Home / Journal / Yarn dye lots

Yarn Dye Lots: Same-Lot, Lot Change and Replenishment

Updated 17 November 2026 · Yarn dye lots, lot-change control, replenishment traceability

A dye lot is not a colour. It is a batch identity created on one day, in one vessel, from one fibre base, and it cannot be recreated — which is why the number printed on a cone label matters more than the shade card it is compared against.

Two crochet orders can be made from the same yarn, the same count, the same twist and the same nominal colour, and still be visibly different when the panels sit next to each other. Nothing went wrong. The yarn was dyed in two batches, and two batches are two colours as far as the eye is concerned, even when a spectrophotometer says they are close. The buyer who receives both halves of a split order sees a seam; the factory that dyed them sees two conforming deliveries.

This article is about the three decisions that sit behind that seam: which lot covers the order, how a lot change is declared when it happens anyway, and how a re-order is traced back to something that no longer exists. It is written for buyers placing crochet and soft-toy programmes and for the people who have to answer the email six months later when a re-order arrives and the shade is "slightly off".

In this article

1. A lot is an identity, not a shade · 2. One lot for one order, and what it costs · 3. The tolerance is a contract, not a number · 4. Four places where lots actually break · 5. Declaring a lot change before you need to · 6. Metamerism: right under one lamp · 7. The re-order that can never match · 8. The retained standard is the only defence · 9. What belongs in the purchase order · 10. What we do before the first cone is wound · 11. Questions

1. A lot is an identity, not a shade

A dye lot is a quantity of yarn dyed together in one operation. Everything in that operation shares a history: the same fibre base, the same water, the same dye recipe weighed out on the same day, the same temperature ramp, the same vessel, the same operator, the same drying conditions. Change any one of those and you have a new lot, even if the recipe card is identical and the target colour is identical.

That is the part buyers underestimate. A colour is a target; a lot is an event. The target can be repeated indefinitely. The event cannot, because the inputs that made it are not fully controllable — the natural fibre base varies from bale to bale, water chemistry varies, dye uptake varies with the substrate, and the correction steps taken during dyeing are a response to what the bath was actually doing that morning rather than a fixed script.

Practically, this means a lot number is the only honest identifier of a colour in production. A shade name is a marketing label. A Pantone or RAL reference is an approximation agreed for convenience. A lab dip is a prediction. The lot number is the record of what actually happened, and it is the only one of the four that can be traced, audited and repeated in a conversation about a specific shipment.

This is also why the number has to travel. If the lot number lives only in the factory's dyehouse ledger, then the carton, the packing list and the invoice describe a shipment whose colour history is invisible to everyone who receives it. The traceability chain that matters is short and unglamorous: lot number on the cone label, lot number on the carton label, lot number on the packing list, lot number referenced on the invoice or the delivery note. When all four agree, a colour complaint becomes a solvable question. When any one is missing, it becomes an argument about perception.

2. One lot for one order, and what it costs

The default instruction from most buyers is "one order, one lot" — and it is the right default. A single lot removes the seam, removes the need for shade banding within the order, and removes the most common cause of a partial rejection: half a shipment that matches the approved standard and half that is close but not close enough.

What the instruction usually omits is that one lot is a reservation, not a wish. Yarn for a single lot has to be dyed in one vessel, and the size of that vessel is a hard ceiling. An order large enough to exceed it will be split across lots whether or not anyone planned for it. The choice is therefore not between one lot and several; it is between several lots that were planned and several lots that happened.

DecisionWhat it commitsWhat it protects against
Reserve one lot for the whole orderYarn is dyed ahead of consumption and held; the mill carries stock it has not been paid for yetA mid-order lot change with no notice, discovered at packing
Reserve one lot plus a top-up allowanceWaste, rework and short-shipment yarn all come from the same bathThe 3 per cent rework that forces a second lot for 40 pieces
Accept two planned lots, bandedThe order is split deliberately, with the seam placed where it will not be seenAn unplanned seam in the middle of a retail run
Accept lots as they comeNothing is committed; nothing is reservedNothing — this is the option that produces the complaint

The top-up allowance is the part most purchase orders miss. A crochet programme consumes yarn twice: once in the pieces that ship, and once in the pieces that do not — the rejects, the re-works, the samples pulled mid-run, the shortfall when a dye lot yields fewer good pieces than the yield calculation predicted. If the allowance is not built into the lot reservation, the rework is dyed separately, and the rework is precisely the material most likely to end up next to the original in the same carton.

Where an order genuinely has to run across two lots, the professional answer is not to hide it but to band it. Banding means the two lots are kept physically separate through packing, the cartons are marked with which lot they contain, and the split is placed where the consumer will never see both at once — one lot per colourway, one lot per size, one lot per retail channel, or one lot per delivery window. A deliberate band is a managed risk. An accidental band is a defect.

