FFmpeg Segment Audio
AI-generatedSummary
Segment audio binary data into overlapping chunks of specified duration using FFmpeg within a workflow.
Inputs
- Input Data Field Name (required) — Name of the binary field containing the source audio to segment.
- Output Data Field Name (required) — Name of the binary field where each output audio chunk will be stored.
- Chunk Size in Seconds (required) — Target duration for each audio chunk; must be at least 30 seconds.
- Overlap in Seconds (required) — Duration of overlap between consecutive audio chunks; must be less than the chunk size.
- Segmentation Timeout in Milliseconds — Maximum allowed runtime for FFmpeg to segment each input item; defaults to 900000 ms (15 minutes).
Output shape
a list of items where each item contains one audio chunk as binary data
Each output item contains the chunk's binary audio under the given output field name, plus updated JSON metadata including chunk index, timing info, chunk and source durations, and strategies used. Items preserve pairing to their source input item. If an error occurs and 'continue on fail' is enabled, the output item includes an error message instead.
Examples
Example 1: Workflow that breaks a long audio file into 10-minute overlapping chunks for processing or transcription.
Set the input binary field name to the field containing the source audio, set output binary field name to desired chunk field, configure chunk size and overlap in seconds as needed, optionally adjust segmentation timeout.