Perler, Hama and Artkal Color Codes: Palette Sizes and Cross-Brand Substitution

Palette size decides how closely a pattern can follow the original image. The nine brand palettes modelled by this tool range from 18 to 221 colors: Perler 117, Hama 89, Artkal 176, Artkal C (Mini) 174, Artkal Hard 176, Artkal Soft 176, Nabbi 30, MARD 221, and IKEA Pyssla 18, plus a separate 48-color universal palette.

Adding the nine brand palettes gives 1,177 color cards. De-duplicating by code leaves 988 unique codes. The difference comes from lines inside one brand reusing the same numbering, as when several Artkal lines share S-prefixed or M-prefixed codes. The 48-color universal palette is excluded from both figures because 21 of its codes already exist in the brand palettes.

As a rough measure, Perler's 117 colors cover about an eighth of the 988 unique codes, Hama's 89 slightly less, and MARD's 221 close to a quarter. The smaller the palette, the more aggressively the tool must merge colors, and gradients and near-identical skin tones are usually the first casualties.

Every manufacturer uses its own numbering, usually a brand prefix followed by digits. Perler uses P (P01, P96, P961), Hama uses H (H01, H101), the main Artkal line uses S (S01), Artkal C (Mini) uses AC (AC01), Artkal Soft uses MA (MA01), Nabbi uses N (N01), and IKEA Pyssla uses I (I01). MARD differs, using a letter plus a number (A1, A10).

A prefix tells you which numbering scheme you are looking at; it carries no color information. P01 and H01 both sit first in their sequences but are not the same color. The order of numbers does not follow the color wheel either, because many brands add codes in batches, which leaves gaps in the sequence.

Artkal deserves a separate warning. Several of its lines start with S or M, so the codes look identical while the beads and colors behind them are not. Whenever you see S01 you must also confirm which line it belongs to, or your substitution will silently pick the wrong bead.

Codes are internal manufacturer numbering. There is no cross-brand industry standard, so no authoritative table of one-to-one conversions exists. Any chart claiming P01 equals H01 is really an approximate match made by comparing colors, not a code conversion, and it holds only as far as the person who made it judged the two beads to be close enough. Two such charts can therefore disagree without either being wrong in its own terms.

The more useful question is whether a substitution is even possible. Three things decide that: whether the diameters match (otherwise the beads will not fit the same board), whether the target code exists at all (otherwise you cannot buy it), and whether the finish matches (pearl, glow, and translucent beads have no plain-color equivalent).

So the correct order is diameter first, then color, then finish. Skipping the first step produces a pattern that will not even sit on your board, however good the color match is. Skipping the last produces a piece whose colors are right but whose surface looks nothing like the original, a disappointment that usually reads as a color failure even though the colors themselves were matched correctly.

Step one, check the diameter. Perler, Hama, Nabbi, and Artkal Hard are midi 5 mm. Artkal C (Mini) and MARD are mini 2.6 mm. Converting a 5 mm pattern to 2.6 mm beads does not merely swap codes: the finished edge length becomes 52 percent of the original, so confirm this before you start.

Step two, match colors. List every code the original pattern uses and find the closest equivalent in the target palette, writing down each replacement. For patterns using 10 to 20 colors, comparing physical swatches by eye is enough. For larger or gradient-heavy patterns, an automated color-difference match is faster and more consistent than manual comparison.

Step three, match finishes. If the original used pearl, glow, translucent, or glitter beads, either find the same effect in the target brand or accept the change in surface quality when substituting a plain color. A color-matching algorithm compares color values only and will not make that judgement for you.

Automated matching converts every source code and every candidate in the target palette into CIELAB space, computes the perceived difference ΔE00 with the CIEDE2000 formula, and takes the smallest value. This tracks human perception much more closely than comparing raw RGB distances, especially in greys, skin tones, and dark areas, where a raw RGB distance tends to call two visibly different colors a close match.

The ΔE00 value is a useful risk indicator for a substitution. Close to zero means the difference is essentially invisible. A clearly larger number means the target palette has no good match, so you should either merge into a neighbouring code or stay with the original brand. The ΔE threshold setting in the tool controls exactly this tolerance.

Keep a replacement table once you have built one. The reason is practical: the same pattern may be rebuilt or restocked later, and a table guarantees the second version uses the same colors as the first. Sharing one table across several patterns also keeps a given color consistent between different finished pieces.

Metallic, pearl, glow, translucent, and glitter beads are excluded from automated difference-based matching. Their visual character comes from reflection and light transmission, while a difference formula compares a single color value, so the nearest match it returns may be a plain bead with a completely different surface. Treat that substitution as an artistic decision, not a technical one.

The practical approach is to list effect beads separately and confirm by hand whether the target brand offers an equivalent line. If it does not, decide the substitute during design: a common solution is to use a bright plain color in place of pearl, or a pale fill plus a dark outline in place of translucent.

There is a related caution about de-duplication. The 988 unique codes count numbering across the brand palettes. Where an effect bead shares a numbering scheme with plain colors, it collapses at the code level even though the physical bead is different. Buy by the material stated on the packaging, not by the code alone.

An 18-color IKEA Pyssla palette and a 221-color MARD palette produce structurally different results from the same photograph. With 18 colors the tool must merge heavily, compressing sky, grass, and skin into a handful of codes with visible banding. With 221 colors the same regions may split across a dozen close codes.

For the person building the piece, more colors is not simply better. Every extra color means another bag to buy, fewer beads per color, and a higher chance of misreading a cell. On small projects, deliberately limiting yourself to 10 to 20 colors usually produces a cleaner piece that is easier to finish. Conversely, if you own only a 30-color Nabbi set, do not pick patterns written for a 221-color palette; choose high-contrast designs with large flat areas and few colors.

When restocking, buy by brand plus code, never by color name. Color names are translated inconsistently between brands, so the same pale blue can be a completely different code elsewhere, and small batch-to-batch variation exists even within one code. Buy the whole project at once when you can. If you substitute across brands regularly, write the replacement relationships down and reuse them, so the next pattern is a lookup rather than a fresh comparison.