videobeaux/docs/_site/programs/utilities/silence_extraction.md

silence_extraction

Description

Extracts sections of silence from a video’s audio track based on duration thresholds.
Useful for identifying dead air, isolating non-dialogue segments, or preparing silence-aware edits and analysis.

Purpose

The silence_extraction program is designed to detect, isolate, or extract moments of silence within a video’s audio.
This is useful for:

• cutting silent gaps out of recordings,
• analyzing pacing or speech density,
• preparing regions for time compression,
• generating metadata for editors or automation pipelines.

How It Works

1. Silence Detection
   FFmpeg’s silence detection logic identifies quiet sections based on amplitude thresholds.
2. Duration Filtering
   • min_d defines the minimum silence duration to be considered meaningful.
   • max_d defines the longest segment to extract or label.
3. Adjuster Logic The adjuster parameter allows tuning how tolerant the detection should be, adjusting thresholds or trimming surrounding audio depending on implementation.
4. Output Behavior
   Extracted silence segments may be exported individually, compiled, or used to generate metadata depending on how videobeaux handles downstream processing.

Program Template

videobeaux -P silence_extraction \
  -i input.mp4 \
  -o output.mp4 \
  --min_d VALUE \
  --max_d VALUE \
  --adjuster VALUE

Arguments

• min_d — Minimum silence duration (in seconds) to count as a silence event.
• max_d — Maximum silence duration to extract or annotate.
• adjuster — Fine-tuning parameter for silence threshold sensitivity or trimming behavior.

Real World Example

videobeaux -P silence_extraction \
  -i myvideo.mp4 \
  -o silence_extraction_styled.mp4 \
  --min_d 1.5 \
  --max_d 12.0 \
  --adjuster medium

Technical Notes

• Silence detection is typically amplitude-based using FFmpeg filters (e.g., silencedetect).
• min_d is useful for ignoring tiny pauses or breath sounds.
• Very large max_d values may capture irrelevant long stretches; tune for your content.
• adjuster may influence thresholding; examples include “strict,” “medium,” or “loose” depending on your implementation.

Recommended Usage

• Removing silent gaps in interviews or podcasts.
• Locating pauses in lectures for automatic chaptering.
• Creating pacing analytics (speech vs silence ratio).
• Identifying dead air in archival footage.

Quality Tips

• Use smaller min_d values (0.3–0.7s) for fast speech.
• Use larger min_d (1–2s) for natural conversations or interviews.
• Fine-tune adjuster to avoid misclassifying quiet music or soft ambience as silence.
• Always review extracted segments before batch processing removal or compression.