Translators who localize software or games know the horrors of sticking to restrictions in segment length or having to cut and squeeze their sentences in order to fit the character limit. It is tough on many levels, and perhaps this is why some well-known experts in the field still tremble when thinking about segment length. When the feature first appeared in the market it soon became a game changer, and the ability to automatically count the number of characters was a revelation to some.
This was the time when mobile games started to grow in popularity, and the first thing you needed to prepare for when localizing mobile games are small screens. In games localization, a small screen equals a low and strict character number, and even when you’ve managed to give all your interface text a perfect length you still have have more work to be done because all appstores have strict length restrictions for app names and app descriptions.
This also applies to the art of subtitling, voice over and dubbing translations. Ever played a game that was originally made in another language? Than you have probably seen clumsy unsynchronised lip-syncing. The same goes for foreign movies and TV series voiced by enthusiasts.
If the translator cannot squeeze the text within the specified time frame, the character will continue talking while the scenery changes or the other way around — the lips of the character will continue moving in complete awkward silence.
Even artistic translation may require line length limiting. Yes, I’m talking about equirythmic translation. I’ve translated some songs over the years and it’s been a struggle — no exceptions. And I had to limit the length of each line because the text would be printed on an advertising material.
Of course, one can argue that equirythmic translation limits the number of syllables, not letters, as syllable length tends to differ from language to language. But imagine the number of saved trees if we always paid attention to the translation line length.
Why set the maximum length of a segment?
Software and Mobile UI projects always require that the translation can fit the placeholder completely and correctly without any truncation. Sometimes, abbreviations are needed to make sure that the translations fit in the space. It is very difficult for a translator or linguist to guess if the translations fit in the space with just visual help. Even with certain guiding principles, some languages by nature are longer than others. So, if there is an automated segment length control system in place, the translator knows from the very beginning how long the text needs to be so that it can fit the placeholder.
Our customers often demand that the number of characters in a segment is not exceeded. This has different reasons. Significant causes are, for example, the limited number of characters in an operator display or the space limitation in a document, where the length is limited in the target language (catalogs, brochures, etc.).
As lots of translated phrases should fit in certain buttons and windows in the interface (we are translating the interface of a trading platform) and in ad banners which have limited dimensions.
In dealing with game or app localization there are a lot of factors other than making sure the translation makes sense that we have to consider — and that’s the length of the translated segment.
What might go wrong if the length of translation exceeds the allowed limit?
The most typical problem is text truncation. This may render the UI looking shoddy, inappropriate and meaningless. If the long translations are not truncated then there could be the problem of text overflow to the next line or beyond the UI boundaries. Operationally, also, it is time-consuming, requires rework and several iterations to finally arrive at the correct length.
Non-observance of the length restriction can have considerable consequences: operator texts are incomplete (risk of incorrect operation!), descriptions in the target language are invisible because of window frames, etc. Simple example English — German:
It doesn’t work (16 chars)
Das funktioniert nicht (22 chars).
The negation in German follows only in the part that would be cut off after 16 characters. The translation would be until then: It works — what exactly the opposite means.
The translation doesn’t fit the space allowed for it and then it overlaps with other texts around or just cannot be read in full.
If the segment is longer than what we’re allowed then it will spill out of the area that it’s being input into, and look terrible. In general, in a Japanese to English translation — English is around 1.5 times longer than Japanese, so we try to keep it within that length. But in most cases we have a very specific character limit that we have to adhere to — and localize accordingly.
Can it be done with Smartcat?
On July 3, 2018, Smartcat introduced this feature to clients with some companies finding it quite useful in their daily operations. The feature was called “target length limit” and literally allowed any company specializing in game or software localization to limit the length of translation units to fit them into the UI or mobile app description field.
It was a joyful day, indeed. From that time on anyone can literally set and change the maximum allowed length (i.e. the number of characters) of the target segments in the editor. You also no longer have to worry about whether the localized content fit into the UI or mobile app description field.
You can restrict the length of multiple segments by clicking on them while holding the Shift key. If the project is multilingual, the language length restriction will be applied to all languages.
To set the length of a segment in Smartcat just:
click the “segment length limit” button,
enter the preferred length for selected segment/segments,
Also, Smartcat will import the maximum target length settings from the XLIFF files you upload and apply them to the corresponding segments.