The framing device for this guide
Say it
Brief the colour in a language the factory can’t misread
Understand drift
Know why screens and fabrics shift the same colour
Approve it
Sign off the lab dip under the right light
Hold it
Keep colour steady across every fabric lot
Lock it
Seal the standard before bulk dyeing starts
Color is the single thing buyers notice first and forgive least. A hoodie that’s two shades too warm doesn’t feel “close enough.” It feels wrong on the shelf, wrong in the photos, and wrong to the customer who ordered the color they saw online.
Here’s the uncomfortable truth: color rarely goes wrong because someone was careless.
It goes wrong because color has to travel. It passes from your head to a screen to a code to a dyed swatch to a vat of dye to roll after roll of fabric.
Every one of those handoffs is a chance for the color to shift.
We call that journey The Color Chain.
This guide walks through all five links, in order.
By the end, you’ll know how to communicate a color so it can’t be misread, why the same color looks different on a screen and on cotton, how to approve a lab dip that actually holds, and how to lock the color down before bulk so the run comes back matching what you signed off.
This is a foundation guide for founders placing custom manufacturing orders. It’s written for people who don’t work in a dye house; no chemistry degree is required.
Chapter 1 · Link in the chain
Before a color can be matched, it has to be named in a way both sides read identically. “Burnt orange” is not a color. It’s an argument waiting to happen.
There are three levels of precision for telling a factory what color you want. Only one of them is safe for production.
“Forest green.” A hex code like #0F4C81. A screenshot. A photo from your phone.
These are fine for a mood board. They are useless as a production standard.
A hex code describes light coming out of a screen.
Fabric color is light reflected off dye. They are two different physical things, and no screen color translates cleanly onto cloth.
More on exactly why in the next chapter.
Pantone is the standard color language of the apparel trade. Instead of “navy,” you specify a number that both you and the factory can look up on an identical physical card.
For clothing, the system you want is Pantone TCX Textile Cotton eXtended. Pantone’s Fashion, Home + Interiors library has around 2,800 colors, each one physically dyed onto 100% cotton fabric, so what you see is close to what the fabric will actually do.
A TCX code looks like this: 19-4052 TCX. The digits aren’t random; the first pair is lightness, the middle pair is hue, and the last pair is chroma (how vivid or dull).
TCX is dyed on cotton. It’s the one to specify for garment fabric. TPG (Textile Paper Green) is the same numbering printed on cheaper paper, useful for trims, plastics, and desk reference, but paper reflects light differently from cloth, so the same number looks a touch brighter on TPG than it will on your hoodie. If you approve on paper and the factory dyes to cotton, you’ve built a mismatch from the start. Older TPX paper guides were retired years ago don’t work from a legacy fan deck.
One more rule: give the code and a physical reference wherever you can.
A Pantone number plus a real swatch you’ve both held removes almost all the guesswork.
Screens can’t be trusted for the final call; one study found monitors drifting up to ±2.8 in color-difference units on reds alone, which is well outside a normal production tolerance.
Note: TCX is calibrated for cotton. Polyester and nylon take dye differently and have their own Pantone swatch systems.
If your fabric is a poly blend, tell the factory the base fiber; the same code behaves differently on it.
A Pantone code gets you close. A lab dip gets you exactly that.
A lab dip is a small swatch of your actual fabric, dyed by the mill to hit your target color.
You approve the physical swatch before a single meter of bulk is dyed.
It’s the moment the color stops being a number and becomes a thing you can hold in your hand.
The parallel for prints and all-over patterns is a strike-off, a test print of your artwork on the real fabric, so you can approve the ink color, the placement, and the hand-feel before bulk printing. Embroidery has its own version, called a sew-out.
From our dye house
When a founder sends us a TCX code and a physical swatch, our first lab dip usually lands within tolerance. When we get only a phone photo or a hex code, we’re guessing at a target and guessing costs everyone a round. The clearer the brief, the fewer rounds you pay for.
