Start by defining the textile route, not just the colour name
Do not write "melange fleece" as a standalone material description. For sourcing and claims control, specify whether the product is a knitted polyester fleece blanket, brushed on one or both sides, and how the heather effect is created. Typical routes are: fibre-dyed heather blend before spinning, dope-/solution-dyed mixed fibres, cationic/disperse differential dye route on polyester, or printed/heather-look effect. These routes do not behave the same in colour repeatability, pilling, or reorder matching.
For a standard 260gsm retail throw, the most common construction is a warp knit or circular knit polyester fleece that is brushed and sometimes lightly sheared after dyeing, then cut and sewn into a blanket. If the supplier uses cationic heather yarn or differential dye polyester, note that separately because colour build, lot-to-lot repeatability, and lab-dip logic differ from fibre-blend heather.
If recycled polyester is used, add a traceability note in the spec because feedstock variability can shift undertone, lustre, and fibre fineness. That risk is manageable, but only if the mill keeps fibre supplier, lot code, and colour recipe records tied to each production lot. If the heather effect comes from solution-dyed fibres, lot stability can be better on light fastness and reorder continuity, but shade drift can still appear if the masterbatch source changes. Buyers dealing with recycled or dope-dyed routes should ask for retained fibre lot references and transaction records where relevant. See rPET fleece documentation and solution-dyed 260gsm polyester fleece trade-offs.
Contract language should therefore identify: fibre composition, knit type, brushing/shearing state, heather generation route, and whether shade approval applies to face only or both sides. On most fleece blankets, approval is taken on the show face; if the reverse is also consumer-visible, state separate criteria.
Lock the approval state: finished blanket, after normal finishing, before shipment
A frequent source of dispute is that the buyer approves a lab dip or dyed fabric swatch, but bulk is judged on the finished blanket after brushing, shearing, heat-setting, and sewing. Those are not equivalent states. For fleece, colour must be approved in the same physical state in which acceptance will be judged.
A practical rule for 260gsm fleece is: lot approval is based on the finished blanket after all wet processing and finishing, after any pre-shrink or heat-setting applied in normal production, and before packing for shipment. If the style is washed after sewing, approval must occur after that wash. If the blanket is not laundered in production, do not invent a post-laundry colour tolerance unless the retail channel specifically requires wash-after-purchase shade retention checks under a separate test plan.
State clearly whether measurement is on finished face only or both finished sides. For single-face brushed fleece with a less critical reverse, face-only control is common. For reversible or double-brushed styles, measure both sides and define which side governs release if the readings differ. Buyers should also state whether pile should be conditioned before reading, for example by laying the blanket flat for a set period and smoothing pile in one direction with light hand pressure to reduce random gloss variation.
If you approve a pre-production sample, identify its exact state on the approval form: greige knit, dyed knit, brushed fabric, sheared fabric, or finished blanket. If approval is not on finished blanket, write what conversion allowance is accepted between approval state and shipment state. Without that clause, the supplier and buyer are comparing different standards. For broader fleece care and laundering context, see blanket care washing guide.
Use ΔE00 correctly: define what is being compared, how many samples are read, and how pass/fail works
Colourimetry needs tighter wording than many blanket POs use. State whether the ΔE00 tolerance is for single lot acceptance versus approved master, inter-lot matching between current lot and prior released lot, or lot-to-master and lot-to-lot both. These are different controls. A lot can be within tolerance to master and still look noticeably different beside the latest shipment if undertone drift accumulates.
For most retail blanket programmes, use two separate rules. Rule 1: lot-to-master acceptance. Every production lot is compared to the retained approved master. Rule 2: inter-lot adjacency check. Reorder or split-shipment lots are also compared with the latest released lot, or with all lots intended to ship together, under visual side-by-side review. This prevents gradual drift across a season.
Write the measurement condition precisely. D65/10° is an instrument measurement condition, not a viewing condition. It defines the illuminant/observer setting used by the spectrophotometer. Visual approval should be stated separately, for example under a light box using D65 daylight simulation and, if store metamerism risk matters, TL84 as a second visual condition.
A workable production protocol for a 260gsm fleece blanket is: sample 5 blankets per lot up to 5,000 pcs, or 8 blankets for larger lots; take 4 read points per blanket on the face from a defined map excluding seams, labels, folded creases, and edge zones within 10 cm of the perimeter; orient pile consistently before reading; record individual point values and blanket averages; calculate the lot average ΔE00 from all points; also record the maximum single-point ΔE00. If the product is double-sided visible, repeat on reverse and report separately.
