calibre Images Missing After Conversion: How to Fix It

When images disappear after an e-book conversion, the cause is usually not a general calibre failure. The converted book may contain broken image references, the source package may omit image files, remote images may never have been downloaded, or the selected output format or reading device may not support the resulting image as expected. Covers add another complication because the library cover, the embedded cover, and an inline cover page are not always the same resource. The safest approach is to determine exactly where the image disappears, then change only the part of the workflow responsible for that stage.

1. Confirm the Symptom With a Small Safe Test

Start by preserving the original book. Do not delete formats, replace your library, or perform a bulk conversion while troubleshooting. Select one affected title and create a controlled test from it.

1. Note the current input format and intended output format.
2. Open the source format in calibre's viewer and visit several known image locations.
3. Convert that title to EPUB unless you specifically need to test another format.
4. Open the new output in calibre's viewer before sending it anywhere.
5. If the output looks correct locally, transfer that exact file to the target device and check it again.

This sequence divides the problem into three useful categories. If the source already shows empty boxes, the resource is probably missing or broken before conversion. If the source works but the converted file fails in calibre's viewer, investigate packaging, references, and conversion settings. If both files work in calibre but the target device fails, concentrate on device compatibility, transfer behavior, and image limits.

Use an image in the middle of a chapter as your test, not only the cover. Covers can be generated from metadata or treated specially during conversion, so a successful cover does not prove that ordinary diagrams and chapter artwork were packaged correctly.

Success looks like this: you can identify the first stage where the image disappears. Once you know that stage, stop changing unrelated settings.

2. Check the Image Resources and Conversion Inputs

2.1 Inspect the book with Edit book

For EPUB, AZW3, and other supported editable formats, right-click the book and select Edit book. The File browser on the left separates text files, stylesheets, images, and other resources. Expand the Images section and look for the missing cover, diagram, or chapter art.

If the expected image does not appear in the File browser, it is not embedded in that book. A paragraph may still contain an image tag, but the tag cannot display anything when its target file is absent. Use the editor's book-checking tools to find broken references, missing files, malformed paths, and case mismatches.

If the image exists, open it directly in the editor. Then open the HTML file that should display it and inspect the corresponding image reference. Common problems include:

• A filename that differs by one character from the embedded file.
• A path pointing to the wrong folder level.
• A reference to a file on the creator's computer rather than inside the e-book.
• Uppercase and lowercase differences that become significant on Linux or some reading systems.
• Spaces or unusual characters handled inconsistently by the source application.
• An image declared in the markup but omitted from the book's package metadata.

You can import a legitimately available missing resource through the editor and update the HTML reference. If you rename or replace a file using the editor's built-in commands, calibre can update relevant references more safely than a manual archive edit.

Success looks like this: the image opens from the File browser and appears in the editor's live preview. Save a copy, reconvert it, and stop if the new output works.

2.2 Package HTML and its images together

Raw HTML is a frequent source of the calibre images missing after conversion problem. An HTML page is only one part of a web-style document. Its pictures may live in an adjacent folder, elsewhere on the computer, or on a website. Moving or importing only the HTML file can leave those resources behind.

Place the main HTML file and every required local resource in one self-contained folder structure. For example, the HTML could sit at the top level while its images are stored in an images subfolder. Confirm that the paths in the HTML match that structure. Then create a ZIP archive containing the HTML and image folders, add the ZIP as a book format, and select ZIP as the input during conversion.

Do not ZIP only the HTML file. Open the archive before converting and verify that the image files are actually present. Also avoid references that climb outside the package, such as paths leading to a parent folder. calibre normally limits which external local files HTML input can collect, partly because unrestricted local-file access can create a security risk.

Success looks like this: the ZIP contains the HTML, images, and related files, and the converted EPUB lists those images in Edit book.

2.3 Replace remote image references with local resources

An HTML document may display correctly in a web browser because the browser downloads images from remote addresses. That does not mean the images are embedded in the document or guaranteed to enter the converted e-book. The remote server may reject automated access, require authentication, block hotlinking, or become unavailable during conversion.

If you are authorized to use the images, download them to the source folder, change the HTML references to local relative paths, and package them with the rest of the book. Do not rely on an internet connection to make an offline e-book complete.

Success looks like this: disconnecting from the internet does not make the source package lose its images, and the converted book contains local image files.

