Fabric That Reads Gray Under Warm LEDs: Approval Conditions That End the Dye Lot Argument

When a fabric reads gray in the client’s room but matched in your showroom, you are looking at three different causes wearing one coat: the light source, the dye lot, and the room around the fabric. Set approval conditions that let each one be proved or dropped on its own. Approve a signed physical cutting from the intended production lot, not a showroom swatch. Name the illuminants you will judge under, and put the client’s own lamps on that list. Record the lamp specification, the sample size, the viewing angle and the adjacent finishes. Agree a tolerance together with the method that produces it, and agree what happens if delivered goods fail it. Then the dye lot stops being an opinion. It becomes a lot number, a measurement and a signature.

Separate the light from the lot before you write anything

Two fabrics can agree under one light and disagree under another. That is metamerism: the measured color coordinates match, but the spectral curves behind them do not, so changing the light source pulls the pair apart. A showroom running neutral daylight-type lamps and a living room running warm LEDs are two different light sources, and a metameric fabric will move between them. Nothing was dyed badly for that to happen, and in a large share of these complaints the dye lot was never the cause.

Dye lot variation is real, and it is a separate problem. Batches dyed at different times can drift a little because of dyebath chemistry, temperature, fiber absorbency and machine conditions, which is why mills keep retained samples and hold lots apart. The only control a buyer has is contractual: name the lot the goods must come from, and require that every visible panel in a run comes from it. When a complaint arrives, the useful question is not whether dye lot variation exists as a phenomenon. It is whether the delivered goods measure outside the tolerance against the master that both sides signed.

The third cause gets blamed on the lot because it leaves no paperwork: context. Color is judged by comparison. A warm gray sits cooler beside oak and warmer beside marble, pile and nap direction change how a surface throws light, and the same fabric reads differently flat on a table, upholstered over foam, and hanging in folds. So run the one test that separates the causes before you redraft anything. Take a single cutting, view it in the showroom condition, then carry that same cutting into the client’s room under the client’s lamps. If the shift follows the light while the lot stays the same, the light is the cause. A textile dye lot cannot travel between rooms with you, but a swatch can, and that comparison usually settles the question in ten minutes.

What goes on the approval sheet, field by field

Most color arguments are not knowledge failures, they are record failures. Explaining metamerism to a client is easy and does not help at all when the approval file says “approved under showroom lights”. Build the sheet so it answers the argument before the argument starts. An industry standard fabric approval process, whatever a mill calls it on its own form, comes down to these fields:

• Fabric identity: supplier code, color name, fiber content, construction, finish, pile or nap direction, roll width.
• The physical standard: one retained cutting, signed and dated by both parties and initialled across the face so it cannot be swapped later, with a copy held by each side.
• Dye lot control: the lot number the order must come from, the roll numbers listed on the delivery documents, and the rule that every visible area of one run comes from one lot. If more than one lot cannot be avoided, each lot is approved separately before cutting.
• Lighting conditions: every illuminant you will judge under, named. A controlled daylight source such as D65 for the shop check, and the client’s installed lamp by make, model, correlated color temperature and rendering information, plus the dimmer setting.
• Viewing conditions: the sample size, which should be a full width or installation sized panel rather than a card; the viewing distance and angle; whether the check happens in daylight, artificial light or both; and whether the fabric is seen flat, upholstered or draped.
• Surroundings: the adjacent finishes the fabric will sit beside, listed, because an isolated match is not a match in a room.
• Tolerance: the pass or fail rule, in visual language and, where the project warrants it, in instrument numbers with their method attached.
• Expected shift: one sentence the client signs, recording that the fabric will look different under light sources that were not part of the approval.
• Remedy and timing: the inspection window, the evidence needed to raise a claim, the testing route, and who pays for what.

Photographs belong in the file as support and never as the standard, because two phones will white balance the same wall into two different colors. The signed cutting is the standard. Everything else is commentary. If the decision is leather rather than a woven fabric, run the physical checks before the color conversation, the way the bench tests for look alike leather swatches do, so the swatch a client signs is one that has already survived something.

Approve under the client’s real lamps, not a showroom proxy

Showroom light is usually chosen to flatter product, which makes it a poor proxy for a bedroom at nine in the evening. The fix is not to make the showroom gloomier. It is to add the destination light to the approval and to write down what that light actually is. “Warm white” is not a specification. A 2700K lamp and a 3500K lamp can push the same fabric two different ways, and two lamps at the same temperature can still render it differently when their spectra differ. That is the same mechanism behind late metamerism surprises in fabric selection and interior design handovers, and it is why specifying lighting conditions for fabric color matching belongs in the approval sheet rather than in the memory of whoever stood in the room.