Getting the colour brief right sits inside your wider sampling and brand work. If you’re still setting up how your samples move from proto to sealed standard, our guide to the clothing sampling process shows where lab dips fit into the ladder. And if you’re trying to make sure your brand’s colours survive the jump from screen design into physical product, that starts with building a brand identity that translates to manufacturing.
Chapter 2 · Link two of the chain
This is the chapter that answers the title. Colors don’t match for reasons that are physical, not sloppy.
Understand the five culprits, and most “the color is wrong” surprises stop being surprises.
Here’s the core idea to hold on to. Color is not a property that lives in an object.
It’s what happens when a light source, a surface, and your eye meet.
Change any one of the three, and the color changes with it, even when nothing about the dye has moved.
Your monitor mixes red, green, and blue light and shines it at your eye. Fabric has no light of its own; it soaks up some wavelengths and bounces the rest back. There is no reliable way to turn “glowing pixels” into “dyed cloth. That’s why a hex code can never be a production standard.
Fiber, weave, and finish all change how dye sits and how light bounces. The same formula reads deeper on brushed fleece than on a smooth, flat weave and different again on cotton versus polyester because each fiber grabs dye its own way. This is exactly why a lab dip must be on your fabric, not a stand-in.
This is the big one. Two colors can match perfectly under one light and clearly clash under another. A gray that matches in daylight can split apart under warm shop lighting. So “does it match?” is an incomplete question. The real question is: match under which light?
Dyeing is a wet, hot, timed reaction. Water, temperature, dye batch, and run time all nudge the result. A good dye house controls these tightly, but “tightly” is not “identically.” Two dye baths of the same recipe will land very close, not pixel-perfect. That margin is normal, and it’s why we set a tolerance instead of chasing zero.
Many fabrics carry optical brightening agents that glow faintly under the UV in daylight to look “whiter than white.” Under shop lighting with little UV, that glow disappears, and the color can look duller or warmer. Whites and pale brights are the most affected.
Culprit two deserves a closer look, because it’s the one founders trip over most. Different fibers take dye through completely different chemistry. Send one Pantone code to a cotton fabric and a polyester fabric, and you will get two different-looking results, not because anyone erred but because the dye classes are different.
| Fibre | Usual dye class | What to expect on colour |
|---|---|---|
| Cotton | Reactive | Deep, rich, matte. Bonds into the fiber, so good wash fastness. The TCX standard is built on cotton, so codes read most predictably here. |
| Polyester | Disperse | Can read brighter and cooler than the same code on cotton. Needs its own color standard; don’t approve a poly color off a TCX cotton chip alone. |
| Cotton/poly blend | Two dyes, two steps | Each fibre is dyed by a different class, so a blend can look slightly two-toned or “heathered” up close. Common in fleece; expect it, spec for it. |
| Nylon | Acid | Takes color readily and can look vivid; fastness varies by dye choice. Common in trims and technical panels. |
| Viscose/rayon | Reactive | Deep and lustrous wet, but it can shift a touch after the first washes. Wash-test the color, not just the shrinkage. |
The practical rule: tell the factory your exact fabric and base fiber when you brief a color, and always approve the lab dip on that fabric.
A color is a promise about a specific cloth, not a number in the abstract.
You can’t remove physics from a dye house. What you can do is pin down the conditions under which color gets judged so both sides are always looking at the same thing.
That means one agreed light, one physical standard, and one written tolerance. The next three chapters turn that into a process.
Plain-English takeaway
If a color “doesn’t match,” the first question is never “Who messed up?” It’s “Under what light, on what fabric, against what standard were we comparing?” Nine times in ten, the answer is hiding in those three words.
Chapter 3 · Link three of the chain
Approval is where color gets decided. Do it under a phone camera in your kitchen, and you’ve approved nothing.
Do it properly, and you’ve built a standard the whole run answers to.