2.4 Distinguish the metadata cover from an embedded cover page

The cover shown in calibre's library interface can come from calibre's metadata database. It may not be the same image embedded inside the selected book format. As a result, the library can display a cover even when the EPUB or other source format lacks a working cover page.

Open Edit metadata and verify the desired cover, but also inspect the book in Edit book. Look for a cover image and a title-page HTML file. In the File browser, calibre can identify resources marked as the cover and title page. If the wrong image is marked, correct the designation and save the book before converting again.

A cover that appears in the device's library view but not inside the book, or vice versa, is usually a cover-role problem rather than evidence that every image resource failed.

3. Check Output Settings and Device Limitations

3.1 Reset saved conversion settings for the test

calibre can remember conversion choices for an individual book. That is useful in normal operation but confusing during calibre troubleshooting because an old output profile or format-specific option may continue to affect each attempt.

For the test book, open the conversion dialog and restore the relevant options to their defaults. Confirm that the intended input format is selected in the upper-left area and that the desired output format appears on the right. If the book has several source formats, calibre may otherwise convert a different format from the one you inspected.

Pay particular attention to the output profile and format-specific image options. Output profiles can influence how images are resized for a device. For EPUB and KEPUB output, maximum-image-size controls can also resize image files. Resizing normally should not remove an image, but an extreme setting, malformed source image, or reader limitation can produce an unusable result.

Success looks like this: a conversion with default settings displays all expected images in calibre's viewer. Retain those defaults unless a specific device requires a carefully tested change.

3.2 Test a broadly compatible image format

Some older devices and reading applications have limited image-format support. A file can be properly embedded yet display as an empty box on the target reader. SVG, WebP, unusually encoded PNG files, very large images, or images using uncommon color profiles can expose these limitations.

If one image fails while ordinary JPEG or PNG pictures work, export the affected resource from Edit book and make a test copy in a conventional RGB JPEG or PNG format. Replace the image through the editor so its links remain consistent, save the book, and convert again. Preserve the original file outside the test copy.

Large maps, scanned pages, and diagrams may also exceed a device's practical memory or dimension limits. Resize a copy to dimensions appropriate for the reader while preserving its aspect ratio. Avoid shrinking text-heavy diagrams so far that they become unreadable.

Success looks like this: the converted test works on the device with a standard-format, reasonably sized image. At that point, there is no reason to modify unrelated library, email, or server settings.

3.3 Separate conversion problems from transfer problems

If the output works in calibre's viewer but not on the device, do not keep reconverting blindly. Delete only the test copy from the device, safely disconnect it, reconnect it in the correct USB or file-transfer mode, and send the known-good output again.

On Android-based hardware, make sure the connection is in a file-transfer mode rather than charging-only mode. On Windows, confirm the device and storage appear normally in File Explorer. On macOS, check whether calibre has permission to access removable volumes or connected devices. On Linux, verify that the mounted storage is writable and that the current user can access it.

If you deliver the book through email, the Content server, a browser, cloud storage, or another reading application, download the delivered file back to the computer and open it in calibre. Compare it with the known-good local output. A service may have served an older format, the reader may have cached a previous copy, or a plugin may have performed an automatic conversion during delivery.

Firewall and antivirus rules matter mainly when the file is transferred through the Content server or another network route. They do not explain missing images inside a locally opened conversion. Cloud synchronization is relevant when the library or working folder is only partially downloaded, locked, or being changed during conversion.

4. Use Logs and Conversion Debug Output

4.1 Read the conversion job details

After a conversion finishes, open the Jobs area and inspect the job details. Search for the missing filename, image warnings, inaccessible paths, unsupported data, or messages about resources that could not be opened. A completed job can still contain warnings that explain missing content.

If a filename appears in an error, return to the source and verify that the file exists at that exact path. On case-sensitive systems, Diagram.PNG and diagram.png are different names.

4.2 Generate conversion debug output

calibre's conversion dialog includes an option to save debug output to a folder. Enable it for one test conversion and choose a new empty folder. calibre writes intermediate versions of the book into stage folders commonly named input, parsed, structure, and processed.

Check the missing image at each stage:

• Missing from input: the input plugin never received or collected the resource. Repair the source package or HTML paths.
• Present in input but missing later: inspect the intermediate HTML and job log for a transformation or reference problem.
• Present through processed but absent from output: suspect the output format, output plugin behavior, or a malformed resource.
• Present in the final file but invisible on the device: suspect reader compatibility, image dimensions, caching, or transfer behavior.

