Fabric Quality Is a Whole: Defect and Color Control Together
Updated: 3 days ago
Catching a hole, a stain or a knitting fault is only half of quality control. The other half is often overlooked: color.
When people think of fabric quality control, defects come to mind first: holes, drop stitches, weft faults, stains, broken needles. Catching these is critical — but quality is not only that. Even a fabric without a single hole will cause problems when it reaches the customer if it is not consistent in color from roll to roll and from side to side. In the cutting room, two panels sewn next to each other won't match in tone, different rolls of the same order look different, and the batch gets rejected at intake.
Yet in most mills, color is still managed by a subjective decision. The operator cuts a piece, carries it to the light booth and compares it against a reference by eye. This method is nothing more than sampling; the meters of fabric between two samples flow past without ever being measured. In this article we focus on a single question: can color consistency, just like defect detection, be tied to a measurable, recorded and reportable standard?
Color: The "Invisible" Risk in Measurable Color Consistency
A hole or a stain is obvious to the eye — it is either there or it isn't. Color is far more insidious, because the drift is slow and relative. A tonal shift that goes unnoticed across a few meters suddenly appears when the beginning and the end of a roll are placed side by side. Color risk builds up in three main forms:
Roll-to-roll difference: Different rolls of the same order diverge because of different dye-bath and batch conditions.
Side-to-side (wing) difference — right/center/left: A tonal difference forms between one edge of the fabric and the other, and on wide fabric this is one of the most frequently missed problems.
Within-roll drift: Color shifts slowly from the start of a roll to its end, and the difference becomes visible only when the two ends are compared.
When all three of these shifts are managed by eye and by sampling, they are among the hardest quality problems to catch. The eye tires, the light changes, and the decision varies with the operator's experience. What's more, the fabric between two sample points is never measured at all.
Is the Light Booth Enough? Sampling vs. Continuity
The light booth is not the wrong tool; but what it does is an instantaneous sample. Pieces cut from specific points are compared against a reference under standardized light. By its very nature this method has two limits. First, there is no continuity: only the cut points are measured, and the fabric in between is left to assumption. Second, the decision is subjective, because it depends on the operator's eye and the lighting at that moment — it varies by shift.
This picture is exactly the transformation defect detection went through. Moving from finding defects by eye to a numerical standard like the 4-Point system took quality out of the realm of debate and tied it to data. The same leap is possible for color — and the name of that leap is ΔE: subjective comparison gives way to a measurable number.
LIGHT BOOTH | COLORTRON | |
Measurement method | A piece is cut and carried to the booth | On the line, in the flow |
Coverage | Point sampling | Continuous along the roll |
Decision | Operator's eye / experience | ΔE — a numerical standard |
Wing (side-to-side) difference | Usually not measured | Right–center–left, along the roll |
Record | Manual / incomplete | Lotted + reported (XML) |
Threshold exceeded | Noticed afterwards | Immediate stop + alert |
Light booth (sampling) vs. Serkon.AI Colortron (inline)
Measurable Color: What Is ΔE?
ΔE (Delta E) is the numerical measure of the difference between two colors. The larger the value, the more pronounced the difference; the closer it is to zero, the more identical the colors become. It is generally accepted that the difference the eye can distinguish begins around ΔE 1; below that lies a region the eye cannot yet resolve but an instrument can measure. This is exactly the critical point: a measurable standard sees beyond the limit of the eye. The difference between saying "looks fine to me" and saying "the ΔE value is below threshold" turns color quality from an opinion into data, and ties the quality decision to the system rather than to a person.
Colortron: Inline Color Measurement, Continuous Along Every Roll
Serkon.AI Colortron measures color on the fabric line as it flows, without carrying it to a booth. A traversing spectrometer moves across the fabric width and reads color continuously along the roll. Its key features:
Measurement runs without stopping the flow; no need to cut samples or carry them to a booth.
Repeatable accuracy down to ΔE 0.05.
A LED-based D65 (6500K) standard light source, so the measurement condition is always the same and unaffected by ambient light.
Spectral range of 400–700 nm: color is read across the entire visible spectrum.
Comparison across a wide color space with ΔE CIE, ΔE CMC 1:1 and ΔE CMC 2:1 calculations.
For each roll, side-to-side (right/center/left) difference and color difference are measured; the traversing head reads across the width, follows these values continuously along the roll and plots them on a live ΔE graph.
Adaptive edge detection automatically finds the fabric edge and locks the correct measurement point, so the wing reading is always taken from the right place.
Temperature compensation and automatic calibration work together: an ambient sensor offsets temperature-induced drift, and the system recalibrates at regular intervals to hold the measurement condition constant.
All rolls are lotted and reported against a ΔE reference.
When the defined threshold is exceeded, the system stops and issues an alert — so the problem is caught the moment it forms, not meters later.
In short, Colortron takes the objective color comparison that the light booth performs and moves it out of sampling into continuity. There is no longer any need to cut a piece and walk to the booth; color is followed continuously along every roll, inside the flow.
Defect Detection + Color Control: Holistic Quality on One Line
The real power emerges when these two layers combine. Serkon.AI defect detection (Q2) catches surface holes, stains and weaving/knitting faults; Colortron measures color on the same line — so surface, wing and color are monitored together. Colortron can also be added to an existing conventional inspection machine: whatever the winding type (fold-to-roll, roll-to-roll, batch-to-roll), it works as a plug-in without changing operating habits.
Moreover, Q2 and Colortron are monitored from a single control screen. Because the process of both machines is gathered in the same interface, the operator manages the entire quality picture from one point — and this makes process management noticeably easier.
Color control stops being an interpretation tied to the operator's eye and becomes measurable, reportable data.
Frequently Asked Questions
Does it completely eliminate the light booth?
Colortron measures color continuously on the line; it moves the instantaneous sampling the light booth performs into continuity. You can keep using the booth for your reference and approval processes, but the color control of all rolls now becomes recorded data rather than sampling.
What exactly does ΔE measure?
The measurable difference between two colors. For each roll, Colortron calculates the side-to-side and color difference in ΔE (ΔE CIE, ΔE CMC 1:1 and 2:1), then lots and reports against the tolerance threshold you set.
Does it work with my existing machine?
Yes. Colortron can be added to an existing conventional inspection machine or to the stenter (ram) exit; the winding type doesn't matter, and it works as a plug-in without changing operating habits.
Does it work together with defect detection?
Yes. Serkon.AI defect detection (Q2) handles surface faults and Colortron measures color; on one line, surface + wing + color are monitored and recorded together.
