What makes subtitle text readable
Readers of subtitles have a harder job than readers of a page. The text appears for a few seconds, over a moving image whose brightness and color keep changing, often on a small screen or from across a room. Readability comes down to a handful of factors that work together: letterforms that are recognized instantly, a size large enough for the viewing distance, and a clear edge between the text and whatever is behind it.
None of these choices fixes text that is too long or flashes by too fast; that is a matter of segmentation and timing. Where subtitles sit on the screen is covered separately in subtitle positioning. This article is about how the characters themselves look.
Choosing a font
Most subtitle styles use a clean sans-serif font with open shapes and generous spacing. Fine serifs and thin strokes tend to break up when video is compressed or scaled down, and decorative or condensed faces slow reading.
Useful tests for a subtitle font:
- Characters that are easily confused stay distinct: capital I, lowercase l and the digit 1; capital O and zero; rn and m.
- Strokes are thick enough to survive compression without disappearing into the background.
- Letters have open counters, the enclosed spaces in a, e and o, that do not fill in at small sizes.
- The font covers every script you subtitle in. A font without Devanagari, Arabic or Chinese characters falls back to another font or shows empty boxes.
- Regular or medium weight works better than light weights, which look elegant in an editor but thin on a phone.
For burned-in subtitles, also check that the font's licence allows embedding it in video. For player-rendered subtitles, the font is usually chosen by the player or the viewer, and missing glyphs show up as garbled text, a problem explained in subtitles showing strange characters.
Size and weight
Subtitle size is best thought of relative to the height of the picture rather than in points, because the same video plays on a phone held at arm's length and on a television across a room. Text that is comfortable on a laptop in the editor is often too small on a phone and fine on a TV, so judge it on the smallest screen your audience is likely to use.
Bigger is not always better. Very large subtitles cover more of the picture, fit fewer characters per line and force more frequent cue changes. Most style guides aim for a size that fits a typical full line comfortably within the width of the frame while remaining legible on small screens; the exact figure depends on the font, the platform and the language.
Color, outline and shadow
White text is the long-standing default because it stands out against most footage, and light yellow is a common alternative on some television services and older caption styles. Whatever the color, light text needs a dark edge to stay readable over light parts of the picture.
- Outline
- A dark border around each character. Works on most footage; too thin disappears on busy backgrounds, too thick makes letters blobby
- Drop shadow
- A dark offset copy behind the text. Softer looking, but on its own can be weak over bright, detailed scenes
- Outline plus shadow
- A common combination for burned-in subtitles, giving a clear edge from every direction
- Semi-opaque box
- A dark rectangle behind the text. The most robust option on any background, at the cost of covering more of the picture
- Opaque box
- Maximum contrast, common in broadcast caption styles; can feel heavy on cinematic footage
Colored text is used in some captioning traditions to identify speakers, but it reduces contrast and depends on viewers being able to tell colors apart, so it should never be the only cue.
Contrast against busy footage
The hardest scenes for subtitles are bright and detailed: snow, sky, white walls, sunlit water, light clothing, screen recordings with white interfaces, and fast-moving textures. An outline that works over a dark interior may fail entirely over a snowy slope.
A creator burns white subtitles with a thin black outline into a 9-minute ski video. In the lodge scenes they are crisp, but on the slopes the white text sits over snow and the outline is the only thing separating it, which a phone screen in daylight makes nearly invisible. They switch to the same font with a thicker outline and a semi-opaque dark box, re-export, and check the snowiest shots on a phone outdoors. The box covers a little more of the picture but every line now reads at a glance.
For text that is part of the picture, accessibility guidance such as WCAG includes contrast criteria for text and images of text, commonly cited as a 4.5 to 1 ratio for normal-size text. Burned-in subtitles count as images of text over a constantly changing background, so a background box is often the practical way to keep contrast consistent from shot to shot. How AI-generated captions fit into accessibility more broadly is discussed in AI and video accessibility.
Viewer caption settings
Player-rendered subtitles have a major readability advantage: viewers can adjust them. Operating systems on phones, computers and smart TVs typically offer caption settings for font, size, text color, edge style and background, and many video players and platforms offer their own controls. A viewer with low vision can enlarge text; another can switch to yellow on black.