3. The tolerance is a contract, not a number

Sooner or later someone writes "ΔE ≤ 1.0" into a specification and believes the colour question is closed. It is not, because a ΔE figure without three other pieces of information is not a measurement — it is a number that happens to be computable.

The first missing piece is the equation. ISO 105-J03:2009, the current edition of the colour-difference calculation standard in the ISO 105 colour fastness series, computes the difference between two specimens of the same material measured under the same conditions and expresses it as ΔE CMC(l:c), with a tolerance that depends only on how close a match the end use requires — not on the colour itself and not on the nature of the difference. That is a genuinely useful property: the same tolerance number can be applied across a colour card. But it is not the same number as a ΔE*ab from CIELAB, or a ΔE00 from CIEDE2000, and the values are not interchangeable.

ASTM D2244 makes the same point from the other direction. Its scope states plainly that colour differences and tolerances obtained in different colour-scale systems are not likely to be identical, and that there is no simple factor that converts one system's numbers into another's across all colours. The same practice also warns that the CIELAB equation is not recommended for small and moderate differences — under 5.0 ΔE*ab units — and recommends the newer equations, including CIEDE2000, for the 0.0 to 5.0 range. Which is exactly the range a textile programme lives in. In other words: the equation that gets quoted most often is the one the standard least recommends for the differences that actually cause disputes.

ASTM D2244 is explicit that this is a contract term rather than a technical constant: for product specification, purchaser and seller are to agree both the permissible tolerance and the procedure for calculating it. That is the sentence to bring to a colour meeting. There is no universal pass value for a dyed yarn, and any supplier who offers one without asking what the product is has skipped the thinking.

The second missing piece is the measurement condition. A spectrophotometer reading is only comparable to another reading taken on the same instrument geometry, the same aperture, the same illuminant, the same observer, the same specular setting, the same backing and the same conditioning. ASTM E1164 covers the instrumental side of that; if two labs measure the same sample and disagree, the disagreement is usually in the settings rather than the yarn.

The third missing piece is the visual check. Instrumental and visual assessment are complementary, not competing. ASTM D1729 standardises the visual appraisal of colour under controlled illumination, and it exists because a number and an eye can legitimately disagree — a difference that is numerically small can read as obvious on a textured surface, and a crochet stitch is nothing but texture.

Put together, a usable colour clause has four parts: the equation, the illuminant and viewing condition, the tolerance value, and who adjudicates when the instrument and the eye disagree. A clause with only the tolerance value is a clause that will be renegotiated at the first complaint.

4. Four places where lots actually break

Lot changes are not random. In crochet production they come from four predictable places, and each has a different remedy.

Yarn supply at the spinning stage

Yarn is dyed in lots, but it is spun in lots too, and the two are not the same batch. A dyer working from two spinning batches will produce two dye lots even from one recipe, because the substrate takes up dye slightly differently. This is the break that surprises buyers most, because the order was placed as a single quantity and nobody mentioned that the raw yarn arrived in two deliveries.

Vessel capacity

The dyeing vessel sets a ceiling on lot size. Above that ceiling the order is split by physics. Where the split falls can often be chosen — by colourway, by size, by delivery — but only if it is discussed before the yarn is scheduled rather than after the first batch comes out.

Rework, shortage and yield

Every order produces some material that has to be made again: pieces that fail an inspection point, sample units pulled from production, and the shortfall between calculated yield and actual yield on a hand-worked product. If those pieces need yarn, and the reserved lot is exhausted, they get a new lot. This is the most common source of the "one carton looks different" complaint, because rework is packed last and often packed together with original production.

Split shipment and staged delivery

An order delivered in two windows may be produced in two windows, and two production windows usually mean two yarn lots. Seasonal programmes make this structural rather than accidental: a replenishment placed in peak season is competing for dyehouse capacity with everyone else's peak season.

The remedy differs by cause. Supply breaks are managed by reserving dyed yarn early and holding it. Capacity breaks are managed by designing the split. Rework breaks are managed by the top-up allowance. Split shipments are managed by treating the lot as a delivery attribute rather than an order attribute, and telling the buyer which lot is in which container before the container sails.

5. Declaring a lot change before you need to

Most lot-change disputes are not caused by the change. They are caused by the change arriving as a discovery. The buyer opens a carton, sees a difference, and the first information they receive is a justification — which reads as an excuse regardless of how true it is.

The alternative is a standing clause that costs nothing to write: any change of dye lot within an order will be notified in writing before the affected goods are packed, stating the lot number, the quantity affected, the cartons affected and the measured difference against the retained standard. Notice before packing converts a complaint into a decision. The buyer can accept, reject that portion, ask for it to be shipped separately, or ask for it to be held — all of which are cheaper than a return.

