The steps
Write or generate the text from the spoken words. Split it into short lines that can be read at the speed of the voice. Time each line to appear when the speech starts and leave when it ends. Then export the text and the timing together.
- Write or transcribe the text.
- Split it into short lines.
- Time each line to the speech.
- Choose a file or burned-in text.
- Check every line against the picture.
Subtitle file or burned-in text
A subtitle file sits beside the film. The viewer can switch it on or off and, on many platforms, choose the language. Burned-in text is part of the picture: it always shows, looks the same everywhere and cannot be switched off.
Use a file when the platform supports it and the viewer should choose. Use burned-in text for social films, which are often watched without sound.
Make it readable
Keep lines short, put them in the lower part of the frame and use a plain, large font with enough contrast against the picture. Do not cover a face, a product or on-screen text.
Check before you publish
Read all the lines against the speech. Look for names, numbers and terms that automatic tools often get wrong. Watch the film on a phone, with the sound off, and make sure the story still works.
Subtitles in another language
A translation that is correct word for word can still be too long to read in time. The text has to be adapted to the length and rhythm of the picture. If the market also expects a new voice, see the guide on how to translate a video.
