How to add captions to a video

Burned-in captions or a caption file

Burned-in captions are drawn into the picture. Everyone sees them, they look the same on every app, and nobody can turn them off. Most short vertical videos use them, because plenty of people watch with the sound off.

A caption file, usually SRT, is plain text with timecodes that you upload next to the video. The platform draws it, and viewers can switch it on or off. YouTube videos often use one. You can do both: burn captions into a Short and upload an SRT with the long version.

What makes captions readable

• Short chunks. One to three words at a time suits vertical video. Horizontal video can carry a full line.
• Breaks at natural phrase ends, such as after a comma, rather than in the middle of a name.
• Strong contrast: white text with an outline or a dark box behind it.
• A safe position. Keep them off your mouth and above the bottom of the frame, where the app puts its buttons and caption text.
• Correct spelling of names, brands and jargon. Auto transcription gets these wrong most often, and they're the words people notice.

The manual way

1. 01Transcribe the video with your editor's auto captions or a separate service.
2. 02Read along and correct every misheard word.
3. 03Split the text into caption blocks and set the in and out time for each.
4. 04Style them and set the position.
5. 05Export with captions burned in, or export an SRT.
6. 06If you cut the video after captioning, retime every caption after the cut.

Why it drags

Timing captions by hand is slow, and it breaks every time the edit changes. Remove one um and every caption after it is now early. Most people caption last for that reason, which means captioning happens when you are most tired of the video.

The Cutroom AI way

1. 01Drop your video in on the Projects page. Transcription runs on your device in 16 languages.
2. 02Open the Captions tab and switch on “Show captions”.
3. 03Pick a “Caption style” (8 to choose from) or a template: Word pop, Karaoke, Keywords, Keywords and emoji, Boxed, Outline or Two lines.
4. 04Set “Words at a time” to 1, 2 or 3, and turn on “Highlight the spoken word” if you like the look.
5. 05Choose Top, Middle or Bottom, or drag the captions on the preview.
6. 06Fix a misheard word: double-click it in the transcript and choose “Fix spelling”. “All” changes every copy of that word in the video, and new edits in this browser spell it your way too.
7. 07Click “Download captions (.srt)” for a caption file, or export the MP4 with captions burned in.

Why the captions stay in sync

Captions are word-timed from the transcript, so a cut moves them with it. Remove an um and it disappears from the captions as well. Restore it and it comes back.

A new caption starts after a comma, so phrases stay together. Keyword captions show numbers, names and strong words in the accent color. You can edit any caption's words without touching the cut.

Picking a caption style for the video

Match the captions to how the video is watched. A quick tip in a vertical clip suits one or two words at a time with the spoken word highlighted, because the eye follows the voice. A slower teaching video reads better as two lines that stay up longer, so a viewer can read along without chasing words.

Keyword styles pull the eye to numbers and names, which helps in a recipe or a listing walk-through where the numbers matter. Use one style per channel and stick with it. Viewers start to recognize your videos by it.

Mistakes to avoid

• Trusting the transcript with names. "Kaitlyn" and "Caitlin" sound the same to any speech model.
• Long captions on vertical video. A full sentence on a phone screen is a wall of text.
• Captions in the bottom strip, where the app's username and buttons will sit on top of them.
• Burned-in captions plus the platform's auto captions. Viewers get two sets on screen.
• Expecting translated captions. Cutroom AI captions in the language you spoke. Caption translation is turned off.