Lot-to-lot consistency does not mean every delivery is chemically identical — it means every deviation is measured against an approved sealed sample and stays inside tolerances you agreed before the order. The control system has four layers: a sample approval chain (lab dip → head-of-dyeing → pre-production sample → sealed golden sample), instrumented colour assessment (spectrophotometer Delta E under ASTM D2244 plus multi-light visual checks under AATCC EP9), roll-level sampling (head/middle/tail of each roll), and a retain-sample system that keeps every stage traceable by dye lot, laminate batch and roll number.

The rest of this article shows how each layer works, and where bonded fabrics need extra controls that ordinary dyed cloth does not.

The most expensive colour complaint in garment production is not “wrong colour.” It is “right colour — in this roll.” The approval sample sailed through, the first container looks perfect on the receiving table, and then the cutting room finds that panel pieces from roll 14 sit half a shade cooler than roll 3 in the same order. Two dye lots, one purchase order, and a garment line that now has to be cut lot-by-lot or shipped with visible side-panel mismatch.

Bulk fabric is produced in batches — dye lots, finishing runs, adhesive batches — and every batch carries small differences in material, machine and moment. The buyer’s job is not to banish variation (nobody can) but to bound it, measure it, and keep it traceable. Here is the system that does that.

Why Lots Drift: The Four Sources of Shade Variation

Understanding where difference comes from tells you where to demand control:

  1. Dye lot chemistry. Dye weighing, liquor ratio, heating curve, hold time, pH, salt/alkali dosing rhythm and even water quality vary between dye batches — each nudges exhaustion and final shade.
  2. Raw material differences. Fibre batch, yarn lot, greige construction density, pretreatment whiteness and moisture regain all change how the same recipe reads on cloth.
  3. Finishing and setting drift. Stenter temperature, overfeed, width, padder pressure, softener or resin pick-up shift surface reflection and apparent depth — and change hand feel with them.
  4. Observation conditions. Two fabrics can match under daylight and disagree under store lighting. That is metamerism: different reflectance curves that coincide under one illuminant. A colour approved under a single light source is only half approved.

Some materials are simply harder: pale shades, fluorescents, bright reds/blues/purples and elastane-containing cloths are commonly treated as higher-risk for reproduction. That is a risk-ranking rule of thumb, not a law — your control level should scale with the actual colour and construction.

The Sample Chain: Locking One Variable at a Time

A working approval sequence narrows variables step by step, so a late-stage failure always points at a known stage:

Stage What it locks What it does NOT lock
Master standard The reference itself — buyer colour card or approved board, recorded with base cloth, construction, weight, colour number, dye system, finish and applicable light sources Everything downstream
Lab dip Dye recipe and shade direction against the master Bulk hand feel, width, shrinkage, bonding performance — lab dips are small, and lots are not
Strike-off / bulk trial (prints, coatings, sensitive shades) That the recipe survives production-scale equipment, paste, screens and finishing Final lot uniformity
Head-of-dyeing / first-bulk sample That the actual production lot reproduces the approved dip — the bridge between “recipe right” and “production stable” The rest of the order
Pre-production sample (PPS) Everything at once: colour, hand, weight, width, function and appearance, made from planned bulk materials and processes Shipping condition of later rolls
Golden / sealed sample The physical contract. Sealed from the approved PPS or first conforming bulk, one copy per party, labelled with order, date, batch, version and validity period —
Shipment sample Final gate: goods about to ship still match the sealed sample and test file —

Two disciplines make the chain real. First, every approval is written — an emoji on WhatsApp is not a lab-dip approval. Second, the sealed sample is the only comparison object for disputes; specs describe it, but the sealed cloth defines it. The same logic applies to your fabric RFQ: the RFQ starts the paper trail that the sample chain then physically seals.

Measuring Colour: Instrument Plus Eye, Under Named Lights

Serious lot control uses both methods — they answer different questions and neither replaces the other:

  • Spectrophotometer + Delta E. Instrument readings convert colour difference into L*a*b* coordinates and a tolerance figure. ASTM D2244 is the standard practice for calculating colour difference and tolerances from instrumentally measured coordinates. Key detail: tolerances exist in several formulas (CIELAB, CMC, CIE94, CIEDE2000) and the numbers are not interchangeable — a contract that says “Delta E ≤ 1” without naming the formula, observer geometry, aperture, backing and specimen layers is unenforceable. For small differences in the 0–5 range, CIEDE2000 is the commonly recommended formula.
  • Multi-light visual assessment. AATCC EP9 covers visual assessment of colour difference against a standard under controlled conditions. The practical minimum is three illuminants: D65 (daylight), TL84/CWF (store/office fluorescence) and Illuminant A (warm/incandescent). If a lot passes under one light only, it has not passed.
  • Grey scales for fastness results. Colour-change and staining after fastness tests are graded against greyscales (ISO 105-A02 / A03), typically 1–5 with half grades.

What tolerance number is right? It depends on the shade, construction, panel placement, brand standard and supply-chain reality. Common commercial figures like “Delta E ≤ 1.0” or “CMC(2:1) ≤ 1.0” are examples, not defaults to copy into every contract — set the number with your supplier against the sealed sample, per colour.

Hand Feel: Managing the “Subjective” Variable

“Soft hand” is a moving target unless you pin it down. Hand drift comes from fibre batch, yarn twist, construction, pretreatment, softener type and pick-up, setting temperature, calendering, brushing and residual chemistry from bonding. The control pattern:

  • compare only against the sealed sample, same direction, same folding layers, side by side;
  • back the subjective check with objective proxies: GSM, thickness, width, stretch and recovery, shrinkage, pilling grade;
  • require process records: softener grade/category, dosage, setting temperature, speed, overfeed;
  • where a brand wants instrumented hand testing, agree the instrument, specimen orientation, conditioning and limits in advance — scores across different instruments or fabrics are not comparable.

