Short answer
Because a file only fixes colour if it names the printing condition it was prepared for; otherwise each printer's driver, profile, paper and process makes its own decisions. RGB files are the most exposed. Even with identical instructions, printing runs to tolerances rather than to zero, so some difference between suppliers is normal — the aim is to shrink it to an agreed, measurable size.
1. Different papers
Test: Ask each printer which paper they used, or print both on the same sheet type. Compare the unprinted paper white first; a bluer or creamier white changes every colour on it.
Paper white, coating and optical brighteners differ between stocks, and each one shifts the whole image.
2. Different colour conversion: each device or driver converts RGB its own way
Test: Send a CMYK file converted to the printer's own named profile, or a PDF/X with an output intent, instead of RGB, and compare again.
An RGB file lets each printer's driver or RIP decide how to convert; different profiles, rendering intents and black generation give different results.
3. Different printing processes and ink sets
Test: Ask whether each job was printed by offset, toner or inkjet, and how many inks. Different processes have different gamuts.
A six- or eight-ink inkjet reaches colours a four-colour press cannot, and toner behaves differently again.
4. Normal press variation within tolerance
Test: Compare several sheets from the same run. If sheets vary among themselves by a similar amount, the difference is within normal process variation.
Printing is controlled to tolerances, not to zero; two conforming runs can still differ visibly on sensitive colours such as greys and skin tones.
5. They are being viewed under different light
Test: Put both prints side by side under the same lamp, then carry them to a window together.
Prints with optical brighteners change more between lamps; comparing one at the office and one at home is not a comparison.
Causes are ordered by how often Colourwise expects each to be the explanation — an editorial judgement, so test them all.
Lay the two prints side by side under one lamp — comparing one in the office with one at home proves nothing. Then look at the unprinted paper. If the paper whites differ, that alone shifts every colour and is the first thing to equalise. If the papers match but colours differ throughout, the conversion differs: the file was probably RGB, or CMYK prepared for a different condition, and each printer converted it their own way. If only the most saturated colours differ, the processes differ: an inkjet with extra inks reaches colours a four-colour press cannot. If differences are small and similar to the variation within each stack of prints, you are looking at normal process tolerance.
An RGB file describes colour for screens; to print it, someone must convert it to the printer's inks. When the file goes to two suppliers, two different conversions happen — different profiles, different rendering intents for out-of-gamut colours, different black generation. Even profiles built from the same characterisation data can differ in how they build black, so the same colour becomes different ink recipes. The remedy is to decide the target yourself: convert to, or tag the file with, a named printing condition that both printers can meet, and send a PDF that carries that intent. Then both suppliers are aiming at the same thing, and the question becomes how closely each hits it.
Papers differ in whiteness and in how much optical brightener they carry, and that changes both how they look and how they measure. Brightened papers look bluer in daylight and less so under lamps with little ultraviolet, so two jobs on different stocks can match in one room and not in another. The industry measurement condition known as M1 was introduced to account for brighteners, and instruments agree much better in that mode on brightened papers than older hand-held ones did. For proofing, the standards ask that the proof paper sit in the same brightener category as the production paper. Name the paper, or at least its whiteness class, in the order.
Paper white matters as much as ink: in the standard offset data sets the aim for paper white moves from neutral to distinctly bluish between older and newer printing conditions.
Printing is controlled to tolerances. The proofing standard, for example, allows process-colour solids on a contract proof to sit up to 3.0 ΔE00 from their aims and warns that tolerances tighter than about 2.5 are impractical because different instrument models disagree by that much. Two suppliers can therefore both be in specification and still differ visibly on sensitive colours such as neutrals and skin. When runs from different suppliers must match, write the tolerance into the order: the printing condition, the paper, the colour-difference formula and the permitted ΔE00, and ask for measurements. Without that, a difference is a matter of opinion.
Why: A less brightened paper, or a different conversion of an RGB file
Fix: Specify the paper and supply a file prepared for a named printing condition
Why: Different processes, papers and driver colour settings
Fix: Accept the office printer as a draft device; approve colour on the production process
Each statement is labelled by kind — established fact, a standard’s requirement, observed market data, a convention, or Colourwise’s own interpretation or analysis — with the strength of the evidence behind it.
FactStrong evidence
More than one ICC profile can be generated from a single characterisation data set, differing for example in black generation, so the same colour can become different CMYK values.
Source: CMYK Characterization Data (registry); ECI offset profiles (colour standards: offset)
StandardStrong evidence
ISO 12647-7:2016 requires proof substrates to match the production substrate within 3.0 ΔE00 and process solids within 3.0 ΔE00 of their aims, and states that tolerances below 2.5 ΔE00 are impractical because of poor inter-instrument agreement.
Source: ISO 12647-7:2016 … — Part 7: Proofing processes working directly from digital data
FactModerate evidence
A comparison of ten commercial spectrophotometers found much better agreement in ISO 13655 M1 mode than between legacy M0 hand-held instruments, especially on media with optical brighteners.
FactStrong evidence
The published paper-white aims differ between printing conditions, from L* 95, a* 0, b* −2 in FOGRA39 to L* 95, a* 1.5, b* −6 in FOGRA51.
Source: FOGRA39 characterization data (coated offset, ISO 12647-2:2004/Amd 1); FOGRA51 characterization data (premium coated offset, ISO 12647-2:2013)
Reviewed 29 September 2026. Colourwise summarises its sources in its own words and does not reproduce standards text or proprietary colour data. Spotted an error? Tell us.