Find a tool
Home/Encoding recovery
Free · 100% browser based

Fix garbled subtitle encoding

Can’t read your SRT subtitles because characters look corrupted? Open the file, compare text encodings, and save a clean UTF-8 copy. No uploads or account required.

Try the tool

No signup, no upload to a server, no file stored.

Private on-device processing
Choose an SRT, VTT or text subtitle fileDrag a file here, or choose one below · up to 15 MB

No file selected

Upload a file to see the detection hint.

Paste or modify text
Your output will appear here.

Add a file or paste subtitles to get started.

How to use it

  1. Upload an SRT, VTT or text subtitle file.
  2. If the letters look wrong, select another original encoding. Common choices include GBK for Simplified Chinese, Big5 for Traditional Chinese, and Shift-JIS for Japanese.
  3. Check the decoded preview. If needed, edit the text directly.
  4. Download the UTF-8 file and test it in your video player.

Why do subtitle characters become unreadable?

Text files are stored as bytes that must be interpreted using a character encoding. An old subtitle file may use Windows-1252, GBK, Big5 or Shift-JIS, while your player assumes UTF-8. Using the wrong decoder creates strange symbols or replacement characters.

This tool lets you reopen the original bytes using different decoders rather than repeatedly saving already-garbled text. The result downloads in UTF-8, the format most modern apps expect.

Important: Encoding detection is not always certain. If the original bytes have already been overwritten with question marks or replacement characters, the lost text may not be recoverable. Always verify the preview.

Frequently asked questions

Does this upload my subtitles to a server?

No. The file is read and decoded with your browser’s File API and TextDecoder. The page does not send the subtitle contents anywhere.

Can this repair Chinese or Japanese subtitle mojibake?

Often, yes, if the original file still contains the original bytes. Try GB18030, Big5, Shift-JIS or EUC-JP and inspect the decoded preview.

What if the original text contains ??? or �?

Some information may have been irreversibly discarded. Switching encodings cannot restore bytes that no longer exist.