Short answer: fabric colour varies between batches because colour is the result of a chemical reaction, not a fixed recipe. Fibre origin, yarn lot, dyebath chemistry, machine conditions, fabric structure and finishing each shift the final shade by a measurable amount. Inside an accepted tolerance that drift is normal and commercially invisible; outside it, the batch fails inspection. The industry quantifies the shift in ΔE and controls it with lab dips, dye-lot reservation and spectrophotometer QC.
If you have ever signed off a sample and then opened a bulk carton six weeks later to find a shade that is subtly, maddeningly different, you have met batch colour variation. It is one of the most common quality disputes in apparel manufacturing — and one of the least well explained to brands. This article breaks down what actually causes it, what tolerance you should be writing into your spec, and how a factory should be controlling it.
What “the same colour” actually means
Outside of a digital file, there is no such thing as an identical colour. When you approve a Pantone chip or a lab dip, you are approving a reference — a target that production will move toward, with some permitted variance around it. Dye houses do not reproduce a colour; they run a process until the output lands inside a window.
That window exists because dyeing a textile is a wet chemistry process with dozens of inputs, and unlike printing a flat colour on paper, the substrate itself is part of the reaction. Cotton from two growing regions takes dye differently. Yarn spun in two batches has different absorbency. Water, salt, alkali and temperature all move the result. The goal is never zero variation; the goal is controlled, quantified, and imperceptible variation.
The seven things that shift a shade
1. Fibre base and maturity
Cotton is a natural fibre. Fibre length, maturity and micronaire vary by origin, season and gin. Immature fibres absorb dye more readily and read lighter; more mature fibres read deeper and duller. When a mill switches cotton bales between lots — which happens constantly, since bales are blended to a spec rather than sourced from one farm — the base absorbency shifts.
2. Yarn lot and blend ratio
Even within a single cotton type, yarns spun in different lots differ in twist, hairiness and residual wax. In blends — cotton/polyester, cotton/viscose, cotton/elastane — the ratio itself drifts within a permitted band, and because each fibre takes a different dye class, a 2% shift in polyester content can move the shade visibly. This is why blends are harder to match on repeat than 100% cotton.
3. Dyestuff lot and affinity
Dyestuff arrives in batches too. Two lots of the same reactive black can have slightly different strength and substantivity. Reputable dye houses standardise every incoming dye lot against a reference and re-calibrate their recipe — a step that is invisible in a quotation but decisive in a repeat order.
4. Dyebath chemistry and control
pH, salt concentration, alkali dosing, liquor ratio, auxiliaries and temperature ramp are the levers a dyer uses. A pH drift of 0.3, an uneven alkali feed, or a ramp that runs hot will shift shade and levelness. This is the single largest source of within-lot variation, and the reason dye-house process control matters more than the dye recipe itself.
5. Machine, load and liquor ratio
The same recipe produces a slightly different shade on a 50 kg sample dyeing machine and a 500 kg production machine. Load density, liquor circulation and the physical handling of the rope all change how the fabric interacts with the bath. Reputable dye houses run a bulk trial on production equipment, not just a lab recipe.
6. Fabric structure and surface
Colour is read by the eye as reflected light. A smooth, compact single jersey reflects differently from a brushed fleece, a loop-back French terry or a rib. This is why the same dyed yarn can look different across two constructions in one collection, and why a brushed fleece hoodie often appears half a shade lighter than the matching jersey tee.
7. Finishing, heat and moisture
Stenter drying, calendering, brushing, enzyme treatment and compacting all change surface reflectance and moisture content. Fabric also reads darker when wet or damp. If the garment or sample you are judging has a different moisture regain from the reference, you will see a difference that is not actually a dye difference.
How the industry measures the difference: ΔE and L*a*b*
Colour difference is expressed as ΔE — a single number describing the total distance between two colours in the CIELAB (L*a*b*) colour space, where L* is lightness, a* is the red–green axis and b* is the blue–yellow axis. A spectrophotometer reads a sample and returns those coordinates; ΔE is the distance to the approved reference.
The practical advantage of using ΔE instead of the human eye is repeatability. Two trained assessors can disagree on a borderline shade, especially in different lighting. A spectrophotometer with a fixed aperture and illuminant will give the same number every time.
| ΔE (CMC 2:1) | What it looks like | Typical commercial position |
|---|---|---|
| 0 – 0.5 | Indistinguishable to the eye | Premium solid-dye programmes, matching sets |
| 0.5 – 1.0 | Only detectable side by side under controlled light | Standard accepted tolerance for most apparel |
| 1.0 – 2.0 | Visible on direct comparison, acceptable apart | “Commercial match” — bulk vs sample |
| 2.0 – 3.5 | Clearly different when placed together | Usually rejected for multi-panel styles |
| > 3.5 | Reads as a different colour | Rejected; re-dye or remedy required |
Tolerance is not one universal number. A tracksuit where jacket and pant must match is judged far more tightly than a single hoodie sold alone. A white or a pale pink is far more unforgiving than a black, because the eye is extremely sensitive to small deviations in light shades, while dark saturated shades hide them.
