How to Build a Fabric Defect Photo Standard: Names, Measurement, Severity & Roll Decisions
Share
A useful fabric defect photo standard does four things separately: it names what is observed, shows how it is measured, records the viewing conditions and states the order-specific decision rule. A photograph alone cannot decide whether a roll passes. Buyers and suppliers must agree how the observation affects grading, cutting, repair, segregation or rejection.
This distinction matters for textured and patterned home-textile fabrics. For example, FAB-0561 printed polyester–spandex velvet has print, pile direction and extensibility to control, while FAB-0120 cotton jacquard has a woven pattern and easy-care finish. The same photographing setup will not reveal every risk on both surfaces.
Why is a defect name not the same as a rejection decision?
ISO 8498:1990 supplies vocabulary for describing woven-fabric defects, including general, yarn, warp-direction, weft-direction, dyeing, printing, finishing and selvedge-related defects. Its official abstract also makes an important boundary clear: the presence of a described characteristic does not automatically mean that the fabric is sub-standard. The commercial decision still depends on the buyer’s product, location, frequency, severity and agreed acceptance rule.
A practical defect manual should therefore keep three fields separate:
- Identity: what is visible—for example, a hole, snag, print-registration shift, pile disturbance or selvedge damage.
- Extent: size, length, area, recurrence, direction and position within the usable width.
- Disposition: mark, cut around, repair, downgrade, segregate, investigate or reject under the agreed contract.
Do not call every unusual appearance a “defect.” On velvet, reversed nap or a changed viewing angle can create apparent shade variation. On jacquard, intentional weave relief can resemble a surface irregularity. First compare the area with the approved construction and visual reference under controlled conditions.
What information should appear in every reference photograph?
| Field | What to record | Why it matters |
|---|---|---|
| Reference ID | Unique code, version and date | Prevents old images from governing a new order |
| Product context | SKU, colourway, pattern, finish and face | The same appearance can mean different things on another fabric |
| Direction | Machine/warp, cross/weft and pile direction where applicable | Shows whether the issue follows yarn, print or nap orientation |
| Scale | Ruler or neutral scale card in the same plane | Allows size to be understood without perspective guesswork |
| Location | Distance from usable edge and position along the roll | Supports cutting and recurrence decisions |
| Viewing setup | Lighting, camera angle, distance and relaxed/extended state | Makes later comparison more repeatable |
| Decision rule | Relevant grading option, zone and buyer-approved disposition | Connects evidence to the contract instead of personal judgement |
Use an overview image to show location and a close image to show structure. Do not crop away the usable-width edge, direction marker or scale. Keep original files and avoid filters that alter colour, contrast or texture. If a defect only appears under grazing light, record that setup and add a normal-view photograph rather than presenting the enhanced view alone.
How should printed velvet be photographed without confusing nap and shade?
For FAB-0561, identify the face and mark pile direction before photographing. Take the comparison with all pieces oriented the same way, then deliberately rotate one reference only if the purpose is to demonstrate nap-related appearance. Photograph print registration and pile disturbance as separate observations.
If stretch affects the motif, record whether the specimen is fully relaxed or held at an agreed extension. The companion article, Printed Polyester–Spandex Velvet for Bedspreads, explains why construction, stretch direction, print geometry and nap require distinct approval fields. A defect manual should not convert an unverified “colorfast” or “warm” catalogue phrase into a performance result.
How should grading rules be connected to photographs?
ASTM D5430-26 provides test methods for visually inspecting and grading fabrics and supports numerical designation from visual inspection. ASTM notes that different point-assignment options can produce different reported values and that the same option should be used when purchaser and supplier results are compared. The buyer must therefore name the selected option, inspection width, calculation basis and acceptance limit; “use the four-point system” without the governing details is incomplete.
The photo library should illustrate the agreed grading rule, not quietly replace it. For each example, state whether the picture represents an identity example, a measurement example, an acceptance boundary or a clearly unacceptable condition. Boundary examples are especially useful for calibration, but they must be approved by both parties.
Use the pre-shipment roll-inspection guide for inspection coverage, roll records and shipment decisions. The photo standard is narrower: it aligns observers on how evidence is captured and interpreted.
When does ISO 2859-1 apply—and when does it not?
ISO 2859-1:2026 defines acceptance-sampling plans for inspection by attributes. It does not automatically determine how many metres of every roll should be visually inspected, nor does it create an AQL for Oulai fabrics. If a buyer applies attribute sampling to cartons, finished pieces or another defined lot, the lot, unit, characteristic, inspection level, AQL and switching rules must be stated.
Do not mix an attribute-sampling decision with a continuous roll-grading score unless the relationship is written into the specification. A roll can have a numerical point result, mapped defect locations and a separate lot-level sampling decision; these are different records.
How should inspectors be calibrated before bulk inspection?
- Select representative accepted, boundary and unacceptable examples for the actual product family.
- Confirm terminology, measurement and point-assignment rules with buyer and supplier.
- Ask each inspector to grade the same hidden sample set independently.
- Compare disagreements by defect identity, extent and disposition—not only final pass/fail.
- Revise unclear images or rules, then repeat the exercise.
- Record the manual version used for each production lot and inspection report.
When a dispute or delivery claim occurs, retain original images, roll IDs, timestamps, scale references and the governing manual version. The fabric quality-claim workflow explains quarantine, evidence preservation, retesting and CAPA. The textile lot-traceability guide helps connect those photographs to the correct roll and shipment.
Buyer checklist
- Is each defect term defined and, where relevant, aligned with ISO 8498 terminology?
- Do photographs show product, direction, scale, location and viewing conditions?
- Are identity, severity and disposition recorded separately?
- Is the ASTM D5430 option or other agreed grading rule named precisely?
- Are print, pile, shade and structural observations kept distinct?
- Are attribute-sampling rules separated from roll-grading calculations?
- Does every report identify the manual revision and roll or lot?
Nantong Oulai Trading Co., Ltd. can use a buyer-approved photo standard with the relevant live fabric, sealed sample and roll records. The standard should be customized to the ordered product; no photograph establishes a universal severity level across all constructions and end uses.
FAQ
Does ISO 8498 decide whether a woven defect is acceptable?
No. It provides defect vocabulary. Acceptance depends on the order-specific grading and disposition rules.
Can a defect photograph replace a physical approved sample?
Usually not for colour, hand, pile or subtle construction effects. Use photographs for location and evidence, while retaining a physical reference where those attributes matter.
Is the four-point system the same everywhere?
No. The point-assignment option, calculation basis and acceptance limit must be identified. ASTM specifically warns that different options can produce different results.
Does an AQL define how much fabric to inspect on every roll?
No. ISO 2859-1 addresses defined attribute-sampling plans. Roll inspection coverage and fabric grading need their own stated rules.