For heather fleece, averaging rules matter because local fibre distribution can create minor point variation. A practical rule is: lot passes only if both lot average and maximum point are within spec. Do not accept lot average alone. Otherwise a lot with several obvious local shifts can still pass mathematically.
Recommended starting tolerances for finished fleece blankets against approved master under instrument reading are below. These are sourcing ranges, not universal law, and should be tightened or relaxed based on channel, colour depth, and texture visibility.
Suggested lot-to-master ΔE00 bands for 260gsm heather fleece:
- Premium retail / catalogue continuity: lot average ≤ 0.8 to 1.0; max point ≤ 1.5 to 1.8
- Mainstream retail / club / chain store: lot average ≤ 1.0 to 1.2; max point ≤ 1.8 to 2.0
- Promotional / event / donation: lot average ≤ 1.2 to 1.5; max point ≤ 2.0 to 2.5
Adjust by shade family:
- Light greys, beige heathers, pale pastels: tighter control is usually needed because warmth/coolness drift shows quickly; consider reducing average tolerance by about 0.1 to 0.2
- Mid heathers: baseline tolerances above usually work
- Dark charcoal, navy heathers, black-based marls: instrument numbers can look good while visual pile gloss differs; keep visual adjacency mandatory even if tolerances are slightly looser numerically
If you need lot-specific inspection controls in receiving and final random inspection, see AQL inspection checklist and blanket quality control inspection.
Separate instrument measurement from visual approval
Do not merge spectro settings and light-box review into one sentence on the PO. The instrument condition may be D65/10°, while visual approval is typically performed in a controlled viewing cabinet under D65 and optionally TL84 for retail environments. That distinction should appear in the approval form and in third-party reports.
A practical visual procedure is: compare the retained master, the current lot sample, and if relevant the latest released lot sample side by side on a neutral grey background, pile aligned in the same direction, viewed first flat and then folded as sold. Record pass/fail under D65, then repeat under TL84 if store-light metamerism is a concern. If a lot passes instrumentally but fails obvious adjacency under D65 or TL84, escalate for buyer sign-off rather than auto-release.
Repeatability matters. The same operator, or at least trained QC staff using the same SOP, should perform readings. Spectrophotometers should be calibrated per maker instruction at the start of the shift and whenever readings appear unstable. For textured fleece, pressure and aperture placement can affect the result; note instrument model, aperture size if relevant, and whether a specular component setting is fixed by the mill SOP. Buyers do not always need that on the PO, but they should require it on the mill test record if colour is commercially sensitive.
For textured brushed fabrics, visual review is not optional. Fleece can show pile gloss, density shift, or shearing streaks that are commercially visible but only weakly captured by ΔE00. Treat instrument control as a gate, and visual adjacency as the commercial release check.
Control the finish: GSM, brushing, shearing, and heat-setting
On 260gsm fleece, surface finishing changes perceived shade almost as much as dye recipe. If the finish drifts, the colour signature drifts with it. Buyers should therefore set measurable process controls, or at minimum require the mill to lock and document them lot by lot.
A practical blanket spec may state 260gsm finished weight ±5%, measured after conditioning in the standard mill atmosphere, with test pieces taken away from hems and labels. Weight spread outside that range often correlates with pile density changes, which can make a lot read darker or lighter even with similar dye values.
For pile and finish, common commercial targets are finished pile height about 1.5 to 2.5 mm for a regular retail micro/polar fleece throw, with the exact number approved against a sealed sample. Ask the mill to state whether the fabric is single-sided brushed, double-sided brushed, and whether it is lightly sheared. If you approve one finish route and bulk uses another, shade continuity risk rises immediately.
Objective control points help. A reasonable internal mill route for polyester fleece may involve 1 to 3 brushing passes per side depending on handfeel target, followed by light cosmetic shearing if required, then heat-setting in a controlled window. Typical polyester heat-setting may sit roughly in the 170 to 190°C range with dwell commonly around 30 to 60 seconds, depending on machinery, width, and overfeed. These are process ranges, not buyer-mandated recipe values, but the mill should document actual settings and notify the buyer before changing the approved route.
Failure modes are predictable. Under-brushing can leave the face flatter and visually darker or streakier. Over-brushing can bloom the surface, making the shade look lighter and reducing anti-pilling margin. Over-shearing can flatten the heather effect and create gloss bands. Excessive heat-setting can increase sheen or compact the pile, shifting apparent depth. If the programme is sensitive, require lot records for brushing pass count, shearing setting, and heat-set line settings as part of release documentation.
