Add Captions to Lecture Videos

Captions as part of the lesson

In a lecture hall a student can ask you to repeat a word. In a recording they can rewind, but rewinding doesn't help when the word is unfamiliar: eigenvalue, habeas corpus or a theorist's surname is just a sound until it is written down. Captions put every spoken term on screen at the moment you use it. That helps students taking notes, students working in their second or third language, anyone watching on a noisy bus, and deaf and hard-of-hearing students who would otherwise wait for a separate accommodation.

It also changes the accuracy bar. A misheard filler word doesn't matter; a misspelled key term gets copied into every set of notes in the class. Aim your review at the vocabulary and treat the rest as a light proofread. For the course-design side of accessibility, see accessible online courses.

Two ways to get captions in front of students

How you publish decides which output you need from the editor.

Caption file (SRT or VTT)

Upload it alongside the video wherever the player supports caption tracks, such as YouTube, Vimeo or a course platform's built-in player. Students can turn it off, and you can correct the text next term without re-rendering.

Captions in the picture (MP4)

Render the lecture with the captions drawn in. Use this when the video lives somewhere that ignores caption files: a shared drive, a file embedded in a slide deck, a class messaging group, or a course page that simply links the raw MP4.

Whether a particular course platform shows a caption track depends on how the video is embedded, so test with a student account before assuming. Exporting the SRT costs nothing, so keep one even when you render the video; it is your editable master.

Getting the subject vocabulary right

Speech recognition is trained on broad, everyday speech, so its weak spots are exactly where lectures are dense: discipline jargon, Latin and Greek terms, chemical names, acronyms, names of researchers and places, and numbers read aloud. Before you review, pull the key terms from your slides or reading list and keep them beside you. Then:

• Play at normal speed while watching the line list rather than the picture, and stop on any term you know is wrong to correct it in place.
• Write symbols and numbers the way the course materials show them. If you say "x squared plus two x", decide whether the caption reads as words or as an expression, and stay consistent.
• Keep acronyms in the form your course uses and spell them out once if students might not know them.
• Drop false starts and repeated words when they make a line too long, but never alter the wording of a definition.

The spoken language is detected automatically. Strong accents, a distant microphone and audience questions the lapel mic barely caught are where errors gather. If you lecture in Hindi, the captions are written in Devanagari. Proofreading an AI transcript describes a review routine that transfers well to captions.

Keeping captions clear of your slides

Most lecture recordings show slides, a document camera or a screen share, and the bottom of a slide is often where the citation, the page number or the last bullet sits. Before rendering, scrub through a few typical slides in the preview and pick the placement that hides the least:

• Bottom with a small upward offset suits a talking-head recording or slides with generous margins.
• Top works for screen recordings where a taskbar or an editor's status line fills the bottom edge.
• A translucent box, as in the YouTube or Modern presets, keeps text readable over busy diagrams while letting the slide show around it.

Caption size is relative to the shorter side of the video, so proportions hold whether you recorded at 720p or 4K; if students mostly watch on phones, go a step larger. Position applies to every caption in the video, so choose what suits most slides rather than the one awkward exception. Subtitle safe areas explain how much margin to leave, and subtitling a screen recording covers the extra pitfalls of small on-screen text.

Long recordings and whole courses

AI subtitles handle one video of up to 2 hours, which covers nearly any single session; cut anything longer into parts first. A 50-minute lecture costs 50 credits (5¢). Signed-in accounts also receive 10 free credits a day, building up through the month until the 1st, so a weekly lecture can often be captioned from the free allowance alone.

Rendering the captioned MP4 runs on your own computer and is free, but a long, high-resolution lecture takes a while and needs free disk space, and the tab must stay open until the download appears. The editor keeps a draft of your lines and style in this browser for up to five videos, which lets you review a lecture across two sittings. It isn't an archive, though: export the SRT for every lecture and store it with the recording. Captioning long videos has more on recordings of an hour or longer.

Three teaching scenarios

Before you publish a captioned lecture

• Every key term on your list appears correctly spelled at least once.
• The line list shows no warnings, or the ones left are deliberate.
• Audience questions are captioned or summarised, for example [Student asks about the exam date].
• No caption covers anything students need to read on a slide.
• The SRT is saved next to the video, so a correction next term doesn't mean starting over.

If what you need is the lecture as text for notes rather than captions on screen, lecture transcription is the right fit, and the subtitle generator produces a caption file without opening the video editor.

About the subtitle editor on this page

Opens

MP4, MOV, M4V, WebM, MKV videos your browser can decode

Subtitles from

AI (from the speech, optionally translated into one of 21 languages), an SRT or VTT file, or typing them in

You download

An MP4 with the subtitles drawn into the picture at the original resolution and frame rate, plus SRT and VTT files

Cost

Editing, styling, rendering and SRT/VTT export are free, with no account and no watermark. AI subtitles cost 1 credit per minute (5-credit minimum per video); signing in with Google gives 100 free credits and accounts get 10 free credits a day

Privacy

The video never leaves your device and is rendered in your browser. Only for AI subtitles is the audio sent, over HTTPS, and it is deleted within 30 minutes of delivery

Rendering needs

Chrome or Edge 94+, Safari 16.4+, or Firefox 130+ on a computer. Editing and SRT/VTT export work in any current browser