PDF/A error: font is not embedded
PDF/A requires every font used in the document to be embedded in the file itself, because a font referenced by name renders differently — or not at all — on a machine that does not have it. Fix it by re-exporting with font embedding enabled, or by converting the document so the missing font programs are embedded for you.
The Fonts check group covers 6.3.x embedding, subsets, metrics, encoding. The exact sub-clause is reported on the issue itself when the validator raises it.
What causes it
- The producing application relied on one of the standard 14 fonts (Helvetica, Times, Courier and friends), which older PDF writers were allowed to reference by name rather than embed.
- Font embedding was switched off in the export settings, often to keep file size down.
- The font's license forbids embedding, so the writer silently skipped it.
- The document was assembled from several sources and one of them contributed an unembedded font.
How to fix it
- Re-export from the source document with font embedding enabled — this is always the best outcome, because the original font is still available there.
- Convert the document: gopdfrab embeds the font programs it can supply and reports anything it could not as a residual issue.
- If the font cannot be licensed for embedding, substitute a font that can be before exporting. A substitution made deliberately is better than one made by whatever reader opens the file in 2050.
Frequently asked
Why do the standard 14 fonts fail PDF/A?
Because "Helvetica" is a name, not a font program. Two readers can resolve it to different metrics and different glyph coverage, so the page is no longer guaranteed to look the same — which is the one thing PDF/A exists to guarantee.
Does embedding a subset count?
Yes. PDF/A allows subset embedding, as long as the subset contains every glyph the document actually uses and the font dictionary describes it correctly. Subsetting is normal and keeps the file small.
Can conversion embed a font that is not on the machine?
Only if a suitable font program is available to it. Where none is, the conversion reports the issue as residual instead of claiming success — a converted file that is still missing a font is not conformant and gopdfrab will not say it is.