To pull the sound out of a clip on its own — keep the voice and lose the music, or take just the score — use Audio separation, inside Extract on the DOTV video node's toolbar. Split audio and video itself lands two nodes at once: an audio track (an audio node) and a silent video (the clip with its sound taken off); Voice only and Background only land just the audio node. This section covers what each of the three ways of splitting keeps, and where the results end up.
Select a video node that has generated, then hit Extract on the toolbar above the canvas — the Audio separation group sits at the end of that menu:
Click the arrow on that button:
| Option | What you get |
|---|---|
| Isolate voice → Voice only | Just the speech; music and ambience are gone |
| Isolate voice → Background only | Just the music and ambience; the speech is gone |
| Split audio and video | The whole original audio track, nothing pulled apart |
It starts as soon as you pick; the new node shows its own progress. When it's done, nodes land to the right of the original one: Split audio and video lands two — a silent video on top (an ordinary video node that plays with no sound) with the audio track (an audio node holding the whole original sound) right under it; Voice only and Background only land just that audio node. The audio node is no different from any other audio node: it plays right there and it downloads, but it feeds nothing — no kind of node takes a piece of sound, so there is no connection point on its right at all (see Chapter 5).
The original clip is left alone.