Wednesday, 28 October 2015

"TM Update" batch task is back again: time to upgrade to Studio 2015

Here it is. The problem I wrote about in my last 2 posts has now been solved: Studio is able to update translation units properly as in the past, i.e. by overwriting used TUs instead of overwriting or duplicating them.

The new standard option is nothing other than the old pre Studio SP2 CU8 default option: if a segment has been translated with a certain translation unit (TU) from a given TM, the ID of the TU is saved in the sdlxliff file. If a correction is made offline to this segment, for example through the Export for bilingual review, the batch task knows which TU has to be corrected in the TM. Good to see it back.

So this is certainly the time to upgrade now. 2015 has one very important neat feature which may transform your correction process: the retrofit function. Translations which have been reviewed in their original layout have always been difficult for translators and translation providers. Most of the time, each correction had to be entered in the CAT tool manually which was very time-consuming. The new function should change this situation completely as it is now possible to enter the corrected document as input for this function and get the correction entered automatically with track changes at the right place in the sdlxliff.

This may sound like magic for those like me who never heard of such possibility. In the background, this is pretty simple and very clever: The corrected document is "prepared", i.e. segmented exactly as the original document was. The segments are then aligned with the segments of the original translation. The differences are imported in the sdlxliff and shown as track changes in the editor to allow the translator to check their validity. For more on the retrofit function, see Emma Goldsmith's blog here

2015 is also interesting from the point of view of translation evaluation: A translation quality assessment (TQA) module has been developed which allows translation providers to organize the evaluation of their translators. The new module is fully customizable and you can create your own metrics.

Other new features: AutoCorrect, customizable ribbon, new interface languages and a new word filter which helped us to complete projects as the old one in Studio 2014 was not able to clean up the files.

More information about 2015 can also be found on the SDL blog as well as information about the service release

Wednesday, 20 May 2015

SDL Studio CU10: “Update Translation Memory” batch task may duplicate translation units

In my recent post about SDL Studio, CU8 and a conflict regarding the "Update Translation Memory" batch task and "Add as New Translation" function (Ctrl+Shift+U), I wrote about the new default setting in the "Update Translation Memory" batch task: "Overwrite existing translation units". If you haven't read it yet, please start there to fully understand the situation.

New default setting

Unfortunately, the old behavior of Studio of only overwriting the existing translation unit (TU), as Ctrl+Enter would do in the Editor, has not been reintroduced as a batch task option yet: in CU10, the default setting has been changed to "Add new translation units".
The good thing is that you cannot erase any TUs from your translation memory (TM) anymore (see former post); the bad thing is that the batch task will systematically create a duplicate TU as soon as the target text has been changed without a Ctrl+Enter validation in the translation editor, for example while doing a review in Microsoft Word or an offline review in SDL Studio.

"The overwrite option is fine for me"

I have heard from many users that the "overwrite" setting was a good thing for them and their translation projects: if an unwanted duplicate TU has been added during the translation process, the batch task deletes it as well as all other TU with the same source and saves only one occurrence-- the (last) one found in the sdlxliff.
To use this "overwrite" option with CU8 and CU9, it was not necessary to change anything in the project templates.
Now, as the default setting has been changed in CU10, it is necessary to select the "overwrite" option explicitly in all templates and running projects. Running the batch task with CU10 without having made such changes will use the "Add new translation units" and new duplicate TUs will be created.
In such a scenario, it is fortunately possible to delete immediately the created duplicates just by running the batch task again with the "overwrite" setting. All added TUs will be deleted and replaced by the (last) TU found in the sdlxliff file.

Tuesday, 10 March 2015

SDL Studio CU8: Conflict between “Update Main Translation Memory” batch task and “Add as New Translation” function

With the cumulative update 8 for SDL Studio 2014, the batch task “Update Main Translation Memory” offers new functionalities and, above all, a new default setting which may change the way translators work.

New options for the “Update Main Translation Memory” batch task and new default setting

In the batch task prior to CU8, it was possible to “Always add new translation units” so that existing translation units with the same source segment remained in the memory and were not overwritten when the target segment differed. In this case, Studio created a duplicate translation unit.

Now, when you want to update a segment in the TM where the source is identical but the target is different in the sdlxliff file, it is possible to:
  • “Add new translation units”: a duplicate translation unit will be created. This is equivalent to the old option “Always add new translation units” in the former version.
  • Overwrite existing translation units: all existing translation units with the same source segment will be systematically deleted and replaced by the one in the sdlxliff. This option is the new default setting for the batch task.
  • Leave translation units unchanged: the batch task will only add translation units if the source segment is a new translation in the TM.
  • Keep most recent translation units: the last modification date of all existing translation units with the same source segment will be compared with each other, and the newest unit will remain in the TM. All others will be deleted.
The old behavior –only one translation unit is overwritten when duplicates are present – is no longer available. Either you overwrite all TUs which have the same source and keep only one, or don’t update any existing TUs at all.

