Guide
How to
time animation to narration.
To time animation to narration, record the narration first, find the moment each key word is spoken, then start each animation a beat before its word so it lands as the word is heard. Word timestamps from a transcript make this fast.
First release: Flowmotion, Cutroom
-
1. Lock the voice first
Record and edit the narration before you finish the animation. Re-timing pictures to a new take is slow; re-timing a take to pictures sounds unnatural. If you must animate first, use a guide recording at the speed you'll really speak.
-
2. Get a timestamp for every word
Transcribe the recording with a speech-to-text tool that gives word-level timings. You then know, to a fraction of a second, when each word starts, without scrubbing through the waveform by ear.
- Check the transcript against the script; fix misheard words
- Mark the cue words: the noun or number each animation shows
-
3. Cue each move to its word
Tie each animation to its cue word rather than to a fixed time. Start it slightly before the word, because the eye needs a moment to find the movement; a few frames early usually feels right, a second late feels like a mistake.
- Entrances lead the word slightly
- Long moves finish as the sentence ends, not after
- Leave a short still moment after a key result
-
4. Captions and sound from the same timings
The same word timings drive captions, so they appear as each word is spoken, and they can place quiet sound effects under the moves. Keep effects subtle and well below the voice.
-
5. Check at full speed
Watch the whole scene in real time with sound. Anything that arrives after its word, or moves while the viewer should be listening to something else, gets nudged. In Flowmotion's development builds, animations can be timed to the narrator's words this way; it isn't released yet.
Apps used
The apps behind it.
The apps are designed to open the same project, so nothing has to be exported between them. You'll install only the ones you use.
Questions
Good to know.
There's no single rule. A few frames to a quarter of a second early is a common starting point for entrances; trust what looks right at full speed.
To the recording. Spoken pacing differs from how a script reads, and timing to the real take is what makes motion feel connected to the voice.
Not strictly; you can mark words by ear on the waveform. Word-level timestamps just make it much faster, especially in long videos.
Early access
Every move,
on its word.
Flowmotion is in development. Nothing can be downloaded or bought yet. Join the waitlist and we'll write when it opens.
We only use your email to tell you about early access. Privacy policy