ON-SCREEN TEXT BLOG / 8 MIN READ

Are my captions actually readable?

Most of your viewers watch on mute, on a phone, in bad light. If your captions are too small, too low-contrast, or changing too fast, the video is failing for the people most likely to share it. Here is how to check.

~80%watch on mute
4–6%frame height for text
2 linesmax on screen
arm's
length
the real test

By Thomas, founder of CutScore · Updated June 2026

CAPTION CHECK · reel_v3.mp4
An editing timeline open on a monitor while captions get their final readability pass before the video is published.
CRAFT SCORE
FIXES ADVISED
are the captions readable on a phone?
Caption size small · 3.1% of frame height00:04
Low contrast over sky · add backing00:22
Inside safe zone · clear of UI
The 30-second answer Your captions are readable if you can read every line in one glance, on a phone, held at arm's length, with the sound off. Concretely, that means text around 4 to 6 percent of the frame height, a solid or semi-solid backing so the words survive both a bright sky and a dark doorway, no more than two lines on screen at once, and a change rate slow enough to finish reading before the next line appears. Keep all of it inside the safe zone, clear of the platform's buttons. If checking that on every clip sounds tedious, that is the exact job CutScore does in one pass.
WHY THIS IS EASY TO GET WRONG

Here is the trap. You read your own captions on a big bright laptop, sitting close, with the audio playing, while already knowing every word that is about to appear. Of course they look fine. You are not reading them. You are confirming a sentence you wrote three hours ago. The text could be half the size it needs to be and you would never feel it.

Your viewer has none of that. They are on a phone the size of a playing card, one hand, maybe on a train, in sunlight that turns the screen into a mirror. A huge share of them have the sound off, scrolling past in near silence. For that person the captions are not a nice extra. They are the video. If the words are too small or they vanish into a white sky for two seconds, your viewer does not squint and persevere. They flick to the next clip.

I have shipped clips with captions I was sure were big enough, then opened them on my own phone outside and could not read my own words. So this is not a lecture, it is a confession with a fix attached. Readability is not a taste call. It has targets, and they are short.

THE FIVE THINGS THAT DECIDE IT

What makes a caption readable.

Five properties carry almost all of it. Each one has a target you can check by eye, and each one is something a viewer notices the instant it fails.

PropertyTarget to hitWhat goes wrong if you skip it
Size≈ 4–6% frame heightToo small and mobile viewers cannot read it before the line changes.
Contrastbacking or outlineWhite text alone disappears the moment the shot behind it goes bright.
Line count1–2 lines, shortThree packed lines is a wall of text nobody finishes mid-scroll.
Timingreadable, then changeCaptions that flick by faster than you can read them are worse than none.
Placementinside the safe zoneText under the platform's buttons gets covered and partly cut off.
The one most people forgetFont choice. A thin, condensed, or decorative typeface reads beautifully on your monitor and turns to mush at phone size over moving footage. A medium or bold sans-serif at a generous size beats a stylish font every single time.
SKIP THE MANUAL CHECK

Checking caption size, contrast and timing frame by frame is slow. CutScore measures all five on every line and hands back the timestamps that fail, so you fix the text instead of inspecting it.

Join the waitlist
HOW TO CHECK EACH ONE

Are my captions readable, line by line.

1. Size: read it on a phone, not a laptop

The single most useful test costs nothing. Put the video on your phone, hold it at arm's length, and try to read a line in one glance. If you lean in, the text is too small. As a working target, caption text wants to be roughly 4 to 6 percent of the frame height, so on a 1080-tall vertical video that is around 45 to 65 pixels of letter. Bigger rarely hurts on mobile. Too small almost always loses someone. The same logic feeds into what we analyze, because on-screen text is read before a single word is heard.

2. Contrast: survive a white sky and a dark doorway

White text looks safe until your shot turns white. A snow scene, an overcast sky, a bright shirt, and suddenly the words are gone for two seconds. The fix is to stop relying on the text colour alone. Give the captions a solid or semi-solid backing box, a thick outline, or a soft drop shadow so they read over anything moving behind them. If you want a number, aim for clearly more than 4.5 to 1 contrast between the text and whatever is behind it at the worst frame, not the easiest one. You can check text contrast in your video in the same pass you check the captions.

A tablet and a phone side by side showing the same video, the everyday screens where captions either stay readable or quietly fall apart.
The real test is the small screen, not the editing monitor. Photo: Pixabay / Pexels.

3. Length: two lines, short, in plain language

A caption is not a transcript dump. Keep it to one or two short lines on screen at a time, and break the line where a person would naturally pause for breath. Three dense lines is a paragraph, and nobody stops mid-scroll to read a paragraph. If a sentence is long, split it across two cards rather than cramming it. Shorter lines also mean bigger text fits without touching the edges, so length and size quietly help each other.