The standards worth citing in a tech pack
- ISO 105-J03 — calculation of colour difference (the international method for ΔE).
- AATCC 173 — CMC colour-difference evaluation, widely used by US buyers.
- ISO 105-A02 / AATCC EP1 — grey scale for assessing change in colour.
- AATCC 61 / ISO 105-C06 — colourfastness to laundering.
- AATCC 8 / ISO 105-X12 — colourfastness to crocking (rubbing).
- ASTM D2244 — instrument-based colour-difference calculation.
- D65, TL84 and A (incandescent) light sources — the three sources used to catch metamerism.
Where batch variation actually hurts
Shade drift is a manageable cost in some products and a commercial disaster in others.
- Multi-panel styles. A hoodie with a contrast panel, a tracksuit, a two-tone jacket — any product where two pieces sit next to each other on the body will expose a difference that would be invisible on a single-piece garment.
- Cut-and-sew across dye lots. The classic failure: one style, one PO, but the cutting room has two lots on the floor and cuts panels from both. Left sleeve and right sleeve end up from different lots.
- Repeat orders. Your bestseller reorders three months later. New dye lot, new shade. The customer reordering the same SKU expects continuity.
- Matching sets sold together. Tops and bottoms are frequently dyed separately and even in different mills. Getting them to match requires deliberate planning, not luck.
- Blends and garment-dyed items. Garment dye produces more variation than piece dye, and blends can shift where the two fibre types meet.
How DOLUXE controls colour on every batch
Colour control is a process we design around, not a step we check at the end. The workflow that keeps bulk output inside tolerance looks like this:
What brands should put in their specification
Most colour disputes are preventable at the tech pack stage. These eight clauses do the heavy lifting:
- State a numeric tolerance — for example “ΔE ≤ 1.0 (CMC 2:1) against the approved lab dip” for solid premium programmes, ≤ 2.0 for general apparel.
- Approve a physical lab dip, never a screen. A monitor, a phone and a printer will all show your colour differently, and none of them is calibrated to your factory's light booth.
- Approve on the correct substrate. A lab dip on plain cotton tells you nothing about how the colour behaves on brushed fleece.
- Reserve the dye lot. Write it in, or at minimum require written approval before a lot is split.
- Use the phrase “commercial match” for bulk against sample, and define what it means numerically.
- Plan for a rejection allowance. Best practice is to produce 3–5% over on shade-critical solid programmes so a rejected lot does not push you past your delivery date.
- For repeat orders, run a continuity programme. Send the retained control swatch, or specify that the mill must match against the previous production lot.
- Check your own lighting. Before you reject a shipment, look at it under the same light source your factory used. Warm hotel lighting and a window at dusk have caused a great many false alarms.
Frequently asked questions
Is colour variation between batches normal?
Yes — some movement is inherent to wet dyeing and cannot be eliminated. What matters is whether the shift stays inside the tolerance you specified. A batch at ΔE 0.8 is a normal production outcome; a batch at ΔE 4 is a defect.
What ΔE is acceptable for clothing?
For most apparel, ΔE ≤ 1.0 is a tight, premium-grade tolerance and ΔE ≤ 2.0 is a standard commercial match. Styles where two pieces are worn together should be specified at 1.0 or tighter. Always define the formula — ΔE CMC 2:1 and ΔE*ab are different numbers for the same pair of colours.
Why does the same black look different on cotton and polyester?
Different fibre classes require different dye chemistry, and each reflects light differently. Cotton is dyed with reactive dyes, polyester with disperse dyes under high temperature. Even when both are matched to the same L*a*b* target under D65, they can drift apart under other light sources — this is metamerism, and it is the reason blends are harder to match than single-fibre fabrics.
Can you guarantee the exact same colour on a repeat order?
Exact is not achievable; controlled is. We retain a control swatch from every lot for twelve months and dye repeat orders against that physical reference, plus the original numeric L*a*b* values. On a continuity programme, repeat bulk typically lands within ΔE 1.0 of the original.
Why can’t I judge colour from a photo?
Because a photograph passes through a camera sensor, white balance correction, compression and a monitor — four colour transformations you do not control. Two phones photographing the same fabric will disagree. Approve lab dips in person, or ask your factory to send physical swatches by courier.
Conclusion
Batch colour variation is not a sign that your factory is careless. It is a property of dyeing that has to be actively managed, measured and specified. Brands that get consistent colour are not lucky — they write a ΔE tolerance, approve physical lab dips, reserve dye lots, and require lot segregation at the cutting table.
DOLUXE dyes and finishes in-house with spectrophotometer verification, D65/TL84 light-booth assessment, per-lot fastness testing and twelve-month control sample retention. If you need a colour-critical programme — a matched tracksuit, a repeat-run bestseller, a solid premium capsule — send us your Pantone reference and we will return lab dips on the correct fabric construction. MOQ starts at 50 pcs per style, with lab dips typically turned around in 5–7 days.
DOLUXE — Do Luxury for Your Brands