Getting the most out of PDFs

Why does translated text never fit the original layout?

Translation changes the length and the height of the text but not the boxes it has to sit in, so a faithful translation almost always overflows or under-fills the original layout — which is why translated text is best taken out, read, and placed deliberately rather than substituted in place.

3 min read

Everybody's first request is the same: translate this PDF and keep it looking exactly as it does now. It is a reasonable thing to want and it is close to impossible, for reasons that have nothing to do with the quality of the translation.

Text expansion is real and large

Translating the same sentence out of English typically changes its length by a lot:

  • German — often 30% longer, sometimes more. Compound nouns are single long words that cannot be broken.
  • French and Spanish — 15–25% longer.
  • Bangla and Hindi — similar length in characters, but the script sets taller, with marks above and below the line, so the same line height crowds.
  • Chinese and Japanese — usually much shorter in characters, and often wider per character.
  • Arabic and Hebrew — run right to left, so alignment, indentation and anything anchored to an edge all mirror.

A button that fits "Save" does not fit "Speichern". A table column sized for "Date" does not fit "Fecha de vencimiento".

Why a PDF makes it worse than a web page

A web page reflows. A box grows, the text below moves down, and the layout absorbs the change.

A PDF does not reflow. Every glyph is placed at a coordinate. There is no box to grow — replacing a line with a longer one means it runs over whatever is to its right, or off the page. Keeping the original layout therefore means one of:

  1. Shrinking the translated text until it fits, which produces a document with a dozen different type sizes.
  2. Truncating it, which loses meaning.
  3. Re-laying-out the page, which is no longer the same design.

Every "translate PDF, keep formatting" service is choosing one of those three, usually the first, and usually without saying so.

What is more useful

Take the text out, translate it, and place it deliberately.

Translate PDF returns the translation paragraph by paragraph with the original beside it, so you can read and check it. From there:

  • If you only needed to understand the document, you are done — that was the actual goal most of the time.
  • If you need a translated document, put the text where you want it with the PDF editor, which lets you resize and reposition it as the language requires.
  • If the original is available as a Word or InDesign file, translate that and re-export. A layout that reflows is worth going back for.

The check nobody does

Whatever route you take, look at the longest item on the page — the widest table cell, the longest heading, the tightest button. Those overflow first, and they are what tells you whether the layout survived.

Common questions

Can any tool translate a PDF and keep it looking identical?

Not faithfully. Something has to give, because the translated text is a different size and the page cannot reflow — tools that claim otherwise are usually shrinking the text without telling you.

Which languages expand the most?

German is the usual worst case at around a third longer, with French and Spanish behind it. Bangla and Hindi are similar in length but taller, which crowds fixed line heights.