Ribbon OEM Colour Approval Governance: Lab Dip, Golden Sample, Tolerance Matrix and Dye-Lot Changeover Rules for Brand Procurement (2026)

Core principle: a colour that is approved is not the same thing as a colour that is controlled. Every point where a decision can silently drift — shade room lighting, dye-lot changeover, finishing heat, packaging time under film — is a point where a signed-off colour can go out of specification without anyone having made a new decision.

Why colour approval fails as a single-step event

Most brand-buyer colour programmes are structured as a linear approval: artwork proof, then lab dip, then a golden sample, then production. Each stage is individually reasonable. The failure is structural rather than procedural — the programme treats colour as a property of a sample, when colour is in fact a property of a process running over time.

A lab dip is a small piece of fabric, dyed on a different machine from the production line, viewed under different light, and assessed against a physical reference that itself has a shelf life. None of those four differences is a defect. All four accumulate. When the bulk roll arrives and the buyer opens the carton expecting Pantone 186 C, what they are actually holding is the compound of every difference — and the reaction, in most programmes, is to blame the mill.

The mill, in most of these disputes, is being handed a colour target it never fully received. If the programme never specified how the colour is to be judged, then the mill is not failing to hit a standard; it is guessing at one, and being judged against the guess.

The five control points most programmes omit

Across the programmes we support, the same five control points account for the large majority of colour-related claims. Each is cheap to specify and expensive to discover late.

1. The viewing condition

A colour decision made under showroom lighting and a colour decision made under a north-facing window are different decisions, even when the same swatch book is in the hand. Specify the illuminant, the geometry and the background. The commercial standard is a D65 daylight source at a defined angle against a neutral grey-to-white card; retail lighting is a different and legitimate condition, and if the product will live under retail lighting, the approval should be reviewed under retail lighting too. Programmes that specify only the physical reference have no way to settle a dispute, because both parties can point to a swatch that genuinely matches under their own conditions.

2. The instrumental tolerance

Eyeball approval is a screening step, not a measurement. A programme needs a numeric limit that defines the boundary between "within specification" and "out of specification", expressed as a ΔE value with a stated illuminant and observer basis (for most work, ΔE*ab at D65/10°). Common commercial thresholds for a single material are around ΔE 1.0 to 1.5; for a multi-material assembly where several components must read as one colour, the effective allowance needs to be tighter per component, because the assembly's perceived colour is a weighted combination. Fixing the number up front, before anyone argues about it, is worth more than any amount of post-shipment photography.

3. The dye-lot changeover

This is the most common point of silent drift. A bulk order split across two dye lots is normal, and a buyer who has approved one lot has not necessarily approved the other. The programme needs to state explicitly: what is the ΔE tolerance between dye lots within the same order, and what triggers a fresh approval? The answer is not always "re-approve everything" — if the mill maintains a retained production sample from the first lot and measures the second against it within the stated tolerance, that is usually sufficient. But the programme has to say so, because in the absence of a rule, the default in practice is "no rule at all", and the discrepancy surfaces at the customer's filling line rather than at the mill's gate.

4. The finishing step

Printing, foil stamping, hot-stamping, flocking and heat-setting all change colour. A dye that is perfectly on target going into a hot-stamp operation may not be on target coming out of it. The correction the mill applies to compensate — a slightly different dye, a different foil, a different temperature band — is invisible to anyone who never saw the pre-finishing state. Specify the finishing parameters alongside the colour target: the substrate, the ink or foil system, the temperature, the dwell, and the resulting ΔE against the same reference used for the base material. This is the single most under-specified area in most programmes, and it is where a majority of the "the bulk colour was wrong" claims actually originate.

5. The packaging time

Small saturated colours shift measurably in the first weeks of light and heat exposure. A ribbon approved against a reference and then photographed for the catalogue, or stored in a hot warehouse before dispatch, may present differently to the buyer than it did at the gate. This is not a mill failure and should not be treated as one. Where the colour is critical to the brand, the programme should specify how long the approval remains valid, what the reference retention conditions are, and whether the buyer or the mill is responsible for colour-protective packaging. Where a brand uses a colour as an identity element, this is often where the identity quietly degrades — slightly, cumulatively, and in a direction nobody is measuring.

A workable colour approval matrix

The following matrix is the format that most consistently survives contact with a real programme. It assigns each stage a defined artefact, a defined instrument, a defined tolerance and a defined authority.

Two features of this table do the real work. The first is that every row names a measurement method, which removes the interpretive argument from the dispute. The second is that the golden sample is sealed and retained, so that both parties are comparing against the same physical object for the entire programme rather than against a description of it. Most colour disputes are not about whether the mill hit the target; they are about which target the two sides think they were measuring.

Retained samples are the cheapest insurance in the programme

A retained sample from every production lot — sealed, labelled with lot number, shade reference and date, stored away from light — costs almost nothing at the point of production and is decisive in every later dispute. It answers three questions that no amount of recollection can answer: what did this lot actually measure, what did the adjacent lot measure, and how far apart were they?

We hold retained samples by default on every ribbon programme we run, and we keep them for a defined period rather than indefinitely, because a retained sample that nobody can locate or read has no evidential value. The label carries the lot, the shade, the measurement value and the date. It is a small discipline with an outsized effect on how colour conversations end.

