I have a 25-page German PDF for work that I need translated into English, but the formatting keeps completely breaking. What do you all use to translate whole documents without messing up the layout?
Honestly, translating a German PDF straight to English is annoying when the layout falls apart. If you have Microsoft Word, I would open the PDF there, let it convert to a document, then use Translate under Review. That keeps tables and headings better than a quick copy and paste. To be fair, I would still check the first few pages before trusting all 25.
Is it a real text PDF or scanned? For digital ones I convert to docx first, then run it through DeepL’s document translation, it keeps tables and headings mostly intact. Scanned pages are messier, you need OCR before anything. For a 25-page work doc I’d skip the free tools that cap file size.
I use DeepL’s document translation for German PDFs, and it usually keeps the layout better than copying the text into a translator. If the PDF is scanned, run OCR first, then check every table manually. Avoid converting to Word before translating, as that tends to wreck the columns.
I usually open a copy in Word, tidy the styles and section breaks, then use Review > Translate > Translate Document. Honestly, tables and text boxes still need checking page by page against the original. If the PDF is scanned, run Word’s text recognition first. Avoid expecting any automatic translation to preserve a complex layout perfectly.
@guitar_mouse is spot on about checking page by page, because translating always seems to ruin the margins. I deal with a lot of vendor schedules, and opening them in Word first is definately the safest bet for keeping the layout intact. What device are you using to edit these?
@Rando72836 Windows laptop here, Word copes fine with that. The shifting I see is usually fonts though. If the German file uses one that isn’t embedded, the translation swaps it and lines rewrap. Have a look at the fonts before you start, and keep a backup of the original PDF.
If it’s scanned, the layout breaks because the translator is reading garbage text from a bad OCR pass first. Run it through proper OCR (text recognition) to get clean text, then translate, layout holds up way better.
if it’s a scanned PDF, run OCR first or the translation will usually scramble the page. I use PDFcap to OCR it and export to Word, then translate the text and save back to PDF after checking tables and headers. is yours a scan or a selectable-text PDF?