Practical sequence: ask for the client’s lamp details before the order, not after delivery. If the lamps are not chosen yet, record the intended temperature and dimming range and treat that as the provisional condition, then re-check the panel once the lamps are in. View a full width panel in the room at the hour the room is used and at the dimmer setting it is used at. A hotel project needs this area by area, because a lobby, a corridor and a guest room rarely run the same lamp, and one approval signed under lobby light will not cover the floor above it. A lightbox on a D65 setting can stand in for daylight when the client’s windows face the wrong way, but it cannot stand in for the client’s evening scene.

Turn the visual decision into a number, then into a remedy

A visual decision is where the argument starts, because perception is easy to dispute and hard to recall. Instrument readings are harder to argue with, provided the method travels with the number. A spectrophotometer reads the fabric and reports coordinates in a standard space, usually CIE L*a*b*, plus the difference against the master as a delta E value. That is useful. A bare number is not. The formula used, the illuminant, the observer, the instrument geometry, the number of fabric layers and the backing all change the result, so a line such as “delta E under one” is not a specification. It is a trap that both sides will read their own way six months later.

Two rules keep it honest. Name the method: which instrument, which formula, which illuminant and observer, single or multiple layers, on what backing, measured by whom, and where the report is filed. Then pair the instrument with a visual rule so the two have to agree, such as “no objectionable difference under any named illuminant”, or an agreed grey-scale grade for color difference of the kind the fastness standards describe. Best practice for fabric sample approval is to let the instrument and the eye disagree in the open, so if the instrument passes a lot and the eye rejects it, that conflict surfaces before the yardage is cut rather than after it is upholstered onto thirty frames.

Tie failure to a remedy, because a tolerance without a consequence is only a description. The order should say that goods failing the agreed limits may be rejected or replaced at the supplier’s cost, including the labor that unwinds the mistake: cutting already done, freight both ways, removal from frames, reupholstery and reinstallation. It should bar installation of disputed material before the dispute is resolved, give a defined inspection window after delivery, and keep batch-wide or concealed non-conformity claimable when it only becomes visible once the pieces sit together in the room. If a fabric has to change part way through an order, that change needs its own written window rather than a verbal yes, which is the lesson behind a mid production fabric change that was never documented. And refuse the sentence “color may vary by dye lot” when it stands alone. That phrase is a control to be used with a lot number and a tolerance, not a waiver that cancels both.

What a Foshan sourcing trip should confirm about color control

Buyers sourcing upholstered furniture in Foshan are usually three light environments away from a clean answer. The fabric is dyed and finished at a mill, cut and upholstered in a workshop, and presented in a showroom, and each of those rooms has its own lamps. The approval sheet has to survive the whole chain, which is the practical reason a Foshan-based source factory with its own showroom, such as EverGoCasa, is worth walking rather than photographing: you can put the same cutting under the showroom lamp and under a warm lamp in one visit, with the people who will control the lot standing next to you.

Six questions turn a factory visit into evidence. Where is the approved cutting kept, and who holds it when the repeat order comes round next year? Which lamp do the showroom samples sit under, and can it be switched to a warm source for ten minutes? Was the color approved after finishing, or only on the greige or the lab dip, given that coating, backing, FR treatment and calendering all move the shade? Will the first completed production lot be presented as a full width cutting before any panels are cut from it? What lot and roll numbers will appear on the delivery documents, and what happens if the mill runs out mid-order? At the cutting table, is the operator judging shade under the same light as the showroom, or a different one?

If the trip is built around a Canton Fair visit, timing matters. Fair weeks compress decisions, and approvals made in a hotel meeting room under cool ceiling fluorescents are exactly how a warm-room failure gets signed off. Carry the physical standard, the client’s lamp specification and the room’s likely hour of use with you, and treat the showroom visit as the place to run the two-light check rather than the place to fall in love with a color that only exists under showroom light.

Write the clause, and strike out the escape hatch

A clause that does the work, in plain language you can adapt: “The approved color standard is the signed physical cutting attached to this order, identified by fabric code, color, construction, finish and direction. The supplier shall deliver fabric from the recorded dye lot, matching the standard under the agreed daylight source and under the client’s installed lamps as specified, using the stated visual and instrumental methods. No change to dye lot, dye formulation, fiber, finish, mill or process is permitted without written approval, and any replacement lot shall be re-approved under all agreed illuminants. A statement that a difference is caused by dye lot, lighting or metamerism shall not by itself constitute conformity. Goods failing the agreed limits may be rejected and replaced at the supplier’s cost, including removal and reupholstery.”

Then strike three phrases out of any supplier document before signing. “Shade may vary” hands the outcome back to adjectives. “Approval of a lab dip constitutes acceptance of production” moves your standard from the fabric on the roll to a small piece that was never on the roll. “Photographs are final” turns a phone camera into the judge of a color question. Keep the signed cutting, the lot number, the lamp specification and the tolerance instead, and sign once, in the client’s room, with the light on. That is the version of approval that ends the argument, because it leaves no gap for the words “dye lot” to fill.