---
title: >-
  PDF Inspector Suite, Smart OCR Triage, Layout-Aware Conversion & Academic
  Research Tools
published_at: 2026-07-31T23:30:00.000Z
version: v2026.07.31
slug: v2026-07-31
canonical_url: 'https://basaltnotes.com/changelog/v2026-07-31'
markdown_url: 'https://basaltnotes.com/changelog/v2026-07-31.md'
---
# Version 2026.07.31 - PDF Inspector Engine, Smart OCR & Academic Research Suite

This major release introduces the **PDF Inspector Engine**. Basalt Notes now performs instant text-based vs scanned PDF page classification, layout-aware PDF-to-Markdown conversion with GFM tables and LaTeX math formulas, Web Clipper PDF capture, batch PDF folder importing, and an advanced suite of research tools including DOI/arXiv citation fetching, RAG semantic vector chunking, audio read-aloud TTS, table CSV/JSON exports, and PDF reading progress tracking.

---

### PDF Inspector & Smart OCR Triage
- **Smart OCR Triage Pipeline**: Inspects PDF pages in ~20ms. Native text pages bypass Tesseract.js WASM canvas rasterization, speeding up indexing time by ~95% while selectively running OCR only on scanned image pages.
- **Layout-Aware PDF Conversion**: Converts multi-page PDFs into clean Markdown preserving heading hierarchy (`#`, `##`, `###`), structured GFM tables (`| Col 1 | Col 2 |`), lists, and reading order.
- **LaTeX Math Formula Auto-Extraction**: Detects mathematical symbol streams in scientific PDFs and formats them into renderable LaTeX blocks (`$$ ... $$`).
- **Chrome Web Clipper PDF Capture**: Captures PDF URLs directly from browser tabs and converts them into structured Markdown notes with frontmatter metadata.
- **Batch PDF Folder Importer**: Import entire folders of PDFs into vault notes via the Migration Wizard.

---

### Advanced Academic & Research Tools
- **Academic Citation & DOI Fetcher**: Automatically detects DOIs (`10.xxxx/...`) and arXiv IDs (`arXiv:2401.xxxx`) to generate formatted APA and BibTeX reference callouts.
- **RAG-Ready Semantic PDF Chunking**: Splits extracted PDF text into ~500-token semantic chunks tagged with page numbers and section headings for precise AI Q&A.
- **Audio Read-Aloud (TTS) for PDF Previews**: Built-in Web Speech text-to-speech reader integrated into PDF preview toolbars (`[▶ Read Aloud]`).
- **PDF Reading Progress & Position Tracker**: Automatically persists and restores scroll position and progress (`Page N of M - X%`) per PDF document.
- **Export PDF Tables to CSV & JSON**: One-click `[Export CSV]` and `[Export JSON]` hover toolbar buttons above rendered PDF tables.
- **Bi-Directional Attachment Backlinks ("Citing Notes")**: View every note in your vault that references or deep-links to a specific PDF attachment (`basalt-attachment://<id>#page=N`).
- **Multi-PDF Comparative Matrix ("Synthesize Papers")**: Generates side-by-side comparative Markdown matrix tables contrasting methodologies, DOIs, and findings across multiple PDF papers.
- **Interactive PDF Outline Sidebar & Deep Links**: Document outline TOC sidebar with `#page=N` deep-linking to scroll directly to target sections.
- **Vault Tree Quality Badges**: Visual `[Text]` (cyan) and `[OCR]` (amber) badges on PDF items in the vault file tree.