Why do images get lost or look wrong when I convert a PDF to ePub?

Two reasons. First, the conversion has to move from PDF's fixed layout (images pinned at exact coordinates) to ePub's reflowable model (images flow inline with text), so a figure that sat in a precise spot now floats near the paragraph it followed — and a weak converter can drop images it cannot place, or detach them from captions. Second, image quality: a converter may re-compress images aggressively to shrink the ePub, introducing artifacts. Good conversion preserves the images, keeps them with their captions, and does not over-compress. The fix is choosing a tool that handles images well and then verifying every figure survived.

How do I keep image quality high in the ePub?

Start from the highest-quality source — a born-digital PDF with full-resolution images beats a pre-compressed one — and use conversion settings that do not crush the images. ePub images are typically JPEG or PNG; photographs suit JPEG at a good quality level, while diagrams and line art are sharper as PNG. There is a genuine trade-off with file size: an image-rich ePub can get large, and some stores cap e-book file sizes, so you balance fidelity against size rather than maximising either. Keep a master and, if needed, produce a lighter version, but never let the convenience copy be the only one.

My book is mostly images (art book, cookbook, comic) — is reflowable ePub right?

Maybe not. Reflowable ePub is designed for text that rewraps to any screen, which is ideal for novels but works against highly-designed, image-led books where the exact layout is the point — an art book, a photography collection, a comic, or a cookbook with carefully placed photos. For those, a fixed-layout ePub (which preserves the exact page design like a PDF, but as an e-book) or simply keeping the PDF is often better than a reflowable conversion that scatters the carefully-placed images. Match the format to the content: reflowable for text-led books, fixed-layout or PDF for design-led ones.

What happens to captions and the text around figures?

In a good conversion, a caption stays associated with its image and travels with it as the text reflows; in a poor one, the caption can end up orphaned a paragraph away or the figure can land in the wrong place relative to the text that references it. After converting, this is the first thing to check — read through and confirm each figure sits sensibly near its reference and keeps its caption. Where the converter got it wrong, you may need to reposition images or re-group caption text in the ePub. For figure-heavy documents, budget time for this cleanup.

Can I convert a scanned (image-only) book to ePub with images intact?

The images are intact by definition — a scan is already images — but there is no selectable text, so the ePub would be a series of page images with no reflowable, resizable text, which defeats much of the point of ePub. To get real text you OCR first, which adds a text layer while the figures remain as images; you then have a reflowable text-plus-images ePub, after proofreading the OCR. So for a scanned illustrated book: OCR to recover the text, keep the images, convert, then verify both the text accuracy and that the images survived.

How do I verify the images survived the conversion?

Open the finished ePub on a real device or reader app and page through the whole thing, not just the first chapter — converters often handle early pages well and degrade later. Confirm every figure is present, reasonably placed near its reference, captioned correctly, and sharp at the reader's zoom levels. Check a few on a small screen (phone) and a larger one (tablet) since reflow differs. If images are missing, blurry, or detached from captions, fix those before publishing. Treating it as "convert, then read the whole thing as a reader would" is the only reliable check.

Is it safe to convert a confidential or unpublished book online?

Unpublished manuscripts and licensed content are sensitive, and you should only convert books you have the right to convert. For privacy, prefer a tool that processes the file locally. ScoutMyTool converts PDF to ePub in your browser tab, so the book never leaves your machine. For anything unpublished or under license, confirm the tool does not upload before converting, and respect any DRM and rights.

6 min read

How to convert a PDF to ePub with images intact

By ScoutMyTool Editorial Team · Last updated: 2026-05-21

The first illustrated guide I converted to ePub came out the other side as clean text with every figure missing — the converter had simply dropped the images it could not place, and a how-to book with no how is just a list of captions for nothing. Converting a text-and-image PDF to ePub is harder than converting plain prose, because the images have to move from fixed positions into a reflowing layout without getting lost, mangled, or detached from their captions. This guide covers doing it with the figures intact: why images break in conversion, keeping quality and placement, when reflowable ePub is the wrong target for an image-led book, and how to verify every figure survived.

How different image content behaves in conversion

Content	Reflow risk	Tip
Inline figures with captions	May separate from caption	Keep image + caption grouped
Full-page / full-bleed art	Cropped or shrunk to fit	Consider fixed-layout ePub for art books
Charts / diagrams	Fine if high-res	Export at adequate resolution
Photographs	Over-compression artifacts	Balance quality vs. file size
Scanned pages	No text; image-only	OCR for text; images stay as images
Decorative backgrounds	Often dropped	Usually fine to lose in reflow

Step by step — convert an illustrated PDF to ePub

Decide reflowable vs. fixed-layout. For text-led books with supporting figures, reflowable ePub is right. For art books, comics, or cookbooks where layout is the point, consider fixed-layout ePub or keeping the PDF.
Start from the best source. Use a born-digital, full-resolution PDF; if it is a scan, OCR it first so you get reflowable text while the images stay as images.
Convert with image preservation. Run PDF to ePub and choose settings that keep images and do not over-compress. See the general PDF-to-ePub guide for the text side.
Check figure quality and size. Photographs as good-quality JPEG, diagrams as PNG; balance fidelity against ePub file size, since image-rich books get large — see quality vs. size.
Fix captions and placement. Confirm each figure sits near its reference and keeps its caption; re-group or reposition where the converter erred.
Reuse assets if needed. Pull originals from the PDF with Extract Images (see extracting images) to re-insert a figure at full quality.
Verify on real readers. Page through the whole ePub on a phone and a tablet, confirming every image is present, placed, captioned, and sharp before you publish or distribute.