Conflict with the “Add as New Translation” function 

The "Add as new translation" function is important when a segment should be translated differently depending on the context. Translators can store 2 different versions with the shortcut Ctrl + Shift + U. When you encounter such a segment, it will be regarded as a 99% fuzzy match because of the 1% default penalty for "Multiple Translations". The pre-translation process won't insert any translation in the sdlxliff file, and the translator will be able to consciously choose between both possibilities in the editor. The translator will get paid for this work just like for a "normal" 99% fuzzy match.

In a scenario where translators use this “Add as New Translation” function, the batch task “Update Main Translation Memory” with the default setting will delete all duplicates that were stored in the TM for the given segment. As the batch task is available to both translators and project managers, being unaware of this new setting can have big consequences for the translation resources: segments as described above will be recognized as 100% matches, which can mean not being paid for the segment. In addition, the segment may also have been potentially incorrectly pre-translated.

Here an example

File to update in the TM

TM before “Update Main Translation Memory”. It contains 2 duplicate pairs: TUs #1 and #3 have been intentionally stored by pressing Ctrl+Shift+U. TU #7 should have overwritten TU #5 but it was unexpectedly added in the TM.

With the old batch task, the TM would have remained as it is. With the new batch task and its default setting, the TM after “Update Main Translation Memory” looks like this:

Both TU #5 and TU #1 have been deleted. The deletion of TU #1 is counterproductive.

In the case you prefer not to take any risks with the deletion of important data, take time to set the default value on “Add new translation units” in your project templates.

The default setting will be re-discussed at SDL and could be changed through the cumulative update 10.

Monday, 25 August 2014

Use SDL Trados Studio for your GRIPS translations

In a recent project with one of our key customers, we managed to migrate our GRIPS data translation process from Transit to SDL Trados Studio. Our new workflow can now make use of all the advantages offered by SDL products without compromise.

GRIPS is a well-known content management system from the company Star, based in Switzerland. Star is also the developer of the translation program Transit and offers its customers a well-established workflow for the translation of GRIPS contents.

Until now, we used this standard workflow to translate GRIPS data, in the past with Transit XV and now with Transit NXT. The decision to look for a solution with SDL products was motivated by our extensive know-how in handling SDL projects and by our preference for using translation memories, which Transit does not support.

The political decision made by SDL to open their development to third-party developers triggered our migration and, as a result, we were able to develop a perfect alternative to using Transit:
  • Programming a file type filter to handle the sgm files which includes previews (for 2011 and 2014)
  • Migration of Transit file-based reference material (Transit “memories”, if you will) to our TM-Server so that the existing translations can be used without loss of the 100% matches
  • Migration of the termbases from TermStar to MultiTerm for term recognition during the translation in Studio
  • Adaptation of the QA rules from Transit NXT in Studio.

The entire workflow has been tested and successfully implemented.

As the file type filter has been developed for a specific customer, we can’t ensure that it will work for other projects without modification. For that reason we won’t publish it until we have assured ourselves that it works reliably. Don’t hesitate to contact us if you are interested in migrating your own projects to another CAT tool or if you want to test our filter for your project.

Thursday, 15 May 2014

Tips to work with the alignment editor - SDL Studio SP1

This may help if you have some difficulties to set your alignment project correctly.

1.    If the alignment is not perfect, set some connections manually and use the “Realign” button. Make some tests to get an idea how the “Realign” function works and use it instead of connecting manually each segment pair. You will be faster.

2.    Studio doesn't check if you choose the right translation memory (TM) with the right languages. Be careful when setting up your alignment project.
If you want to check the selected TM before importing your aligned translation units, click on the "Import into Translation Memory" text (with the down-arrow) and select "Advanced Import"

In the opening window, you can change the TM and also the fields and values that have to be saved with the translation units (TUs).

3.    If the segmentation in the source or the target text is not correct, there are 2 possible reasons for this:
a.    The file type settings are not correctly set. In this case you will see the same problem in both source and target sides and you will have to troubleshoot the corresponding file type filter
b.    The segmentation rules of the TM are not correctly set. Segmentation rules have to be set for each single language.

Note that the Realign Button won’t change the segmentation of the files to align: after a change of the segmentation rules, you will have to start over the alignment of your files to get new *.sdlalign files.

4.    Studio tries to align each segment of the source file with each segment of the target file. If one file has a paragraph more (which has not been translated yet for example), erase this part before aligning. This will help the alignment work and the final result.

