n8n FFmpeg Silence Detect Workflow: Speech & Pause Segments

Instantly detect silence and map speech segments in your audio with n8n + FFmpeg

This n8n workflow uploads an audio file, runs FFmpeg’s silencedetect filter to measure pauses, and returns a JSON report showing silence intervals plus derived speech vs. pause totals (including percent silent).

What this workflow does

• Receives input via an n8n form: audio file upload, a silence threshold (dB), and a minimum silence duration (seconds).
• Creates a per-run temporary folder and normalizes settings (file path, threshold dB, minimum silence seconds) for reliable FFmpeg access.
• Saves the uploaded audio to disk so FFmpeg can read it by filesystem path.
• Runs FFmpeg with the silencedetect filter to log silence_start and silence_end events without rendering an output audio file.
• Parses FFmpeg stderr to extract the total duration and all detected silence intervals.
• Calculates derived speech segments as the non-silent gaps between silences, then returns a summary JSON including:
  • silence count
  • silent seconds vs. speech seconds
  • percent silent

Use cases

• Analyze recorded calls, interviews, or podcasts to quantify how much time is spent paused.
• Pre-check audio quality before transcription by identifying long silence gaps.
• Support media indexing or analytics by segmenting “speech vs. pause” based on configurable dB threshold and minimum pause length.

Technical details

• Self-hosted n8n required (workflow relies on Execute Command, not available on n8n Cloud).
• Install FFmpeg on the same machine/container and ensure the ffmpeg binary is on your PATH.
• Use file access settings so n8n can write to /tmp (e.g., set N8N_RESTRICT_FILE_ACCESS_TO to include /tmp).
• Nodes/logic used: form trigger, code, sticky note, read/write file, and execute command.

Requirement: standard FFmpeg with the silencedetect filter (available in mainline builds).