A scanner reads your Data Matrix code. A verifier gives it a grade. Only the grade decides whether your parts ship.
Short answer
- Ask your customer which standard they use. ISO/IEC 29158 for marks on the part. ISO/IEC 15415 for labels. AS9132 and MIL-STD-130 add their own limits.
- Check your material and hardness. Above 60 HRC, dot peen is out. Black on aluminium needs MOPA, not standard fiber.
- Map what happens to the part after marking. Blasting, heat treatment and coating all change the grade.
- Only then compare machines — on module size, contrast method and repeatability. Not on wattage.
Many buyers work the other way round. They compare machines first and think about acceptance later. The code reads fine in their workshop. It fails at the customer.
You do not have to choose between Data Matrix and QR. The standards already chose. MIL-STD-130, AS9132 and FDA UDI all ask for Data Matrix ECC200. We explain why in Data Matrix vs QR code. This page is about picking the machine.
Which standard will your code be graded against?
Five standards cover most industrial work, and they measure different things. Ask your customer which one applies before you look at any machine.
This matters more than it sounds. The same mark can get different grades under different standards, because each one lights the part in a different way.
| Standard | Applies to | What it measures | Who asks for it |
|---|---|---|---|
| ISO/IEC 16022 | The code itself | Data Matrix ECC200 structure. From 10 × 10 modules, with a one-module quiet zone | Everyone. It defines the code |
| ISO/IEC 15415 | Printed codes and labels | Print quality, graded A to F under fixed lighting | Label and packaging supply chains |
| ISO/IEC 29158 (AIM DPM) | Marks made directly on the part | The same parameters, but with lighting and aperture rules made for direct marks | Automotive, aerospace, medical, defense |
| AS9132 | Aerospace Data Matrix geometry | Cell fill 60–105%. Contrast difference 20% or more. Graded A / B / F | Aerospace primes and their suppliers |
| ATA Spec 2000 Ch. 9 | Commercial aviation parts | What data goes in the code, and in what format, for the airline supply chain | Commercial aerospace and MRO |
| MIL-STD-130 | US defense item marking (UII) | Sets the pass mark: grade B or better | Defense supply chain |
No machine can be “MIL-STD-130 certified” or “AS9132 compliant.” These standards judge the mark, not the equipment.
A machine can only be able to produce marks that pass, on a stated material, with stated settings. If a supplier says their machine is certified, they are describing something that does not exist.
Not sure which standard your customer uses? Tell us your industry and we will name it — ask a marking engineer.
Why does a code that scans fine still fail?
A scanner only has to read the code once. A verifier scores each parameter on its own — contrast, modulation, fixed pattern damage, axial non-uniformity.
It then reports the lowest score as your grade. One weak parameter sets the whole result.
This is where most disputes start. The workshop scans the part with a handheld reader. It reads instantly. The parts ship. The customer runs a verifier and returns the batch at grade D.
ISO/IEC 29158 exists because direct part marks have no ink. A dot peen mark is a field of small dents. It creates shadow, not colour contrast. So the standard lights it at 30° to catch that shadow, and grades it at two aperture sizes, reporting the better one.
ISO/IEC 15415 does none of this. If you grade a dot peen mark under 15415, you are using the wrong standard.
If your codes are already graded and you need to know why one parameter came back low, our guide to readable versus verifiable codes goes through the grade report line by line.
Does your material rule out dot peen?
Above 60 HRC, yes. Below that, hardness is not the problem. Contrast is.
Carbide styli stop biting into hardened surfaces and wear out fast. Below the limit, dot peen makes shadow, never colour contrast. That is a physical limit, not a quality problem.
| Constraint | Dot peen | Fiber / MOPA laser | UV laser |
|---|---|---|---|
| Hardness limit | 60 HRC | None | None |
| Wavelength | — | 1064 nm | 355 nm |
| Mark depth | 0.1–1 mm. Portable pneumatic to about 2 mm | Surface to under 1 mm | Very shallow. Cold process |
| How it makes contrast | Shadow from dents | Oxide, melt or colour change. MOPA pulse width 2–550 ns adjustable | Chemical change. Smallest heat-affected zone |
| Good for | Steel, castings, forgings, oily or rough surfaces | Steel, aluminium, brass, titanium. Colour marks on medical stainless | Glass, ceramics, PCB, medical plastics |
| Not for | Hardened parts. Thin walls. Fatigue-critical surfaces | Glass, most ceramics | Deep marks. High volume on steel |
True black on aluminium, and colour on stainless. Both need adjustable pulse width. Standard fiber sources cannot do it. You need MOPA.