Two details make the clause work. First, "notified" has to mean a named person on each side, not a message to a general inbox; a notice nobody reads is not a notice. Second, the notice has to carry a measurement, not an adjective. "Slightly deeper" is not information. "ΔE CMC(2:1) 0.8 against the retained standard, measured on the same instrument under D65" is information, and it lets a buyer decide in a minute instead of a week.

There is also a threshold question worth settling in advance: at what point does a lot change require approval rather than notification? A common and workable split is that a change within the agreed tolerance is notified, and a change outside it requires approval before packing. Without that split, every lot change becomes a negotiation, and negotiations at packing time are settled by whoever is more afraid of the delay.

6. Metamerism: right under one lamp

Two yarns can match under the lamp in the dyehouse and separate under the lamp in the shop. This is metamerism: different spectral curves that produce the same perceived colour under one illumination and different colours under another. It is not a defect in either sample, and it cannot be detected by a single measurement under a single light source.

For a crochet programme the practical consequence is that a colour approved under one illuminant is only approved for that illuminant. A buyer checking a shipment in a fluorescent-lit office, a warehouse under sodium lighting and a retail floor under warm LED are running three different tests. ASTM D4086 addresses the visual evaluation of metamerism specifically, and the working method is simple: compare the candidate against the retained standard under at least two substantially different sources, one of them daylight, and record which sources were used.

Metamerism is also the reason a "perfect" instrumental match can still generate a complaint. The instrument reported a small difference; the customer is standing under a light the instrument was not asked about. The remedy is procedural rather than technical: state the illuminants in the colour clause, keep the retained standard physically present at both ends, and treat "matches under D65 but separates under the store lighting" as a finding to be recorded rather than an argument to be won.

7. The re-order that can never match

The hardest lot question is the one that arrives months later: the style sold, the buyer re-orders, and the request is "same colour as last time". Last time no longer exists. The lot was consumed, the dyed yarn is gone, and the dyehouse that made it is now making something else.

What can be reproduced is the target, not the batch. A re-order is therefore a new colour approval, and the honest way to run it is to treat it as one: the retained standard is pulled, a new lab dip is submitted against it, and the buyer approves or adjusts with the knowledge that a small step is possible. Buyers who understand this build a tolerance into their programme rather than expecting an exact match. Buyers who do not understand it discover the issue at the third re-order, when the accumulated drift has become visible across a shelf.

Two habits make re-orders manageable. The first is a stored physical standard per colourway per season, held in controlled conditions at both ends, because a standard that has faded in a warehouse drawer is worse than no standard — it silently redefines what "correct" means. The second is a recorded colour history: lot number, date, measured values and the approved standard for each production run, so that a re-order starts from data rather than from a photograph.

There is a commercial point buried here. A programme that re-orders the same colourway repeatedly should reserve capacity rather than hope: the same dyed yarn held for a planned replenishment is cheaper than a fresh dye lot plus a fresh approval round plus the risk of a shade that the buyer rejects on arrival. That is not a favour to the supplier; it is the buyer buying certainty at the only point where certainty is still cheap.

8. The retained standard is the only defence

Everything above depends on one physical object: the retained standard. Not a photograph, not a Pantone chip, not a file of numbers — a piece of the actual dyed yarn, or the actual finished piece, kept as the reference against which every subsequent lot is judged.

A retained standard has four requirements, and failing any one of them quietly destroys its value. It has to be from an approved lot, not from a promising trial. It has to be large enough to measure and to compare by eye, and to survive being handled. It has to be stored away from light, heat and humidity, in a sealed bag, because dyed textile standards drift. And it has to be recorded: which lot, which date, which approval, which tolerance, held by whom at each end.

The failure mode is always the same and always invisible until it matters. The standard sits in a drawer for eight months, fades slightly, and is then used to judge a new lot. The new lot is rejected for being different from a standard that is itself no longer correct, or accepted for matching a standard that has moved. Either way the discussion is now about two unknowns.

A retained standard is also what makes an inspection meaningful. Without it, an inspector can only report that the goods look consistent with each other — which is not the question. With it, the inspector can report a measured difference against an agreed reference, and that report is a decision document rather than an opinion. This is the same logic that governs sealed samples and retention periods, applied to colour rather than to construction.

9. What belongs in the purchase order

Almost all of the control discussed here is contractual rather than technical. A colour clause that covers the following eight points converts a recurring argument into a process.

