Reference
Eight symbologies, the exact ZPL command for each, the data rules that decide whether the printer accepts your value, and the two settings — module width and quiet zone — that decide whether the finished label actually scans.
^BY sets the narrow-bar width in dots, the wide-to-narrow ratio, and a
default height. It is state: it applies to every barcode after it until the next
^BY, which is why one stray value can quietly change three barcodes.
Make^BY and the barcode command — the last one wins.
| Resolution | Minimum | Comfortable | Narrow bar, comfortable |
|---|---|---|---|
| 203 dpi | 2 | 3 | 0.38 mm |
| 300 dpi | 3 | 4 | 0.34 mm |
| 600 dpi | 6 | 8 | 0.34 mm |
Note what the last column shows: the same physical bar width needs a different
^BY at every resolution. Copying a 203 dpi label onto a 600 dpi printer
without changing ^BY produces a barcode a third of the intended size.
Sizes and dpi covers the rest of that conversion.
Every snippet below is exactly what MakeZPL emits for that barcode type. Drop one between
^XA and ^XZ and it will print.
^BCData rules: Any ASCII character from space to ~ (32–127).
^FO30,60
^BY3,2,90
^BCN,90,Y,N,N
^FD4125500001^FSDense, self-checking and variable length. If you do not have a reason to pick something else, pick this.
^B3Data rules: Digits, uppercase A–Z, and - . space $ / + %. Lowercase will not encode.
^FO30,60
^BY3,2,90
^B3N,N,90,Y,N
^FDBIN-C14^FSAbout 30% wider than Code 128 for the same data. The second parameter turns the mod-43 check digit on.
^BUData rules: 11 or 12 digits. Give 11 and the printer calculates the check digit.
^FO30,60
^BY3,2,90
^BUN,90,Y,N,Y
^FD07253100640^FSFixed length and fixed proportions — do not squeeze it below about 80% of nominal size or scanners at the till will struggle.
^BEData rules: 12 or 13 digits.
^FO30,60
^BY3,2,90
^BEN,90,Y,N
^FD501234567890^FSThe 13th digit is the check digit; supply 12 and the printer works it out.
^B8Data rules: 7 or 8 digits.
^FO30,60
^BY3,2,90
^B8N,90,Y,N
^FD9638507^FSOnly use it when EAN-13 genuinely does not fit — the shorter number space is allocated separately.
^B2Data rules: Digits only, and an even number of them.
^FO30,60
^BY3,2,90
^B2N,90,Y,N,N
^FD00104822000037^FSAn odd digit count is the classic ITF bug: the printer pads it and the scan comes back wrong. MakeZPL blocks the export instead.
^BQData rules: Any text, including URLs and multi-line payloads.
^FO30,60
^BQN,2,6
^FDQA,https://example.com/asset/41255^FSThe Q in ^FDQA, is the error-correction level (H, Q, M or L) and A is automatic input mode. The third ^BQ parameter is magnification, 1–10.
^BXData rules: Any text. Denser than QR at small sizes.
^FO30,60
^BXN,6,200
^FD41255-A^FSThe second parameter is the module size in dots and the third is the quality level; 200 is the modern ECC 200 encoding.
| If your data is… | Use | Command |
|---|---|---|
| Letters and digits, any length | Code 128 | ^BC |
| Uppercase only, legacy scanner | Code 39 | ^B3 |
| A retail product, North America | UPC-A | ^BU |
| A retail product, elsewhere | EAN-13 | ^BE |
| An even-length carton number | Interleaved 2 of 5 | ^B2 |
| A URL or a long payload | QR code | ^BQ |
| A payload on a tiny part | Data Matrix | ^BX |
MakeZPL validates the value against the symbology before it lets you export. Feed 13 digits to a UPC-A barcode or an odd number of digits to Interleaved 2 of 5 and you get a blocking error with the reason, not a label that prints and then fails at the scanner.
A barcode needs blank label at both ends — at least ten narrow-bar widths, so
30 dots at ^BY3. Butting a code against a box drawn with
^GB, or against the edge of the label, is the most common cause of a code
that looks perfect and reads never.
Excess heat spreads each bar sideways. The bars merge, the narrow:wide ratio drifts
out of tolerance, and the scanner rejects the code. Turn ^MD down until
the bars look crisp rather than bold.
A barcode rotated with R or B still needs its full length.
Rotate a 500-dot code on a 406-dot-tall label and the printer clips it silently
— MakeZPL flags it as an out-of-bounds error before you print.
Symbology choice, check digits, sizing and why codes fail to scan.
Code 128 unless something forces your hand. It encodes the full ASCII range, is denser than Code 39, and every modern scanner reads it. Use Interleaved 2 of 5 for ITF-14 carton codes, and UPC-A or EAN-13 only for retail point-of-sale.
Nine times out of ten the narrow bar is too thin for the print resolution, or there is not enough white space around the code. At 203 dpi keep ^BY at 2 or more — 3 is safer — and leave a quiet zone of at least ten narrow-bar widths at each end. Excess darkness also fattens the bars until the ratios go out of tolerance.
Tall enough for the scanner to find a clean scan line: about 15% of the code's length, and rarely less than 60 dots at 203 dpi. Height costs you nothing but label space, so err upward.
For UPC-A and EAN you can supply the value without its final check digit and the printer calculates it. Code 128 and Data Matrix carry their own internal check characters. Code 39 has an optional mod-43 check digit that is off by default.
Yes. Each is its own ^FO plus barcode command plus ^FD block. Remember that ^BY is state: it applies to every barcode after it until you set it again, which is a common cause of one code printing at the wrong module width.
Both are 2D codes holding roughly the same kinds of data. QR is more widely recognised and reads well from phone cameras; Data Matrix packs the same payload into a smaller mark, which is why it dominates small-part and electronics marking.