Aug 2, 2013

Translating SDL Trados Studio SDLXLIFF files & more in memoQ!

My latest demonstration video actually covers a number of memoQ features so that I would have an excuse to create this video index:
Time  Description
  Importing the first SDLXLIFF file to memoQ
1:12  Exporting the finished translation
1:27  Viewing the translation in SDL Trados Studio 2009
1:40  Re-importing the edited translation for a TM update
3:24  Saving the translation in a LiveDocs corpus for later reference
3:55  Importing a new version of the text in an SDLXLIFF source file
4:25  Comparing source text versions
5:55  Document-based pretranslation ("X-Translate")
7:11  Examining a "warning" for forgotten tags
7:46  Results of the second translation in SDL Trados Studio

That is the sort of thing I was talking about in a recent blog post about new approaches for online instruction. Many times I have wished for just such an index for long webinars or even much shorter reference videos like this one.

This tutorial was inspired by a Skype chat with a colleague in the US a few days ago. She uses memoQ but works with a number of others who use various versions of SDL Trados Studio, and there were some questions about about how one might deal with TM updates after a translation as well as the inevitable new versions that legal and financial translators often encounter. 

I have also noticed that quite a number of people are not up to date on SDLXLIFF compatibility with memoQ; this video also shows that former issues with preserving segment status have been taken care of, and everything now works well.

What is not obvious in the video is that one can also change the segmentation of the SDLXLIFF in memoQ; this happens only in the memoQ environment to allow better translation and more sensible translation memory content, and when the SDLXLIFF file is exported from memoQ, the original segmentation from Trados is preserved in the Trados environment.

Also not shown in the video is how I imported a third version of the source text, this time as a Microsoft Word file, not an SDLXLIFF. The document-based pre-translation (X-Translate) worked perfectly, and the target file was exported in the proper format (DOCX).

There are, of course, many other ways one could handle a "project" like this, but the procedure shown is not unlike what I sometimes do in projects myself.


I apologize for the quirky click animation in this tutorial; Camstudio had some problems I have never encountered before, and I'll have to get to the bottom of that if I keep using that tool. Otherwise, the video quality is probably the best I have achieved so far, and I would like to thank the friend who revealed the "secret" of better quality video for YouTube.


  1. Nicely done!

    Are you saying that it is no longer necessary to prepare the sldxliff segmentation? I have had problems when trying to subsequently join segments in the memoQ grid in order to stretch two sentences into one or whatever.

  2. Chris, it is always advisable to clean up segmentation in the original environment as much as possible. But sometimes people who prepare and send the files are lazy or they simply don't know what their doing or, even if they are super-competent, they overlook something. Or I want to combine some sentences so that the words simply flow better. There are, of course, implications to working like this which I won't discuss here. Let's just say that there are some benefits to taking these steps to improve translation quality, but not everyone may see it that way. I have a better TM then in any case and tools that can make use of it.
    Maybe I'll post a screenshot of what that changed segmentation looks like. It gets really interesting if you rework the word order in a big way, which I did not. Now that I think about this a bit, I could have real fun with such a blog post....

  3. Just did a little more testing with another set of SDLXLIFF files after someone suggested that status from SDL Trados Studio and the match percentages from there would not be respected in memoQ. That is mostly not true. I tested with memoQ 2013 Build 10. I did find two little glitches. The tool tip over the lock icon on locked segments suggests that these can be unlocked, and that is in fact not true. Also, "signed off" status is read correctly by memoQ, but upon export of the SDLXLIFF file it is changed to "approved" instead. A minor oversight which can be fixed easily enough in a future build. Everything worked perfectly otherwise, and the fuzzy match percentage from pretranslation in SDL Trados Studio was preserved in memoQ and was left intact in the final, exported file. What does not get sent back to SDL Trados Studio in the SDLXLIFF file is any information on fuzzy matches in memoQ while translating the file. But that isn't a bad thing - it's nobody's business what resources of my own I may have and how much these might assist me.

    If anyone wants a test file with different status settings, one of my test files can be downloaded here. It's an EN-DE file.


Notice to spammers: your locations are being traced and fed to the recreational target list for my new line of chemical weapon drones :-)