Skip to main content
Audio Split

Audio Split · screenshot coming soon

​
What this node does

Audio Split places a single cut point (split_s) on a waveform timeline and produces two audio snapshots: audio_a (before the cut) and audio_b (after it). Like Audio Trim it accepts either an audio input or a video input (whose audio track is split). Both parts persist to the project history, so either half can feed a different branch of your graph.

​
When to use it

  • Split a long music generation into intro and body, using each separately
  • Cut narration at a scene boundary and route the halves to different videos
  • Keep both sides of a cut instead of discarding one (which Audio Trim would)

​
How ComfyTV designed this

  • One cut, two outputs: the split point is dragged on the same waveform timeline used by Audio Trim; both resulting files are written and stored.
  • Stage + ▶ Run: re-splits the upstream snapshot; downstream nodes read whichever half they’re wired to.
  • No GPU: PyAV on disk.

​
Types (COMFYTV_* vs native ComfyUI)

​
Parameters

​
audio (input, optional)

Source COMFYTV_AUDIO. Wire this or video.

​
video (input, optional)

Source COMFYTV_VIDEO; its audio track is split.

​
split_s

The cut position in seconds, dragged on the waveform.

​
Outputs

​
Step by step

  1. Wire an audio (or video) source.
  2. Drag the cut point on the waveform; preview both sides.
  3. ▶ Run; wire audio_a and audio_b to their destinations.

​
Repository and workflows

​
FAQ

Q: I only need one half.
A: Use Audio Trim instead — one range in, one output out. Q: split_s at 0 or beyond the end?
A: One side ends up empty; place the cut inside the source duration.
  • Audio Trim — keep one range
  • Audio Crossfade — join parts back with a blend
  • Mux · Audio — attach a part to video