We review dozens of spec sheets every month — for our own sourcing and for buyers who ask us to sanity-check a machine they found online. The pattern is always the same: a table stuffed with impressive-looking figures, most of which have almost nothing to do with what the machine will actually do on your shop floor. Learning to ignore the right numbers was one of the more useful skills we developed in this business.

This guide covers both laser engravers and FDM/SLA 3D printers, because the spec-sheet tricks overlap more than you'd expect. The goal isn't to make you an engineer — it's to stop you from paying for performance that doesn't exist.

Laser power: peak vs. average, and why "60W" might mean 8W

The single most abused spec in the laser world is wattage. For diode lasers (the kind sold as desktop engravers), many sellers advertise peak optical power or even input electrical power, while the useful number is average optical output at the workpiece. A "60W" diode module can easily deliver only 8–10W of continuous optical power once you account for duty cycle and thermal throttling. That's not fraud — it's just a measurement taken under conditions you will never reproduce during actual engraving.

Fiber laser cutting metal
Fiber lasers quote average power honestly because industrial buyers demand it. Diode marketing is looser.

What to look for instead: ask the supplier for the continuous-wave (CW) optical output measured at the nozzle after warm-up, ideally with a third-party test report. If they only provide "max" or "peak," treat the real working power as roughly 15–20% of that figure for diodes. For CO₂ and fiber lasers the gap is smaller, but you still want the rated average power at operating temperature, not the cold-start burst.

Also check the wavelength. Diode lasers (450 nm blue) mark plastics and coated metals well but barely touch bare steel. CO₂ (10,600 nm) handles organics and glass. Fiber (1064 nm) marks metals efficiently. Picking the wrong wavelength is a bigger mistake than buying too few watts — no amount of power fixes a physics mismatch.

Engraving speed: mm/s without context is meaningless

You'll see claims like "10,000 mm/s engraving speed." This is usually the galvo scan speed in empty air — the mirrors moving with the laser off. Actual marking speed depends on material, mark depth, hatch spacing, and required contrast. On anodized aluminum at readable contrast, a galvo fiber system that advertises 10,000 mm/s might effectively produce parts at 300–800 mm/s of useful line speed.

The honest question to ask: "At what parameters did you achieve this speed, and what does the resulting mark look like?" A supplier who answers with specific settings (material, hatch, frequency, power percentage) is giving you real data. One who repeats the headline number is giving you brochure filler.

For gantry-style CO₂ machines, the limiting factor is often mechanical mass, not laser power. A heavy head on a belt-driven axis won't hold tight corners at top speed regardless of what the controller says. Ask about acceleration (mm/s²) and positioning accuracy under load — those predict real throughput better than max velocity.

3D printer build volume: usable ≠ advertised

Printers advertise the travel limits of the axes. What you actually get is smaller by the width of the nozzle carriage, clamps, sensor offsets, and any purge/bucket zones. A "300 × 300 × 300 mm" printer might realistically fit a 260 × 260 × 280 mm part. Some manufacturers already subtract these margins; others don't.

Verification: ask for the maximum printable bounding box for a solid object, not the axis travel. Better yet, request a sample file sliced for the machine and measure the gcode boundaries yourself. It takes two minutes and has saved us from three bad purchases over the years.

Spec claimWhat it often meansAsk for instead
"Max print speed 250 mm/s"Travel move, no extrusionVolumetric flow rate (mm³/s) at quality target
"±0.05 mm accuracy"Single-axis repeatability, unloadedDimensional tolerance on a printed test part
"Compatible with ABS/Nylon/PC"Nozzle temp reaches setpointChamber temp + verified material profiles
"Auto-leveling"Probe existsLeveling method, mesh size, first-layer reliability data

Volumetric flow: the 3D printer spec that actually predicts throughput

If you're buying a printer for production rather than prototyping, ignore mm/s and ask for volumetric flow rate in mm³/s. This combines nozzle diameter, layer height, and extrusion capability into one number that directly tells you how fast you can fill a part. A machine doing 250 mm/s with a 0.4 mm nozzle at 0.2 mm layers moves 20 mm³/s. A machine doing 120 mm/s with a 0.6 mm nozzle at 0.3 mm layers moves 21.6 mm³/s — slower on paper, faster in practice.

High-flow hotends (Voron StealthBurner clones, Mosquito Magnum-class) push 35–50 mm³/s reliably. Standard setups plateau around 12–18 mm³/s. When comparing machines for batch work, this single number matters more than any other line on the sheet.

Resin printers: resolution claims and the XY trap

DLP/LCD resin printers advertise pixel resolution (e.g., "4K," "8K") as if it were a direct proxy for detail. But the relevant metric is pixel pitch at the build plate — the physical size each pixel projects. An 8K screen on a large build area can have coarser effective resolution than a 4K screen on a small one. Calculate: build plate width ÷ horizontal pixels = pixel pitch. Anything under 35 µm is excellent for jewelry and dental; 50–70 µm is fine for general prototyping; above 100 µm starts showing visible layer artifacts on smooth surfaces.

Also verify the Z-axis lead screw backlash specification and the light-source uniformity across the build plate. Edge-of-field dimming causes inconsistent cure depth, which shows up as soft edges on large builds. Suppliers rarely publish this, so ask directly.

Certifications and safety marks: read the fine print

CE, FCC, RoHS logos on a spec sheet sound reassuring. In practice, self-declared CE (no notified body number) for laser equipment is common and nearly worthless. Class 1 laser enclosure certification requires testing to IEC 60825-1 by an accredited lab — ask for the test report number and issuing body. For 3D printers, UL or ETL listing for electrical safety matters if you're installing in a commercial facility; CE alone doesn't satisfy most Western insurance requirements.

We keep a short list of accredited test labs we trust. If a supplier can't provide verifiable certification paperwork within 48 hours, assume it doesn't exist. This filter alone eliminates about 30% of the suppliers in our inbox.

A practical checklist before you sign anything

Print this out or save it. Before approving a purchase based on a spec sheet, confirm each item:

  • Laser: CW optical power at nozzle, wavelength, spot size at focus, galvo field size, f-theta lens model.
  • 3D printer: usable build volume (not travel), volumetric flow rate, chamber temperature range, tested material list with profiles.
  • Both: real-world cycle time for a representative part you supply, maintenance schedule with consumable costs, warranty scope including response-time SLA.
  • Safety: valid third-party test reports matching the exact model being quoted, not a similar product family.

None of this is exotic knowledge. It's just the result of reading too many datasheets and watching too many machines fail to match their promises. The suppliers who welcome these questions are usually the ones worth buying from. The ones who deflect are telling you something important — listen to it.

— the GERMET export desk

Need help evaluating a machine?

Send us the spec sheet and your intended application. We'll flag the red flags for free — whether or not you end up sourcing through us.

Share Your Spec Sheet