Automatic PDF translator is a useful tool for many situations: when you don’t understand a manual in foreign language, when you want to read a foreign language article or small book, when you want to translate a guide for foreign customers, when you don’t have the time nor skills to translate PDF document manually etc. Check our article about suitable situations for machine translation to learn more about the issue.
Although PDF translation is a valid and sometimes even the best choice for certain situations, there are still some challenges related to PDF format that make automatic PDF translation more or less imperfect:
First of all, PDF file is a visual format. Every PDF file have its own layout and it’s a great format for documents which have images and charts. We all know how sophisticated and simple it is to read a document in PDF format. The challenge is that most PDF documents which contain images and tables include some references from text to these visual elements. If the automatic PDF translator represent the translated text without images, it is very difficult to see “the big picture”. The correct layout is crucial especially with all sorts of manuals and user guides. Thus, the first challenge is to preserve the layout.
Another challenge is closely related to the first one. Images and charts in PDF files are stored differently than the actual text. While it is no problem for a PDF translator to recognise the text, it is basically impossible to identify any text within the visual elements. This means that those pieces of text won’t be translated. Likewise, scanned files are like one huge, untranslatable image to PDF translators. The second challenge is to identify as much text as possible in a PDF document.
Thirdly PDF file handles some fonts differently than others. This causes harm to PDF translators because all fonts don’t follow any universal encoding. If the font is encoded in the PDF file in a proprietary way, PDF translators cannot “understand” the text. This happens even with Acrobat Reader sometimes; the text displays correctly but when copying it to clipboard, the text becomes nonsense. Thus the third challenge is to cope with all the fonts, and there really are many of them!
Fourthly any PDF file can be secured or locked by the person who has created the PDF document. Naturally it is reasonable to give a change to secure one’s own intellectual output from any possible misuse. PDF translators are not meant to be used to violate any legal rights, and thus those locked files cannot be translated. Even if the document is not locked, you should always pay attention to this issue of copyright. In most cases it is OK to translate for your own use if you don’t try to benefit commercially or take advantage of the translation without proper permissions. Anyway all the moral issues are on the people’s responsibility and from the PDF translators’ point of view, the fourth challenge is to respect all the security features in PDF documents.
These are the most common challenges related to PDF files. As a summary, it can be said that PDF is not the easiest format for automatic translators. Luckily there are also more advanced PDF translators available and together with constantly developed machine translation, which can already produce good quality, they are very handy tools for most PDF documents.