Fonts in PDF Files: Embedding and Subsetting

Why fonts matter in a PDF

A PDF says “draw the letter A using font X at this position”. If font X is inside the file, every device draws the same A. If it isn’t, the reader must guess a replacement.

Three ways a font can be stored

Status	Meaning	Result
Embedded	The whole font file is inside the PDF	Always correct, larger file
Embedded subset	Only the characters actually used are included	Correct and small. The most common setting
Not embedded	Only the font’s name is stored	Looks right only if the reader’s computer has that font

Subset fonts show up with a random prefix like ABCDEF+Calibri. That’s normal.

What goes wrong when fonts aren’t embedded

• Text appears in a different font (often Arial, Helvetica or Courier).
• Lines overflow, overlap or wrap differently.
• Symbols, Asian characters or special glyphs turn into boxes (☐) or random characters.
• Print shops and PDF/A validators reject the file.

The “copy gives gibberish” problem

Even with an embedded font, copying text can produce nonsense like ÿþ¤¢. That happens when the PDF lacks a ToUnicode map, the table that tells software which real letter each glyph represents. The page looks fine but the computer doesn’t know what the characters mean. See Can’t copy text from a PDF.

How to make sure fonts are embedded

• Word (Windows): File > Options > Save > tick Embed fonts in the file. Saving as PDF from Word embeds fonts by default.
• LibreOffice: Export as PDF embeds fonts automatically.
• Print to PDF drivers (Microsoft Print to PDF, macOS Save as PDF) embed fonts.
• For print or archiving, save as PDF/A or PDF/X, which require embedding.