What is PDF/A, and how is it different from a normal PDF?

You upload a document to a court portal, a university repository or a government system, and it comes back rejected: "file must be PDF/A". The file opens perfectly. It looks fine. So what is wrong with it?

Nothing, today. PDF/A is about whether it will still open, and look the same, in thirty years.

A PDF that has to outlive the software that made it

An ordinary PDF is allowed to lean on things outside itself. It can use a font it expects your computer to have. It can describe colours in a way that depends on the device showing them. It can run scripts, play video, be locked with a password, or point to other files.

All of that works now. For a document that has to be kept for decades, each one is a way to fail later: the font is gone, the password is lost, the video player no longer exists.

PDF/A, published by ISO as ISO 19005, is a stricter version of PDF that removes those dependencies. A PDF/A file is still a normal PDF, and any PDF viewer opens it. It is simply one that follows extra rules.

What PDF/A requires, and what it forbids

The versions (1, 2 and 3) and the letters after them (b, u, a) are explained in PDF/A-1b, 2b, 2u, 3b or 3u: which to choose.

Does a PDF/A file look different?

It should not. Preserving the look is the whole point.

When a file only needs fixing (a colour profile added, archival metadata written, a script removed) the pages themselves do not have to change. When they do have to be rebuilt, usually to embed a font that was missing, the closest available font is embedded. The spacing of every line stays the same, but letter shapes can differ slightly from the original.

Why the PDF/A file is bigger

Because the fonts and the colour profile now travel inside the file instead of being borrowed from whatever computer opens it. A text document that relied on standard fonts can grow noticeably. A file that already embedded its fonts barely changes.

Can you edit a PDF/A file?

You can open and edit it in any PDF editor. The catch is that editing software does not necessarily keep to the PDF/A rules when it saves, so an edited, compressed or merged PDF/A file may no longer be PDF/A, even if it still says it is.

The practical rule: convert last, after every other change, and convert again if anything changes afterwards.

How to tell whether a PDF really is PDF/A

A PDF/A file carries a label in its metadata saying which version it follows. Some viewers read it: Acrobat, for instance, shows a bar saying the file claims to comply with PDF/A.

A claim is not proof. Any program can write that label, and a file edited after conversion keeps the label while breaking the rules. The only real test is to check the file against the rules of the standard with a validator.

KovaPDF notices when a file you upload already says it is PDF/A and tells you, but it does not treat that as evidence. Every file it converts is checked against the rules of the standard you chose, with an open-source validator, before the download button appears. If it does not pass, you get no file and a plain explanation instead. Nothing is ever just renamed PDF/A.

How to convert a PDF to PDF/A

1. Open [PDF to PDF/A](/pdf-to-pdfa) and choose your file.
2. Read the compatibility check. Before anything changes, the tool lists what will happen to this document under five headings: fonts, colour, interactive content, attachments and security.
3. Choose the standard. PDF/A-2b is recommended. The others (PDF/A-1b, 2u, 3b and 3u) are one click away.
4. Convert. The document is converted, then checked against the standard. You can download it only if it passes.
5. Read what changed. The result screen lists every change, and whether the pages were compared before and after.

What conversion cannot fix

• A password. PDF/A cannot be encrypted, so a file that needs a password to open has to be unlocked first with Unlock PDF. Printing and copying restrictions are removed automatically.
• A digital signature. Converting changes the file, so a signature cannot stay valid. It is removed and its visible stamp kept. Keep the original as the signed record.
• Unsearchable scans. A scan becomes PDF/A as it is, with no text recognition added. To make it searchable, run it through OCR PDF first, then convert.
• Form fields, sometimes. When the file can be fixed in place, fields stay fillable. When the pages must be rebuilt, they are flattened: the typed values stay visible but can no longer be edited.
• Accessibility tags, when pages are rebuilt. Tags are kept when the file is fixed in place, and lost when it has to be rebuilt.