Color consistency is managed by measuring the difference between a reference color and the produced color in the CIELAB color space and reducing it to a single Delta E (ΔE) number; the smaller the number, the better the match. There are several methods for calculating Delta E (CIE76, CMC(2:1), CIEDE2000), and a tolerance value is meaningless unless you know which one was used. A dye house measures color with a spectrophotometer, confirms it visually in a standard light booth, and carries it consistently from lab-dip approval to bulk production.
What is the CIELAB color space and why is it used?
CIELAB (L*a*b*) is a perceptually approximately uniform color space that positions a color along three axes. These axes turn a color into numerical coordinates:
- L* — lightness (0 black, 100 white).
- a* — the red (+) to green (−) axis.
- b* — the yellow (+) to blue (−) axis.
The same color can also be expressed in cylindrical coordinates: C* (chroma, a color's saturation/vividness) and h (hue, the color tone angle). This is exactly where the value of color management begins: instead of subjective phrases like "a bit bluer," the target and the sample each settle into an L*a*b* coordinate and the difference between them becomes measurable.
How is Delta E calculated? The difference between CIE76, CMC(2:1) and CIEDE2000
Delta E gives the total difference between two colors, but the method for calculating that difference has evolved over time. The critical point is this: these methods do not produce interchangeable numbers on the same scale; when specifying a tolerance, you must also specify the formula.
- CIE76 (ΔE*ab) — The oldest method. It is the simple Euclidean (resultant) distance of the differences along the CIELAB axes (L*, a*, b*). It is easy to compute but does not match human visual perception in every region; particularly with saturated colors it can produce numbers larger than what is actually seen.
- CIE94 and CMC(l:c) — They improve perceptual agreement by adding different weights to lightness, chroma and hue differences. CMC(l:c) was developed by the SDC (Society of Dyers and Colourists) and was adopted as an ISO standard in 1995. In the CMC(2:1) variant, the classic workhorse of textiles, twice as much tolerance is allowed for lightness as for chroma (l=2 lightness, c=1 chroma); this is based on the logic that lightness deviation in fabric is less objectionable than tone deviation.
- CIEDE2000 (ΔE00) — Today the standard regarded as the most perceptually accurate. It is not a Euclidean distance; in addition to lightness/chroma/hue weights, it includes terms such as the interaction between hue and chroma and an a* correction near neutral colors. It is tuned to give results closest to visual perception, especially in blue and dark color regions.
Practical upshot: for the same two colors, CIE76 and CIEDE2000 give different numbers. That is why, when a laboratory says "ΔE 1.0," it must specify which formula it used (for example CMC(2:1) or CIEDE2000); otherwise the value cannot be compared. For the relationship between Delta E and fastness tests, see the color fastness and Delta E page.
What does a Delta E value mean? Visual threshold interpretation
Delta E is a number; what makes it "good" or "bad" is how that number is perceived by the human eye. The table below summarizes the visual threshold ranges commonly used with trained observers. These ranges are a general reference; the acceptance criterion can tighten depending on the product, customer and color region.
| Delta E range | Visual perception | Typical interpretation in textiles |
|---|---|---|
| <= 1 | Imperceptible to most eyes | Tight color target; no difference seen when placed side by side |
| 1 - 2 | Only a trained eye distinguishes it | Noticeable under controlled conditions, accepted in most applications |
| 2 - 3.5 | Distinct difference on careful inspection | Commercial acceptance limit; accept or reject depending on product/customer |
| > 5 | Clearly different color perception | Counted as a different color; generally rejected / re-dyed |
Why are these thresholds given as ranges rather than absolutes? Because perception depends on the region the color occupies and on the observer; the same ΔE value may be imperceptible in a light pastel yet visible in a dark tone. Formulas like CIEDE2000 add weighting terms for exactly this reason.
What is metamerism and why are multiple illuminants needed?
Metamerism is when two samples match each other under one light but do not agree under another light. This is the point where color management breaks down most often when the observation condition is not controlled. When comparing two colors, two things must be fixed:
- Illuminant (light source): CIE standard illuminants are used — D65 (daylight), A (incandescent lamp/tungsten), F2 (fluorescent). Samples are checked separately under each of these sources.
- Observer: a 2 degree or 10 degree standard observer angle is selected; color values come out differently for small and wide fields of view.
That is why approval and inspection are done in a standard light booth and under multiple illuminants: if a sample that matches under D65 diverges under A light, the match is metameric and will cause problems in the field. The definitive solution to metamerism is to use the same colorants in the reference and in production; when the same color is achieved with a different colorant combination, the risk of metamerism is highest. For the relationship between colorant selection and the color–fiber bond, see the reactive and disperse dyeing page.
How is lab-dip approval carried over to bulk production?
Color is first approved at small scale (lab-dip), then transferred to bulk production; the real challenge is preserving the same color across these two scales. A lab-dip-approved recipe may not give exactly the same result when the fabric type, weight, machine and process parameters change; for this reason bulk production lots are also measured against the standard with Delta E. For details of the lab-dip flow, see the lab-dip approval process page.
In bulk production, color deviation appears in two typical patterns and is caught by measurement:
- Tailing (head-to-tail difference): A color shift between the start and the end along a roll or lot. In continuous processes it is detected by tracking the ΔE between start and end.
- Listing (edge-to-center/edge-to-edge difference): A color difference across the width of the fabric, between the edges and the center or between the two edges. It is controlled by taking measurements at multiple points across the width.
Catching these deviations early makes it possible to correct the color error before it is carried into the subsequent cutting-and-sewing stage. Color consistency is monitored together with other quality parameters such as dimensional stability; for related topics you can see the fastness tests (ISO and AATCC) and dimensional stability and spirality pages.
What does KARCEM's DeltaE < 1 target mean?
The DeltaE < 1 color control target stated by KARCEM aims for the produced color to be so close to the reference color that, when placed side by side, it cannot be distinguished by most eyes; in the threshold table above, this corresponds to the "imperceptible" range. This target is not a single-point measurement but expresses a control discipline: a recipe kept consistent from lab-dip to bulk production, multi-illuminant control in a standard light booth, and keeping the lot homogeneous in color through head-to-tail (tailing) and edge (listing) measurements. When a single coordinator owns the chain — in-house knitting plus a vetted contract network for dyeing, printing and finishing — it is easier to track and correct the source of color deviation within the coordinated process rather than searching for it in a fragmented supply chain. For the advantage of integration, see the coordinated contract network advantage page.