Kerf & Press
Search⌘K

Lab status: every product page is marked researched until a machine we bought has been through the test protocol.

Methodology

How we test

Every number we publish will come from a written, versioned protocol with calibrated instruments and raw data. We have published the protocols first; testing starts with our first purchased machines.

Last updated

Published protocols (v1.0)

Principles behind every protocol

  • Measure what sellers care about. Minutes per finished item, cost per print and how a machine behaves in a small room matter more than headline specs.
  • Same conditions for every machine. Same room, same material batches, same test files, room temperature and humidity logged.
  • Repeat, then report the median. At least three runs per measurement. Failed and aborted runs are published too.
  • Calibrated instruments. We list the instrument class and minimum specification in each protocol, and name exact models on results pages.
  • Retail units. We buy machines like you do. We test the EU-voltage version where one exists and say so; if a US-only version is tested on a converter, we note it.
  • Raw data or it did not happen. Every result links to downloadable data and photos of the test pieces.

How scores work

Each protocol produces a sub-score from 0 to 10 using the weights and formula printed on its page. Scores compare machines within the same group (for example, diode lasers with diode lasers). We never show a score without the measurements behind it, and a product gets no overall verdict until it has been tested. No test, no review.

Status labels

Researched
Specs and facts gathered from cited sources. Not tested by us.
In lab
We own the machine and testing is underway. Partial results may appear, clearly marked.
Tested
At least one complete protocol run, with data published.
Discontinued
No longer sold new; kept for reference.

Protocols in draft

These measurements are planned but their protocols are not finished:

Challenge our methods

If you run a shop and think a protocol misses what matters, tell us. Good suggestions become the next protocol version, credited in its changelog. Contact us.