What's PDF/A and when should you convert a PDF for archiving?

PDF/A in brief

PDF/A is a family of international standards, published as ISO 19005, for the long-term preservation of electronic documents.

• The central rule: Its central rule is that a file must be self-contained: everything needed to display it the same way in the future is embedded inside it. Fonts must be embedded, colors must be defined device-independently, and document metadata is stored in a standard XML form. Features that depend on external resources or could change the appearance aren't allowed, including encryption, JavaScript and links to external content.
• PDF/A-1: The family has several parts. PDF/A-1 is the oldest and most restrictive.
• PDF/A-2 adds support for more modern PDF features like transparency, layers and JPEG 2000 compression.
• PDF/A-3 also allows files of any kind to be embedded.
• PDF/A-4: And PDF/A-4 is based on PDF 2.0.
• Conformance levels: Parts also define conformance levels, of which level b, which guarantees the visual appearance, is the most widely used.

When PDF/A is needed or worth it

• When it's required: Some courts, government agencies, national archives and records-management policies require or recommend PDF/A for filings and deposits, and many universities ask for it when theses are submitted to their repositories.
• Electronic invoices: Structured electronic invoices in some countries use PDF/A-3 with machine-readable data embedded in the file.
• When it's worth choosing: Outside these cases, PDF/A is worth considering for any document you must keep readable for many years: signed contracts, property records, final reports and personal papers.
• When it's the wrong choice: It isn't needed for everyday sharing, and it's the wrong choice for files that must stay editable, carry a password or include interactive features.

What VelloDoc's conversion does

• The engine and the target: PDF to PDF/A uses Ghostscript to rewrite the document as PDF/A-2, aiming at the basic level that preserves visual appearance.
• What it embeds: It embeds an sRGB color profile as the file's output intent, converts colors to RGB, embeds fonts where they're available and builds the standard XML metadata.
• Unsupported features: Ghostscript is instructed to leave out features that PDF/A doesn't permit and to continue, instead of stopping at the first incompatible element.
• Protected source files: If the source file is password protected, you enter its password so it can be read, and the result is saved without encryption, because PDF/A doesn't allow it.

What can change during conversion

Because the goal is a self-contained, predictable file, some things may not survive.

• Print colors become RGB: Colors defined for print, like CMYK, are converted to RGB. It can shift some shades slightly.
• Scripts, actions and form fields: JavaScript, actions and multimedia are removed, and interactive form fields may no longer behave as fields.
• Passwords and permissions are dropped.
• Attachments: Files embedded as attachments aren't carried over.
• Substituted fonts: A font that's missing from the source and can't be embedded may be substituted. It can change line breaks.
• Compare before you rely on it: Always compare the converted file with the original, page by page if the document matters, before you rely on it or delete the original.

Validate the result

A conversion tool can aim for PDF/A, but only validation shows whether a file actually conforms.

• veraPDF: The open-source veraPDF validator, supported by the PDF Association, checks a file against the PDF/A rules and lists any problems it finds.
• Validate before submitting: If a receiving court, archive or system requires PDF/A, validate before submitting, and keep the report if the recipient may ask for it.
• When validation fails, the cause is often in the source file: damaged structure, unusual fonts or elements that couldn't be converted. Running Repair PDF first and converting again can help.
• If errors remain, dedicated archiving software may be needed.

Preparing a document for archiving

Order matters.

• Repair first: Repair damaged files first.
• Make scans searchable: Make scanned documents searchable with OCR PDF, because PDF/A preserves what's there but doesn't add text to images.
• Set the metadata: Set a clear title, author and subject with View / Edit Metadata before converting, not after: editing properties afterwards removes the XML metadata that PDF/A requires.
• Complete and flatten forms: Complete any forms, and if their appearance must be fixed exactly, Flatten PDF turns pages into images, at the cost of selectable text.
• Convert last, then validate: Convert to PDF/A as the last step, validate the result, and avoid running other tools on the archived file, since any later change can break conformance.
• Keep the original alongside the archival copy.

Questions

Tools for this topic