For pilling and fleece surface stability in adjacent programmes, see anti-pilling test requirements.
Required approval documents and sign-off chain
If colour continuity matters, ask for more than a verbal "shade approved". The supplier release pack should include documents that let the buyer verify what standard was used and which lot was judged against it.
Recommended colour approval file for each production lot:
- PO number, style number, colour name/code, and SKU
- Retained master ID and issue date
- Fabric lot number, dye lot number where applicable, fibre lot references if heather route uses blended inputs
- Production quantity per lot
- Instrument report listing measurement condition, read-point count, individual readings, blanket averages, lot average, and max single-point ΔE00
- Visual approval record under D65 and, if required, TL84 with pass/fail remarks
- Finishing-state declaration confirming readings taken on finished blanket after normal production finishing
- Photograph of retained master and current lot sample under controlled cabinet as a reference only, not as sole approval evidence
- QC manager sign-off at mill level
- Buyer or nominated third-party approval if the programme requires pre-shipment release authority
The sign-off chain should also be explicit. A common sequence is: mill lab prepares data → finishing QC verifies production state → internal QA manager signs release recommendation → buyer QA or third-party inspector confirms commercial acceptance → shipment booking only after release. If shipment is allowed before buyer colour release, say so clearly and assign the risk.
Keep the retained standards traceable. Minimum good practice is one physical master held by buyer, one mirror standard held by mill, and one archive swatch in the lot file. These should be sealed, dated, and protected from light. If the master is replaced, old and new master IDs must both be recorded so reorders are not matched to the wrong reference.
If the order uses FOB, FCA, or DDP terms, colour-release timing should be linked to the shipping milestone so there is no dispute over who bears the cost of a hold. Related planning points are covered in custom blanket lead times and shipping.
Inter-lot segregation: operational rules for factory and warehouse
Many POs promise lot control but then allow mixed cartons, mixed pallet builds, or no physical tag system. That defeats the control. Inter-lot segregation needs a factory rule, a packing rule, and a warehouse rule.
At factory level, each dye/finish lot should receive a physical lot tag linked to the fabric roll, cutting bundle, sewing bundle, and finished-carton batch. Cartons should carry at minimum style, colour, lot number, carton count, production date, and destination. If multiple lots are produced for one PO, carton marks must make them separable without opening the box.
At pallet level, apply a no-mix pallet rule unless the buyer has approved mixed-lot distribution. If mixed pallets are unavoidable for urgent shipping, each pallet should have a visible pallet card listing lot numbers and carton ranges, and the packing list should show exact carton-to-lot mapping. For retail replenishment, the cleaner rule is one lot per pallet and one lot per inbound warehouse location where practical.
At warehouse receipt, require the supplier to provide a lot summary sheet and keep lots physically segregated until buyer release. If a shipment contains more than one lot, buyers should decide whether the lots may be allocated to different doors, different channels, or different production months. That choice should be made before shipping, not after goods arrive.
A workable carton marking clause is: Do not mix lots within carton. Do not mix lots within retail inner pack. Do not mix lots on pallet unless pre-approved in writing. Carton and pallet labels must show production lot number in human-readable text at minimum 15 mm high. This is basic, but it prevents a large share of shade complaints.
Reorder matching rules: master-only, latest-lot, or shade-family
Do not treat reorder colour approval as the same decision for every channel. The right rule depends on how the product is sold and whether prior lots remain in stock.
Master-only matching is usually required for core retail programmes, catalogue continuity, and any SKU likely to be sold beside earlier receipts. Here, every new lot must match the retained approved master first, and then pass visual adjacency against any open inventory lots if overlap exists. This is the most conservative approach and the safest for long-running programmes.
Latest-released-lot matching can be acceptable where inventory turns fast, old lots are fully depleted before replenishment lands, and the buyer is more concerned with continuity to what is currently shipping than with the original launch lot. This should not be used if multiple historical lots are still in stores, in e-commerce returns stock, or in regional DCs.
Shade-family matching is acceptable for some promotional, off-price, donation, or non-adjacent channels. In that case, the buyer approves a master plus upper/lower visual limits and numeric limits, and any lot inside that family may ship. This approach reduces rejection risk but only works if the channel will not place visibly different lots side by side.
A buyer should also decide what happens after a master revision. If the colour has been intentionally updated, state whether old and new lots may coexist under the same SKU or whether the SKU code must change. Many avoidable claims come from changing the master while leaving replenishment identifiers unchanged.
