ERROR - Updated 2026-06-25
Shift-JIS to UTF-8 CSV Conversion
Convert a Japanese Shift-JIS or CP932 CSV to UTF-8 safely by previewing the original bytes, checking columns, and choosing Excel or upload output.
Open CSV Encoding FixerQuick Answer
To convert a Japanese Shift-JIS or CP932-style CSV to UTF-8, keep the original file unchanged, preview a small sample with CSV Encoding Fixer, confirm Japanese text and column boundaries, then export the copy your destination expects. Excel often benefits from UTF-8 with BOM, while web uploads and APIs often prefer plain UTF-8.
Pre-Conversion Checklist
Do these checks before editing the CSV. They prevent the most common failure: saving a file after it already displayed garbled text.
| Check | Why It Matters | Safe Action |
|---|---|---|
| Original file is preserved | A garbled save can destroy the evidence you need | Work on a copy |
| Japanese text previews correctly | Confirms the likely source encoding | Compare Shift-JIS-style and UTF-8 previews |
| Column count is stable | Delimiter issues can look like encoding issues | Check comma, tab, and quoted values |
| Destination is known | Excel and upload forms may need different copies | Create separate output files |
Destination-Specific Copy Guidance
| Destination | First Output to Try | Watch For |
|---|---|---|
| Excel reviewer | UTF-8 with BOM | Excel opening the file directly instead of importing |
| Web app upload | Plain UTF-8 | Upload forms that reject BOM or expect a fixed delimiter |
| API or database import | Plain UTF-8 | Required schema, delimiter, and line ending rules |
| Legacy Japanese system | Keep a Shift-JIS copy if required | Round-trip loss for characters outside the legacy repertoire |
Practical Workflow
- Save the original CSV as an untouched backup.
- Open a short sample in CSV Encoding Fixer.
- Compare Japanese text under UTF-8 and Shift-JIS-style decoding.
- Confirm row count, column count, and delimiter signals.
- Export one UTF-8 copy for upload and one UTF-8 BOM copy if Excel users need to open the file directly.
- Reopen the output in the actual destination before sending the full file.
Example
If 東京, 商品名, or 顧客番号 appears as broken symbols, do not correct those symbols manually. The visible text is already the result of the wrong decoding. Go back to the original bytes, preview the encoding again, and export from the readable preview.
Common Mistakes
- Overwriting the original CSV after Excel shows mojibake.
- Assuming every Japanese CSV is Shift-JIS. Some exports are already UTF-8.
- Fixing commas and quotes after conversion without checking whether the delimiter problem existed first.
- Sending one file to both Excel reviewers and upload systems when the destinations expect different BOM behavior.
Practical FAQ
Is Shift-JIS the same as CP932?
Not exactly. Many Windows Japanese CSV files are CP932-compatible rather than pure Shift-JIS. In practice, treat the label as a signal to preview carefully instead of assuming a perfect round trip.
Should I always convert Japanese CSV files to UTF-8?
UTF-8 is usually the best final format for web apps, APIs, and modern data pipelines. Keep a legacy copy only when a receiving system explicitly requires it.
Why does Excel still show garbled text after conversion?
Excel may choose an unexpected encoding when a CSV is opened directly. Try a UTF-8 BOM copy or use Excel's import flow where the encoding can be selected.
Can a converted CSV still fail upload?
Yes. Upload failures can come from headers, delimiters, line endings, quotes, required columns, or BOM handling even when the text encoding is correct.
Should I repair mojibake characters manually?
Only as a last resort for a tiny file. For real work, return to the original CSV and decode it correctly.
Related Tool
Related Guides
Updated
Last reviewed: 2026-06-25.
Official references
These official references were used to verify the criteria discussed in this article.
- Import or export text (.txt or .csv) filesMicrosoft Support - Checked: 2026-06-25
- Encoding StandardWHATWG - Checked: 2026-06-25
- TextDecoderMDN Web Docs - Checked: 2026-06-25
- The Unicode StandardUnicode Consortium - Checked: 2026-06-25