4. Timing: long enough to actually read

Captions that change faster than you can read them are worse than no captions, because they tease information and then snatch it away. Each card needs to stay up long enough to read at a calm pace, with a beat of breathing room. Karaoke-style word-by-word highlighting can look slick, but if the whole line never sits still, slower readers and non-native speakers lose it. The honest test is the same one again: can you, a stranger to this clip, read every card the first time through?

5. Placement: keep it clear of the interface

This is the one that survives the edit and dies on the platform. On vertical video the bottom roughly 15 to 20 percent of the frame is buried under the platform caption, your username, and the row of action buttons. Park your text down there and it gets covered, or partly cut off, on the exact apps it was made for. Keep captions in the middle band, inside the safe zone, then preview inside the real app before you decide it is fine. The desktop preview is not the truth. The phone in your hand is.

RATHER SEE IT THAN READ IT?

Here is a real CutScore coaching report for an everyday reel: caption size, contrast, timing and safe zones, scored, with the exact timestamps that need a fix.

See a sample report
SHORT ON TIME

If you only fix three things.

Most of the jump from "unreadable on mobile" to "anyone can follow it on mute" comes from these three. Fix them first.

1
2-MIN FIXTEXT
Make the text bigger than feels comfortable
On a laptop your captions feel correctly sized at about half what a phone actually needs. Push them toward 4 to 6 percent of the frame height. It will look slightly oversized on your monitor, which is exactly right for the device people are really using.
How Open the clip on your phone at arm's length. If you read it in one glance, the size is right. If you lean in, go bigger.
2
EDITCONTRAST
Add a backing so text never disappears
Plain white text dies the instant the footage behind it goes bright. A solid or semi-solid box, a thick outline, or a strong shadow keeps the words legible over a white sky and a dark room alike. Test it on the brightest frame, not the easiest one.
How Scrub to your lightest shot. If the caption is hard to read there, your backing is too weak. See checking text contrast.
3
QUICKLAYOUT
Move captions out of the platform's buttons
The bottom of a vertical frame belongs to the app, not your text. Username, caption and action buttons sit there and will cover or crop anything you place underneath. Lift your captions into the middle band so they survive on the exact platform they were made for.
How Preview inside the real app, not just your editor, and confirm nothing important sits under the interface.
THREE WAYS TO CHECK READABILITY

By eye, by device, or in one pass.

OPTION 01

By eye on your editor

Free, and the least reliable. The big bright preview flatters small, low-contrast text, and you already know every word, so you cannot read your captions like a stranger would. Better than nothing, but it is the version that quietly lets unreadable text through.

OPTION 02

On a real phone, on mute

The honest manual method. Send the clip to your phone, turn the sound off, and read it at arm's length in normal daylight. You will catch size and contrast problems fast. The cost is doing it carefully on every single clip, and remembering to preview inside the real app for safe zones.

OPTION 03

With a coach in one pass

Hand the file (or a link) to CutScore. It measures caption size, contrast, line count, timing and safe zones across the whole video, and gives you a 0 to 100 score with the exact timestamps that fail and the fix for each. See a sample report.

How CutScore checks your captions for you CutScore is an AI video quality coach for pre-publish QC. For captions it measures the text height as a share of the frame, the contrast against whatever is moving behind each line, how many lines sit on screen, how long each card holds, and whether any of it strays into the platform's interface. You get one score, the timestamps that fail, and a clear fix for each, before anyone else sees the video. It judges the craft of the video itself, not your tags or thumbnails, so it sits next to a growth tool rather than competing with one. More on the method and the standards.
QUESTIONS

Frequently asked.

Watch your video on a phone, held at arm's length, with the sound off. If you can read every line in one glance without squinting, your captions are readable. If you have to lean in, lose the line before it changes, or the text fights a busy background, they are not. The fix is almost always bigger text, stronger contrast, and a slower change rate.
There is no single pixel value, because it depends on resolution and aspect ratio. A safer rule is that caption text should be roughly 4 to 6 percent of the frame height, so on a 1080-tall vertical video that is about 45 to 65 pixels tall. Bigger is rarely a problem on a phone; too small almost always is.
Enough that the text never disappears into the footage behind it. White text alone is not enough, because your shot has white moments too. Add a solid or semi-solid backing box, a thick outline, or a soft drop shadow so the words stay legible over both a bright sky and a dark doorway. Aim for clearly more than 4.5 to 1 contrast against whatever is behind the text.
Keep them inside the safe zone, away from the platform's interface. On vertical video the bottom roughly 15 to 20 percent is covered by the caption, username, and action buttons, so park your text in the middle band, not at the very bottom. Then preview inside the real app, because every platform crops and overlays differently.
EARLY ACCESS

Stop guessing whether anyone can read it.

CutScore checks caption size, contrast, timing and safe zones on every line and tells you exactly what to fix, with the timestamps to prove it. Join the waitlist for early access.

Join the waitlist