Professional color approval happens in a light box, a small viewing cabinet with switchable, standardized lamps. You place your lab dip next to your reference and look under each light in turn. The industry-standard set is worth knowing:
| Light source | What it simulates | Why you check under it |
|---|---|---|
| D65 | Average daylight (~6500K) | The primary reference. Also, the white point of most screens and web images, so it keeps fabric and photos aligned. |
| TL84 / F11 | Store and office fluorescent (~4000K) | How your product will look in a shop. Catches metameric shifts under retail light. |
| A (Incandescent) | Warm home/tungsten (~2800K) | How it looks in a living room or a warm-lit photo. A common place for grays and navies to break. |
| UV | Ultraviolet component of daylight | Reveals optical brighteners essential for whites and pale colors. |
You don’t need to own a light box as a founder. You need to ask the factory to approve it under one and to state your primary light. Most apparel work is judged with D65 as primary, then checked under A and TL84 for metamerism. If a lab dip matches under D65 but splits under store light, you flag it now, not after 500 units are dyed.
“Close enough” is subjective. The trade measures color difference with a single number called Delta E (written ΔE). Roughly, a ΔE around 1 is a difference most people can’t see; the bigger the number, the more obvious the gap.
Typical apparel practice is to accept shell fabric within about ΔE 1.0–2.0 of the standard for visible parts, sometimes looser on linings or contract goods.
You don’t have to calculate this yourself, but you should agree to the number in writing so “pass” and “fail” aren’t a matter of opinion. Confirm GIBBEN CLOTHING’s working tolerance and put it on the PO.
Vague feedback triggers a guess, and a guess costs a round. Anchor every comment to the standard and a direction.
A clear brief with a good TCX reference often approves in one to three lab-dip rounds. Tricky colors, bright fluorescents, deep blacks, exact greys, or an unusual fiber blend can take more.
Each round has a real cost and adds days to the calendar, which is the honest reason a strong Chapter 1 brief pays for itself here.
From our sample room
The cheapest color correction is the one you make on a swatch. The most expensive is the one you notice on a pallet of finished garments. Every hour spent on lab dip approval is buying down risk on the whole run.
Chapter 4 · Link four of the chain
You approved one perfect swatch. Now the run needs hundreds of meters of fabric.
The new risk isn’t matching your standard; it’s matching itself, roll to roll.
Fabric is dyed in batches. Each batch is a dye lot.
Because dyeing has that natural chemical margin from Chapter 2, two lots of the same color will be extremely close but not identical.
Mills print a lot of numbers on every roll for exactly this reason.
The danger shows up when two lots end up in the same garment, a hood cut from lot A and a body from lot B, or side by side in the same drop. Individually, each is “in tolerance.” Next to each other, the eyes catch the step.
Color tolerance is measured against the standard, but the human eye judges color against whatever is next to it. Two rolls can both pass versus the sealed swatch and still look mismatched beside each other. Managing lots is about controlling those neighbors.
The cleanest fix is to not create the problem. If your whole order can be dyed in a single lot, color variation between rolls nearly disappears.
This is easiest on smaller runs and is one of the underrated advantages of ordering fabric for the full quantity up front rather than in dribs.
Mills usually have a minimum dye quantity per color, often in the region of a few hundred kilograms of yarn or fabric (commonly ~200–300 kg), so very small colors can force a compromise. Confirm GIBBEN CLOTHING per-color dye minimums.
Each incoming lot gets compared back to your approved standard, under the agreed light, before it’s cut.
A lot that drifts out of tolerance gets rejected or re-dyed; it never quietly enters the run.
When a run does span multiple lots, the factory keeps lots separated on the cutting table so a single garment is always built from one lot.
Rolls can also be shade-banded and grouped by micro-shade so the mildest variation is spread across the order rather than concentrated in one pocket of it.
If this is a color, you’ll reorder a signature brand shade the sealed standard from this run becomes the continuity reference for the next one.
It’s how a brand keeps “our black” looking like our black across drops, seasons, and even different fabrics.
From our dye house
On a repeat brand color, we keep your sealed standard on file and dye every future lot back to it, not to the last batch we made. Matching to the previous lot lets color creep a little each time until, five drops later, it’s visibly moved. Matching to the original standard every time is what stops the drift.
