A colour dispute rarely starts on the line. It starts at the brief, where a brand names a pantone reference & a printer hears a target that the substrate will not deliver. By the time the first packs arrive the two sides are comparing different things: the brand is looking at a chip under its own office lighting, the printer is looking at a film of ink on a curved, possibly frosted, possibly coloured surface. Both are right, & the run has already been paid for.
This article sets out how colour is actually agreed in cosmetic packaging: what a pantone reference does & does not settle, how the substrate moves the result, how to read a drawdown, the proofing stages worth paying for, tolerance bands & metamerism, the approval environment, & what a signature on a proof commits. It is written for the buyer who signs the sheet, not for the press operator. The print techniques behind the colour are set out in our print techniques guide, & the material choices that move a colour are covered in our packaging materials guide.
Why colour disputes start at the brief
Most disputes trace back to one missing sentence: the brand named a colour, & nobody named the surface it would sit on. A pantone reference describes a colour in the abstract. A pack is not abstract. It is amber glass, or frosted glass, or a natural polyethylene tube, or a metallised board, & each of those pushes the same ink somewhere different.
The second missing sentence is about lighting. Two people can look at the same pack under the same lamp & disagree, & they can look at it under two lamps & reach opposite conclusions from the same print. Without a stated viewing condition, an approval is a memory rather than a record, & a memory is what a reorder will be measured against by someone who was not in the room.
The practical consequence is that the brief has to carry three things before any ink is mixed: the substrate, the viewing condition, & the tolerance. A brief with a pantone number & nothing else is a request for a surprise, & the surprise costs a reprint rather than an email.
What a pantone reference does, what it does not
A pantone reference is a standard. It tells both sides what colour is meant in terms that can be ordered, quoted & checked, & that is genuinely useful. It removes the argument about whether the brand meant a warm gold or a cool gold, which is the argument that eats the most time.
What it does not do is guarantee a match on a given surface. A pantone swatch is printed with a defined ink on a defined paper under a defined condition. A cosmetic pack uses a different ink system, on a different substrate, applied by a different process, & often through a mesh or under heat & pressure. The reference is the destination; it is not the road.
The reference also ages. Chips fade, get handled, pick up dirt, & get separated from the book they came from. A brand that approves against a chip should record the chip's code & its condition, & should replace the book on a schedule, because an old chip quietly becomes a different standard while everyone continues to call it by the same name.
Substrate: the same ink on three different packs
The substrate changes a colour in three ways at once. Its own colour shows through the ink film, its surface texture scatters the light that reaches the eye, & its surface energy decides how the ink lies down in the first place. A single ink on clear glass, on frosted glass & on white coated board will read as three colours to most people in the room.
Opacity is the first lever. On a dark or amber substrate the ink film has to hide what is underneath, which usually means a white underbase beneath the colour. An underbase changes the perceived colour, so a brand that approves a colour without an underbase & then receives it with one has approved one thing & bought another. The underbase belongs in the brief.
Texture is the second lever. A frosted or soft-touch surface diffuses the reflected light, which lowers the apparent saturation & dulls a metallic. A gloss surface reflects directly, which makes the same ink look deeper & slightly darker. This is why a foil that looks correct on a polished cap looks flat on a frosted bottle, & why there is no ink change that fully fixes it.
Drawdowns: how to read one
A drawdown is a sample of the ink pulled across the actual substrate so both sides can see what the ink does on that surface rather than on a swatch book. It is the cheapest honest artefact in the whole colour process, & it is the one most often skipped because it takes a day & looks informal.
Reading one takes discipline. A drawdown should be made with the same ink, at the same film thickness, on the same substrate, & hardened the same way the production pack will be hardened. A drawdown made with a hand applicator lays down a different film from the one a mesh or a blanket will lay down, so it proves the colour & not the process.
The useful habit is to make two drawdowns at different film thicknesses, & to view both. Ink is translucent; a thicker film reads deeper & a thinner one reads lighter. If the two sit far apart, the process has to be controlled on film thickness or the run will drift from pack to pack, & that drift is what the customer eventually notices on a shelf.
