PDF/UA-1 Validation Errors

It means the PDF is missing a valid language tag (Lang entry) at the document, structure element, or content level. The fix is to set a correct language identifier in the document metadata or the affected structure element. Covered under clause 7.2-34.

Each validator implements ISO 14289-1 independently and applies different logic, heuristics, and rule coverage. In our study of 155 PDF test files, the four leading validators produced inconsistent results in 50.3% of cases – meaning the same file passed one tool and failed another. This does not mean one tool is wrong. It reflects genuine differences in how each tool interprets ambiguous clauses, which errors it prioritizes, and how it handles edge cases. The practical consequence: relying on a single validator can produce incomplete compliance conclusions. Cross-checking with at least two tools is strongly recommended for any document requiring formal compliance. See also: PDF/UA Validator Comparison (155 files tested).

Many can. Automated tools – including PDFix Desktop’s Auto-Fix – resolve all machine-checkable errors: missing metadata, font embedding, DisplayDocTitle, language tags, and role mapping. However, approximately one third of PDF/UA criteria require human verification and cannot be automated. Errors that need manual judgment include whether alt text meaningfully describes an image, whether reading order reflects the document’s logical structure, and whether heading levels match the actual content hierarchy. A document can pass every automated check and still fail real-world accessibility testing. Best practice is to run automated fixes first, revalidate, then manually review remaining errors.

The fix depends on whether the image carries meaning – and no automated tool can decide this for you.

– Informational image – add meaningful Alt text describing what the image communicates, not just what it looks like.

– Decorative image – do not add empty alt text. Mark it as an Artifact instead. An artifacted element is excluded from the tag tree and ignored by screen readers. Adding empty alt text to a decorative image is itself a PDF/UA error.

In PDFix Desktop, use the Tags panel to set the Alt text or change the element type to Artifact depending on which applies.

Related but not the same. Both are font errors that frequently appear together – but they describe different problems.

“Font not embedded” (clause 7.21.4.1-1) means the PDF references a font that is not included in the file itself. Screen readers cannot reliably read text when the font is missing, which makes the document inaccessible and causes it to fail PDF/UA validation.

“Character cannot be mapped to Unicode” (clause 7.21.7-1) means the font is present but lacks the data needed to translate its characters into readable text. The document may look correct on screen but fails when a screen reader, search tool, or copy-paste tries to extract the text.

In PDFix Desktop, use Embed Fonts to fix the first error and Add Missing Unicodes for the second. For complex cases – such as broken encoding tables or CID font errors – PDFix also offers an automated Font Fix action that detects and repairs both issues in one step.