Fabric choice sits underneath all of this; some fibers and constructions hold color far more predictably than others. If you’re still choosing your base cloth, our guide to sourcing fabric for streetwear covers how fiber and construction affect what color will do.
Chapter 5 · Link five of the chain · The Detailed Checklist
Everything so far has been preparation
This is the gate. Once bulk dyeing starts, color is expensive to fix.
Cross this checklist first, and then the run comes back matching what you signed.
Locking color isn’t one action; it’s a short sequence of confirmations that turn your approved lab dip into an enforceable production standard.
We call it The Lock-Before-Bulk Gate. Walk it once, deliberately, before you release the dye order.
the deliverable
Every item is a confirmed “yes” before bulk dyeing begins. Don’t release the order on a maybe.
The approved lab dip is physically sealed, labelled with the color name/code and date, and signed by both sides. A counter copy stays with the factory.
The standard was dyed on the actual bulk fabric and fiber, not a substitute swatch.
Primary light named (usually D65), plus the secondary lights for metamerism checks. Written down, not assumed.
The acceptable ΔE (or “commercial match to sealed standard”) is stated on the PO, so pass/fail isn’t an opinion.
Single lot where possible; if multiple, lots are cut separately, and each is checked back to the standard.
Strike-offs, embroidery sew-outs, and dyed trims (rib, drawcord, zip tape, and labels) are approved against the same standard color that lives on more than the shell.
Colorfastness checks if the color survives washing without bleeding, crocking (rubbing off), or fading beyond tolerance.
A unit from early bulk gets pulled and compared to the sealed standard before the full run continues.
Notice what this gate really does. It removes every “we assumed” from the process. After it, there is no debate about what the color should be there’s a signed physical object, an agreed light, and a written tolerance. Disputes get settled by holding a garment next to the sealed standard in the D65 box. That’s it.
Matching is only half the job. A color also has to stay.
A perfect navy that bleeds in the first wash or rubs off onto a white sofa fails the customer just as badly as a mismatch.
That property is called colorfastness, and it’s the reason “Wash test passed” sits on the gate above.
Fastness is checked with standardized lab tests and scored on a simple 1–5 grayscale; 5 means no change, and 1 means severe change.
You don’t run these yourself, but you should know what to ask a factory to test and pass:
| Fastness test | Answers the question | Standard (ISO / AATCC) |
|---|---|---|
| Wash | Does it fade or bleed onto other garments in the laundry? | ISO 105-C06 / AATCC 61 |
| Rubbing (crocking) | Does color rub off onto skin or other surfaces dry and wet? | ISO 105-X12 / AATCC 8 |
| Light | Does it fade in sunlight over time? | ISO 105-B02 / AATCC 16 |
| Perspiration | Does sweat change or lift the color? | ISO 105-E04 / AATCC 15 |
| Water | Does plain water exposure shift it? | ISO 105-E01 / AATCC 107 |
Typical commercial targets sit around wash 4+, dry rub 3–4, wet rub 2–3, and light of 4+ for visible apparel, with tighter numbers for children’s wear. Confirm the grades GIBBEN CLOTHING tests as standard. Two practical rules keep this clean:
From our dye house
A color that matches beautifully but crocks onto a customer’s white tee generates returns and one-star reviews. We’d rather flag a fastness risk at the lab dip stage and adjust the dye or add a fixative than have it surface after the drop ships. Ask your factory to fastness-test before bulk not after.
Color is one lane of a bigger pre-production approval.
The sealed color standard rides alongside your fit, construction, and trim approvals into the pre-production sample, the last full rehearsal before bulk.
Our guide sample, the production approval, and PP samples show how color folds into that final gate, and once bulk is running, AQL inspection is where finished units get checked back against everything you sealed, color included.
The one-line version
Communicate the color precisely, understand why it drifts, approve it under the right light, hold it across lots, and seal it before bulk. Do all five links of the color chain, and “Why doesn’t it match?” stops being a question you have to ask.
