A buyer approves a swatch. Twelve weeks later five hundred pieces arrive, and two of them read wrong next to the others.
Nobody was careless. The dyehouse hit its target. The problem is that “the same color” in textile production is never a single value — it is a range. Whether that range is acceptable is a decision someone has to make, in writing, before bulk dyeing starts.
This is about how that decision gets made: the three documents that define color, what Delta E actually measures, and where sleepwear specifically fails.
Why the Same Recipe Produces a Different Shade
Dyeing is a chemical process run on a substrate that is never perfectly uniform. Neither side of that equation is fully reproducible.
- Fibre lot: cotton from a different growth region absorbs dye differently.
- Yarn lot and twist: same fibre, different uptake.
- Water quality and mineral content at the dyehouse.
- Temperature ramp and hold time in the dye vessel.
- Machine loading — a full vessel dyes differently from a half-full one.
- For synthetics and blends: heat setting, and the ratio of fibre types that take dye at all.
None of this is sloppiness. It is the physics of the process. The practical consequence is that a dyehouse works to a tolerance band by design. Your job is to define that band rather than discover it three weeks after the goods land.
The Three Documents That Define Your Color
| Document | What it is | Issued by | When it matters |
|---|---|---|---|
| Color standard | The physical target — a retained reference swatch plus a named color reference | Agreed between you and the mill | Baseline for everything downstream |
| Lab dip | A small laboratory dyeing submitted for approval before bulk | Dyehouse / mill | Before bulk dyeing starts |
| Bulk shade band | The production dyeing, plus the range of variation you will accept within and across lots | Mill, approved by you | During and after bulk dyeing |
The most common and most expensive confusion is treating lab dip approval as bulk approval.
A lab dip is dyed in a beaker on a small sample, often by a technician. Bulk is dyed in a vessel holding hundreds of kilograms. Clearing the dip does not clear the lot. You approve bulk separately, against a retained sample, and you keep that sample.
What Delta E Actually Measures
Delta E (ΔE) expresses the distance between two colors in a defined color space. Lower means closer. It is the standard way to put a number on color tolerance instead of arguing about it.
Three things buyers get wrong.
It summarises without explaining. Two submissions can carry the same ΔE with entirely different problems — one is off in lightness, the other in chroma. An identical number is not an identical error. If direction matters for your shade (for deep and neutral tones it usually does), the report must give you L\, a\, b\* values, not just the total.
The formula has to be named. CIE76, CMC, CIE94 and CIEDE2000 return different numbers for the same pair of samples. “ΔE under 1.5” is not a specification until it says which formula. The calculation methods are defined in standards such as ISO 105-J03 and ASTM D2244.
The limit is yours, not universal. There is no industry-wide acceptable ΔE, because acceptable depends on the shade, the substrate and the end use. Deep saturated shades are harder to match than pale ones. Small parts read differently from large panels.
What buyers commonly write into specs looks roughly like this. Treat it as a starting position for a conversation, not a rule:
| ΔE (CIEDE2000, indicative) | Typical commercial reading |
|---|---|
| ≤ 1.0 | Tight match; often required for solid-dyed programs and adjacent panels |
| 1.0 – 2.0 | Commonly accepted for bulk against an approved standard, shade dependent |
| > 2.0 | Usually visible side by side; needs a decision, not a default |
Where Sleepwear Fails Specifically
Sleepwear has three characteristics that make color harder than it looks. See our fabric specification library for the parameters that sit alongside color.
Multiple knit structures in one garment. A pajama set is often jersey for the body, rib for the cuffs and neckband, and a different knit for the shorts. Those structures take dye differently even in the same bath. If you specify one color, you must also specify whether the rib may sit at the edge of the tolerance band or must be matched exactly.
Trims are a different fibre. Elastic, drawcord, thread and labels are frequently polyester or a blend while the garment is cotton or modal. They will not match. The fix is not to demand a match — it is to state the exemption in writing, with its own target, so nobody is surprised at final inspection.
Metamerism. Two samples can match under one light source and diverge under another. This bites hardest on navy, charcoal, olive and warm neutrals, and on blends. Specify the light source for evaluation — a D65 viewing condition is the usual reference — and evaluate every submission under that light, every time.
How to Build a Color Approval Workflow
| Step | What happens | Who signs off |
|---|---|---|
| 1 | Agree the color standard; both sides retain a physical reference | You |
| 2 | Request lab dips — typically 2–3 options, one light, one dark, one closest | Mill |
| 3 | Evaluate dips against the standard under the specified light source | You |
| 4 | Record the approved dip, with ΔE, formula and direction | You |
| 5 | Bulk dyeing; mill submits a shade band with L\, a\, b\* values | Mill |
| 6 | Approve bulk; seal and retain a sample against future lots | You |
| 7 | Inline check at cutting and rolling; final check at inspection | Mill + your QC or third party |
Seven steps, and only two of them are the mill’s to decide. The rest sit with you — which is why color problems are usually specification problems.
What to Put in Your Tech Pack for Color
A tech pack line reading “color: navy” has specified nothing. The line needs to carry five things:
- The target reference — a physical standard, plus a named color reference
- ΔE tolerance, the formula used, and whether direction is constrained
- The light source for evaluation
- Which components must match the main body, and which are explicitly exempt
- How many submissions are expected at dip stage, and who carries re-dip costs
Put those five lines into your tech pack and most color disputes never start.
One Honest Note
Color matching has a floor set by physics, and no factory can go below it.
When a buyer asks two different knit structures in the same deep saturated shade to match within a ΔE the process cannot deliver, the honest answer is to widen the tolerance or shift the shade — not to promise the number.
We would rather say that at the swatch stage than have you reject a container.
FAQ
Q: How many lab dips should I expect at the first submission?
A: Two to three is normal — a slightly light, a slightly dark and a closest match. Fewer than two usually means the dyehouse chose for you instead of letting you choose. Re-dips after approval are typically charged.
Q: Can I approve color from a photo?
A: No. Photos depend on camera, lighting and screen, none of which are controlled. Use photos to catch gross errors. Approval needs the physical submission against your standard, under your specified light.
Q: What if the bulk arrives outside the tolerance?
A: It depends on what your spec and contract say. If you wrote a tolerance, a formula and a direction, you have a defensible position: re-dye, replace, or accept with a price adjustment. If your spec said “navy”, you have a conversation.
Q: Does one shade band cover multiple dye lots?
A: No. Each dye lot gets its own band. Within a lot the variation is usually small; across lots it is larger, and that is where most complaints originate. On a repeating program, require each new lot to be matched to the retained standard — not to the previous lot.
CTA
Send us your color standard and target shade. We will return a lab dip submission with ΔE values, the formula used and the direction of any deviation, so your team approves a number instead of a feeling.