Why a PDF conversion fails, and which failure you are looking at

Five failures that look identical from outside

From a visitor's side every failed conversion looks the same: you waited, and then it did not work. Underneath they are not remotely the same event.

• Encrypted. The document has a password and the tool cannot open it. Nothing about your file is broken.
• No text layer. The pages are images. There is nothing to convert to text, so the output would be empty.
• Too large or too long. The work genuinely cannot finish inside the time a web request is allowed to live.
• Not the file you think. A .pdf extension on something that is not a PDF, or a PDF truncated by a failed download.
• Our end. The conversion service was restarting, overloaded, or has a bug.

The reason this matters is that only the last one is ours to fix, and only the first three have anything you can do about them. Telling them apart is the whole job of a good error message, which is why our tools report a cause rather than a generic failure wherever the cause is actually known.

The file is password protected

A PDF can carry two different passwords and they fail differently. An open password means the file's contents are encrypted and nothing can read a single page without it — every converter fails immediately, and correctly. An owner password leaves the file readable but marks it as restricted, and some tools refuse to process it on principle even though they can read it perfectly well. What a PDF password actually protects explains the difference properly.

If you know the password, remove it first and then convert the result. If you do not know it, no tool can help you — that is the encryption working as designed, and anything claiming otherwise on a modern AES-encrypted file is either guessing passwords or lying.

There is no text in the file

This is the most common surprise, because a scanned page looks exactly like a typed one on screen. A photograph or scan of a document contains a picture of words and no words. Ask it for Word or Excel output and the honest answer is that there is nothing to extract.

You can check in two seconds without any tool: open the PDF and try to select a line of text with your cursor. If you get a selection rectangle over the whole page instead of a line highlighting, it is an image. Page count and file info will also tell you whether a text layer exists.

The fix is OCR, which reads the picture and writes a text layer underneath it. Convert the OCR'd file, not the original. Be aware that OCR is a very good guess rather than a transcript — OCR and PDF to Word are not the same job, and the digits it confuses are the ones that matter most.

It is too large, or there is too much of it

Server-side conversions run inside a time limit, because a web request that never ends is worse for everybody than one that stops and says so. A six-hundred page document, a file full of high-resolution scans, or a page with tens of thousands of individual vector objects can all exceed it legitimately.

Three things that usually work:

• Convert a range rather than the whole file. Most tools here take a page selection, and most people only need part of the document.
• Split it first and convert the pieces.
• Compress it first if the size is coming from images. Compress PDF often takes a scanned document from unworkable to routine.

A timeout is reported as a timeout here rather than as a generic failure, precisely so that this is the paragraph you end up reading.

When it is us

Some failures are ours, and the useful thing is that they look different: they happen to files that worked yesterday, they happen to several tools at once, or they happen on the first try and not the second.

If a conversion fails and the same file succeeds a minute later, you hit a service restarting. If it fails twice with a file you are confident is fine, that is worth telling us about — say which tool, roughly how many pages, and what kind of document. Please do not send the file. We do not want it and almost never need it; what the document is narrows a bug down far faster than the document itself.

Browser-based tools fail differently again. Those run entirely on your own machine, so an older browser, a very large file against limited memory, or a blocked script will stop them while every server tool keeps working.