Paste-ready PO clause for colour and lot control
Below is a practical clause buyers can adapt for a 260gsm heather fleece blanket programme:
Colour approval and lot release clause:
Product: 260gsm 100% polyester heather brushed knit fleece blanket, finished blanket approval state, face side governing unless otherwise stated. Colour acceptance shall be based on comparison to approved retained master ID [insert ID]. Instrument measurement shall be CIEDE2000 (ΔE00) under D65/10° on finished blanket after normal production finishing and before packing. Sampling: 5 blankets per lot up to 5,000 pcs; 8 blankets per lot above 5,000 pcs. Read 4 points per blanket on face according to approved sampling map; exclude hems, labels, creases, and edge zones within 10 cm. Pile to be aligned consistently before reading. Lot passes only if lot average ΔE00 is ≤ [insert target, e.g. 1.0] and no individual point exceeds [insert target, e.g. 1.8]. In addition, lot shall pass visual side-by-side approval against retained master under D65 light box, and against latest released lot under D65 and TL84 where shipments may overlap in market. Measurements and approvals must be recorded with PO no., colour code, lot no., read values, average, max point, date, and sign-off by mill QA manager. No lot mixing within carton or inner pack. No lot mixing on pallet without buyer written approval. Carton and pallet labels must show production lot number. Reorders for this SKU shall match approved master; matching to latest released lot alone is not acceptable unless buyer confirms depletion of prior inventory in writing.
If the channel is lower risk, you can edit the clause to allow a wider shade-family window or permit latest-lot matching after inventory depletion confirmation.
Where buyers should relax or tighten the rule
Not every programme needs the same colour discipline. Tighten the spec where the blanket is sold folded with multiple units visible together, photographed for e-commerce against prior stock, or distributed to regional DCs over several months. Tighten again for pale heathers and neutral greys, where undertone drift is easier to spot.
Relaxation may be commercially sensible for short-life promotions, donation programmes, or one-off event blankets where lots will not be visually adjacent. In those cases, a wider average ΔE00 band and shade-family rule may cut reject risk without hurting sell-through.
Be careful with claims about colour permanence. Shade continuity at receipt is different from colourfastness after use. If the buyer also needs wash fastness, crocking, or light fastness controls, add those as separate tests rather than assuming the shade-approval protocol covers them. Related guidance: ISO 105-C06 wash fastness, ISO 105-X12 rubbing fastness, and solution-dyed fleece light fastness.
Frequently asked
What is a realistic ΔE00 tolerance for 260gsm heather fleece blankets? For finished blanket lot-to-master acceptance, many mainstream retail programmes start around lot average ΔE00 ≤ 1.0 to 1.2 with max point ≤ 1.8 to 2.0, measured on the finished face. Premium retail often tightens to about average ≤ 0.8 to 1.0. Promotional channels may accept average up to about 1.5. The right number depends on shade family, pile texture, and whether lots will sit together.
Should D65/10° be written as the viewing condition? No. D65/10° is an instrument measurement condition for the spectrophotometer. Visual approval should be specified separately, normally in a controlled light box under D65, and often TL84 as a second condition if store-light metamerism matters.
Should colour be approved on fabric or on the finished blanket? For fleece blankets, approval on the finished blanket is usually the safer commercial rule because brushing, shearing, heat-setting, and any production washing can shift visual shade and gloss. If approval is done at an earlier state, the PO should define the allowed conversion from that state to the shipped state.
How many blankets and read points should be measured per lot? A practical mill-release plan is 5 blankets per lot up to 5,000 pcs, or 8 blankets for larger lots, with 4 read points per blanket on the face from a fixed map. Record individual points, blanket averages, lot average, and maximum single-point value. High-risk programmes may sample more heavily.
When is matching to the latest released lot acceptable instead of matching to the original master? It can be acceptable if earlier inventory is fully depleted before replenishment lands and the buyer confirms that old and new lots will not be adjacent in stores or DCs. For core retail continuity, master-only matching plus adjacency review against open inventory is usually the better rule.
What documents should a supplier provide for colour approval? At minimum: retained master ID, PO and SKU, lot numbers, production quantity, instrument report with measured values, visual approval record under stated light conditions, finishing-state declaration, and QA sign-off. For recycled or fibre-blend heather routes, add fibre lot references and traceability records.
Have a project in mind? Send us your spec — we'll reply within one business day with indicative pricing and a sample plan.