Why Denim Shade Variation Appears Between Sample and Bulk
The first cartons of a mid-blue style have landed, and the question in the Monday trading meeting is why the pairs on the rail do not match the approved reference, when the pilot looked right and the sealed sample was signed off. This is denim shade variation, and for experienced buyers the useful question is not who to blame but which control was missing: the reference, the fabric lot, the laundry recipe, the pilot or the batch check.
Denim is unusually exposed to this problem. Indigo is a surface dye that sits mostly on the outside of the yarn, so every wash step that removes, moves or adds colour changes the result visibly. A sealed sample is made once, by skilled hands, from one piece of fabric, in one small wash. Bulk is made across several fabric rolls, several wash loads and several days. Without a written standard, denim shade variation is the expected outcome rather than a surprise.
This page follows a fixed order: set the reference and the band, control the inputs, approve a pilot, monitor batches, then record decisions and corrective action. A worked illustration with a lot-by-lot shade table shows what a controlled bulk and an inconsistent bulk look like, and a KPI section explains how to track the result.
Where Denim Shade Variation Comes From
Shade drifts at specific points in the process, and each point has a different owner. Naming the source matters because the fix for a fabric problem (a mill or supplier control) is different from the fix for a laundry problem (a recipe or batch control). Our guide to laser and ozone denim finishing covers newer finishing routes that reduce some manual variables, but none of them removes the need to control the base fabric.
Sources of denim shade variation and who controls them
| Source | Where it arises | What the buyer sees | Primary owner |
|---|---|---|---|
| Fabric lot | Indigo dyeing at the mill: bath strength, number of dips, oxidation, finishing | Whole lots start lighter, darker, redder or greener before any wash | Mill and supplier fabric control |
| Within-roll shading | Along the length of a roll, or side to centre to side | Panels in one pair differ, so the leg and yoke do not match | Fabric inspection and cutting-room sorting |
| Laundry recipe | Time, temperature, liquor ratio, chemical dosage, pH | A steady drift across a day or across a whole lot | Laundry, through a written and frozen recipe |
| Wash load | Load weight, machine, water quality, chemical batch | Different results from loads that used the same written recipe | Laundry, through batch records |
| Manual finishing | Spray, sanding, scraping, localised treatments | Piece-to-piece differences in contrast and placement | Finishing line, against a reference piece |
| Drying and conditioning | Dryer temperature, moisture level at inspection | Damp pieces look darker and can pass or fail wrongly | Laundry and quality control |
| Viewing conditions | Light source, background, who judges | The same pair passes in one room and fails in another | Agreed between buyer and supplier |
General sources of variation in indigo denim processing. Which ones dominate depends on the fabric, the wash and the factory.
The last row is often overlooked. Two people judging the same pair under different lights will disagree, which turns every dispute into an argument about opinion. The first control, covered next, is therefore a physical and written reference that both sides can check against.
Set the Shade Band and Golden Sample
A golden sample is the approved physical garment that defines the target shade, finish and hand-feel, signed and dated by the buyer. A shade band is the range of acceptable difference around it. The band is what turns a subjective comment such as "slightly dark" into a decision rule: inside the band the pair is accepted, at the edge it is reviewed, outside it is reworked or rejected.
Define the band physically first. Alongside the golden sample, approve a light limit and a dark limit, which are two garments at the edge of what you will accept. Some programmes add instrument readings from a spectrophotometer, reported as a colour difference (often written delta E) with an agreed formula and an agreed limit. There is no universal limit: the figure depends on the wash, the customer and the price point, so agree it with the supplier in writing, and treat instrument data as support for visual judgement on heavily washed denim rather than a replacement.
Setting the golden sample and shade band
- 1Approve the sampleSign and date the golden sample after the final fit and wash sample, not an earlier round.
- 2Fix the viewing conditionsAgree the light source, background, angle, dry and conditioned state, and fabric direction.
- 3Approve the limitsApprove a light and a dark limit garment, so the band exists as objects, not words.
- 4Freeze the recipeAttach the wash recipe and the unwashed fabric swatch with its lot number to the standard.
- 5Distribute and sealHold matching sets with buyer, supplier and any third-party inspector, labelled and stored away from light.
A sequence for building a shade standard. Order matters: the recipe and fabric lot must be recorded at the moment the sample is approved.
