottom	No, description and audio credit
Right column	0 to about 150 from the right	No, action buttons
Left margin	0 to about 60 from the left	Avoid, edge cropping on some phones
Middle lower third	About 1150 to 1450 from the top	Yes, our default position

That gives you a usable text box roughly 870 pixels wide. Centre the caption inside that box, not inside the full frame, so it sits slightly left of true centre and clears the button column. If a speaker's mouth or a product sits in that band, move the caption up to the area just below the face rather than dropping it lower.

Exact interface positions change with app updates, so test with a screenshot. Export a five second clip, post it as a private draft, take a screenshot on your phone, and check it against your template. That takes three minutes and saves a reupload.

Line length, font size and timing on a phone

Our widescreen house style allows two lines of 42 characters. For vertical captions we use these numbers instead:

Characters per line: 20 to 26, two lines at most.
Font size: 56 to 70 pixels at 1080 width for a clean sans serif. Below about 48 pixels, small phones lose legibility.
Outline or box: a 4 to 6 pixel dark outline, or a semi transparent box at around 60 percent opacity. Drop shadows alone vanish on busy footage.
Minimum duration: about 0.8 seconds for a single short phrase, longer for two lines.
Reading speed: stay under roughly 17 characters per second, the same ceiling we use in our guide to caption reading speed and line length.

Shorter lines mean more caption events per minute. A 60 second clip that needs 12 subtitles in widescreen might need 25 in vertical. Break lines at natural phrase boundaries: after a comma, before a conjunction, never between an article and its noun. "I left the studio / at six" reads better than "I left the / studio at six."

Burned in or uploaded caption files

For widescreen work we nearly always prefer a separate caption track. On short form platforms the decision is less clear. Some apps accept caption files, some offer only their own auto captions, and clips get downloaded and reposted with no track attached at all.

Our practice is to burn in captions for vertical clips that will travel across several apps, and to keep a clean master without text plus an SRT file in the archive. That way a later edit, a translation or a correction does not mean rebuilding the video from scratch. If you post to only one platform that supports caption files, upload the file and skip burning in. Viewers can then resize text in their accessibility settings, which they cannot do with pixels baked into the picture.

The word by word style is usually a mistake

A lot of creator advice says to use animated captions that pop up one word at a time, with each word scaling or changing colour as it is spoken. The argument is that motion holds attention.

It may hold attention, but it hurts reading. Readers take in phrases, not single words, and the eye needs a stable block to scan. One word captions force the viewer to wait for the speaker instead of reading ahead, and they double or triple the number of flashes on screen. For deaf and hard of hearing viewers who rely on the text, that is tiring over anything longer than 20 seconds. Fast cuts make it worse.

If you like the look, compromise: show a full short phrase of three to six words and highlight the current word inside it. The block stays still, the eye can read ahead, and you keep some of the movement. Keep colour changes subtle, because a bright yellow highlight on a white word can drop contrast below what low vision viewers need.

A quick workflow that holds up

Transcribe the clip with your usual tool and correct names and jargon by hand.
Set your editor's caption style to the vertical template above: 22 character lines, 60 pixel text, outline on.
Reflow the captions so each event is one complete phrase.
Move any caption that collides with a face or key object.
Export a private test, screenshot it on a phone, and fix collisions.
Export the final with burned in text, and archive the clean master and the SRT alongside it.

Frequently asked questions

Where should captions go on a TikTok, Reel or Short?

Keep them in the middle lower third, roughly 1150 to 1450 pixels from the top of a 1080 by 1920 frame. Stay clear of the bottom 400 pixels and the right 150 pixels, where the app interface sits. Test with a private post because layouts shift with updates.

What font size is best for vertical video captions?

Between 56 and 70 pixels at 1080 width works well for most clean sans serif fonts. Pair it with a dark outline or a semi transparent box. Anything under about 48 pixels becomes hard to read on smaller phones.

Should I use the app's auto captions instead of my own?

Auto captions are better than nothing, but they mishear names, technical terms and accents, and you often cannot control their position. If accuracy matters, write your own and burn them in or upload a corrected file. At minimum, review and edit the auto captions before posting.