Why Your EPUB Looks Weird: 15 Formatting Fixes

Most formatting issues come from PDF extraction, heavy Word styling, inconsistent headings, and reader rendering differences.

If the source isn’t structured, no converter can invent perfect structure. The best fix is often upstream: improve the source.

1) Random line breaks: enable paragraph reflow or merge lines after conversion.

2) Missing TOC: ensure Heading styles or real H1/H2 tags are used.

3) Weird spacing: remove manual spacing and rely on CSS margins.

4) Font inconsistency: avoid embedding fonts unless required.

5) Oversized images: compress images before conversion.

6) Blurry images: avoid upscaling; export at target size.

7) Broken italics/bold: ensure source text uses real styling, not fake characters.

8) Chapter breaks missing: insert headings or explicit page breaks.

9) Tables broken: flatten to text or recreate with simple HTML tables.

10) Footnotes out of order: move notes inline or convert to endnotes.

11) Lists collapsed: use real lists in DOCX/HTML.

12) Hyphenation oddities: set language metadata.

13) Reader differences: test in two readers and simplify CSS.

14) Cover not showing: ensure cover metadata + cover page.

15) File too large: downsize images and remove unused resources.

The cleanest EPUBs start with clean sources—DOCX/HTML/MD beat PDFs for structural fidelity.

Adopt a preflight checklist: real headings, minimal styling, optimized images, language metadata.

When an EPUB renders poorly, debug in this order: source structure, heading hierarchy, paragraph spacing, image sizing, then CSS overrides. Most issues are upstream in source semantics, not in the final package.

Create a “known good” sample EPUB and compare styles side by side. This baseline speeds up root-cause analysis and prevents regressions when updating conversion logic.

Track recurring defects as named patterns (for example: “double line break drift” or “fake heading TOC miss”). Pattern-based QA reduces random fixes and helps scale editorial consistency.