Appendix · Keep this handy
The words you’ll hear on a color call in plain English so nothing in an email catches you off guard.
| Term | What it means |
|---|---|
| Lab dip | A swatch of your real fabric dyed to your target color, approved before bulk dyeing. |
| Strike-off | A test print of your artwork on the real fabric to approve print color, placement, and hand before bulk. |
| Sew-out | The embroidery version of a strike-off is a test stitch to approve thread colors and density. |
| Pantone TCX | Textile Cotton eXtended Pantone colors dyed on cotton. The reference system for garment fabric. |
| Pantone TPG | Textile Paper Green: the same numbers printed on paper. Handy for trims and desk reference, it reads better than cloth. |
| Hex / RGB | Screen color values. Fine for digital and mood boards; never a production standard for fabric. |
| Substrate | The material a color sits on: cotton, poly, or paper. The same dye reads differently on each. |
| Metamerism | When two colors match under one light but differ under another. The main reason colors “don’t match.” |
| Light box | A viewing cabinet with standardized, switchable lamps for judging color consistently. |
| D65 | The standard daylight illuminant (~6500K) is used as the primary reference for apparel color. |
| TL84 / A | Store-fluorescent and warm-incandescent lights are used to check for metameric shifts. |
| Delta E (ΔE) | A single number for color difference. Around 1 is barely perceptible; bigger is more obvious. |
| Dye lot | One batch of dyed fabric. Different lots of the same color are very close but not identical. |
| Shade band | Grouping rolls by micro-shade so variation is spread evenly, not concentrated in one part of a run. |
| Colourfastness | How well a color stays put, resisting fading, bleeding, and rubbing. Scored 1–5. |
| Crocking | Color rubbing off onto another surface, tested dry and wet. |
| OBA | Optical brightening agents make whites glow under UV and can shift color under low-UV shop light. |
| Sealed standard | The approved, signed, sealed color reference that bulk is measured against. Your color in physical form. |
| Continuity standard | The sealed standard is kept on file so repeat orders match the original, not the last batch. |
A lab dip is a small swatch of your actual fabric that the mill dyes to hit your target color. You approve the physical swatch before any bulk fabric is dyed, so color is confirmed on real cloth, not on a screen or a paper card. Once approved and sealed, it becomes the standard the whole run is matched against.
It’s the right starting language, but not a guarantee on its own. A Pantone TCX code gets the factory close, and it removes vague words like “navy.” But the final color still depends on your specific fabric, the dye chemistry, and the light it’s judged under. That’s why you confirm a Pantone code with a physical lab dip on your real fabric before bulk.
This is called metamerism. A color is the result of a light source and a surface, and your eye changes the light, and the color can change with it. Two colors that match in daylight can visibly clash under warm store or home lighting. It’s why professional color approval checks a sample under several standard lights, not just one.
TCX is dyed on 100% cotton fabric and is the reference to use for garment color. TPG is the same numbering printed on paper, cheaper, and handy for trims or desk reference, but it reads slightly brighter than the cloth. For approving fabric color, always work with TCX so you and the factory are looking at the same substrate.
A dye lot is a single batch of dyed fabric. Because dyeing is a chemical process with a small natural margin, two lots of the same color come out extremely close but not identical, and mills number every roll for that reason. The fix is to dye in one lot where possible, cut each garment from a single lot, and check every lot back to your sealed standard.
Delta E (ΔE) is a single number for how different two colors are. A ΔE of around 1 is a difference most people can’t see; larger numbers mean a more obvious gap. Apparel work commonly accepts visible fabric within roughly ΔE 1–2 of the standard. Agreeing on a tolerance number in writing turns “Does it match?” from an opinion into a measurable pass or fail.
No not as your final standard. Screens emit light while fabric reflects it, and no two screens are calibrated the same, so a photo or hex code can be off by more than a normal production tolerance. Use screens for early mood-boarding only. Your production standard has to be a physical lab dip, approved under controlled light.
A clear brief with a good Pantone TCX reference often approves in one to three rounds. Difficult colors, bright fluorescents, deep blacks, exact greys, or unusual fiber blends can take more. Since each round costs money and adds days, a precise color brief at the start is the cheapest way to keep the count low.