Medical stainless. Ablation damages the chromium oxide layer that stops the steel rusting. Annealing is the only route, and parts still need re-passivation afterwards.
Which material are you marking?
Material decides how the mark gets its contrast, and contrast is the parameter that fails most often.
| Material | Start with | What decides the grade | Watch out for |
|---|---|---|---|
| Steel and stainless | Fiber above 60 HRC. Dot peen below it | Surface finish, more than power. A brushed or blasted surface grades better than a polished one. Polished stainless reflects the verifier’s own light straight back | Never use the same stylus on carbon steel and stainless. Iron particles stay in the dent and rust |
| Aluminium | MOPA, not a standard fiber source | Pile-up. Aluminium is soft, so dot peen pushes metal up around each dot. The raised ring joins neighbouring cells and cell fill leaves the 60–105% window | Check anodised parts after anodising, not before. The coating changes both colour and fine detail |
| Copper and brass | Fiber, but test your own alloy first | Reflectivity at 1064 nm. Results vary more between suppliers on copper than on any other common metal | A demo on the supplier’s brass sample proves nothing. Copper content, temper and surface oxide all change the result |
| Bronze | Dot peen on cast surfaces. Fiber on machined ones | Alloy mix. Tin content changes how the surface reacts to a laser. Two bronzes that look the same can grade one letter apart | Cast bronze is rough and porous. Dot peen is steadier there, because shadow copes with rough surfaces better than colour contrast |
| Plastics and PCB | UV at 355 nm | Which way the plastic reacts. Dark resins go lighter. Light resins go darker. The two behave differently under a verifier | Fiber can melt or burn the surface, which ruins the cell edges. Dot peen cracks hard plastics and dents soft ones |
One rule covers all five. Ask the supplier to mark your material, in your surface condition, and send you the verifier report.
A sample plate hides the alloy, the temper, the finish and the curve. Every one of those changes the grade.
Send us your part. We will mark it and send you the grade.
Reading about your material is one thing. Seeing a verifier report on your own part is another.
- Your part back, marked, with the settings we used
- A verifier grade report showing every parameter, not a pass or fail
- Measured mark depth
- If a different process suits your part better, we will tell you
Laser Data Matrix marking: what decides the grade
A fiber laser can make a Data Matrix code three different ways. The three do not grade the same.
Picking the wrong one is the most common reason a laser data matrix marking process fails, even on a material the laser handles well.
| Method | How it works | How it grades | Speed |
|---|---|---|---|
| Annealing | Heat below melting point makes an oxide layer under the surface | The most even contrast. No metal is removed, so it is the only choice when surface strength matters | Slowest |
| Oxide colour | The surface oxidises. Pulse width and frequency set the colour | Good contrast on stainless and titanium. Drifts as the laser warms up | Medium |
| Ablation | Metal is removed or re-melted to make a rough cell that scatters light | Highest contrast, but the texture varies cell to cell. That shows as poor axial non-uniformity | Fastest |
Why a laser code can be harder to pass than a dot peen code
This surprises people. A laser mark has real colour contrast, so it gets graded under 90° on-axis lighting. That lighting shows every small difference between cells.
A dot peen code is read as shadow at 30°, which forgives a cell that came out a little light. So the process that looks better is the one judged more strictly.
Three specs that decide it
- Spot size against module size. A module needs to be several spot widths across, or one pass defines the cell edge and modulation drops. Ask for spot size in microns, not just wattage.
- Fill pattern. Single-line fill is fast but leaves cell fill at the low end of the AS9132 window. Cross-hatch fill holds it better but takes longer. Ask which one the quoted cycle time assumes.
- Flat field limit. The marking area on any data matrix code marking machine is a flat area. Wrap a code around a diameter and the outer cells go out of focus. A matrix code marking machine for round parts needs a rotary axis or 3D dynamic focus.
Where does the mark sit in your process?
A code can leave your workshop at grade B and arrive at your customer at grade D, with nobody doing anything wrong.
Heat treatment, shot blasting, anodising and coating all change contrast or shape. So your process order decides the marking method. Shops that buy the machine first often miss this.
| Operation | What it does to the code | What to do |
|---|---|---|
| Shot blasting | Wears away shallow marks. Rounds off dot edges and kills the shadow | Mark deeper with dot peen, or mark after blasting |
| Heat treatment | Scale and oxide change how the surface reflects. Hardness rises past the dot peen limit | Mark before hardening, or use laser |
| Anodising | Changes surface colour and can fill fine detail | Check the grade on the anodised part, not the bare one |
| Paint or powder coating | Covers shallow marks completely | The mark must be deeper than the coating |
| Passivation (medical) | Needed after laser marking stainless. Ablated marks may already have damaged the protective layer | Use annealing, then re-passivate |
On load-bearing parts a deep mark is a weak point. NASA-STD-6002 and PRC-9003 both say marks must sit in low-stress areas.
