Fix Japanese CSV Mojibake: Shift-JIS to UTF-8

Quick Answer

Japanese CSV mojibake happens when bytes from one encoding are decoded as another encoding. The most common production case is a Japanese Windows or legacy export that needs Shift-JIS / CP932-style handling, but it is opened as UTF-8 or passed through Excel without an explicit import step.

Start from the original file, compare UTF-8 and Shift-JIS previews, check whether columns still line up, and export a clean copy only after the Japanese text is readable. If the CSV is for Excel review, a UTF-8 BOM copy may be the safest handoff. If the CSV is for a web upload or API, plain UTF-8 without BOM is often safer.

Start Here

If you searched for mojibake utf-8 shift_jis, do not edit the broken characters. Open a small sample in CSV Encoding Fixer, compare UTF-8 and Shift-JIS, then choose the output by destination.

Why This Matters

Japanese product names, addresses, company names, and customer messages often move through CSV files. If mojibake is saved back into the file, the original text can be hard to recover. This is especially common when a Japanese Windows export is opened on another machine or uploaded to a UTF-8-only service.

The practical goal is not to make one universal CSV. The goal is to preserve the original, identify the correct reading of the source bytes, then create a destination-specific copy.

Common Causes

Practical Workflow

1. Do not save the file while it is garbled.
2. Check whether the file came from Japanese Windows, Excel, a SaaS export, or an API.
3. Preview the file as UTF-8 and Shift-JIS.
4. Check that the delimiter and column count are stable after the text becomes readable.
5. If UTF-8 looks correct but Excel is the problem, create a UTF-8 BOM copy.
6. If Shift-JIS looks correct, convert a copy to UTF-8 before uploading to modern systems.

Use the Encoding Comparison Panel

The fastest way to avoid guessing is to compare the same bytes through several decoders.

1. Open CSV Encoding Fixer.
2. Load a small sample file, not a private full customer export.
3. Compare the UTF-8, EUC-KR / CP949, and Shift-JIS cards.
4. Prefer the candidate with readable Japanese, fewer mojibake signals, and stable columns.
5. Use the selected decoding, then download an Excel-friendly UTF-8 BOM copy only when the destination is Excel.

The score is a triage signal, not proof of encoding. Human preview still matters because names, product codes, and half-width kana can look plausible even when one character is wrong.

Symptom Checklist

Use the visible broken characters as a clue before changing the file. If the text only breaks in Excel but looks correct in a code editor or browser, the file may already be UTF-8 and Excel may need an explicit import step or a UTF-8 BOM copy.

Choose a Fix by Destination

FixData Tool

CSV Encoding Fixer compares encoding candidates, checks delimiter and column stability, and creates a UTF-8 BOM version for Excel. Use a small sample when the file contains private customer data.

Example

Mojibake-like preview:

company,address
譁ｰ讒倥し繧､繝医Ο繧ｰ,譚ｱ莠ｬ驛ｽ蜊区

Expected:

company,address
株式会社サンプル,東京都千代田区

The fix is not to edit the broken text by hand. Reopen the original with the right encoding, then convert a clean copy.

Common Mistakes

• Saving the CSV after Excel shows mojibake.
• Assuming UTF-8 BOM fixes every Japanese CSV.
• Ignoring CP932 / Shift-JIS files from older Japanese systems.
• Uploading an Excel-only copy to an API that rejects BOM.
• Fixing delimiter problems before confirming the characters are readable.
• Comparing only the first row when the Japanese text appears later in the file.

Practical FAQ

Is mojibake utf-8 shift_jis always a Shift-JIS problem?

Not always, but it is a strong clue. The safest workflow is to reopen the original file and compare UTF-8, Shift-JIS, and CP932-style previews before exporting a clean copy.

Should I use UTF-8 BOM for every Japanese CSV?

No. UTF-8 BOM is useful when the recipient opens the file in Excel, but some APIs and upload forms prefer plain UTF-8 without BOM. Keep separate copies for Excel review and system upload.

Can I repair a file after it was saved while garbled?

Sometimes, but recovery is much harder. If the broken text was saved over the original, you may have lost the byte sequence needed to recover the Japanese text. Look for the original export first.

Why does the delimiter matter if the problem is mojibake?

Encoding decides which characters appear. The delimiter decides where columns split. A file can have readable Japanese text but still be unusable if semicolons, tabs, or quoted line breaks are parsed as plain commas.

Why does the tool compare EUC-KR / CP949 on a Japanese page?

It is mainly there to expose false positives in mixed Korean/Japanese operations. If a supplier file contains both Korean and Japanese text, comparing legacy Korean and Japanese decoders helps you avoid choosing a candidate that only fixes part of the file.

Related Tool

• CSV Encoding Fixer

Related Guide

• CSV / Excel Encoding Center
• Excel character encoding for CSV

Updated

2026-06-24