What to write into the technical pack

If a brand buys ribbon regularly and wants colour claims to resolve quickly, four items in the specification do most of the work:

1. A named reference. A physical Pantone chip plus a mill-produced lab dip that has been jointly approved. The Pantone chip alone is not a sufficient reference for a woven product, because it is a coated paper chip and does not predict a woven textile.
2. An illuminant and observer specification. "D65 / 10°" for instrumental work; a defined booth for visual work. Without this, ΔE numbers from two labs are not comparable.
3. A numeric tolerance. A single number, stated for the base material and, if relevant, a separate number for the decorated assembly.
4. A changeover rule. What triggers a fresh approval and what does not. This is the clause that prevents the most common late-stage dispute.

With those four in place, the large majority of colour disputes become arithmetic rather than argument — which is, in the end, what both sides actually want.

Common failure patterns we see

• Approval by photograph. A photo of a swatch under office lighting is treated as a colour decision. Screens are uncalibrated; phone cameras auto-shift white balance. If photographic approval is genuinely required for speed, it should be explicitly labelled as provisional and followed by an instrumental confirmation against a retained sample.
• One approval for every colour in the programme. A brand that approves a dark navy and then applies that approval across a range of eight colours has approved nothing. Each colour needs its own reference and its own tolerance, however similar the shades look on paper.
• Re-approval fatigue. Where every dye-lot change forces a full re-approval cycle, programmes slow down and buyers start skipping the approval step altogether — which produces exactly the drift the process was meant to prevent. This is why the changeover rule matters commercially, not just technically.
• Colour specified after the material is chosen. Some shades cannot be achieved on all constructions. Specifying colour before construction forces the mill to find a route; specifying it afterwards forces the buyer to accept the nearest achievable result, which is not the same thing and is where a lot of quiet dissatisfaction originates.

Working with your mill on colour

The strongest colour programmes in practice are collaborative rather than supervisory. The mill's colourist and the buyer's QC engineer agree the reference, the instrument settings and the tolerance before the first lab dip is cut, and both sign the golden sample. The buyer's role is to specify the commercial requirement and the market tolerance; the mill's role is to tell the buyer, early, which of their intended shades will be difficult on the chosen construction and substrate — and to say so at the point where a change is still cheap, not after the artwork is locked and the tooling is cut.

We have seen the difference repeatedly. Where colour governance is written down and shared, a colour question that used to take three weeks and a letter of intent now takes an afternoon and a spectrophotometer reading. Where it is not written down, the same question takes a season.

Requesting a colour approval pack from us

If your current programme does not have a documented colour tolerance, we can produce a colour approval pack for your next ribbon order at no charge. It includes a jointly approved lab dip on your specified substrate, a retained golden sample, a measurement record with instrument settings, and a dye-lot changeover clause you can paste straight into your specification. Send us your Pantone references, substrate and construction, and the end application, and we will cut the lab dips against it.

Ribbon OEM Colour Approval Lab Dip Pantone Matching Quality Control Brand Procurement

Frequently asked questions

What Delta E tolerance should I specify for ribbon colour approval?

For most single-material woven ribbon programmes, Delta E*ab of 1.0 to 1.5 against an approved lab dip measured at D65 with a 10-degree observer is the usual commercial range. For decorated assemblies where several components must read as one colour, the per-component allowance should be tighter because the assembly's perceived colour is a weighted combination of its parts. Specify the instrument basis alongside the number, because Delta E values measured under different illuminants or observer angles are not directly comparable.

Is a Pantone chip enough as a colour reference for ribbon?

No. A Pantone chip is a coated paper chip and does not predict how a woven textile will read, particularly in saturated shades and in metallic or lustre finishes. The reference should be a mill-produced lab dip on your actual substrate and construction, jointly approved and retained. Use the Pantone chip as the starting intent, not as the approval artefact.

When does a dye-lot change require a fresh colour approval?

That is a decision the programme should specify in advance rather than discover mid-order. The workable rule is that a retained production sample from the first dye lot is measured against the new lot; if the difference is within your stated tolerance the change is accepted with notification, and if it is outside it, a fresh approval is triggered. In the absence of a written rule, discrepancies tend to surface at the customer's filling line rather than at the mill's despatch gate.

Does finishing really change the ribbon colour enough to matter?

Yes, and it is the most under-specified area in most colour programmes. Printing, foil stamping, hot-stamping, flocking and heat-setting each shift the appearance of the base colour, and mills typically compensate with a slightly different dye or process adjustment that nobody outside the mill sees. Specify the substrate, ink or foil system, temperature, dwell and the resulting Delta E against the same reference used for the base material, and most late-stage colour claims disappear.

How should colour be approved for photographs, packaging and store displays?

Approve the physical sample, and treat photography as documentation rather than as the decision. Screens and phone cameras are uncalibrated and auto-shift white balance, so a photo-based approval records how the swatch looked on one device on one day. If speed requires a provisional photographic sign-off, label it as such in the specification and follow it with an instrumental confirmation against a retained sample before bulk production is released.

Who should hold the retained samples and for how long?

The mill should retain a sealed, labelled sample from every production lot, carrying the lot number, shade reference, measured value and date, stored away from direct light. Both parties benefit: the mill can evidence what a lot actually measured, and the buyer can compare an adjacent lot against it rather than against a description. Retained samples should be kept for a defined period rather than indefinitely, since a sample nobody can locate or read has no evidential value.