Proofing when you cannot afford a proofing system

Four things called a “proof”, and they are not equal

• A soft proof: your screen, simulating the destination profile. Free. Tells you which colours are unreachable; tells you almost nothing about how the sheet will look unless your screen is profiled.
• A digital proof: an inkjet proofer calibrated to simulate a specific press profile, printed on proofing stock. This is what “contract proof” usually means. It carries a colour bar and often a signature line. Costs money; catches nearly everything.
• A press proof or wet proof: a short run on the actual press, actual ink, actual paper. Expensive, usually reserved for packaging and long runs. The only thing that shows you the real substrate and the real ink film.
• A machine proof or “first-off”: the first good sheet of the actual run, approved at the press. Free if you can be there; the strongest thing available to a small job.

If your budget allows one, take the machine proof and go to the press.

Calibrating a screen without buying anything

A profiled screen is the cheapest large improvement available, and there are three tiers.

1. Borrow a colorimeter. An X-Rite i1Display or Datacolor Spyder takes ten minutes. Many print shops, colleges and photography clubs have one. DisplayCAL with ArgyllCMS is free, open source, and better than most bundled software.
2. Rent or share one. A group of freelancers buying one between six is entirely normal.
3. If you cannot: set your monitor to its sRGB preset, set brightness to about 120 cd/m² (roughly 40 to 50% on most panels, which will feel too dark at first), turn off any dynamic contrast or “vivid” mode, and work in a room with consistent light. This is not calibration and it removes the worst of the variation.

Do not use a phone or a laptop with an automatic brightness that changes through the day for any colour decision.

The phone-photo trap

A client photographs the printed sheet and sends it saying the colour is wrong. That photograph has been through: the room’s illuminant, the camera’s automatic white balance, its tone curve, HDR processing, JPEG compression, and their screen. It is evidence of almost nothing.

The reply that resolves this, politely:

Photographs of print are not reliable for colour. Can you hold the sheet next to the printed reference we agreed and tell me whether they differ to your eye, under daylight or a 5000K lamp?

That reframes the conversation onto the only comparison that means anything: two physical objects under one light.

Getting a free printed reference

Shops will generally give you, if you ask:

• Their CMYK tint chart on the stocks they run. The single most useful thing in this lesson.
• A paper dummy: the actual stock, folded and trimmed to size, unprinted. Free, and it settles arguments about weight and feel before anything is printed.
• Samples of previous jobs with colours near yours.
• A “gang run” slot: some shops let a small job ride along on a larger sheet at low cost, which gets you a real printed sample for a fraction of a normal run.

Approving a proof properly

Whatever the proof is, approve it under a known light, mark it, and keep it.

1. View under D50, or failing that north daylight, never under a warm tube.
2. Check against the reference, not against your memory of the screen.
3. Mark corrections on the physical proof, in a colour that is not in the job.
4. Sign and date it, photograph it, and keep both the photograph and the exact PDF you supplied.

That last step is not bureaucracy. When a run comes back wrong, the question is always whether the file was wrong or the press was, and the only way to answer it is the artefact you both agreed on. Handling that conversation is a later lesson; having the evidence is this one.

What you can honestly promise

You can promise: the file is built correctly, the colours are specified against a named reference, and a supplied proof matched. You cannot promise an exact colour across two suppliers without spot inks, or an exact match between a screen and a sheet. Saying this at the start costs nothing and removes the single most common source of dispute in print work.