Fix subtitle encoding

What "mojibake" is and why it happens

Mojibake is the garbled text you get when a file is read using the wrong character encoding, so café turns into café, caf? or caf�. The dialogue is not actually corrupted; the bytes are fine. The problem is interpretation. A character encoding is simply the agreed table that maps bytes to letters, and if a file is saved with one table but read with another, every non-English character comes out wrong.

Subtitles are especially prone to this because so many were created years ago, on Windows, in regional encodings rather than today's universal UTF-8. A Spanish subtitle saved in Windows-1252, a Russian one in Windows-1251, or a Greek one in ISO-8859-7 will look perfect on the machine that made it and break the moment a modern, UTF-8-expecting player opens it. This tool reads the raw bytes, detects the encoding they were really written in, and re-decodes them correctly into UTF-8 that works everywhere.

When you'd use it

• Accented letters (é, ñ, ü, ç), curly quotes, em dashes or ellipses show up as strange symbols, boxes or question marks.
• Cyrillic, Greek, Arabic, Hebrew or East-Asian text appears as nonsense rather than the right script.
• A subtitle displays fine in one player but turns to gibberish in another, or on a different device or TV.
• You are tidying a subtitle collection and want every file standardised to UTF-8.

The most common culprits

The encoding a broken file came from is usually tied to its original language.

• Windows-1252 / ISO-8859-1 covers Western European languages (French, Spanish, German, Portuguese). By far the most common source of subtitle mojibake.
• Windows-1251 covers Cyrillic scripts (Russian, Ukrainian, Bulgarian, Serbian).
• ISO-8859-7 / Windows-1253 covers Greek.
• Big5 and GBK cover Traditional and Simplified Chinese.
• Shift_JIS covers Japanese.
• EUC-KR covers Korean.

You do not have to know which one you have. The detector works it out from the byte patterns. For a deeper, plain-English walkthrough, see the guide on fixing garbled subtitle characters.

How the repair works

The file's bytes are analysed statistically to guess the source encoding, then decoded with the browser's built-in, standards-compliant decoder and re-encoded as UTF-8. The before-and-after panels let you confirm the result looks right before you download. If the "after" text reads correctly in its language, the detection was right. If the file is already valid UTF-8, the tool says so and leaves it untouched rather than risking damage from a needless re-decode.

Avoiding mojibake next time

If you edit subtitles yourself, always save as UTF-8 (UTF-8 without a BOM is the safest choice for the widest player support). When you download a subtitle, prefer the UTF-8 version if the site offers a choice. Sticking to UTF-8 end to end means this whole class of problem simply never appears.

Is my file uploaded anywhere?

No. Detection and re-decoding happen entirely in your browser. Your subtitle file is never sent to a server, and there is no sign-up or limit.