5.    If you have imported aligned TUs in a false TM, you can delete them with a filter and the batch delete function in the "Translation Memories" view:

First set a new filter like illustrated below (Value is the name of the origin sdlalign file) and save it.

Then run the batch delete with the new filter selected.

You can start again…

Wednesday, 30 April 2014

Review of Studio SP1

The first service pack of Studio 2014 has now been published and with it the brand new alignment editor. SDL developed a new alignment interface to replace the old WinAlign which couldn't cope neither with new formats from Microsoft, Adobe… nor with the "new" bilingual standard .sdlxliff. The first step was already done without editor but many feedbacks have come to describe specific use cases where the alignment tool was not appropriate, for example when source document contains texts which are missing in the target document because of a difference in the document version. 
So SDL made it:

The new editor gives now a view in the alignment results before segments are imported in a selected TM as well as basic edition features to realign the segmented text. This way, it is possible to see if there are important problems in the alignment and to change erroneous connections.

The most interesting function in the new editor is the "Realign" function which permits to rework the alignment after having manually set or confirmed some connections. Studio won't change any of the confirmed segments and will "re-think" the alignment with these new "100% trustable clues". This function helps to correct several false connections at once and at the end to get a completed proofed alignment quicker than with the old tool.
The editor is in its first version and does not include all the functionalities of WinAlign yet like "join" and "split". This will be added in the next updates/service packs.

Another new function in Studio SP1 - actually a redevelopment - is the embedded content processing for XML file type filters. Since SDL Studio 2009, it was possible to enter regular expressions in the XML file type filter to tag embedded content, for example HTML tags in so called "CDATA section".

In SP1, you have now the possibility to select an HTML4 or HTML5 processor in the new embedded content section of the XML file type filter to automatically tag the HTML elements embedded in your XML files without having to write any regular expressions at all. You just have to check the box "Process embedded content using the following processor" and choose the HTML version.

The settings of the processor can be found at the bottom of the left-hand pane.
The plain text processor allow to write regular expressions as the old concept proposes.

This is a great improvement especially for users without experience with regular expressions and who used to copy-paste untagged HTML elements from source to target (i.e. in plain text), which is quite "dangerous”. For advanced users, this is a relevant gain of time in setting XML file type filters.

SDL chose to offer both solutions: old embedded content in the file type settings and new embedded content processors. It remains easy to adjust old settings without having to think about the new concept and to migrate all project templates. 

In this first version, the new processor misses some features that could be important for you:
-          No possibility to join HTML tagging and plain text regular expressions,
-          Preview of CDATA does not work yet.
The major advantage of the new processor is that it is possible to import and export embedded content settings. No need to search inside the “.sdltpl” file anymore to use them in other templates or project settings.

To me, 2 important enhancements: the first to replace an alignment tool that I couldn't use anymore and the second to make embedded content easier to handle.

The SP1 installer is a free upgrade available for those who have a valid license of Studio 2014 through the personal account on

 Another review here:

Monday, 14 October 2013

Review of Studio 2014 - translate faster and safer

The new version of Trados Studio was published last week on the SDL website. I was able to test Studio 2014 during a 4-month-long beta-test phase in which a lot of remarks from translators and project managers were taken in consideration.
Here is my personal review of the 2 main improvements that impressed me much in this new version.

In my opinion, the main work done on this new version was done to some algorithms that made Studio 2011 very slow in many use cases. Opening, saving and closing files could last a very long time as well as some file "preparation" including pre-translation and analysis.
In Studio 2014, file handling is definitively faster. In some extreme cases, by as much as 70% according to SDL.
Translators can also work quicker: in Studio 2014, it is now possible to open several files on the same tab. With this new function called "virtual merge", the files are displayed on a single tab of the editor, so you don't have to switch between editor and file view or between the file tabs in the editor. The files are displayed as in a merged file. The big advantage is that you will be able to perform a search and replace for all virtually merged files at once, if you detect a typo or a terminological inconsistency, which will let you gain a lot of time for these kinds of project whenever you receive a lot of small files to translate.

...and safer
One of the main expectations of Studio users was to have an autosave function as found in many other programs. Beside the speed improvement of Studio 2014, SDL could implement a proper autosave function with a default delay of 5 minutes between autosaves. The files are stored temporarily and if Studio crashes for one reason or another (which does not happen that much actually!), reopening the current file lets a window open asking if you want to retrieve the last autosaved version.
This is a real improvement. Even if the translation memory function stores all the confirmed segments directly, there have been a lot of scenarios I saw as a support technician, where the user asked me first if he could recover the file or had to start all over again.

SDL Trados Studio 2014 BadgeI think it is worth it to upgrade to Studio 2014 for these 2 main reasons. Just try it to see for yourself. You have a 30-day evaluation period.

For more information about the new features, take a look at the release notes.

And other reviews: