Guides / Finishing

How the Anodised Colour on a Keyboard Case Is Really Decided

September 21, 2026

Anodising is not painting. Nothing is added on top of the metal. The aluminium surface is converted into a porous oxide layer, dye is drawn into the pores, and the pores are then sealed shut. The colour you see is dye trapped inside the metal’s own oxide, which is why it wears so well and also why it is so hard to repeat exactly.

Four things decide what comes out of the tank. Understanding them is the difference between a run that ships and a run that gets remade.

One: the alloy and its temper

The oxide layer grows out of the aluminium, so whatever else is in the alloy ends up in the finish.

6063 is low in silicon and copper, which gives a clear, even oxide and the brightest colours. 6061 has more silicon, so the oxide picks up a slight grey cast that pulls every dye a shade cooler and duller. 7075 carries zinc and copper and anodises darkest and least predictably of the three, with a tendency to blotch on large flat faces.

None of this is a defect. It is the alloy doing what the alloy does. The mistake is specifying 6061 for the top case and 6063 for the bottom because a supplier had stock, then expecting the halves to match. They will not, in any dye, ever.

Two: the blast, before any colour exists

Anodising is transparent. Every mark under it shows through, and the texture you feel when you pick the board up was decided before the part ever saw the tank.

Blasting is specified by media and grit size, not by adjectives. “Matte” means nothing to a shop. A fine glass bead leaves a soft satin that reads almost smooth. A coarser aluminium oxide leaves a deeper, drier texture that scatters more light and makes the same dye look lighter and chalkier.

Two consequences follow. A change of grit changes your colour even with the dye untouched, and a repeat run blasted at a different grit will not match the first one. Write the grit on the drawing.

Three: the dye lot and the bath

Dye is consumed as parts are processed. A bath that has coloured forty cases is weaker than the bath that coloured the first four, so later parts come out lighter unless the operator replenishes and checks. Bath temperature drifts across a shift. Time in the dye is judged by an operator who has done it ten thousand times, which is skill, not measurement.

This is the variable you cannot specify away. You can only contain it, and you contain it by batching. All the parts that have to match each other go through in the same load.

Four: the seal

Sealing closes the pores so the dye cannot leach out. Hot water sealing is the usual route and it slightly softens and lightens the final shade. Cold nickel sealing holds the colour closer to what came out of the dye.

It is a small shift compared with the other three, but it is another reason the part in the tank never looks quite like the part in your hand.

What this means for a specification

Write four things on the drawing and most of the risk disappears:

1. Alloy and temper, the same one for every part that has to match.
2. Blast media and grit, as a number.
3. Colour target, as a Pantone or RAL reference plus a physical sample where one exists.
4. Batch instruction: top case, bottom case and any exterior part in the same anodising load.

Then add the sentence that does the most work of all: the approved golden sample is the standard, not the drawing note. Agreeing what counts as a defect before production starts is what stops a colour argument becoming a remake argument.

Type II against Type III

Type II is the standard decorative anodise and the one nearly every keyboard case uses. The layer is thin, so it holds bright colours and barely touches your dimensions.

Type III hardcoat builds a much thicker, denser layer. It resists wear far better, which sounds ideal for something that lives under someone’s hands for years. In practice it darkens and mutes most dyes, narrows the colour range towards greys, blacks and deep tones, and adds enough thickness to matter on a tight plate pocket or a gasket shelf. If you want hardcoat, design the fits with the build allowance in from the start rather than discovering it on the prototype.

What to do when anodising cannot give you the colour

Some colours simply are not available through a dye bath. Deep saturated whites, true pastels and some very flat blacks are the usual ones. That is not a shop failing, it is a limit of the process, and the answer is a different finishing route rather than more pressure on the anodiser.

The practical checklist

Before you approve production:

• One alloy across every matching part, stated on every drawing.
• Blast grit specified by number.
• A physical sample in the shop’s hands, not a render.
• Written instruction that matching parts share an anodising load.
• A golden sample approved by you and kept by both sides.
• An agreed answer to what happens if a batch misses, before it misses.

None of that costs money. All of it costs a conversation, and it is a much cheaper conversation than the one about a hundred and eighty cases in the wrong blue.

See also: the anodise note on a keyboard case drawing, with a note you can copy and adapt.

Frequently asked questions

Keep reading

• Two Colours on One Case: What Masking Adds to an Anodised Run
• Specifying E-White So Cases Match Each Other and the Keycaps