Who Invented the Printing Machine? What It Taught Me About Buying a Mazak CNC Lathe

Ask “Who invented the printing machine?” and you'll get one of two answers. The common one: Johannes Gutenberg, around 1440. The more precise one, if you care about manufacturing quality: Friedrich Koenig, who built the first steam-powered printing press in 1814 and put it to work at The Times. Gutenberg's press was a hand-operated screw device. Koenig's press could run 1,100 sheets an hour and produce the same result, sheet after sheet. That difference—not the invention itself—is what made printing an industry.

I think about that distinction every time someone hands me a machine spec sheet. I believe most buyers choose machines by nameplate specs when they should choose by evidence of repeatability. A machine that can do a thing once is a prototype. A machine that can do it a thousand times is a production asset.

What a spec sheet doesn't tell you

I'm a quality compliance manager at a machinery and digital printing company. I review every deliverable before it reaches a customer—roughly 200 units a year. In 2024, I rejected 11% of first deliveries. Most rejections weren't caused by major failures. They were caused by gaps between what the sales sheet promised and what the machine actually did in production.

Take “Mazak” as an example. I've tested Mazak lathes, turning centers, and a Mazak 5 axis CNC machine for customers who needed complex parts in one setup. The Mazak machines generally pass because their documentation lists limits as well as capabilities. But even a brand with that reputation needs proof. I've seen a Mazak 5 axis CNC machine come off a truck with a worn spindle taper that should have been caught before it left the factory. The name on the paint didn't matter. The acceptance test did.

A Dutch buyer once asked me why so many searches include “CNC draaibank Mazak.” I told him that I've lost count of how many spec sheets promise “newer control, improved accuracy” without a single test behind it. “Draaibank” just means lathe. The word choice doesn't solve the trust problem. The process does.

The same thing happens on the printing side. An inkjet printing machine might offer 1200×1200 dpi, nine ink channels, and a feeder speed that looks great in a brochure. But if the test print looks perfect and the production run has a half-inch color band every forty sheets, the marketing numbers don't mean much. I've watched this happen on a $75,000 wide-format project. The vendor offered to replace the printheads. The real issue was the firmware's handling of media thickness—a detail hidden in the fine print.

Vinyl sticker printing is an even better case. Customers order a vinyl sticker printing machine for die-cut decals, labels, or signage. They compare speed and resolution. The first question should be: “What is the smallest color error you'll accept before the roll is flagged for waste?” If the answer isn't numeric, you're not buying a production machine; you're buying someone else's learning curve.

For color, the numbers exist. According to Pantone's color matching system, the industry standard for brand-critical color is Delta E < 2, and a Delta E of 2–4 is noticeable to trained observers (Source: pantone.com). In one audit, we measured a sample from a new inkjet printing machine at Delta E 3.8 against a standard graphic—noticeable in a side-by-side comparison and unacceptable for a national retail launch. The printer was “within the manufacturer's tolerance.” We rejected it. Rework cost us $18,000 and pushed a client's launch back three weeks.

Koenig wouldn't have tolerated that ambiguity. His presses were designed for identical output, not for occasional perfection.

Transparent pricing is a proxy for transparent production

This is where my view on pricing comes in. I've learned to ask “what's NOT included” before “what's the price.” A vendor who lists all fees upfront—even if the total looks higher—usually costs less in the end. A vendor who hides the inspection, setup, training, and spare bearings until after the deposit isn't just annoying. They're showing you how they think about accountability.

The old “buy local, inspect in person” thinking comes from an era before standardized acceptance tests and video-run results. Today, an overseas supplier can send you a 20-minute run test with tool breakage data. The data is more useful than a factory tour. I'll still inspect in person when the risk is high, but I no longer believe that geography guarantees quality.

If you don't write rejection criteria into the purchase order, the vendor's quality system is the one that counts. That's a gamble I've stopped taking.

So, who invented the printing machine?

Before I wrap up, let me answer the question directly. If you ask “who invented the printing machine?” the safest answer is still Johannes Gutenberg. He invented the movable-type printing press around 1440. But the machine that made printing a repeatable industry process was invented by Friedrich Koenig in 1814. That difference is the whole point of this article.

I hear the counterargument all the time: “We can't afford to pay for a brand name. We have to take the lower quote.” I'm not here to defend brand premiums. I've got mixed feelings about them. Part of me wants to say brand shouldn't matter, and every machine should stand on its own test results. Another part knows that a Mazak lathe from a good reseller still holds tolerance because the company has invested in verification for decades. I reconcile those two feelings by treating brand as a filter, not a finish line. The brand reduces risk only when the acceptance criteria are written into the contract.

A CNC lathe, a Mazak 5 axis CNC machine, a vinyl sticker printing machine, or an inkjet printing machine: they all need the same treatment. Look at the machine's limits, ask what happens on the 1,000th part, and put the pass/fail criteria in writing. That's how you buy trust. That's also, by the way, how you honor what Koenig actually invented.