TextMesh Pro: localization and fallback fonts is mixing fonts

Unity Engine
1

Hi all,

I am trying to localize my game and I realized for the Russian language I need a fallback font, so all the “special” characters will be displayed. And it works fine on the technical side but it is looking bad sometimes, because only the cyrillic characters are replaced of course.

As an example I got an address like:
Bleeker Street 2a
Localized to
xxxxxxxx, 2a (can’t post the characters in the forum for some reason, see image below)

The original font is of a handwriting type and I looked for something similar for the Russian fallback font but of course, they don’t match exactly. The result is, that the street name is displayed in the Russian font and the “2a” house number is displayed in the original font which is looking a bit odd.

A solution would be only to rely on fonts that have all the characters I need but that is a limitation I don’t want to be honest. Another idea would be to have font objects for every language and toggle them depending on the user language but then I have every string on GameObjects and not in one single resource file for example.

So what I would expect and maybe there is a setting or maybe this is a bad idea and I just don’t see it yet:
if a TextMesh object has to fall back to a certain font then the whole object should be rendered in that font.

How is that problem solved in general? I think it is very common and should have a solution so I would be very glad if you could point me onto that :slight_smile:

5983058--642869--textmeshproblem.jpg
5983058--642869--textmeshproblem.jpg1052×590 93 KB

2

Finding fonts that blend / work well together can be a challenge. Especially with more heavily stylized fonts.

Finding fonts that include all characters / languages is difficult. Finding font families like Google’s Noto font family for which there are font files for all languages is easier.

In terms of the feature you are requesting, this is not the first time it has been requested. It certainly makes sense but does present some challenges in implementation.

In the meantime, you should be able to achieve the result you seek by using the fontAsset.HasCharacter() functions to check (prior to setting the text) if the font asset contains all the requested characters and if not, to switch to another font asset that contains all of them. The challenge with this and with the other feature request is back to having to use a font that contains all characters where in such case why not use that font to begin with.

3

Thank you for your helpful answer!
With your suggestion I should at least be able to have a consistent font in one language, I don’t need it to be too consistent between different languages.