The whole job in one line
Collect the page images, put them into one archive in reading order, and name the archive with a .cbz extension. Any ZIP tool can do it; no comic-specific software is required.
Name pages so they sort correctly
- Zero-pad to a fixed width: 001.jpg through 120.jpg. Without padding, 10.jpg sorts before 2.jpg in plain text order.
- Keep one naming pattern. Mixing page-1.jpg with p10.png and scan_011 forces every reader to guess.
- Use chapter folders only if the reader flattens paths predictably. This one sorts by file name first, then by full path.
- Avoid duplicate numbers. Two files named 012.jpg in different folders are ambiguous by design.
Zip it without wasting time
- JPEG and PNG pages are already compressed. Deflate on top saves a few percent and costs CPU on every page open; choose Store when your tool offers it.
- Archive the images themselves, not the folder that contains them. Most readers flatten a single wrapper folder, but the extra level makes debugging harder.
- Exclude macOS resource forks (__MACOSX), .DS_Store, Thumbs.db, and cover thumbnails you do not want as pages.
- Do not add a password. Browsers have no prompt for ZIP passwords, so an encrypted CBZ is unreadable in a tab.
- Windows: select the images, right-click, Send to, Compressed folder, then rename .zip to .cbz. macOS: select the images and use Compress; compressing a folder instead adds one nested directory.
Add ComicInfo.xml only if it helps
A minimal file at the archive root is enough: a ComicInfo element with Series, Number, PageCount, and Manga. Series and Number become the title, and Manga set to YesAndRightToLeft switches the initial reading direction.
Wrong metadata is worse than none: a stale PageCount or a wrong Manga value changes what the reader shows. Most readers ignore unknown tags, so keep the file small.
Check the archive before you share it
- Rename a copy to .zip and open it: the page count should match, the first and last pages should be present, and no thumbnails should appear in the middle.
- Open it in a reader and jump to a middle page. A missing central directory or a truncated file usually fails at open time, not at page 40.
- Spot-check a spread. Two-page spreads scanned as one image will look wrong in double-page mode.
Common packing mistakes
- Re-saving JPEGs instead of copying them, which lowers quality with every pass.
- Mixing page sizes wildly; the reader has to rescale each page and the reading rhythm suffers.
- Leaving an .exe or readme inside. It is harmless to the reader, but it makes the archive look untrustworthy.
- Assuming a PDF belongs in a CBZ. A PDF is not a page image and will be ignored.