Burn Word-Highlighted Karaoke Captions with FFmpeg in n8n

Turn your narrated videos into word-highlighted karaoke captions—automatically, inside n8n

This n8n workflow takes a video and its full script, estimates per-word timings, generates karaoke-style ASS subtitles, and uses FFmpeg (libass) to burn word-highlighted captions directly into the video—so users can download the finished captioned file.

What this workflow does

• Collect inputs via an n8n form trigger: receives an uploaded video file and the complete caption script text.
• Create temporary workspace: generates a unique temporary folder, saves the uploaded video to disk, and prepares file paths plus a word list from the script.
• Measure the actual video properties: runs FFprobe to read duration and resolution.
• Estimate word timings: computes per-word timing estimates weighted by character length, then groups words into readable caption lines.
• Generate karaoke ASS captions: creates a .ass subtitle file sized to the video using karaoke-tag styling.
• Burn highlighted captions into the video: calls FFmpeg with libass using the subtitles filter, burning the ASS captions while copying the original audio stream.
• Return output and a short report: outputs the captioned video plus a summary (duration, word count, and line count).

Use cases

• Creator workflow: publish narrated explainers with precise-looking (word-by-word) karaoke captions.
• Localization prep: quickly visualize timing alignment for scripts before manual editing.
• Learning content: generate engaging captioned videos for language study and transcript-based training.

Technical details

• Run on self-hosted n8n with Execute Command node access (n8n Cloud not supported).
• Requires FFmpeg and FFprobe installed on the same machine and available in PATH.
• FFmpeg must be built with libass support (plain macOS Homebrew FFmpeg may not meet this requirement).
• Workflow uses: Code, Form Trigger, Read/Write File, and Execute Command (plus supporting notes).