Game Localization · Glossary
Double-Byte Characters — CJK Text Rendering in Game Localization
Double-byte characters are characters that require two bytes of storage rather than the single byte used by standard ASCII characters. Chinese, Japanese, and Korean (CJK) scripts use double-byte encoding, which affects font selection, text rendering performance, file sizes, and UI layout calculations in game localization.
Why Double-Byte Characters Matter for Games
Standard ASCII (covering English and most European languages) uses one byte per character — supporting 256 possible values. CJK scripts contain thousands of unique characters and require two bytes per character in double-byte character sets (DBCS), or three bytes per character in UTF-8. The practical consequences: string table file sizes roughly double when adding CJK languages; font atlas textures must be dramatically larger to cover the full CJK character set; text rendering pipelines that assume fixed character widths break with CJK’s variable-width glyphs.
Engine Support for Double-Byte Characters
Unity handles CJK well using TextMeshPro with proper font atlas configuration — static atlases must include every character needed, while dynamic atlases load on demand but may cause performance hitches. Unreal Engine has built-in CJK support since UE4 with font fallback chains. Older engines, custom engines, and web-based games often require custom text rendering pipelines for CJK. The critical variables: font atlas size limits, line-breaking algorithm (CJK does not use spaces to separate words — standard word-wrap breaks at space characters and will break CJK text incorrectly), and glyph rendering performance at scale.
Common Double-Byte Issues in Game Localization
Glyph atlas overflow: too many unique characters for the font texture atlas, causing some characters to appear as white squares or missing. Incorrect line-breaking: Japanese and Chinese do not separate words with spaces — wrap algorithms must break at CJK character boundaries, not space boundaries. Font fallback failure: a character exists in Unicode but not in the game’s loaded font, displaying as a box. Performance issues: rendering thousands of unique glyphs has GPU overhead not present with 26-character Latin alphabets.
SandVox and Double-Byte Characters
SandVox CJK localization includes a technical audit of your engine’s double-byte text pipeline before translation begins. We identify font atlas size limits, line-breaking behavior, and rendering performance issues at the engine level during scoping — problems caught here cost an engineering afternoon; problems discovered at LocQA require translation revision and re-testing.
Related terms: Game Internationalization · Pseudo Localization · Character Encoding · Localization Qa
Frequently Asked Questions
Does my game engine support double-byte characters?
Most modern engines (Unity, Unreal, Godot) support CJK with proper configuration. The key variables are font atlas size, font selection, and line-break algorithm. SandVox audits your engine setup during project scoping for any CJK localization project.
Will adding CJK languages significantly increase my game’s file size?
Somewhat — primarily due to font file sizes. A CJK font covering the full Simplified Chinese character set can be 5–20MB, compared to under 1MB for a typical European font. Dynamic font loading (loading characters on demand) mitigates this for games where package size matters, at the cost of potential hitches when new characters first appear.
Are Simplified Chinese and Traditional Chinese the same for double-byte purposes?
No. They use different character sets and require separate font files — many characters have different glyphs between Simplified and Traditional. Always treat zh-CN (mainland China) and zh-TW (Taiwan/Hong Kong) as separate localization targets with separate font assets and separate string files.
Need Expert Game Localization?
SandVox provides end-to-end game localization including double-byte characters — for narrative games, mobile titles, webtoons, and interactive fiction.