CadenzaboxHelpv2 · Phoenix

Open CSVs in Excel safely

Oh Excel, how we love you. We know you’re just trying to help, but we do wish sometimes you just wouldn’t! 🫠

Excel is meme tier internet famous for overzealously reformatting data, so here is the (frustratingly fine-grained) Cadenzabox guide to cajole our favourite spreadsheet app into faithfully respecting your data, rather than silently corrupting it.

To sidestep these hurdles entirely, other apps are available with none of these problems and which work well with truly massive CSVs. For anyone wanting to stick with the OG though, we’ve got your back too.

The three main problems Excel introduces are:

  • date conversion
  • changing track version numbers
  • UPC code corruption

From the beginning we’ve blocked date format errors before ingestion, and since September 2026 we now have checks in place to catch and block the other two. However Excel corrupts these fields as it opens a file, so it’s easy not to notice until you reimport the CSV, after hours of work on a damaged file. 😐

The fix is to import the file with every column set to Text, so Excel keeps each value exactly as it is. This page shows how in Excel on Windows and Mac and in Google Sheets, and which CSV editors avoid the problem altogether.

Never double click a CSV to open it in Excel

Double clicking, dragging the file onto Excel, and File → Open all convert the values as the file opens. If you have already done this, close the file without saving: the file on disk only changes when you save it.

Other CSV editors are available

A dedicated CSV editor treats every value as text and saves exactly what you typed, with no import steps to get right. It also opens files far larger than Excel can.

What Excel gets wrong

In the CSVWhat Excel converts toWhat that does
2026-03-133/13/26, 13/03/2026, or 46094Dates are reformatted, or saved as Excel’s internal day count
TRACK: Number
4.10 (old format)		4.1Version 10 of track 4 becomes version 1 and overwrites it
TRACK: Number
4.60 (old format)		4.6Same: version 60 overwrites version 6
CODE: UPC
5060123456789		5.06E+12The code is lost; only the first few digits survive
CODE: UPC
018736598029		18736598029The leading zero goes, and the UPC no longer checks out
A 16+ digit code 1234567890123456712345678901234500Excel keeps 15 digits and replaces the rest with zeros
1-2, MAR12-Jan, 1-MarCodes, versions, and keys turn into dates
é, ’, –é, ’, ‚ÄôThe file was saved in an old format instead of UTF-8

Other than converting date formats, TRACK: Number is the cell Excel damages most, in files that use our older track number format.

Cadenzabox uses ALBUM: Code, TRACK: Number, and TRACK: Version Number to decide which track a row updates. Both numbers are whole numbers: TRACK: Number 4 with TRACK: Version Number 10 is track 4, version 10, and a main track leaves the version number blank or 0. With no decimal in either column, Excel has nothing to round.

New version number column

Before September 2026, a version was numbered in TRACK: Number alone, with the version after a dot: 4.10 for track 4, version 10. That format is still supported, and any file without a TRACK: Version Number column is read that way.

We moved away from this because Excel treats 4.10 as a decimal and “helpfully” removes the trailing zero, leaving 4.1: version 1, whose metadata the row would then overwrite. If you still use the older format, import every column as Text, as shown below.

Exports in the Cadenzabox format only ever have whole numbers in TRACK: Number and TRACK: Version Number. Other export formats, such as Harvest Media, keep the dot.

Excel on Windows

The Text Import Wizard is the one import route that lets you set every column to Text. Current versions of Excel hide it, so switch it on once:

  1. Go to File → Options → Data.
  2. Under Show legacy data import wizards, tick From Text (Legacy) and click OK.

Then, for each file:

  1. Open a new, blank workbook.
  2. Go to Data → Get Data → Legacy Wizards → From Text (Legacy) and choose your CSV.
  3. Step 1: choose Delimited, set File origin to 65001 : Unicode (UTF-8), and tick My data has headers. Click Next.
  4. Step 2: under Delimiters tick Comma only, and leave Text qualifier as ". Click Next.
  5. Step 3: select every column in Data preview: click the first column, scroll to the far right, and Shift+click the last column. Every column is now highlighted. Under Column data format choose Text. The heading above each column should now say Text. Click Finish.
  6. Choose Existing worksheet at =$A$1 and click OK.

Double-check the data preview step

A column left as General is converted like any other. With a wide Cadenzabox CSV, the selection is easy to lose when you scroll; scroll back through and make sure no column still says General.

