ナレッジハブ・技術ガイド
字幕ファイル（SRT・VTT）エディタとタイムコード調整
字幕ファイル（SRT・VTT）エディタとタイムコード調整
本ツールは100%クライアントサイド（ブラウザ内）で動作します。写真、動画、音声、PDFファイルが外部サーバーにアップロードされることは一切なく、お使いの端末のメモリ内で安全に処理されます。
コアアーキテクチャ & 計算式
Subtitle Node = Sequence Number + [Start Time ➔ End Time] + Text Payload
The SubRip (SRT) format is strictly sequential. A single missing arrow `-->` or a malformed timestamp (HH:MM:SS,ms) will break the entire subtitle file during playback.
ベストプラクティスとガイドライン
- Keep Lines Short: Human reading speed is limited. Never exceed 42 characters per line, and never exceed 2 lines per subtitle block. If a sentence is longer, split it into two separate timestamped blocks.
- Use WebVTT for Modern Web: While SRT is the legacy standard for desktop video players, WebVTT is the native standard for HTML5 `<track>` tags. VTT supports advanced styling (colors, positioning) that SRT lacks.
- Include Non Speech Audio: True accessibility requires Closed Captions (CC), not just subtitles. Always bracket significant audio cues, like `[explosion]` or `[upbeat music playing]`, for deaf and hard of hearing viewers.
よくある質問 (FAQ)
Why are my subtitles out of sync?
Desync usually occurs when the subtitle file was created for a video running at 24 frames per second (fps), but applied to a video running at 30 fps. Time drift compounds as the video plays.
Can I embed the subtitles directly into the video?
This tool generates sidecar files (.srt/.vtt) which allow the user to toggle them on or off. To permanently 'hardcode' or 'burn' subtitles into the video pixels, you need a heavy rendering engine like Handbrake.
Does this tool auto generate subtitles using AI?
No. High accuracy speech to text AI requires massive GPU computation. This tool provides a highly efficient manual synchronization interface for precise human editing.