Before the camera existed, there was exactly one way to get the image of a painting hanging in someone else's house — or in a church you'd never enter again, or in a private collection an ocean away. You paid a painter to go copy it, by hand, in the same medium as the original. No shortcuts. No Xerox machine waiting in the wings.

This wasn't some quaint side hustle. In the workshops of the great masters, junior painters got put to work copying finished pictures for two reasons at once: it taught them how the older hand had solved a problem of drapery, or light, or a stubbornly foreshortened arm, and it produced a second, sellable version of a picture the workshop had already been paid for once. Smart business, honestly — the Renaissance version of a re-run. A patron who admired a painting he couldn't buy, because it belonged to a church or a rival family or was simply not for sale, commissioned a copy instead, and expected that copy to be painted, not printed, because printing barely existed yet and couldn't render color at all.

So here's the situation every generation of copyists since has inherited: a buyer wants a picture that looks, at a glance and from across a room, like the picture being copied — but painted with real pigment, by a real hand, at a price somewhere below the price of an original commission. That last condition is the one worth sitting with. It's the condition every generation since has had to solve over again, and most of them cheated.

The problem two generations of copyists could not solve

Once photography showed up, and then color printing right behind it, anyone could get an accurate flat image of almost anything, cheap and fast. Hand-copying, meanwhile, stayed exactly what it had always been: one trained person, one brush, one canvas at a time — no cloning, no assembly line that actually works. A single skilled copyist can only crank out so many finished pictures in a month, and the price of a copy is, when you boil it down, mostly the price of that person's working hours.

That's a real bottleneck, and it produced a real problem. A faithful hand-painted copy stayed, for a very long time, either the province of patrons wealthy enough to commission it one at a time — which kept the price sky-high — or something a workshop had to rush, which is exactly where hand-copying earned its bad reputation: coarse, hurried work a trained eye spots from across the room the way you'd spot a bad toupee. The question anyone selling hand-painted reproductions to an ordinary buyer, rather than a wealthy patron, had to answer was this: how do you keep it genuinely hand-painted — real oil, a real brush, real relief on the surface — while dragging the price down to what an ordinary household pays for a piece of furniture, without turning it into the kind of coarse, assembly-line junk that gives the whole category a bad name?

Looking closely at what a hand-painted copy actually gives you

Let's look at one up close. Not a famous picture on a museum wall — I don't have one in front of me right now, so there's nothing honest I can say about it — but the plainer, more useful object: an ordinary hand-painted reproduction canvas, the kind any competent workshop turns out today, sitting on a table.

Tilt it so light from a window rakes nearly flat across the surface instead of falling straight onto it. That's called raking light, and it's the single most useful trick a person can do with an unfamiliar picture — like checking a used car's paint job under a flashlight instead of the dealership's fluorescent lights. Look along the edge of a tree trunk, or the fold of some drapery, wherever the brush laid down a thicker stroke. The light catches a genuine ridge there and throws a thin shadow beside it, exactly the way it would across the ridges of a fingerprint held up to a lamp. That ridge isn't an optical trick and it isn't painted-on shading. It's paint, standing a fraction of a millimeter proud of the canvas, exactly where a brush loaded with pigment got dragged across the surface and lifted away.

Now run a finger, lightly, across that ridge, and then across a stretch of flat sky beside it. They feel different — the sky is comparatively smooth, the ridge catches under your fingertip like a scar. They feel different because a person built the ridge up, stroke by stroke, judging each pass by eye, and stopped when it looked right — not because a machine hit the edge of a programmed run, which would leave every square inch of the surface, sky and tree alike, exactly as textured as every other, the way a printed placemat never varies no matter how many times you run your thumb over it. Flip the canvas over. The back is bare woven cloth, stretched over a wooden frame and fixed at the edge with staples, sometimes faintly stained where oil has soaked through from the front over months of drying, like a coffee ring on a napkin nobody bothered to wipe. None of this — the ridge, the give of the weave, the unfinished bare back — belongs to a printed sheet, however good the printer that made it. It belongs only to a thing that was, at some point, genuinely wet.

What the workshop system worked out, and what it costs

None of this is actually a new solution, and I find that oddly comforting. The great painting workshops of earlier centuries already answered the scale problem by dividing labor: one hand for drapery, another for landscape backgrounds, another for faces, the master reserving the hardest passages for himself and signing the result like a head chef plating the dish somebody else cooked. A modern reproduction workshop solves the price problem the exact same way — specialization, and a price built from actual inputs instead of a single bespoke negotiation every time.

