AI Video Editor FAQs

Question 1

Does this skill re-encode my video and reduce quality?

Accepted Answer

No, the skill uses FFmpeg stream copying ('-c copy') during the execution phase. This preserves the original video quality and makes the editing process extremely fast as it avoids time-consuming re-encoding.

Question 2

What makes the 'trim-silence' feature unique for Korean content?

Accepted Answer

The dynamic mode analyzes Korean sentence endings and particles to apply different silence caps (e.g., 0.5s for full sentences vs. 0.15s for continuing phrases), resulting in much more natural-sounding narration than a fixed silence gate.

Question 3

Do I need an ElevenLabs account to use this skill?

Accepted Answer

An ElevenLabs API key is only required if you use the TTS narration features. The basic 'analyze' and 'execute' commands for removing silences and retakes work using Whisper and FFmpeg without an external API.

Question 4

How does the skill detect and remove retakes?

Accepted Answer

The skill analyzes Whisper transcriptions to identify segments with matching opening text. It assumes the last take is the best one and automatically marks previous duplicate segments for removal in the edit plan.

Question 5

Can I fix transcription errors before generating TTS?

Accepted Answer

Yes, the workflow includes a 'tts-prepare' step that generates a JSON file. You can review this file and apply a 'corrections.json' map to fix any Whisper misspellings before the final 'tts-generate' step.

AI Video Editor

AI Video Editor

主な機能

ユースケース

主な機能

ユースケース