Why the re-order never looks like the first batch

Colour drift is not carelessness. It is the sum of four variables, each moving inside a tolerance nobody wrote down. Put the tolerance in writing and the second order matches the first.

Published by JPChairs · Chair manufacturer in Foshan, China · 2026-10-16

Where colour drift actually comes from

Four independent variables move before a chair is packed, and each one is normal on its own.

• Dye lot. Fabric is dyed in lots of 300 to 500 m, and two lots of the same reference routinely differ by Delta E 0.8 to 1.5 before a single panel is cut.
• Mill run. One colour reproduced 12 months later, on a different machine with water at a different temperature, lands 1.5 to 3.0 Delta E from the original.
• Substrate. Open-grain oak takes stain 10 to 15 per cent darker than beech in the same bath, and one frame can contain both species.
• Application. Spray viscosity, a room held between 18 and 28 degrees C, and a coat of 80 to 120 microns all shift the final reading.

None of these is a defect. They become one when the tolerance is undefined and two batches end up in the same room.

Delta E: the number to put in writing

Delta E is the measured distance between two colours. It is the only colour tolerance worth specifying, because it is measured rather than argued.

• Below 1.0 - imperceptible to almost every observer. The target for chairs standing in the same room.
• 1.0 to 2.0 - visible to a trained eye side by side, invisible across a dining room. Acceptable for contract work.
• 2.0 to 3.0 - visible to most people in direct comparison. Acceptable only between separate rooms or floors.
• Above 3.0 - a complaint waiting to happen. Reject it.

Measure with a spectrophotometer, not a phone camera and not a swatch book under warehouse lighting. Read at three points on the seat and two on the back, at 4000 to 5000 K, and average the five. A single reading on one leg tells you nothing.

Write Delta E 2.0 or less against the retained reference for fabric and leather, and 3.0 or less for stained wood, where the substrate itself introduces variation no finisher can remove.

Dye lots and batch cards

Fabric arrives in lots and the lot number is printed on the roll end. It is the most useful piece of information in any re-order.

• Fit one lot to one order. A 300-chair restaurant order uses 180 to 240 m of fabric, which fits inside a single 300 m lot if it is planned before cutting.
• Sequence the cutting. Cut all panels for one room from the same roll and mark the roll number on the batch card.
• Hold a reserve. Keep 3 to 5 per cent of the lot, roughly 6 to 12 m, for repairs and for the re-upholstery that will be needed in year three.
• Record it. Date, lot number, machine, operator and a retained swatch pinned to the card. This is what makes a claim traceable 18 months later.

Wood and stain are a different problem

Wood is the harder case, because you are staining a material that varies within a single board.

Ash, oak and beech take the same stain differently. Open-grain oak reads 10 to 15 per cent darker than beech in an identical bath, and variation within one frame can reach Delta E 2.0 before any finishing error at all. Sapwood and heartwood from the same board differ by more again.

Three controls do most of the work: sort frames by timber before finishing and run each species as a separate batch; spray a test panel from every new stain batch and read it before the line starts; and hold the finishing room between 18 and 28 degrees C with humidity under 65 per cent, because a 10 degree swing changes dry time by 20 to 30 per cent and takes the colour with it.

Where a room must be uniform, a pigmented lacquer rather than a penetrating stain roughly halves substrate variation, at the cost of hiding some grain.

Re-ordering a colour 12 months later

Re-ordering the same colour a year on is the most common way to create a mismatch, and it is preventable at almost no cost.

• Quote the batch card reference, not the colour name. Names like walnut are reproduced differently by every mill.
• Send a physical swatch. 10 x 10 cm is enough. Mills match to a physical reference at Delta E 1.5 to 2.5, and to a name at 3.0 to 5.0.
• Allow 15 to 25 days of mill lead time for a colour not held in stock, on top of a 300 to 500 m lot minimum. Below 300 m, buy the lot and store it rather than re-dyeing.
• Plan the join. If new chairs will sit beside old ones, put them in a different zone. A 2.0 Delta E is invisible at 5 m and obvious at 0.5 m.

The four lines for the purchase order

Most colour disputes are settled by four lines that were never written.

• Reference. Retained sample number or swatch, physically attached to the purchase order.
• Tolerance. Delta E 2.0 maximum for fabric and leather, 3.0 for stained wood, measured at five points under 4000 to 5000 K light.
• Lot control. Single dye lot per order, lot number on the batch card, 3 to 5 per cent cutting reserve retained.
• Rejection terms. Who pays for a batch above tolerance, and whether it is re-made, re-covered or credited. Agree it before production, not after.

These tolerances are applied on the solid wood chair finishing line, and the finishing and inspection process behind them is on the factory finishing lines.