That flexibility depends on two things. The subtitles must be delivered as a separate file rather than burned in, and the player must respect system or user settings. Some web players ignore them or offer only a few options, which is one of the criteria covered in accessible video player. When you control a website's player, avoid styling that overrides the viewer's choices without good reason.
Player rendering versus burned-in text
Player-rendered subtitles look slightly different on every device, since each platform applies its own default font, size and edge style. You give up control in exchange for adjustability, crisp text at any resolution, and the ability to correct the file without re-exporting the video.
Burned-in subtitles look identical everywhere and are the usual choice for vertical social video and for platforms that do not support subtitle files. Because viewers cannot change them, choose conservatively: a larger size, a strong outline or box, and a font that copes with compression. Export at a bitrate high enough that the edges of the letters do not smear, since compression artifacts around thin text are a common readability problem.
Readability mistakes and limits of styling
- Choosing a stylish thin or condensed font that falls apart on phones.
- Judging size only in the editor at full screen on a large monitor.
- Relying on a drop shadow alone over bright or detailed footage.
- Using color as the only way to show who is speaking.
- Burning in subtitles at low bitrate, so the letter edges break up.
- Overriding viewers' caption settings in a custom web player.
- Using a font that lacks characters for one of your subtitle languages.
Styling also has limits. No font or box makes up for a cue with too much text, a reading speed that is too high, or a translation that is hard to follow. Readability is the last layer on top of accurate, well-segmented and well-timed subtitles.
How mydubly subtitle files are displayed
mydubly's SRT and VTT files contain plain text only: one single-line cue per recognized speech segment, saved as UTF-8, with no font, color, size, outline, background or positioning information. Their appearance is decided entirely by the player that displays them, or by your editor if you burn them into the picture yourself. mydubly does not burn subtitles into the video, so the viewer's caption settings remain available wherever the player supports them.
Because the cues are single lines, long segments can produce wide subtitles; splitting them into two lines in a subtitle editor before styling makes them easier to read. For accessibility use, mydubly's captions contain recognized speech only, without sound tags, speaker labels or audio description, and they should be reviewed by a person before publishing. The subtitle generator costs 1 credit per minute with a 5-credit minimum, so the 9-minute ski video above would cost 9 credits (0.9¢).
Next step: test on three screens
Pick a scene with the brightest, busiest background in your video and play it with subtitles on a phone, a laptop and, if possible, a TV from normal viewing distance. Note where text becomes hard to read and adjust the edge style, background or size before styling the rest. If the file still needs to go onto a platform, the guide to adding subtitles to a video covers uploading and burning in.
Frequently asked questions
Which font works well for subtitles?
There is no single answer, but most subtitle styles use a plain sans-serif font with open letter shapes, distinct I, l and 1, and strokes thick enough to survive compression. The font must also include every script you subtitle in. Avoid light, condensed or decorative faces.
What size should subtitles be?
Judge size relative to the picture height and on the smallest screen your audience uses, usually a phone. Text that looks right in an editor on a large monitor is often too small on a phone. Too large is also a problem, because it covers more of the picture and fits fewer words per line.
Is an outline or a background box better for subtitles?
An outline, often with a shadow, is less intrusive and works on most footage. A semi-opaque box is more robust over bright, busy scenes such as snow, sky or white interfaces, at the cost of covering more of the picture. For burned-in subtitles on varied footage, a box is the safer choice.
Why are white subtitles hard to read on some videos?
White text disappears over light backgrounds unless something separates it from the picture. A thin outline or a soft shadow may not be enough over snow, sky or sunlit scenes, especially on a phone outdoors. A thicker outline or a dark background box fixes most cases.
Can viewers change how subtitles look?
Yes, when subtitles are delivered as a separate file and the player respects caption settings. Phones, computers, smart TVs and many platforms let viewers change font size, color, edges and background. Burned-in subtitles cannot be changed by the viewer.
Do mydubly subtitle files include font or color styling?
No. mydubly's SRT and VTT files are plain text cues with timing, so the player or your editor decides how they look. That keeps viewers' caption settings in effect wherever the player supports them.