PDF to PDF/A

PDF/A is not a separate file type. A PDF/A file is an ordinary PDF that follows extra rules, and it opens in any PDF reader. What makes it different is a promise: everything needed to show the pages is inside the file, so it does not depend on the fonts, color settings or software of the computer that opens it years later.

The first part, PDF/A-1, was published in 2005 as ISO 19005-1 and is based on PDF 1.4. PDF/A-2 followed in 2011, based on the ISO version of PDF 1.7, and allows newer features such as embedding other PDF/A files. PDF/A-3, from 2012, keeps the rules of PDF/A-2 but allows attachments in any format. A later part, PDF/A-4, is based on PDF 2.0. The tool offers the levels of parts 1 to 3 that archives and portals most often ask for.

Each part has conformance levels, marked by a letter. Level b, for basic, guarantees that the pages will look right. Level u adds that every character of text is mapped to Unicode, so it can be searched and copied reliably. Level a, for accessible, also requires a tagged structure that describes headings, paragraphs and reading order. Level b is the usual target for scanned documents, whose text often comes from OCR rather than from the original file. The Library of Congress keeps a separate record for PDF/A-2b among its format descriptions.

Each rule removes a way in which a document could stop working. A font that is not embedded will be replaced by whatever the reader has, so line breaks shift and characters may go missing. Colors defined only for a particular screen or printer can look different on another device. A password may be lost, a script may stop running in future software, and a link to an outside file may break. PDF/A rules all of these out.

Most PDFs made by office programs are close to PDF/A already. The usual gaps are fonts that were referenced but not embedded, missing color profiles and missing metadata. Files from scanners, older design tools or several rounds of editing can carry more problems, such as transparency in PDF/A-1, broken fonts or embedded media.

The rules also explain what PDF/A is not. It is not a way to protect a document, since passwords are not allowed, and it is not a smaller format; embedded fonts and profiles usually make files a little larger. It is also not a guarantee of accessibility unless the file meets level a, which requires a tagged structure that conversion from an untagged PDF cannot invent.

Who asks for PDF/A? Typical requesters are national and university archives, some courts and public bodies that accept electronic filings, and businesses that must keep records for many years. E-invoicing standards in several European countries build on PDF/A-3. The exact level always comes from the receiving side, so look for it in their filing or upload instructions.

Convert to sets the target: PDF/A-1b, PDF/A-2b, PDF/A-2u or PDF/A-3b. Choose the one the receiving archive, court, tax office or portal names in its instructions. If none is named, PDF/A-2b is a sound general choice, because it allows more modern PDF features than part 1 and has wide support among archive systems.

Color profile decides which output profile is embedded. sRGB suits documents that will mostly be read on screen. The CMYK profiles, FOGRA39 for European print and SWOP for US print, are for files that also go to a print shop. Gray gamma 2.2 is for purely grayscale documents. Embedded files controls attachments: PDF/A-3 allows them, which is how electronic invoice formats such as ZUGFeRD and Factur-X carry their machine-readable XML part inside a readable PDF.

Rasterise problem pages is a safety net. If a page contains something that cannot be fixed, such as a damaged font, it is rendered as an image so the file still passes. The page looks the same, but its text is no longer selectable, so use it when passing matters more than editable text. Make scans searchable adds an invisible OCR text layer to scanned pages in the language you set, which PDF/A-2u needs and which makes any archived scan easier to find later.

Several files in one batch are converted to the same level with the same profile. When documents need different levels, for example invoices with attachments for PDF/A-3b and letters for PDF/A-2b, run them as separate batches.

Remove the password first. A file that is encrypted cannot become PDF/A, so unlock it with the Unlock PDF tool using the password you already know, then convert the unlocked copy. Archives keep their own access controls; the file itself must open freely.

Fill in and finish forms before converting. Interactive features are restricted in PDF/A, and many archives expect forms to be final. Flatten PDF turns filled-in fields and comments into ordinary page content, which converts cleanly. If several documents belong together in one record, combine them with Merge PDF before converting, so the archive receives a single conforming file.

When a portal rejects a file as not PDF/A, run it with Check only first. The report lists what is wrong without changing anything, which helps you decide whether a full conversion is enough or whether the source document needs fixing in the program that created it.

Keep in mind that PDF/A is about how a file is built, not about what it says. A conforming file can still be the wrong document, a poor scan or an incomplete record. Open the result once and check that every page is there and readable before you submit or archive it.