This is one of the most efficient ways to isolate the failure because it shows what calibre had at each point in the conversion pipeline.

4.3 Use device debugging only when transfer is implicated

If calibre cannot detect the target reader reliably, use Preferences > Miscellaneous > Debug device detection. This can help distinguish a device connection problem from a conversion problem. It is not useful when the converted file already fails in calibre's viewer.

Advanced users can also run calibre-debug from a terminal. Its options include device-driver debugging and tools for opening or extracting supported book formats. Most users should begin with job details, Edit book, and conversion debug output because those tools usually reveal missing image resources without command-line work.

5. Run a Clean Temporary Test Before Reinstalling

Reinstalling calibre rarely restores an image that was never embedded in the source. Before reinstalling, deleting the library, disabling security software, or changing many preferences, run a clean test outside the normal workflow.

1. Copy one legitimate source book and its resources to a local folder that is not cloud-synchronized.
2. Use a short folder path containing ordinary letters and numbers.
3. Add the test copy to calibre as a separate temporary entry.
4. Disable optional third-party plugins only for the test if they participate in import, conversion, metadata, or device transfer.
5. Convert with default options to EPUB.
6. Open the result in calibre's viewer and Edit book.
7. Transfer the same verified file to the device without an additional automatic conversion.

If the clean test works, calibre itself is functioning. Compare the working test with the normal workflow one difference at a time. Re-enable plugins individually, restore the original output profile, or return the source to cloud storage only after confirming the previous step still works.

If the clean test fails and debug output shows that the image was absent at the input stage, the source is incomplete. Reinstalling will not repair it. Obtain a complete, authorized source file or add the missing resources yourself when you have permission.

6. Quick Fix Checklist

• Confirm whether the source image works before conversion.
• Verify that the correct input format is selected.
• Open Edit book and check whether the image is embedded.
• Use Check book to locate broken image references.
• Package HTML, images, and stylesheets together in one ZIP.
• Replace remote image addresses with authorized local files.
• Check capitalization, relative paths, and filename extensions.
• Verify both the metadata cover and embedded cover page.
• Reset saved conversion settings for one controlled test.
• Test ordinary RGB JPEG or PNG resources.
• Resize exceptionally large images for the target device.
• Inspect conversion job details for resource warnings.
• Use conversion debug output to find the first failing stage.
• Compare calibre's viewer with the target device.
• Transfer the verified output without another conversion.
• Run a local, non-cloud test before reinstalling calibre.

7. Frequently Asked Questions

7.1 Why does the cover appear in calibre but disappear after conversion?

The library cover may exist only in calibre's metadata while the selected source format lacks a correctly embedded or marked cover. Inspect the source with Edit book, confirm that the image and title page exist, and verify their cover roles. Also make sure you converted the format you inspected.

7.2 Why do images from an HTML file become broken boxes?

The HTML probably references files that were not included with it. Put the HTML and every required image in a self-contained folder structure, use relative paths, create a ZIP containing the complete structure, and convert that ZIP. Remote website images should be downloaded and embedded locally when you are authorized to use them.

7.3 Why do images work in calibre's viewer but not on my e-reader?

The output file is probably intact, so focus on the device. Test a standard JPEG or PNG, reduce exceptionally large dimensions, confirm that the device supports the output format, remove cached copies, and transfer the exact file that worked in calibre. Do not continue changing conversion settings after the local output has been verified.

7.4 Can antivirus or firewall software remove images during conversion?

It is uncommon for a firewall to affect a conversion that uses fully local resources. Security software can matter if it blocks access to source folders, quarantines a resource, or interferes with network delivery through the Content server. Check security logs and permissions rather than disabling protection broadly.

7.5 Should I reinstall calibre when images are missing?

Not initially. Missing image resources, broken paths, incomplete HTML ZIP files, and device limitations are much more likely. Reinstall only after a clean local test with a known-complete source also fails and the debug evidence points to the application rather than the book.

7.6 What should I do if only one book is affected?

Treat it as a source-specific problem. Inspect that book's File browser, references, package contents, and conversion debug output. If other books convert with images intact, global settings and the calibre installation are unlikely to be the cause.