Auto Captions for Videos: How to Add Accurate Subtitles Without Typing
Why captions matter, the main ways to add subtitles to a video, and how to keep them accurate, well-timed and readable on vertical video.
In this guide
Auto captions turn the words in your video into on-screen text automatically, so you never type subtitles by hand. They matter because lots of people watch with the sound off, because captions make video accessible to people who are deaf or hard of hearing, and because text on screen helps everyone follow. The most accurate kind comes straight from your script, timed to your voice.
Why auto captions matter
- Muted viewing. People scroll on buses, in offices and in bed next to someone sleeping. Without captions, a talking video tells them nothing.
- Accessibility. Captions are how deaf and hard-of-hearing viewers follow speech. The W3C's guidance on captions and subtitles explains who relies on them and what good captions include.
- Comprehension. Accents, fast speech, technical terms and second-language viewers all benefit from seeing the words.
- A text version of what you say. A caption file gives platforms like YouTube the words of your video in text form, which is also what makes translations possible.
Four ways to add subtitles to a video
| Method | How it works | Accuracy | Effort |
|---|---|---|---|
| Type them yourself | You write and time every line | As good as your patience | High |
| Speech-to-text auto captions | Software transcribes your audio | Often close, but names, numbers and jargon slip | Low, plus proofreading |
| Captions from your script | Text from your script, timing from your voice | Words match what you wrote | Low |
| Professional captioning | A person transcribes and times the video | High | Highest cost and turnaround |
Most social platforms offer their own speech-to-text captions, and they're a reasonable fallback. The weak spots are always the same: proper names, product names, numbers and anything said quickly. If your video depends on those words being right, a script-based method saves you the proofreading.
How script-based captions stay accurate and synced
If you record from a script, you've already typed every word once. Script-based captions reuse that text, so names are spelled the way you wrote them, and the software's only job is to match each phrase to the moment you say it.
That's how captions work in Cameratel: every take gets captions synced to your speech, generated from your script, with no typing and no transcript to correct. You can export them as an .srt file or burn them into the video in one tap; both exports are part of Cameratel Pro.
One caveat: because the words come from the script, the captions are most accurate when you stay close to it. If you improvise a sentence or change a name on the fly, read through the captions before you post.
Styling and placement for vertical video
Good captions can be read at a glance on a phone held at arm's length:
- Position: slightly below the middle of the frame, clear of the bottom area where apps show buttons, usernames and descriptions, and away from the right edge, where the like and comment buttons sit.
- Length: one or two lines. A common subtitle guideline is no more than about 42 characters per line, and short-form videos often go shorter, a few words at a time.
- Contrast: white or yellow text with a dark outline or background box. Thin fonts disappear over busy backgrounds.
- Size: big enough to read without squinting, small enough not to cover your face.
- Line breaks: break at natural phrases, never in the middle of a name or between a number and its unit.
Line breaks are where most captions go wrong, so here's the difference in practice:
| Awkward break | Better break |
|---|---|
| "Most people film in the / wrong spot of the room." | "Most people film / in the wrong spot / of the room." |
| "It takes about 20 / minutes to set up." | "It takes about / 20 minutes to set up." |
| "Ask for Dr. / Lee at the front desk." | "Ask for Dr. Lee / at the front desk." |
Write scripts that caption well
Captions are only as readable as the sentences behind them. A few habits in the script pay off on screen:
- Keep sentences short. A sentence that fits in two or three caption chunks is easy to follow; one that needs six is not.
- Spell out symbols. "&", "@" and "/" look cluttered in captions and are awkward to read aloud, so write "and," "at" and "or" instead.
- Put the key word at the end of a phrase. "The secret is the light" lands better as a caption than "Light is the secret, really."
Burned-in captions or a separate file?
Burned-in (open) captions are part of the picture. They show up everywhere, in every app, but viewers can't turn them off and you can't edit them after export. A separate caption file, usually an .srt, carries closed captions alongside the video: viewers can switch them on or off, and you can fix a typo or add translations later.
A simple rule: burn captions in for Reels, TikTok, Stories and Shorts, and upload an .srt file for YouTube videos, courses, LinkedIn and Facebook. Our SRT subtitle file guide shows how to upload one on each platform.
How to check caption timing before you post
Captions that run early or late are almost as distracting as no captions. Before posting:
- Watch the first ten seconds with the sound off. Each caption should appear as you start the phrase, not a beat later.
- Watch the whole video once with sound, looking only at the captions.
- Check every name, number and product word.
- Look for awkward line breaks and single words stranded on their own line.
- Play it on your phone at normal size, the way viewers will see it.
If you uploaded an .srt file, preview the video on the platform as well, because some apps restyle or move caption files.
For script ideas that also read well as captions, see our short video script templates, and to see how captions are built in the app, visit the auto captions page.
Frequently asked questions
Are auto captions accurate?
Speech-to-text captions are usually close but often miss names, numbers and technical terms, so proofread them. Captions generated from your own script avoid most of these errors, because the words are already right.
Where should captions go on a vertical video?
Slightly below the middle of the frame, above the bottom area where apps show buttons and descriptions. Keep them away from the right edge too.
How many words should a caption show at once?
One or two short lines at most. Short-form videos often show just a few words at a time, which is easier to follow on a small screen.