Bonded Fabric: The Lots Hiding Inside the Lot

For laminated constructions, colour is only one of several batch variables — and the ones unique to bonding are where returns actually come from:

  • Adhesive batch differences. Different glue batches vary in viscosity, solids, softening point and activation behaviour, and in hydrolysis/heat-ageing resistance. A laminate that peels fine with batch A can underperform with batch B at the same machine settings.
  • Backing shade lots. Face and backing are often dyed in different machines with different chemistries. The face can match perfectly while the backing shifts — visible at rolled edges, seams and thinner or semi-transparent areas.
  • Lamination parameter drift. Coat weight, oven temperature, nip pressure, line speed, cooling and substrate tension all move. Drift shows as local weak bonds, bubbles, ripples, edge lift or dimensional instability.
  • Ageing, not just day-one strength. Heat, humidity and repeated laundering degrade the bond line over time, which is why initial peel alone is not a batch-release criterion. Our delamination testing guide covers the wash-retention design that catches this.

The batch-release package for bonded fabric should therefore include, per adhesive or lamination lot: T-peel results (ISO 11339 or ASTM D1876 — force per 100 mm width, reporting average, minimum and failure mode), post-wash retained peel, edge/seam/coated-area delamination checks, separate colour measurement and sealed samples for face and backing, and full traceability — adhesive batch, membrane batch, face-lot and backing-lot numbers, bonding date, machine and parameters. Peel limits cannot be copied between orders; set the internal floor from the sealed sample’s measured value plus a safety margin.

Roll-to-Roll and Within-Roll Variation

Same dye lot does not mean same roll. Even within one lot, head/middle/tail and centre/edge can differ from padding, setting and tension unevenness — the classic “centre-run” and “end-of-roll” shade bands. A practical roll protocol:

  • sample head, middle and tail of each roll (more often on long rolls, sensitive shades and high-value orders), measured against the master with the agreed Delta E formula;
  • judge rolls side by side, not one roll in isolation — a lot of individually acceptable rolls can still read uneven as a group;
  • quarantine and label out-of-tolerance rolls instead of mixing them with conforming stock;
  • for shade-sensitive styles, plan cut-room pairing: keep major panels of one garment from the same roll or same dye lot. How strictly this is enforced is factory- and style-specific — agree it before cutting, not after.

Fabric inspection systems like the 4-point method grade defects; they do not replace the shade, hand and peel sampling above. Our bonded fabric inspection checklist places both layers at the right checkpoints.

The Retain-Sample System: Your Evidence Chain

Retain samples exist so that any dispute six months later can be resolved with cloth, not memory. A complete set:

  1. Raw-material retains — fibre/yarn, greige, dye, chemical, adhesive and membrane batch samples with lot numbers;
  2. Process retains — lab dips, strike-offs, head-of-dyeing, pre/post setting, pre/post bonding;
  3. Bulk retains — per dye lot, per lamination lot, per roll at defined positions;
  4. Sealed golden sample — written approval, both parties holding one, version-labelled;
  5. Shipment retains — drawn from what actually shipped, kept past the claim window;
  6. Test file — colour data, multi-light visual results, fastness, shrinkage, weight, peel and post-care retention;
  7. Traceability log — order ↔ dye lot ↔ roll ↔ adhesive batch ↔ machine ↔ parameters ↔ inspector ↔ date.

Retention length is a commercial term, not law: 6–12 months after shipment is a common baseline for regular orders, extended toward 24 months or the warranty period for high-claim-risk, repeat-order or functional bonded programs.

A clause you can adapt:

“Bulk fabric shall match the approved sealed sample within the agreed instrumental tolerance and visual assessment requirements under specified illuminants. For bonded fabrics, each adhesive or lamination lot shall be traceable and supported by initial and post-wash peel-strength results.”

FAQ

What causes fabric shade variation between lots? Multiple stacked variables, not one: dye-lot chemistry differences, raw-material batch shifts, finishing/setting drift, and observation-condition differences including metamerism. Control means bounding each with records and measurement, not chasing a single culprit.

What Delta E is acceptable for fabric? There is no universal number. Tolerances are commonly expressed in CIELAB, CMC, CIE94 or CIEDE2000 terms (per ASTM D2244) and are set per colour, construction and brand standard against a sealed sample. Agree the formula and conditions before using any figure.

How many rolls should be sampled for colour? Head, middle and tail of every roll is the safe baseline for shade-sensitive programs; risk-based sampling (every roll on critical colours, spot-check on commodity lots) is common too. Write the sampling plan into the order so both sides count the same way.

Do bonded fabrics need extra batch controls? Yes — adhesive batch, backing dye lot and lamination parameters each create their own lot boundaries. Release per lamination lot with peel data (initial and post-wash), separate face/backing colour records, and batch traceability.

How long should retain samples be kept? Until the claim window closes — commonly 6–12 months after shipment, longer for repeat or functional programs. It is a contract term; agree it with your supplier at order stage.

Next step: send your shade tolerance, test methods and reorder schedule with your enquiry — through the sample request form or the contact page — and we will confirm the approval chain (lab dip → head-of-dyeing → PPS → sealed sample), the sampling plan and the retain-sample terms in writing before bulk starts.