ClauseWhat it should state
Reference standardWhich physical standard governs, its lot number and date, and where the counterpart is held
Equation and toleranceThe exact colour-difference equation (CMC(l:c) with stated weights, or CIEDE2000, or another named method), plus the numeric tolerance
Measurement conditionsInstrument type, illuminant, observer, aperture, and whether the reading is on yarn or on finished piece
Visual conditionsIlluminants for visual appraisal, and the requirement to check a second source for metamerism
Lot reservationOne lot for the order where capacity allows, plus a stated top-up allowance for rework
Band policyIf more than one lot is used, how it is banded, marked and disclosed
Change notificationWritten notice before packing, with lot number, quantity, cartons and measured difference
AdjudicationWho decides when instrument and eye disagree, and what happens to the affected goods in the meantime

Two of these clauses are usually missing from first-draft specifications, and both are the ones that get tested. The top-up allowance is missing because nobody plans to make rejects. The adjudication clause is missing because nobody expects a disagreement about a measurement they have already agreed. In practice, the first lot change on a new programme is where both get discovered.

10. What we do before the first cone is wound

Our own sequence, in the order it happens: the colour reference is fixed as a physical standard rather than a code, and both parties keep one; the tolerance is agreed with the equation and the measurement conditions written down, not just the number; the order is checked against dyeing capacity before it is confirmed, so that a planned split is a decision and an unplanned one is not possible; a top-up allowance is reserved within the same lot for rework and sampling; the lot number is printed on the cone label, the carton label and the packing list, and referenced on the delivery documents; any change of lot inside the order is notified before the affected goods are packed; and the retained standard is stored sealed, dark and dated, so that the re-order nine months later starts from an object rather than a memory.

None of this is dyehouse chemistry. It is the paperwork that decides whether the chemistry gets a fair hearing — and it costs a clause, not a batch. If you are setting up a crochet or soft-toy programme and want the colour control schedule we use — reference standard, equation and tolerance block, band policy and the lot-change notice — send the colourway count and the delivery structure you are planning, and we will return it filled in for your order.

Related reading: specification to bulk tolerance, which is where the same reservation logic applies to dimensions rather than colour; sealed samples and retention periods, which covers the physical standard side in more detail; and pre-shipment inspection booking, where lot numbers are first reconciled against cartons.

What exactly is a yarn dye lot?

A quantity of yarn dyed together in one operation, sharing the same fibre base, water, dye recipe, vessel, temperature profile and drying conditions. It is identified by a lot number, and it is a batch identity rather than a shade: the same recipe run on a different day, or on a different spinning batch, produces a different lot and can produce a visibly different colour.

Can two dye lots be identical if the recipe is the same?

No. The recipe is an instruction, not a guarantee. Natural fibre varies from bale to bale, dye uptake varies with the substrate, water chemistry varies, and the corrections made during a bath are a response to what that bath was doing. A well-run dyehouse can bring successive lots very close, but close is a measured statement, not an identical one.

Is "ΔE ≤ 1.0" a sufficient colour specification?

No, because the equation is missing. A ΔE CMC(l:c) value, a ΔE*ab value and a ΔE00 value for the same pair of samples are different numbers, and ASTM D2244 states there is no simple factor that converts between colour-scale systems across all colours. A usable clause names the equation, the measurement conditions and the tolerance, and ASTM D2244 states that the tolerance and the calculation procedure are to be agreed between purchaser and seller.

Which colour-difference equation should a textile programme use?

ISO 105-J03:2009 calculates colour difference as ΔE CMC(l:c) for two specimens of the same material measured under the same conditions, with a tolerance that depends only on the closeness of match the end use requires rather than on the colour. ASTM D2244 recommends the newer equations, including CIEDE2000, for differences in the 0.0 to 5.0 ΔE*ab range and notes that CIELAB is not recommended for small differences. Many programmes therefore specify CMC for the historical record and CIEDE2000 for the working tolerance — but whichever is chosen, it has to be named.

Why do two matching yarns look different in the shop?

Metamerism: different spectral curves that appear to match under one light source and separate under another. It cannot be detected by measuring under a single illuminant, which is why ASTM D4086 covers the visual evaluation of metamerism and why a colour approval should be checked under at least two substantially different sources, one of them daylight.

What happens if a lot changes in the middle of an order?

It should be notified rather than discovered. The workable clause is written notice before the affected goods are packed, stating the lot number, the quantity, the cartons and the measured difference against the retained standard — with a threshold above which approval, rather than notification, is required. A change disclosed before packing is a decision the buyer can act on; a change found at goods-in is a return.

Can a re-order match the original lot?

No. The original lot was consumed and cannot be recreated. What can be reproduced is the target, judged against the retained standard, so a re-order is properly a new colour approval. Programmes that re-order the same colourway repeatedly should either reserve dyed yarn against the planned replenishment or accept a fresh approval round each time — the expensive option is expecting an exact match and discovering the drift on the third order.

Related: yarn count and consumption in crochet costing — how Ne, Nm and tex convert, why a label and a laboratory disagree on the same cone, and the three errors that move a yarn line.

Want the colour control schedule we use? Send the number of colourways and your delivery structure, and we will return the reference-standard, tolerance and lot-change notice block filled in for your order.

OEM/ODM process · WhatsApp