In short: The RowSlice CSV to UTF-8 converter re-saves a CSV file as UTF-8, with a byte order mark (BOM) by default so Excel shows accented letters and symbols correctly, and can switch the delimiter between comma, semicolon, tab and pipe in the same pass. It recognises UTF-8 and, by its byte order mark, UTF-16 on its own, treats any other file as Windows-1252 (ANSI), and reads ISO-8859-2, Shift JIS or GBK when you pick them. The file is read in your browser a few megabytes at a time and never uploaded, so there is no set size cap; the limit is the free disk space your browser can use. The result is always UTF-8: it cannot convert to ANSI or any other encoding.
How to Convert a CSV File to UTF-8
- Choose or drag in the CSV file. Its encoding and delimiter are detected from the first 256 KB of the file and shown above the options, for example Detected: WINDOWS-1252, semicolon separated.
- Look at the preview. If letters still look wrong, change the Encoding under Advanced Options until they look right.
- Pick the delimiter you want in the result, and keep Add a UTF-8 BOM ticked if the file will be opened in Excel.
- Press Convert File and download the fixed CSV, named
yourfile-converted.csv.
Which Garbled Characters Mean What
| You see | Should be | What happened | Fix |
|---|---|---|---|
Café | Café | UTF-8 file read as Windows-1252 | Convert with BOM ticked |
don’t | don’t | UTF-8 curly apostrophe read as Windows-1252 | Convert with BOM ticked |
M�nchen | München | Windows-1252 file read as UTF-8 | Convert (encoding detected automatically) |
| Everything in column A | Separate columns | Delimiter does not match Excel's regional setting | Change the delimiter |
If you see é and the file has already been saved again by another program, the damage may be written into the file itself, and converting will faithfully keep é. In that case go back to the original export if you can.
When it is not clear which line of that table applies, first run the file past the CSV validator: it names the encoding it detected, says whether the file starts with a byte order mark, and lists the bytes that are not valid in that encoding, all without changing the file.
Why CSV Encoding Goes Wrong So Often
A CSV file is plain text and has no place to say which character encoding it uses. Most modern systems write UTF-8, which can store every language. Excel on Windows, when it opens a CSV by double-click, assumes the older encoding of your Windows region unless the file starts with a UTF-8 byte order mark. Many exports from web apps, databases and Google Sheets do not include one, so accented letters, curly quotes and currency symbols come out garbled.
Which Encodings It Reads
The converter turns ANSI and UTF-16 files into UTF-8, the only encoding it writes. It works out the encoding from the first 256 KB of the file, and you can override its choice:
| Encoding | Detected automatically? | How it is recognised |
|---|---|---|
| UTF-8 with a BOM | Yes | The file starts with the bytes EF BB BF |
| UTF-8 without a BOM | Yes | The text is valid UTF-8 |
| UTF-16 LE or BE | Yes, when it has a BOM | The file starts with FF FE (LE) or FE FF (BE) |
| Windows-1252 (ANSI) | Yes, as the fallback | Anything that is not valid UTF-8 |
| ISO-8859-2 (Central European) | No | Choose it under Advanced Options |
| Shift JIS (Japanese) | No | Choose it under Advanced Options |
| GBK (Chinese) | No | Choose it under Advanced Options |
ANSI is the name Notepad and other Windows programs give the older encoding, and on a US or Western European copy of Windows it means Windows-1252. A file described as Latin-1 or ISO-8859-1 can be read as Windows-1252 too: the two have the same letters in the same places.
Because Windows-1252 is the fallback, a file with no BOM that is not valid UTF-8 is read that way at first. The bytes alone cannot show whether a file is Windows-1252, ISO-8859-2 or another older encoding, so if the preview shows the wrong accented letters, pick a different one from the Encoding menu under Advanced Options and the preview updates. Do the same for a UTF-16 file without a BOM: choose UTF-16 LE or UTF-16 BE yourself. Other encodings, such as Big5, EUC-KR or Mac Roman, are not in the menu.
Converting a UTF-16 file to UTF-8 is also what lets you read it in RowSlice's viewer, which finds rows by their position in bytes and stops on a UTF-16 file rather than showing every character half a step out.
One case to watch in large files: if the first 256 KB of an ANSI file has no accented letters or special symbols at all, it is taken for UTF-8, and accented letters further down come out as �. If you know the file is ANSI, choose Windows-1252 before converting. How to tell which encoding a file uses from its symptoms is covered in UTF-8, UTF-16 and Windows-1252: which one is your CSV?
UTF-8 with or without a BOM
The box Add a UTF-8 BOM so Excel shows accented characters correctly is ticked by default. It adds a BOM (byte order mark), three invisible bytes at the very start of the file that Excel on Windows needs before it will read a double-clicked UTF-8 CSV correctly.
Untick it for a file going to a database, a script or a web app's import screen. Some of those treat the three bytes as part of the first column name, so a column called id can no longer be found by that name. With the box unticked the result has no BOM even if the original had one, so this is also how to convert a CSV to UTF-8 without a BOM. CSV UTF-8 vs CSV in Excel shows the same file opened in Excel with and without one.
Change the Delimiter from Semicolon to Comma
Like the encoding, the delimiter Excel expects depends on your region. In regions where the decimal separator is a comma, spreadsheet programs save and expect semicolons, so a CSV from a US system opens as a single column in a German or French Excel, and a CSV from a European Excel confuses US software.
To change a CSV's delimiter, set Delimiter in the result to Comma, Semicolon, Tab or Pipe before you convert. The switch happens in the same pass as the conversion to UTF-8, and any value that contains the new delimiter is put in quotes so it stays in one column. Which delimiter your Excel expects, and other ways to fix a file that lands in column A, are covered in why your CSV opens in one column in Excel.
What This Converter Changes, and What It Doesn't
The result is always UTF-8. When you keep the same delimiter, each row's text is copied as it is, so quoting and values like 00742 do not change. When you change the delimiter, values are re-quoted only where needed. Line endings can be kept, or set to Windows (CRLF) or Mac and Linux (LF). The file is processed in small chunks on your own computer, so large files are converted without being uploaded.
Last updated .