How to use a spectrophotometer for fabric color quality control
A spectrophotometer converts the light reflected by fabric into numerical colour data, helping production teams compare a strike-off with an approved sample and identify drift before cutting. Used correctly, it makes fabric color quality control faster, traceable and less dependent on who happens to hold the swatch. The result is not a number alone: it must connect to a defined reference, measurement settings and an agreed decision rule.
For Australian designers, converters and garment makers, the instrument is especially valuable when a colour moves from a supplier’s lab dip to local cutting, sewing or retail. Strong daylight, mixed retail lighting and varied production batches can expose differences that a screen or uncalibrated eye misses. Instrument readings should therefore support, rather than replace, trained visual assessment and a clear production record.
Establish what counts as a pass
Start with the business requirement, not the instrument. Identify the master reference: an approved bulk standard, physical colour card or labelled production swatch. Then define the tolerance against that reference. CIELAB’s ΔE*00, often written ΔE00, expresses the distance between two measurements, but a common limit such as 1.0 is only a starting point. Finished fabric, customer expectations and contractual standards may justify another limit.
| Method | Information obtained | Best suited to | Main limitation |
|---|---|---|---|
| Visual daylight assessment | Overall colour, texture and finish under defined lamps | Final approval and unusual finishes | Subjective and lighting-dependent |
| 45/0 spectrophotometer | Spectral reflectance with gloss largely excluded | Woven and knitted fabric | Transparent fabric and backing affect readings |
| d/8 sphere instrument | Reflectance with selectable specular handling | Coated, shiny or mixed textiles | Configuration must stay consistent |
| Fibre-optic probe | Point reflectance on a non-destructive area | Roll checks and garment panels | Window cleanliness and probe pressure matter |
No single method answers every quality question. A sphere reading can be repeatable while still missing a difference in nap, weave or sheen that matters on a garment. Assess the fabric in its intended orientation under an agreed daylight or retail condition, then use the instrument to quantify repeatable differences. If two shades match under one illuminant but separate under another, flag the metamerism rather than declaring the batch acceptable from one reading.
Choose the right measurement configuration
Most routine fabric checks use a 45/0 geometry: illumination arrives at 0° and the sensor views the sample at 45°, reducing the effect of surface gloss. A d/8 sphere offers another common format and can include or exclude the specular component for sheen, coatings and laminated textiles. Select based on the material and customer method, then document it. Changing geometry later can shift readings even when the fabric has not changed.
Record the illuminant and standard observer with every result. D65 approximates broad daylight, while TL 2/10 represents a warm shop or horizon light; a 10° observer often suits modern display textiles better than a 2° observer. Australian shops may combine cool LEDs, warm track lights and window daylight, so a second viewing condition can reveal faults the primary setting misses. Whichever combination is chosen, keep it fixed for routine approvals and supplier comparisons.
Prepare representative fabric samples
Condition specimens before reading them. AS/NZS 2306, the Australian method for determining textile colour, and the method specified by the customer should govern preparation; many workflows condition fabric near 20 °C and 65% relative humidity. This matters in Queensland, where subtropical humidity can leave cotton or linen holding moisture and shift the result. Record temperature, humidity and conditioning time with the test.
Take readings from representative areas, not the most convenient corner. Sample the beginning, middle and end of a roll, plus any panel, seam or fold likely to change appearance. Keep the fabric flat, use the same backing for translucent material and rotate the sample to check directional variation. Record results individually before calculating a batch average; an average can hide a fault that will appear across one garment panel.
Calibrate and take consistent readings
Calibrate on the instrument’s schedule and verify it at the start of each session. Clean the measurement window, use the correct white and black references, and run a stable green or manufacturer-specified check tile. If a check fails, stop and investigate rather than adjusting a suspect reading. Temperature drift, a dusty port or an expired reference can all look like a fabric problem.
Measure each swatch in the same orientation, with light pressure and without stretching the cloth. Capture at least three readings at separate points, then record the mean and spread. A useful record names the fabric and batch, reference standard, date and operator, instrument serial number, calibration status, geometry, illuminant, observer and conditioning conditions. That detail allows a disputed shade to be repeated months later rather than debated from memory.
Convert measurements into release decisions
Treat repeatability and acceptability as separate checks. First confirm that the instrument is stable and the method is repeatable; then apply a documented commercial tolerance. ΔE00 should be calculated against the correct master, not the previous roll, or small errors will accumulate. Grouping readings by colour, fabric and finish also reveals whether a limit fits the product. Loose knits, saturated darks and textured upholstery may require tighter or broader limits than plain shirting.
Create action bands before production, such as pass, review and reject. A reading in the review band triggers examination of the actual garment, a second instrument configuration or a fresh measurement under controlled conditions. Do not average a failed area into a passing batch. If the fabric matches numerically but looks wrong because of lustre, nap or illuminant effects, hold the stock and escalate to the colour owner with photographs and traceable readings.
Fit colour control to Australian production
Colour control must fit local commercial reality. Melbourne’s fashion precinct and Sydney’s Haymarket garment district often combine small runs, rapid changes and contract manufacturers working from imported stock. A retailer may approve digitally, but cutters still need a signed physical reference that travels with the roll. Recording the mill, dye lot, finishing route and any washing or press treatment prevents an approved shade from being confused with a superficially similar one.
Practical material guidance can help teams connect measurements with fabric construction and finish behaviour; Aurora Design Fabrics’ fabric colour insights is a useful companion resource. Place the approved swatch in the same viewing environment used for final sign-off, not beside a window where direct sun can shift the result. Compare one production roll with both the next and the previous roll so drift is visible before it reaches a customer.
Build a routine the whole team can follow
Use a consistent routine and train everyone who handles colour decisions. These habits make the instrument’s output useful across buying, cutting and production:
- Calibrate and verify the spectrophotometer before each measuring session.
- Condition fabric and measurement references under the agreed conditions.
- Measure several representative positions rather than relying on one reading.
- Use the same geometry, illuminant, observer and backing for comparisons.
- Record the batch, operator, settings, readings and any corrective action.
- Review borderline results with both instrument data and a controlled visual assessment.
Make the next approval count: nominate a colour owner, lock the reference and tolerance, and schedule the first instrument check before the next strike-off or bulk delivery. Place the signed swatch at each cutting table, keep calibration records with the batch file, and release stock only after instrument data and visual review agree.