PDF eBook to ePub: the general conversion guide.
eBook to PDF: the reverse direction.
Extract images from a PDF: recovering figures at full quality.
Share without losing quality: balancing fidelity and size.
Embedding fonts: keeping captions and labels intact.
PDF to ePub tool: convert in your browser.
All ScoutMyTool PDF tools: the full toolkit.

FAQ

Why do images get lost or look wrong when I convert a PDF to ePub?: Two reasons. First, the conversion has to move from PDF's fixed layout (images pinned at exact coordinates) to ePub's reflowable model (images flow inline with text), so a figure that sat in a precise spot now floats near the paragraph it followed — and a weak converter can drop images it cannot place, or detach them from captions. Second, image quality: a converter may re-compress images aggressively to shrink the ePub, introducing artifacts. Good conversion preserves the images, keeps them with their captions, and does not over-compress. The fix is choosing a tool that handles images well and then verifying every figure survived.
How do I keep image quality high in the ePub?: Start from the highest-quality source — a born-digital PDF with full-resolution images beats a pre-compressed one — and use conversion settings that do not crush the images. ePub images are typically JPEG or PNG; photographs suit JPEG at a good quality level, while diagrams and line art are sharper as PNG. There is a genuine trade-off with file size: an image-rich ePub can get large, and some stores cap e-book file sizes, so you balance fidelity against size rather than maximising either. Keep a master and, if needed, produce a lighter version, but never let the convenience copy be the only one.
My book is mostly images (art book, cookbook, comic) — is reflowable ePub right?: Maybe not. Reflowable ePub is designed for text that rewraps to any screen, which is ideal for novels but works against highly-designed, image-led books where the exact layout is the point — an art book, a photography collection, a comic, or a cookbook with carefully placed photos. For those, a fixed-layout ePub (which preserves the exact page design like a PDF, but as an e-book) or simply keeping the PDF is often better than a reflowable conversion that scatters the carefully-placed images. Match the format to the content: reflowable for text-led books, fixed-layout or PDF for design-led ones.
What happens to captions and the text around figures?: In a good conversion, a caption stays associated with its image and travels with it as the text reflows; in a poor one, the caption can end up orphaned a paragraph away or the figure can land in the wrong place relative to the text that references it. After converting, this is the first thing to check — read through and confirm each figure sits sensibly near its reference and keeps its caption. Where the converter got it wrong, you may need to reposition images or re-group caption text in the ePub. For figure-heavy documents, budget time for this cleanup.
Can I convert a scanned (image-only) book to ePub with images intact?: The images are intact by definition — a scan is already images — but there is no selectable text, so the ePub would be a series of page images with no reflowable, resizable text, which defeats much of the point of ePub. To get real text you OCR first, which adds a text layer while the figures remain as images; you then have a reflowable text-plus-images ePub, after proofreading the OCR. So for a scanned illustrated book: OCR to recover the text, keep the images, convert, then verify both the text accuracy and that the images survived.
How do I verify the images survived the conversion?: Open the finished ePub on a real device or reader app and page through the whole thing, not just the first chapter — converters often handle early pages well and degrade later. Confirm every figure is present, reasonably placed near its reference, captioned correctly, and sharp at the reader's zoom levels. Check a few on a small screen (phone) and a larger one (tablet) since reflow differs. If images are missing, blurry, or detached from captions, fix those before publishing. Treating it as "convert, then read the whole thing as a reader would" is the only reliable check.
Is it safe to convert a confidential or unpublished book online?: Unpublished manuscripts and licensed content are sensitive, and you should only convert books you have the right to convert. For privacy, prefer a tool that processes the file locally. ScoutMyTool converts PDF to ePub in your browser tab, so the book never leaves your machine. For anything unpublished or under license, confirm the tool does not upload before converting, and respect any DRM and rights.

Citations

Wikipedia — “EPUB,” including reflowable and fixed-layout modes and image support. en.wikipedia.org/wiki/EPUB
Wikipedia — “JPEG” (ISO/IEC 10918), the lossy image format used for photographs in e-books. en.wikipedia.org/wiki/JPEG
Wikipedia — “PDF” (ISO 32000), the fixed-layout source format. en.wikipedia.org/wiki/PDF

Convert without losing the pictures

Turn an illustrated PDF into an ePub with images intact using ScoutMyTool’s in-browser converter — your book never leaves your machine.

Open PDF to ePub →