Proofing stages worth paying for
A digital proof settles layout, type size & position. It is cheap, fast, & genuinely useful for those three things. It does not settle colour, because the device that produced it reproduces colour in a different space from the one a screen press or a foil uses. A brand that signs colour off a digital proof has signed a picture of a colour.
A press proof, or a short run of the real pack on the real process, is the only artefact that answers the question the brand actually has. It costs a set-up & a slot on the line. It is also the only stage at which a problem can still be corrected without scrapping finished goods, which makes it cheap in context.
Between the two sits the pre-production sample: a handful of packs off the line, made under production conditions, before the full run starts. This is where process drift shows up, because one pack can be made carefully & three hundred have to be made repeatably. A brand with a tight launch date is tempted to skip it, & a brand that skips it finds the drift at goods-in.
Tolerance bands, metamerism
No print matches a reference exactly, so the useful question is how far off is acceptable. A tolerance band answers that, & the most practical form is a numerical colour difference agreed in advance, measured with a named instrument under a named illuminant, rather than a phrase such as close enough.
The band has to be stated with the method. A colour value measured on one instrument with one aperture & one light source is not comparable with a value measured on another. Where both the brand & the printer measure, they should agree on the instrument settings first, because a disagreement about the number is usually a disagreement about the method rather than about the ink.
Metamerism is the trap inside the band. Two samples can measure as an acceptable match under one light & look plainly different under another, because the pigment combinations behind them reflect light differently even though they agree at one viewing condition. The practical defence is to view the approval under more than one source, including the lighting the product will be sold under, & to say so in the brief.
Lighting, viewing booths, the approval environment
An approval made on a desk under a mixed lamp is not repeatable. A viewing booth with a stated light source, a neutral surround & a defined viewing angle makes the same decision repeatable next quarter, which is the entire point. Where a brand has no booth, a printer's booth & a written record of the source used is the next best thing.
The retail environment matters as much as the booth. A shade approved under one condition can read wrong under the lighting in the channel the product is actually sold through, & the brand that never checks it finds out from a returns figure rather than from a proof. Where the product has a known channel, that channel's lighting belongs in the approval.
The record is what survives. A signed approval should note the light source, the reference used, the substrate, the instrument settings if a number was taken, & the date. Five lines of writing convert an opinion into a specification, & a specification is what a reorder can be held to.
Ink film, opacity, the white underbase question
An ink film is translucent by nature. The colour the eye sees is the light that passed into the film, was selectively absorbed, & came back out, so what sits behind the film is part of the colour. A white underbase behind a colour raises its brightness & its apparent cleanliness; a dark substrate behind the same colour lowers both.
The decision has cost attached. An underbase is an extra screen, an extra pass & an extra set-up. It also changes the hand feel of a print, because the film is now two layers rather than one, & a raised edge becomes easier to catch on a carton or a neighbouring pack.
The brand should ask whether the quoted colour was priced with or without an underbase. This is one of the most common sources of a quote that rises after approval, because the printer priced the artwork as drawn & the substrate requires a pass the artwork did not mention.
Where colour drifts after approval
Colour drifts at the reorder, & it drifts for ordinary reasons. An ink batch from a different production lot carries a slightly different pigment load. A mesh that has been cleaned many times passes slightly less ink. A lamp that has aged hardens the film to a slightly different degree, which changes the final shade of a UV system.
It also drifts with the pack. A substrate from a different production run can vary in base colour or in surface energy, & a coated board from a new lot can take ink differently. These variations are normal & are usually inside tolerance, but they are only known to be inside tolerance if somebody measures rather than remembers.
The control is a retained reference & a measurement. A golden sample of the decorated pack, stored away from light, plus a numerical reading taken at approval, gives the reorder something to be checked against. Neither is expensive; both are difficult to reconstruct eighteen months later.
Signing a proof: what the signature commits
A signature on a proof is a release. It usually commits the brand to the colour as shown, to the layout as shown, & to the cost of changing either afterwards. Brands sign quickly because the launch date is close, & then discover that a signature also closes the door on a correction the printer would have made for free a day earlier.
