Home › Guides › How to Add Subtitles to a Video

How to Add Subtitles to a Video

Video 8 min read

Most people assume subtitles are an accessibility feature for a small minority. The usage numbers say otherwise: a large share of social video is watched with the sound off, on a phone, somewhere it would be rude to have audio playing. Subtitles are how that audience follows your video at all.

What an SRT file actually is

SubRip Subtitle files are the plain-text standard, and they are refreshingly simple. Open one in a text editor and you see numbered blocks:

1
00:00:04,120 --> 00:00:07,480
The whole format is really just this, repeated.

An index, a start and end timestamp, and the line of text. That is the entire specification worth knowing, and it is why SRT works everywhere — YouTube, Vimeo, LinkedIn, most players, and every editing program.

Separate file or burned in?

Two ways to put subtitles on a video, and the difference matters.

A separate SRT file travels alongside the video. The platform reads it and renders the captions itself, which means viewers can turn them off, search engines can read them, and you can fix a typo by editing a text file. This is what you want in almost every case.

Burned-in (open) captions are drawn permanently into the picture. They cannot be turned off, cannot be searched, and fixing a mistake means re-rendering the whole video. Their one advantage is that they display anywhere, including platforms that ignore subtitle files.

Start with an SRT. You can always burn it in later; you cannot un-burn captions from a video.

How automatic transcription performs

Modern speech recognition on clear, well-recorded speech is genuinely good — frequently better than a distracted human typist. It is not perfect, and its errors follow predictable patterns:

  • Proper nouns. Names of people, companies and places are the biggest single source of mistakes. The model has never encountered your colleague's surname.
  • Technical vocabulary and jargon. Industry terms get mapped to similar-sounding everyday words.
  • Overlapping speakers. Two people talking at once degrades both.
  • Strong accents under-represented in the training data.
  • Background noise and music competing with the voice.
  • Homophones. "Their", "there" and "they're" sound identical; only meaning distinguishes them.

Treat automatic output as a first draft that saves you the typing, not as a finished product. Ten minutes of review on a five-minute video is normal and worth it — particularly for names, which viewers notice immediately.

Recording so the transcription works

Accuracy is decided when you record, not when you transcribe.

  1. Get the microphone close. The single biggest factor. A cheap microphone near the speaker beats an expensive one across the room.
  2. Kill background noise where you can — fans, air conditioning, traffic, and especially background music with vocals.
  3. One person at a time. Overlapping speech is the hardest thing to transcribe.
  4. Speak at a normal pace. Rushed speech loses consonants, which is exactly what recognition relies on.

Timing and readability

Good subtitles are as much about pacing as accuracy. A few conventions that make captions comfortable rather than exhausting:

  • Two lines maximum on screen at once.
  • Roughly 42 characters per line is the traditional broadcast limit.
  • At least one second on screen, even for a short line.
  • Break lines at natural phrase boundaries, not mid-clause. "I went to the / shop yesterday" reads worse than "I went to the shop / yesterday".

Why subtitles help beyond accessibility

An SRT is machine-readable text describing exactly what your video says. Platforms use it for search and recommendation, which means a subtitled video is discoverable in ways an unsubtitled one simply is not. It also gives you a transcript you can repurpose into a written article or show notes with very little extra work.

Try it: Generate Subtitles transcribes your video's speech into a downloadable SRT, with language auto-detect and optional translation. For audio files rather than video, use Audio to Text. Clean up a noisy soundtrack first by pulling it out with Extract Audio and running Noise Reduction.

Common questions

Do I get an SRT file or a video with captions on it?

An SRT file. Upload it alongside your video wherever you publish, and the platform renders the captions — which keeps them switchable, searchable and easy to correct.

How long does transcription take?

Longer than most tools, because every second of audio has to be analysed. Expect a wait proportional to the length of the video rather than its file size.

Can I get subtitles in another language?

Yes — the tool can translate as well as transcribe, between English, Spanish, French and German. Bear in mind that translation errors compound on top of any transcription errors, so review translated output more carefully.

My video has no speech, just music. Will it work?

No. Speech recognition needs speech. Music with lyrics gives poor results, because singing differs considerably from speaking.

More guides