The price of a copy is set, in practice, by two things: the area of canvas to be covered, and the difficulty of what's being covered. The area part is simple enough to write as a rule: a base charge, plus an amount that grows with the surface area of the canvas — though not quite in step with it, because covering a very large canvas isn't proportionally as hard as covering a small one, so price grows a little slower than area does. On this basis, the sizes commonly ordered come out roughly like this, before any frame or options get added:

The difficulty part explains the rest of the price list. A portrait with a second face costs more than the same canvas with one face, because a face is where a copyist's judgment gets tested hardest — a mouth or an eye a hair's-breadth off reads as wrong to anyone alive, in a way a slightly-off leaf on a tree never does. So a workshop typically tacks on a fixed amount for every face beyond the first. Assigning the work to a more experienced painter costs more again, for the plainest possible reason: an experienced hand works faster and makes fewer mistakes, and that's exactly what you're paying for.

And because oil paint doesn't merely dry the way water does — it cures, through a slow chemical reaction with the air that can take anywhere from six months to two years to work all the way through a thick passage of paint, even though the surface feels dry to the touch within a couple of days — a workshop that varnishes too early traps solvent underneath the varnish and stores up a problem for later, the way sealing a deck before the wood's finished drying just locks the moisture in. A careful workshop either ships the canvas unvarnished, with instructions, or waits out the cure before adding a final coat. That's the short version of a chemical process that takes a lot longer to explain properly, but it's enough to show why "the painting will be ready in six months" is sometimes an honest answer, not a slow one.

Three questions worth asking before commissioning one

How long will it actually take, and why won't the workshop ship it varnished right away?

Because touch-dry and cured are two different things. Oil paint feels dry to a finger within a day or two, but the chemical process that hardens it all the way through — oxygen from the air slowly cross-linking the oil into a solid — takes months, sometimes close to two years in a thickly painted passage. Varnish applied before that process finishes traps solvent beneath it and causes trouble later, so a reproduction that arrives unvarnished, or lightly varnished with instructions to finish the job after six months, is not a sign of haste. It's closer to the opposite.

Are we paying for the size of the canvas, or the skill of the painter?

Both, and it's worth separating them before you order. The base price tracks the area — a larger canvas simply takes longer to cover. On top of that sit charges that track difficulty and skill specifically: an amount per additional face in a portrait, and a separate charge for assigning the work to a more experienced painter. If you want the best possible likeness of a face, rather than simply the largest possible canvas, you're better off spending on the second kind of charge, not the first.

How do we know the canvas that arrives is the one we ordered?

Because no two hand-painted copies are ever identical — this is the one real question a printed reproduction never has to answer at all. The plain remedy is to ask, before the canvas ships, to see a photograph of the actual piece rather than a stock image of the picture being reproduced. A workshop confident in its own work will show you this without being pressed, and one that hesitates just told you something.

The new problem this solution creates

Because a person, not a machine, executes every copy, no two copies of the same picture are ever quite identical. Colour mixed by eye on a Tuesday won't be the exact colour mixed by eye on a Friday; a passage might get revised a little further on one canvas than another. For a buyer who wants a warm, competent likeness on the wall, this hardly matters. But it raises a question no printed reproduction ever has to face: ordering from a photograph of a workshop's past work, rather than standing over the painter's shoulder, how do you know in advance what will actually show up at your door?

Where this leaves the next problem

That variability is precisely the problem a printing process, rather than a paintbrush, was built to solve. A printer produces the same image, to the same spec, every single time, because a machine executes it instead of a hand. What you give up in that trade is surface, relief, the singularity of a thing no two copies share.

Related on Mus3ums.com

• A longer case for why the ridge under your finger matters at all: Beyond the Print.
• What a workshop's "museum quality" claim should actually mean before you pay for it: Timeless Elegance.

Sources

• The Six-Month Rule: When Is It Safe to Varnish Oil Paintings? — MontCarta
• How Long Does Oil Paint Take to Dry? — Fine Art Tutorials
• Comprehensive Characterization of Drying Oil Oxidation and Polymerization — PMC / NIH
• Is My Painting Real? Original or Print — Fine Art Restoration Company
• Prints versus Paintings: 8 Top Ways to Tell the Difference — WorthWise Appraisers