PDF/A error: glyph is not present in the embedded font
The document draws a character whose glyph is missing from the embedded font program, usually because the subset was built from an incomplete character list. Fix it by re-embedding the font with a subset that covers every character actually used, or by embedding the full font.
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
- Aggressive subsetting: the writer embedded only the glyphs it thought were used and missed some, commonly on text added after the subset was computed.
- Text was edited or merged after export, introducing characters the existing subset does not cover.
- The encoding maps a character code to a glyph name the font does not define — frequently with accented characters, dashes or currency symbols.
How to fix it
- Re-export from the source with subsetting recomputed, or turn subsetting off entirely for the affected font.
- Avoid post-export text edits on documents that must be archival; re-export instead.
- Convert the document and check the residual issues — if the glyph cannot be supplied, this failure survives conversion and needs fixing at the source.
Frequently asked
Is this the same as "font not embedded"?
No. The font is embedded; it just does not contain everything the page asks it to draw. The failure mode is different too: an unembedded font renders with a substitute, while a missing glyph typically renders as nothing or as a notdef box.
Why does the document look fine in my reader?
Many readers silently substitute a glyph from another font when one is missing. That is exactly the environment-dependent behaviour PDF/A forbids — it looks fine on your machine and is not guaranteed to anywhere else.