Decide where the mark goes before you decide how deep it will be.
What should you send a supplier before asking for a quote?
Six things. Send them and you get a specific quote instead of a generic one. You also find out fast which suppliers cannot answer.
- The standard and the grade you need. “ISO/IEC 29158, grade B minimum” is a spec. “Good quality code” is not.
- Material and condition. Alloy name plus hardness in HRC. Say whether the part is machined, hardened, cast or coated.
- Code size and content. How many characters must fit, and how much space you have in millimetres.
- Surface shape. Flat, curved or round. Give the diameter if it is round.
- The full process order. Every step the part goes through after marking, in order.
- Cycle time and volume. Parts per shift, and seconds available per mark.
Then ask for one more thing: a marked sample on your own material, with a report showing each parameter, not just pass or fail.
Send us those six things and your part. You get the sample back with a full verifier report — request a graded sample.
Getting the order right
The expensive mistake is not buying a machine that is too small or too slow. It is buying one that marks a code nobody checked. Then a customer runs a verifier.
By that point your tooling, your fixtures and your cycle time are all built around the wrong process. A machine costs a few thousand dollars. Re-qualifying a process on a part already in production costs far more.
So work in this order:
- Find the standardAsk your customer which one they grade against, and what grade you have to hit.
- Check the materialHardness and surface finish decide whether you have a choice of process at all.
- Map the processList every step the part goes through after marking, in order.
- Then compare machinesOn module size, contrast method and repeatability. Not on wattage.
Will your code pass your customer’s verifier?
That one question decides which marking machine you need. Tell us your industry and what your customer asks for. Our engineers will name the standard, the grade you have to hit, and what that means for the equipment.
We reply within one business day.FAQs
Which method should I use on copper, brass or bronze?
Test your own alloy first. All three reflect a lot of light at 1064 nm, so results vary more between suppliers than on steel.
On machined surfaces, fiber usually wins. On cast bronze, dot peen is often steadier, because shadow copes with a rough porous surface better than colour contrast does. Tin content matters too — two bronzes that look the same can grade one letter apart.
Is a Data Matrix printer the same as a Data Matrix marking machine?
No. A Data Matrix printer puts a code on a label or a box, using ink or heat. A data matrix code marking machine forms the code in the part itself, with dents or a laser.
Labels fall off, burn, or wash off with solvent. Traceability standards ask for a mark on the part.
Laser or dot peen — which passes verification more easily?
It depends on the surface. Laser suppliers often say dot peen cannot reach high grades, and on a smooth machined face they have a point: laser gives real colour contrast, dot peen only gives shadow.
But on rough, oily or cast surfaces the picture flips. A laser needs a clean, even surface to give even contrast. Dot peen does not care. If your parts come straight from a foundry, dot peen usually holds a steadier grade.
How small can the code be?
ISO/IEC 16022 allows 10 × 10 modules with a one-module quiet zone, so the code itself can be tiny. Your process sets the real limit.
Each module has to stay big enough for the verifier to resolve it, and cell fill must stay inside the 60–105% window AS9132 asks for. On a 4 mm surface that needs very fine control — ask for the smallest module size the machine can hold, not the smallest it can attempt.
My part has almost no flat space. Does the code have to be square?
No. Data Matrix also comes in rectangular sizes. If you have width but no height, a rectangular code fits where a square one will not.
Check two things before you commit: that your customer's scanner reads rectangular codes, and that your marking software supports them. Not all do.
What does it cost to run, after the machine?
Dot peen has one consumable: the carbide stylus. How long it lasts depends on your material hardness and mark depth, so ask for the price and the minimum order rather than a life figure.
Laser has no consumable in normal use. The costs sit elsewhere — fume extraction on most materials, and the laser source itself, which is the expensive part to replace. Ask how long spare sources stay available.
The cycle time is too slow. What can I change?
Three things, in this order. First, check whether marking time or loading time is the real limit — on most stations it is loading.
Second, ask about the fill pattern. Single-line fill is faster than cross-hatch, though it puts cell fill near the low end of the AS9132 window. Third, a code made of dots instead of solid squares marks faster and still grades, as long as each dot stays inside its cell.
