File Formats for Technical Translation

Transline translates documents from all standard Windows, Mac and Linux file formats:

  • Access
  • AmiPro
  • CorelDRAW
  • EPS
  • EXE
  • Excel
  • FrameMaker
  • HTML
  • Illustrator
  • InDesign
  • Interleaf
  • plain text
  • Open Office
  • PageMaker

You can choose whether you would like to get your translation back in the source format or in another structurally related file format. Transline will handle the conversion process for you.

Translatability of File Formats

The various file formats differ in their translatability. Some formats can be translated by the translator directly and without technical preparation. Other formats have to be prepared and, if necessary, also post-processed for the translation process to be able to ensure high quality of translation.

Using translation memory systems and terminology databases, which can result in increased quality and reduced costs in the case of large volumes or translations that are similar, is only possible for most file formats when documents are pre-processed or prepared. The translatability of a file format is thus an important part of pricing at Transline.

The different requirements for processing dissimilar file formats can be illustrated by the following examples.

FrameMaker files

FrameMaker, the application used by many companies as a layout program, uses an atypical coding for text characters. This made life difficult for many translators in the past because automatic translation processes were virtually impossible as a result.

Meanwhile the problem shifted. With the latest FrameMaker patches it is now possible to process the standard Latin characters on a conventional word processor. Still difficult, however, is the situation in the case of languages with uncommon special characters, such as the Baltic languages. Improvisation is often called for here, which can prolong the translation process and make it more expensive as a result. The languages whose text is not based on Latin characters, like Greek, Chinese or Japanese, can also present a problem.

Irrespective of the processable language, FrameMaker files are not easy to translate because the standard copy-and-paste concept does not work. The translated texts cone to where the should only by means of an unconventional detour.

Transline has developed an interface of its own to integrate FrameMaker files into the translation databases where they can be translated using the Across translation memory system. Then the files are reconverted into the original format again. This communication which circumvents the formatting problem described above is made possible by the cross-platform Maker Interchange Format (MIF). All file information is translated into MIF commands and in this way can be read by Across. FrameMaker interprets the commands in the MIF file and reconverts them into its own format and layout settings.

As a result, Transline can quote you translations of FrameMaker files comparatively favourably.

PDF files

PDF (Portable Document Format) is an exchange format for finished documents, which cannot be edited or only to a limited extent and is incommensurable with standard word processing programs. As a consequence, PDF does not lend itself well to translation. Your PDF documents can, however, be read at Transline and transferred to an editable and translatable format.

Optical Character Recognition (OCR) converts an image file or a scanned document into editable text. Transline's OCR system recognises texts in 122 languages. It receives data from a scanner connected to the PC or processes image files in PDF, TIFF, JPEG, PNG, BMP, PCX or DCX format. The OCR system either saves the result of the recognition process as a PDF, text, RTF, HTML, CSV or DBF file, as well as in Microsoft Word or Excel format in versions 95 to 2007.

Files that are read in with the OCR system have to be checked by an employee in any case.

Thanks to a considerable number of functions, the Transline-adapted OCR system can also read column matter, multilingual texts or tables. The OCR system shows itself capable of being taught as regards proper names, ligatures, special characters or ornate type; the teaching effort is all the more worthwhile, the more comprehensive the text is, to which what has just been learned is to be applied.

Difficulties arise in the case of handwriting, very small type, rotated sections of text, white typeface on coloured background and originals scanned at a low resolution.

If in addition to a PDF document original data are still available, e.g. in QuarkXPress or InDesign, such formats are preferable because the process of working with the OCR system is rendered superfluous as a result.