The signature should name what was approved. Approved for colour only, approved for colour & layout, or approved as a production reference are three different commitments, & the difference matters when a later run is compared against the sheet. A proof marked with a single tick leaves the question open.
The brand should also keep a copy of the proof it signed, not only a photograph of it. A scan loses colour fidelity; a print kept flat, dry & out of the light does not. Where a dispute arises, the physical sheet is the evidence, & a photograph of a sheet is a weaker version of the same argument.
Colour control at a glance
The table below sets each proofing artefact against what the brand receives, what it genuinely proves, & the point that is easy to miss. The rows are a working comparison rather than a fixed procedure; the right sequence depends on the process, the substrate & the run size, so agree the stages with the printer at the brief rather than discovering them on the day.
| Proof stage | What the brand receives | What it proves | Easy to miss |
|---|---|---|---|
| Digital proof | A screen or paper render | Layout, type size, placement | Colour is not representative of the press |
| Wet drawdown | Ink pulled down on the real substrate | How the ink behaves on that surface | Film thickness differs from the production pass |
| Press proof | A print made on the actual pack | The process under real conditions | Costs a set-up & needs a slot on the line |
| Pre-production sample | A short run off the production line | Repeatability across more than one unit | Skipped when a launch date is tight |
| Pantone chip | A physical reference standard | The target in terms both sides can order | Chips fade, soil & drift out of date |
| Instrumental reading | A colour value with a stated difference | A number both parties can quote later | Needs the instrument, aperture & illuminant named |
| Signed golden sample | A physical production reference | What every reorder is judged against | Light & handling change it unless stored well |
The rows are a starting map rather than a rule. A brand that names the substrate, the viewing condition & the tolerance at the brief gets a colour it can reorder; a brand that names a pantone number alone gets a colour it can argue about, & the argument is resolved by whoever holds the retained sample.
What to put in the brief
Name the substrate before the colour, because the surface decides whether an underbase is needed & whether a metallic will read at all. Name the viewing condition, since the same print can be approved under one lamp & rejected under another. Then name the tolerance, in a form both sides can measure rather than describe.
- The pack item code, the material, & the surface finish the colour will sit on.
- The pantone reference, or the reference standard the colour is being matched to.
- Whether a white underbase is permitted, & whether it is included in the quoted price.
- The viewing condition for approval, naming the light source & the retail lighting if known.
- The tolerance band, with the instrument, the aperture & the illuminant stated.
- Which proof stages the brand will pay for, & which stage carries the sign-off.
- Whether a golden sample will be signed, retained, & stored away from light.
- Who inside the brand holds the approved reference & answers a reorder query.
Then agree how a reorder is checked. The useful arrangement is a named contact at the printer, a defined response time for a drawdown request, & a confirmation that the retained golden sample is still the reference before a new run is scheduled.
Where our capability fits
We ask for the substrate, the viewing condition & the tolerance before a colour is quoted, & we put the underbase question on the table at that point rather than after artwork approval. Where a finish will not deliver the reference a brand has named, we say so while the artwork can still be changed, & we supply a drawdown on the real pack so the decision is made on the surface rather than on a chip.
We work with brands on private label (your brand on a proven base), OEM/ODM contract manufacturing & full custom formula development, & the colour reference travels with the pack specification from the first sample. Our lines include an emerald facet perfume bottle, a pear drop glass perfume bottle & a citrus glow vitamin face cream jar, each approved against a retained reference rather than a photograph.
Tell us the pack you intend to use, the colour you need on it, & the volume you are planning, & we will come back with what the substrate can carry, which proof stages make sense, & what a workable tolerance looks like. Start at our inquiry page, email capsulebiotech@163.com, or message the right team on WhatsApp: Capsule Series (capsules only) +86 158-7655-5319, Cosmetics Series (serum, hair care, body care, fragrance) +86 158-7655-5319.
General manufacturing information, not regulatory or legal advice. Colour references, proofing stages, tolerance values & the viewing conditions used for approval vary by substrate, by print process & by supplier, & the applicable position should be agreed in writing with the printer before artwork is finalised for production.