Why Does Bottle Colour Shift Between Batches? Delta E Bands, Master Samples, and Re-Tint Rules

Bottle colour drifts between production batches because of dosing, heat and material variables you can measure. The control is a Delta E band around a retained master sample, plus written rules for re-tinting and re-qualification.

A brand approves a colour, the first lot matches, and three lots later the bottles read a shade off. The reflex is to blame the factory. The useful move is to understand that colour on a moulded bottle is a process output, not a fixed property: the same mold, masterbatch and resin code still produce a measurable shift between lots. What separates a manageable supplier from a chaotic one is not the absence of drift, but a defined band, a retained master sample, and agreed rules for a lot that lands outside the band. This guide covers where drift comes from, how to measure it in Delta E, the master-sample protocol, the re-tint decision, the re-qualification triggers, and what to write on the purchase order.

Why Colour Shifts Between Batches

Colour on a moulded bottle comes from masterbatch, a concentrated pigment carrier dosed into the resin at a let-down ratio of roughly 1 to 5 percent. That small fraction is where drift is born: if the dosing slips by a fraction of a percent, the whole lot moves, because there is no second correction downstream. A feeder slightly out of calibration, or a hopper not purged between colours, shows up in every bottle of the run.

Heat is the second lever. Colour develops as the resin melts, so dwell time and melt temperature change how the pigment reads. A hotter, longer dwell deepens a shade; a cooler, faster cycle leaves it pale and streaky. Two machines running the same mold at different settings will not match, which is why a line move is a colour event, not just a capacity event.

Material inputs matter too. Regrind, the reclaimed runner and rejected-bottle content, dilutes the pigment feed, so a high-regrind lot reads lighter than a lot on virgin resin. Resin lots vary slightly in base colour, and the same masterbatch reads differently on clear PET, white HDPE and PETG because the substrate tints the result. None of this is a defect; it is process noise, and the response is a measured band, not an expectation of zero. Our masterbatch versus pre-coloured resin guide explains how the colouring method changes the drift profile.

Measuring Drift: Delta E Bands

You cannot manage what you cannot measure, and the measurement for colour is Delta E, a single number for the distance between two colours in CIE Lab space. Under 1.0 reads as a match to a trained eye; 1.0 to 2.0 is the usual commercial band for packaging; beyond 3.0 the difference is visible side by side. Our Pantone and Delta E specification guide covers the formulas and illuminants, so the numbers here build on that base.

The measurement has to be controlled or it lies. Read the master sample and the lot sample under the same illuminant, with the same instrument geometry, on the same part of the bottle, because a glossy shoulder and a matte label panel do not read the same. Record the reading with the lot, not just a verdict, because a trend of small drift inside the band warns that the process is walking toward the limit.

Bands should be set by product, not by mood. A premium glass-look jar on a white shelf can hold a tight band; a textured, high-regrind body bottle needs a wider one. Write the band as a number, tied to the master sample, before production starts, so the factory aims at a target instead of remembering one. A trend line of per-lot readings beats a single inspection certificate.

The Master Sample Protocol

The master sample is the reference every lot is judged against, and it fails in predictable ways: it gets lost, it fades, or nobody agrees which sample was approved. The protocol fixes all three. At approval, one physical sample is signed, dated and labelled with the shade reference and the Delta E reading taken against the standard. It is stored sealed, upright and out of direct light, because UV and handling both move a retained sample over months.

Two artefacts beat one. Keep the physical sample for visual judgement and the recorded Lab values for instrumental judgement, so a faded reference cannot silently redefine the approved colour. When the physical sample degrades, the numbers still stand, and a new retained sample can be cut from a verified lot. Store a second set off-site, so a fire or a lost box cannot erase the standard.

Tie the master sample to the lot code, not to the product name. Our batch coding and traceability guide covers how lot numbers link a delivered carton back to its production run, which is what makes the master sample auditable instead of anecdotal. The sample approval workflow guide shows where it sits in the wider approval sequence.

When a Lot Fails: Re-Tint, Re-Run or Concession

A lot outside the band has three honest outcomes, and deciding them in advance keeps a colour problem from becoming a schedule problem. The first is re-tint: add masterbatch and re-run to pull the reading back toward the master. Re-tint works for a small miss, but it is not free, because each extra pass adds dwell time and regrind content, which move colour again. One correction with a re-measure is normal; a third means the process is out of control and the batch should be re-run on fresh material.

The second outcome is a full re-run on virgin resin, which resets the regrind dilution and the thermal history. It costs lead time, so it should be a documented decision with a date, not a loop of small corrections. The third is a concession: the lot sits slightly outside the band, the brand accepts it in writing for a specific order, and the deviation is recorded against the lot code.

What is not an outcome is shipping quietly and hoping nobody measures. A lot shipped against a recorded deviation is a business decision; an unrecorded one is a quality escape that surfaces at goods-in or, worse, on shelf. Our incoming-inspection guide shows how to catch the escape at the dock with a measured sample rather than a glance.

Re-Qualification Triggers

An approved colour is only approved for the conditions it was approved under. Change one of those conditions and the approval lapses until it is re-measured. The four triggers that matter are the resin source or grade, the masterbatch batch or let-down ratio, the mould or machine, and the colouring method itself. Each moves colour through a different mechanism, which is why a visual check alone is not a re-qualification.

A new resin lot is the mildest trigger: it usually holds within the band, but it should still be measured against the master before the full run, not after. A new masterbatch batch is stronger, because pigment strength varies between concentrates even at the same nominal dose. A mould or machine move is the strongest, because dwell time and melt temperature change with the equipment, and the same settings on a different press are a different thermal history.

Write the triggers into the supplier agreement so re-qualification is a routine step, not a negotiation. The rule is simple: any change to material, colourant, mould or machine requires a measured comparison against the retained master before the lot ships. That clause prevents the most common colour dispute in the category, where both sides are right about different lots.

What to Write on the Purchase Order

The purchase order is where batch colour stops being a conversation. Four clauses cover it. State the approved shade reference and the Delta E band around the retained master sample, as a number. Require the per-lot Delta E report, measured against that master under the agreed illuminant, delivered with the shipment rather than on request.

Name the retained master sample as the contractual reference, including who holds it, so there is one standard instead of competing memories. Then list the re-qualification triggers explicitly: resin, masterbatch, mould, machine, colouring-method change. Each requires a measured comparison before shipment, and the report travels with the goods.

Add the handling rule for a failed lot: re-tint once with a re-measure, re-run on virgin resin if the second reading misses, and ship only against a written concession that names the lot and the deviation. At ShiJin we mould PET, HDPE and PETG from 1,000+ open molds with MOQ 5,000 units and 15-25 day lead times, and we hold the approved master sample and report the per-lot Delta E against it, so your colour stays inside its band with evidence attached.

Frequently Asked Questions