En bref : Nous avons cartographié 568 couleurs à travers Perler (103), Hama (92), Artkal-S (199) et Artkal-C (174) avec CIEDE2000 color-difference math against a 291-color MARD reference. Result: only 31 of 291 reference colors (10.7%) have an “excellent” match (ΔE < 2) in the best-matching brand — most colors are approximations, and about a quarter have no close equivalent in any other brand at all.
Chaque créateur de perles finit par se heurter au même mur : un motif demande une couleur que votre marque ne fabrique pas, or you’re holding a bag of Hama and the tutorial lists Perler codes. The internet’s answer is usually a blurry photo of two bead bags side by side. We wanted real numbers, so we computed them. This page is the full data — how many colors exist, how they map across brands, and where the gaps are.
Method note: brand manufacturers do not publish machine-readable color values, so our RGB data comes from the beadcolors open community palette (contributor-attributed). Every entry in our dataset is flagged calibration : en_attente_humain until it is verified against physical beads — treat single-color decisions accordingly. The cross-brand math itself is exact: CIEDE2000 computed on those RGB values, validated against the colormath reference implementation to 4 decimal places.
Combien de couleurs de perles existent ?
| Brand | Line | Couleurs dans notre jeu de données |
|---|---|---|
| Perler | 5 mm midi | 103 |
| Hama | 5 mm midi | 92 |
| Artkal | Série S, 5 mm | 199 |
| Artkal | Série C, 2,6 mm mini | 174 |
| Total | 568 |
Artkal’s S series is the largest single midi line at 199 colors — nearly double Perler’s 103. But raw count is misleading: many of those extra colors sit in ranges other brands simply don’t cover (glow-in-the-dark, pearl, and ultra-specific pastels), which is exactly where cross-brand matching breaks down.
Cross-brand matching: the ΔE numbers
We took a 291-color MARD reference set and found each color’s closest match in Perler, Hama, and Artkal-S using CIEDE2000 — the same perceptual color-difference formula used in the print and textile industries. ΔE < 2 means two colors are indistinguishable to most eyes (“excellent”); ΔE 2–4 is a good match; 4–7 is approximate; above 7 is visibly different (“poor”).
| Qualité de correspondance | Perler | Hama | Artkal-S |
|---|---|---|---|
| Excellent (ΔE < 2) | 3 | 12 | 16 |
| Bon (2–4) | 40 | 54 | 73 |
| Approximatif (4–7) | 141 | 106 | 134 |
| Poor (≥ 7) | 107 | 119 | 68 |
Le constat principal : Artkal-S est la meilleure correspondance inter-marques pour l’ensemble de référence, with 89 colors (31%) at good-or-better and the smallest poor-match tail (68 of 291, 23%). Perler is the hardest brand to match into — 107 reference colors (37%) have no close Perler equivalent, and only 3 match excellently. If your pattern was designed in one brand and you bead in another, expect roughly une couleur sur quatre visiblement différente no matter which direction you convert.
Les pires correspondances : couleurs sans bon équivalent
The biggest gaps cluster in two places: very saturated mid-tones (bright teals, lime greens) and specialty finishes. The five worst Perler matches in our data, with their ΔE distances:
| Couleur de référence | Perler le plus proche | ΔE |
|---|---|---|
| MARD D10 | 80-15243 | 14.82 |
| MARD D21 | 80-15258 | 13.79 |
| MARD D17 | 80-19093 | 13.62 |
| MARD H7 | 80-19018 | 13.19 |
| MARD P15 | 80-19091 | 12.29 |
At ΔE 12–15, these aren’t “close enough” — they’re different colors. A sky-blue pattern converted with these matches comes out purple-ish. If a pattern you love uses a color in the poor-match zone for your brand, the practical moves are: substitute two adjacent shades and blend them across the area, or order that one color from the original brand.
Ce que cela signifie pour vos motifs
- Concevez dans la marque que vous possédez. A pattern made in Artkal-S converts worst to Perler; start with the palette you’ll actually buy.
- Moins de couleurs convertit mieux. A 10-color pattern survives brand conversion far better than a 20-color one, because every color matters less individually.
- Surveillez la liste des mauvaises correspondances. Before a big project, check whether any of your pattern’s colors sit in your brand’s ΔE ≥ 7 zone — those are the ones to buy specially.
The éditeur de motifs gratuit applies this same data live: switch the palette between brands and your pattern re-matches every color to that brand’s nearest bead, so you can see the conversion before you bead it. New to the craft? Start with our 25 motifs débutants — all designed with small, conversion-friendly palettes — and iron them with the guide complet de repassage.
FAQ sur les données couleur
What is ΔE?
ΔE (Delta E) measures how different two colors look to the human eye, using the CIEDE2000 formula. Below 1 is invisible to almost everyone; 2–4 is a close match; above 7 is clearly a different color. It’s the standard metric in printing, textiles, and paint.
Pourquoi Perler et Hama ne publient-ils pas de valeurs officielles ?
We asked the same question. None of the three major brands publishes machine-readable RGB or Lab values for their bead lines — every color chart online is community-measured, which is why our dataset flags every entry as pending physical verification and why we publish our method openly.
Puis-je mélanger les marques de perles dans un projet ?
Yes — all standard 5 mm midi beads fuse together fine. The risk is purely color: two brands’ “red” may differ by ΔE 5+. Check our match quality tables and stick to excellent-or-good matches when mixing.
Où puis-je télécharger le jeu de données complet ?
A browsable color library with per-color cross-brand matches is in development. When it ships, every color above links to its own detail page with the full ΔE table.