A PDF is a page-description format, not an image format, so what happens when you convert one to CBZ depends entirely on how the PDF was built.

The good case: one image per page

Most comic PDFs are a thin wrapper. Each page contains exactly one embedded image—the scan or the export—positioned to fill the page. Nothing else is drawn.

When that is true, the right conversion is not a conversion at all. The embedded image can be lifted out of the PDF and written straight into the CBZ untouched. Same bytes, same dimensions, same encoding. This is the case that should be lossless, and if your converter re-encodes here it is throwing away quality for no reason whatsoever.

The awkward case: a page that has to be drawn

Some PDFs are not built that way. A page might be split into several image tiles, or have text and vector art layered over the artwork, or store the image in a form a CBZ cannot carry. There is no single embedded image to extract, so the page has to be rendered—drawn to a bitmap at a chosen resolution and saved.

Rendering is a real conversion and the result is a new image. It can look excellent, but it is not the original file, and any tool that tells you otherwise is not being straight with you.

Why the distinction is the whole thing

The two cases are invisible from the outside. Open two comic PDFs and they look identical; one converts perfectly and one has to be redrawn.

That is why CBZ Forge reports a verdict for every page rather than a single progress bar: copied, or re-encoded, or rendered, and which pages were which. You get to see where the losses actually happened instead of assuming there were none.

Before you convert a whole shelf

Run one book first and read the verdict list. If your PDFs came from the same source they will usually behave the same way, so one book tells you what to expect from the rest—and whether these particular files are worth converting at all.