Reading MARD Bead Color Codes: 221 Colors, Cross-Brand Mapping, Color Drift
The code on every cell — A4, H2 — is not arbitrary: it is a numbered position within a color series. This guide covers how to read the system, why cross-brand mapping can only approximate, and how to avoid ordering the wrong color.
The encoding: a letter for the series, a number within it
The MARD-compatible 221-color palette used here encodes each color as a letter (the series) plus a number (its position in that series):
| Letter | Series | Colors |
|---|---|---|
| A | Yellow / orange | 26 |
| B | Green | 32 |
| C | Blue / cyan | 29 |
| D | Blue / purple | 26 |
| E | Pink / rose | 24 |
| F | Red | 25 |
| G | Skin / brown | 21 |
| H | Black / white / gray | 23 |
| M | Earth tones | 15 |
| Total | 221 | |
Three special codes come up constantly: H1 is transparent (leave it empty — no bead placed), H2 is white, and H7 is black. Together they appear more often than any other codes.
Are 221 colors enough?
For cartoons, icons and pixel art — yes. These styles already speak the language of a limited palette, and 221 colors nearly always contains a close enough match. The real pressure comes from two kinds of images: broad gradients (skies, water), where steps between adjacent codes become visible, and low-saturation midtones.
The latter is where this palette has done its homework: 21 skin/brown codes plus 15 earth tones — 36 desaturated colors covering skin, wood and shadow, the unglamorous range that starter 48/96-color sets handle worst.
Why cross-brand mapping is only approximate
A common question: can patterns coded for COCO, Manman or Panpan be worked in MARD beads? The tool maps codes by nearest-color matching using a color-difference formula (Delta E 2000). Emphasize: this is a mathematical approximation, not an official cross-reference. The MARD data here traces back to a public snapshot (MIT licensed); some other brands simply have no reliable public color data, and the accuracy ceiling of those mappings is real.
Practical rule: treat cross-brand mappings as a starting point and the physical color card as the judge — especially for large background areas, where one code off is conspicuous across two or three thousand beads.
Where screen-to-bead drift comes from
Two sources. First, the medium: screens emit RGB light, beads are solid reflective plastic — the same values look different on each. Second, an uncalibrated monitor skews the input before anything else happens. Internally, nearest-code matching runs in a perceptually uniform space (Oklab / CIELAB), which tracks human judgment better than raw RGB comparison — but it cannot fix a monitor that is already off.
A pre-order checklist
- Verify the top 5 codes by count against physical beads — they set the tone of the whole piece.
- Merge codes used fewer than 10 times into a neighbor, or drop them — a whole pack for a handful of beads is a bad trade.
- Keep 5–10% extra of the background and dominant colors — misplacement, over-ironing and loss all happen there.
- Have substitutes in mind for rare codes — an out-of-stock color should not stop the project.
For counting and margins, see the bead count guide; for more on color choice and drift, see the color guide.