Character Encoding Converter & Mojibake Fixer

Paste garbled text and automatically detect the encoding mismatch that caused it, with ranked repair candidates. Also supports explicit conversion between UTF-8, Shift_JIS, EUC-JP, and JIS (ISO-2022-JP).

Tips

  • When multiple candidates appear, the one marked "Best match" — usually the most natural-looking Japanese or ASCII — is normally the correct one.
  • The classic "縺薙繧薙?..." pattern seen in emails and CSV files almost always resolves with the UTF-8 → Shift_JIS pair.
  • If the text has turned into a string of "?" or "�" characters, part of the original byte sequence has likely been lost, making a full text-level recovery difficult.
  • Use the "Manual conversion" tab for a one-step result when you already know which encoding pair caused the garbling.

Frequently Asked Questions

It happens when the encoding used to save a piece of text differs from the encoding the reader assumes when displaying it. Japanese text is especially prone to this because UTF-8, Shift_JIS, and EUC-JP are all still in use, and mismatches between them are the main cause of mojibake.

No. Depending on the type of mismatch, part of the original byte sequence can be permanently lost during the initial (incorrect) decoding, making full recovery from the string alone impossible. This is especially true when one Japanese multi-byte encoding is misread as another; text that has turned into "?" or "�" characters generally cannot be restored.

UTF-8 is the dominant encoding today, but some older Japanese systems and legacy Windows applications still assume Shift_JIS. Because UTF-8 byte sequences for Japanese characters often also happen to be valid Shift_JIS byte sequences, no information is lost in the process — which is exactly why this particular mismatch can usually be reversed.

No. All conversion runs locally in your browser's JavaScript, so it is safe to use even with text containing passwords or personal information.
ツールくん

Side Note — Why "縺薙繧薙?縺ｫ縺｡縺ｯ" became the face of Japanese mojibake

Search for Japanese mojibake and you will almost certainly run into "縺薙繧薙?縺ｫ縺｡縺ｯ" (which should read "こんにちは," or "hello") — it has become something of a cultural icon among Japanese internet users. It's the textbook result of opening UTF-8 text in an application that assumes Shift_JIS, such as an old version of Windows Notepad or various legacy systems.

This particular pairing shows up constantly because a Shift_JIS decoder forcibly reinterprets the 3-byte sequences UTF-8 uses for Japanese characters as if they were 2-byte Shift_JIS characters. By coincidence, most of those reinterpreted sequences happen to fall within the range of valid Shift_JIS characters, so no error occurs — the text just looks garbled while secretly remaining fully recoverable. That property is exactly why an automatic fixer like this one succeeds so often.

By contrast, opening Shift_JIS text as if it were EUC-JP tends to produce invalid byte sequences that get discarded outright, destroying information the moment it happens — recovery at the string level then becomes impossible in principle. This split between "recoverable" and "unrecoverable" garbling is a big part of why Japanese character-encoding issues have quietly frustrated engineers for decades.

More tools in this category