Subtitles, voice-over, transcription of video and audio. We translate to the video’s timing, so each line fits the time it has on screen or the pause between lines, rather than just translating line by line.
For subtitles we also watch readability — line length and how fast they change — so viewers have time to read the text.