Linearize PDF

An ordinary PDF keeps its cross-reference table at the end of the file, so a reader has to fetch the last bytes before it can locate anything at all. Over HTTP that means waiting for the whole download, or at least several round trips before the first page appears.

A linearized PDF puts the first page and a hint table at the front. A reader that supports byte-range requests can fetch a small prefix, draw page one, and pull the rest in the background. The visitor starts reading while the download is still running. This is what Acrobat has always called fast web view.

The pages, fonts, images and text are untouched. Linearizing is a rearrangement, not a re-encoding, so nothing is recompressed and no quality is lost. A linearized PDF is still a perfectly ordinary PDF: a reader that knows nothing about linearization opens it exactly as before.

It also does not shrink the file, and it will not speed up a PDF that a visitor downloads in full before opening. It pays off for large documents read in a browser, and it costs almost nothing to do, which is why many publishers linearize by default.

The rewrite runs on your device with QPDF compiled to WebAssembly. Your document is never sent anywhere, which matters when the file you are preparing for the web is not public yet.