Why Fonts Break After Translation

A translated document can contain the right words and still look broken. Missing characters, unexpected font changes, disconnected letters, and overflowing text usually come from different problems, so each one needs a different fix.

The Short Answer

Fonts appear to break after translation when the original typeface cannot display the target language, or when it is both unavailable on the system opening the file and not embedded in the exported PDF. Some scripts also need shaping and layout support that a basic font substitution cannot provide. In other cases, the font is working correctly but the translation takes up more space than the original text.

Use the visible symptom to identify the likely cause:

What you see	Likely cause	First fix to try
Empty squares or boxes	The font lacks the required glyphs	Switch to a font that supports the target script
A different typeface	The original font is missing or cannot be embedded	Install, replace, or embed the font
Broken Arabic or Indic letters	The app, font, or export process lacks proper shaping support	Open the file in software that supports the script
Clipped or overlapping text	The translation needs more space	Resize the text frame or adjust the layout
Text looks fine on one computer but not another	The file depends on a locally installed font	Embed the font or use a widely available alternative

A Font Does Not Contain Every Character

Unicode assigns a number to each character, but Unicode is not itself a font. The Unicode Consortium explains that fonts map those character numbers to visible glyphs. If a typeface has no glyph for a character, the application has nothing suitable to draw.

That is why an English document set in a Latin-only display font may turn into boxes when translated into Arabic, Hindi, Thai, Chinese, or another script. Accented Latin letters can fail too when a small custom font includes only basic English characters.

The cleanest fix is to choose a typeface with verified coverage for the target writing system. Google’s Noto collection covers more than 1,000 languages and over 150 writing systems, making it a useful starting point when a brand font does not support the required characters. You do not need to use one font for every language; a documented font map is often better:

Content	Example font strategy
Latin, Greek, and Cyrillic	Use one compatible family if its coverage is complete
Arabic or Hebrew	Choose a family designed for the script and test right-to-left layout
Chinese, Japanese, and Korean	Select the correct regional family and check punctuation and line breaks
Devanagari, Thai, or other complex scripts	Verify combining marks and letter shaping in the final application

Font Substitution Changes the Document

A document can name a font without carrying the font data inside the file. If someone opens it on a computer where that typeface is unavailable, the software selects a fallback. The replacement may be wider, narrower, taller, or visually unrelated to the original.

PDF embedding reduces this risk. Adobe explains that an embedded font travels inside the PDF, preventing substitution when the file is viewed or printed on another system. A PDF may include the entire font or only a subset of the characters used in the document. Embedding can still be restricted by the font vendor’s license.

Subsetting is efficient for a finished PDF, but it can cause trouble when the file is edited later. A subset created from an English source may not contain the new characters needed for the translation. For editable work, keep the original DOCX, PPTX, or design file and export a new PDF after translation instead of treating the final PDF as the master copy.

If PDF formatting is already damaged, use the checks in our guide to translating PDFs without losing formatting.

Some Scripts Need More Than Glyphs

Having the right characters in a font is necessary, but it is not always sufficient. Arabic letters change form according to their position in a word. Many Indic scripts combine consonants, vowels, and marks into shaped clusters. Traditional Mongolian uses positional forms and vertical layout.

The Unicode Consortium notes that a font is only one component in a text-rendering system. The application and its shaping engine must also interpret the font’s rules correctly. This explains a common failure: the same sentence looks correct in a modern browser but appears disconnected or misplaced after export through older software.

When this happens:

1. Paste a short sample into a current browser or word processor. If it renders correctly there, the original application or export process is the likely cause.
2. Test a font known to support the script.
3. Check paragraph direction and language settings.
4. Export again from software with current complex-script support.
5. Inspect the final PDF rather than trusting the editable source alone.

Text Expansion Can Look Like a Font Problem

Translation changes both character count and character width. The W3C notes that English and Chinese are especially compact source languages, so translations from them are often longer. Short labels can expand more sharply than full paragraphs, while some writing systems also need more vertical space.

The result may look like a font failure even when every glyph is correct: words wrap onto extra lines, headings collide with images, table rows grow, or software automatically shrinks the type until it becomes unreadable.

Fix the container before shrinking the font:

1. Widen or deepen the text frame.
2. Allow headings and table rows to wrap.
3. Remove manual line breaks inherited from the source language.
4. Adjust paragraph spacing and line height.
5. Reduce font size only when the layout cannot be changed safely.

For slide-specific checks, see the PPTX translation QA checklist. InDesign users can also follow the font and overflow checks in our IDML translation guide.

A Practical Repair Workflow

1. Identify the symptom

Do not replace every font immediately. Determine whether the problem is missing glyphs, substitution, shaping, or overflow. A broad font change can hide the original cause and create new layout problems.

2. Return to the editable source

Use the DOCX, PPTX, spreadsheet, or design file whenever it is available. Source files preserve paragraphs, tables, and text frames more reliably than a flattened PDF.

3. Create a font map

Record the approved font for each target script and weight. Check regular, bold, italic, and symbol characters separately; a family may support a script in its regular face but not in every style.

4. Translate without flattening the file

Copying document text into a plain text box discards much of the structure that controls styles and layout. When the format is supported, upload the original file to a document translator instead; OpenL Doc Translator is one option. Regardless of the tool, the output still needs a font and layout review.

5. Export with fonts embedded

When the license permits it, embed the required fonts in the final PDF. In Acrobat, open Document Properties and check the Fonts tab to see which fonts are embedded or subsetted.

6. Test the delivered file

Open the result on a device that does not have the source fonts installed. Check every page at normal zoom, search for target-language text, and copy a sample sentence out of the PDF. Visual appearance alone does not prove that the underlying text is usable.

What to Check Before Delivery

• No empty squares, bullets, or replacement characters appear.
• Arabic, Indic, and other shaped scripts join and position correctly.
• Bold, italic, superscript, and symbol characters display correctly.
• Headings, tables, captions, and footnotes do not overflow.
• The PDF uses embedded fonts where licensing allows.
• Search, selection, and copy-and-paste still work.
• The document looks correct on a second device.

The key is to treat font coverage, font availability, script rendering, and layout as separate checks. Once you identify which layer failed, the repair is usually straightforward.

Sources

• Adobe: Embedding fonts in PDFs overview — Explains PDF font embedding, subsetting, licensing restrictions, and substitution.
• Unicode Consortium: Fonts and Keyboards FAQ — Explains the relationship between Unicode characters, fonts, glyph mappings, and rendering systems.
• Google Fonts: Noto — Provides a multilingual font collection covering more than 1,000 languages and over 150 writing systems.
• W3C: Text size in translation — Describes text expansion, character width, line height, and flexible layout considerations across languages.