Guides
How to Import Documents for Audio
This guide covers importing documents into EchoLive's Studio, from supported formats and segmentation strategies to AI-assisted Smart Import.
- Supported: .txt, .md, .docx, .pdf, .html, .gdoc, URLs
- Auto-detection picks the best segmentation strategy
- Heading-based, paragraph, marker, and length-aware splitting
- Smart Import uses AI to refine boundaries and pacing
- Edit and regenerate any segment in the Studio editor
EchoLive supports a wide range of document formats. Upload a file or paste a URL, and the import pipeline extracts text with formatting. PDFs use client-side extraction (PDF.js), Word docs preserve headings and formatting (mammoth.js), and URL imports use a "reader mode" extraction on the backend.
Segmentation happens automatically: the system detects whether your document has headings, markers (like '---' separators), or should be split by paragraph. You can also choose a strategy manually. Length-aware fallback splitting prevents any segment from being too long for TTS processing.
For best results, enable Smart Import. The AI analyzes text complexity and suggests: optimal segment boundaries, pacing adjustments per section, pause placements, style suggestions (e.g., conversational vs. formal), and emphasis candidates for key phrases.
After import, every segment appears in the Studio timeline. Edit text, change voices, adjust SSML, reorder, and regenerate, all before exporting the final audio. It's a full editing workflow, not a one-shot conversion.