Excel on Mac

  1. Open a new, blank workbook.
  2. Go to Data → From Text (Legacy) and choose your CSV. In older versions of Excel for Mac, use File → Import, choose CSV file, and click Import.
  3. Step 1: choose Delimited and set File origin to Unicode (UTF-8). Tick My data has headers. Click Next.
  4. Step 2: tick Comma only and click Next.
  5. Step 3: click the first column in the preview, scroll to the last, and Shift+click it, then choose Text under Column data format. Click Finish.
  6. Choose Existing sheet at =$A$1 and click OK.

Before you edit

The import sets the cells it filled to Text, but not the empty cells around them. So that new rows and columns behave the same:

  1. Click the triangle in the top left corner of the sheet to select everything.
  2. Press Ctrl+1 (Mac: Cmd+1) to open Format Cells, choose Text on the Number tab, and click OK.

This does not change any value that is already there.

While you edit

  • Pasting from another file brings that file’s formatting with it. Use Paste Special → Values (Windows: Ctrl+Alt+V; Mac: Ctrl+Cmd+V) so the cell stays Text.
  • Dragging the fill handle counts upwards. Dragging CBL001_01 down gives CBL001_02, CBL001_03… To copy a value down instead, select the cells and press Ctrl+D (Mac: Cmd+D), or click the Auto Fill Options button that appears and choose Copy Cells.
  • Every track needs its own TRACK: Code. Do not fill a main track’s code down over its versions and stems: the rows would all claim to be the same track.
  • Leave the columns you are not changing alone. Sorting, filtering, and deleting rows are fine; retyping or reformatting a whole column is how values drift.
  • Excel stops at 1,048,576 rows. If an export is larger, export fewer albums or labels at a time.

Save it as UTF-8 CSV

  1. Go to File → Save As.
  2. Set the file type (Mac: File Format) to CSV UTF-8 (Comma delimited) (.csv). Not CSV (Comma delimited), CSV (Macintosh), or CSV (MS-DOS): those are not UTF-8 and turn accented letters and curly quotes into nonsense.
  3. Click Save. If Excel warns that some features will be lost, keep the CSV format.

Cells formatted as Text are saved exactly as they appear, so the file keeps every zero, digit, and code.

Google Sheets

Opening a CSV straight from Google Drive converts it. Import it instead:

  1. In a blank spreadsheet, go to File → Import → Upload and choose your CSV.
  2. Set Separator type to Comma and untick Convert text to numbers, dates, and formulas. Click Import data.
  3. Before you type anything, select all (Ctrl+A, or Cmd+A on a Mac) and choose Format → Number → Plain text, so new values stay as typed too.
  4. When you are done, go to File → Download → Comma Separated Values (.csv). Sheets always saves CSV as UTF-8.

Apple Numbers

Numbers is gentler than Excel: in our tests it kept track numbers such as 4.10, dates, and long codes exactly as they were. It does change three kinds of value as it opens a CSV, though, and we haven’t found a setting that stops it:

  • Leading zeros are dropped. An IPI of 00123456789 becomes 123456789, and a UPC of 018736598029 becomes 18736598029.
  • Some codes turn into money. Codes that start with certain currency codes, such as MTP009 or ALL005, become MTP 9.00 and ALL 5.
  • Codes that look like scientific notation are rewritten. 1E10 becomes 1E+10.

If your file has any of these, use a CSV editor, or Excel or Google Sheets with the steps above. Otherwise, check those columns before you upload.

Excel’s automatic data conversion settings

Recent versions of Excel can switch off some of their conversions: on Windows under File → Options → Data → Automatic data conversion, on Mac under Excel → Settings → Edit (Preferences in older versions). Untick them all as a second line of defence, but do not rely on them: they keep leading zeros and long numbers, but decimals such as 4.60 and dates such as 2026-03-13 are still converted. Importing as Text is the only complete fix.

Check the file before you upload

Open the saved CSV in a plain text editor (TextEdit, Notepad) and search for:

  • In the older track number format, track numbers that have lost a final 0: a 4.1 between 4.09 and 4.11 should be 4.10.
  • Odd characters such as é, ’, or ¬†.
  • #NAME?, #VALUE!, or other # errors.
  • E+: a number in scientific notation.

If you find any, go back to your original export and start again with the steps above: a value Excel has already rounded or shortened cannot be recovered from the damaged file.