Fix garbled text before converting.

Keep an untouched copy of the original file. Repeatedly saving unreadable text can permanently lose characters.

A text file stores bytes. An encoding tells your editor how to turn those bytes into letters. If a file written with one encoding is read using another, you may see nonsense characters or replacement symbols such as �.

How EpuBloom reads a TXT file

1. If the file has a byte-order mark (BOM), the converter recognizes UTF-8, UTF-16LE, or UTF-16BE.
2. Without a BOM, it checks whether the bytes are valid UTF-8.
3. If that check fails, it tries GBK, which is used by some older Chinese text files.

This is a fallback sequence, not a guarantee that every encoding can be detected. Other legacy encodings and UTF-16 files without a BOM may need to be opened and resaved in an editor first.

Make a readable UTF-8 copy

1. Open the original TXT in a text editor that lets you choose the encoding used to read a file.
2. If it looks wrong, reopen the untouched original with the likely source encoding. Older Chinese TXT files may use GBK; use the source application’s information when available.
3. Check several paragraphs, punctuation marks, and chapter headings. Make sure the text is readable before saving.
4. Use “Save As” or the editor’s encoding controls to save a separate UTF-8 copy.
5. Select the new copy in the converter, create the EPUB, and inspect it in a reader.

Reopen and convert are different operations

Reopening with an encoding changes how existing bytes are interpreted. Saving as UTF-8 writes the characters currently shown by the editor into a new file. Saving an already garbled display as UTF-8 can preserve the damage rather than fix it.

If the only available copy already contains lost characters or replacement symbols, changing the encoding cannot reliably restore the original. Look for a backup or obtain a fresh source file.

When the text is readable but the ebook looks wrong

If characters are correct but chapters are missing from the contents, check the chapter-heading formats. If you extracted an EPUB and lost images or table layout, those are plain-text conversion limits, rather than necessarily an encoding problem.

For help, describe the source encoding, browser, and visible error through support. A short non-private example is more useful than sharing a full personal manuscript.