Every approval in the pack (golden sample, light and dark limits, recipe sheet, fabric swatch and viewing conditions) should also appear in the sampling record rather than being agreed by email, and the same discipline of measurable references applies across categories, as we explain in how to improve sample approval rate.
Control Fabric Lots and the Laundry Recipe
Once the reference is fixed, the next job is to limit what can move. Two inputs deserve a written control: the fabric lot and the laundry recipe. A buyer does not run the laundry, but can require that both are named, recorded and frozen before bulk is booked.
Fabric lots: one standard, tracked by lot number
Fabric from different dye lots can differ in shade before it is ever cut. Ask for each fabric lot to be identified, and for the lot numbers to follow the cut work through sewing and into the wash. The practical rules are short: cut each garment from one roll or one matched set of rolls, keep bundles together so all panels of a pair come from the same shade, and compare each new fabric lot with the approved unwashed swatch. If a new lot is visibly different, it needs its own check wash before it enters production, because the same recipe applied to a different base can land in a different place. Fabric-stage inspection, including how defects and shading are recorded, is covered in our guide to the four-point system for denim fabric.
Laundry recipe: written, frozen, and changed only by notice
The recipe is the list of steps that produces the wash: desizing, abrasion, bleaching or enzyme steps, neutralising, softening and drying, each with its own parameters. After the pilot is approved, freeze it. Any change, including a new chemical supplier, a different machine or a different load size, becomes a recorded deviation that the buyer is told about before it is used on bulk. For fashion washes this is the most valuable single control.
Laundry parameters that move shade and how to control them
| Parameter | Why it moves shade | Control to request |
|---|---|---|
| Time and temperature | Longer or hotter steps remove more surface indigo | Recorded set points and actual readings per load |
| Liquor ratio (water to garment) | Changes the mechanical action and chemical concentration | A fixed ratio on the recipe sheet |
| Chemical dosage and batch | Enzyme, bleach and stone effects scale with dose and activity | Weighed dosages, logged chemical batch numbers |
| pH and water quality | Affects enzyme and bleach activity and the final tone | Recorded pH checks and consistent water source |
| Load size and ballast | Fewer garments in a machine changes abrasion per piece | A set load weight, with ballast rules for part loads |
| Machine type and condition | Drum speed and loading differ between machines | The pilot run on the machine type used for bulk |
| Drying method | Dryer heat and residual moisture alter tone and hand-feel | A fixed drying cycle and a conditioning rule before inspection |
A general list of laundry variables. Specific set points belong to the programme and should be confirmed in the pilot wash, not copied from another style.
Pigment, acid and similar fashion washes carry the same logic, and our guide to pigment and acid washed hoodies applies the shade band approach to knitwear.
Approve a Pilot Wash and Monitor Batch Variation
A pilot wash is a small production-condition wash, made from the actual bulk fabric lot, on the machine type used for bulk, with the frozen recipe. It tests the recipe as a repeatable process rather than a one-off craft result. Approve it against the golden sample under the agreed viewing conditions, and check the full set of properties: shade, hand-feel, shrinkage, strength and colourfastness, because a wash that matches the colour can still fail on rub resistance. Our guide to denim colour fastness and crocking explains what to ask for.
After the pilot, shade is monitored load by load. The principle is hold, check and release: no wash lot is packed until it has been compared with the golden sample and recorded. Retaining one reference garment from each load builds a shade history that the buyer can read later.
From pilot to release: a monitored batch sequence
- Pilot washSmall load, bulk fabric lot, frozen recipe. Buyer approves against the golden sample, or the recipe is adjusted and the pilot repeated.
- Recipe freezeApproved recipe issued with machine type, load size and drying rule. Changes now need notice.
- First production loadHeld until checked against the golden sample and the limits. Early deviation is cheaper to fix than late deviation.
- In-line load checksA reference piece from every load is judged dry and conditioned, then rated and logged by lot.
- SortingPairs are sorted into shade groups so pieces destined for the same carton or store delivery match.
- Final inspectionShade is assessed alongside other defects against the agreed sampling plan, and the decision is recorded.
A generic sequence. Timings depend on the wash, the factory and the order; each programme should agree its own checkpoints.
Final inspection normally uses an AQL plan, the acceptable quality limit sampling standard, with shade treated as a classified defect. Agree in advance whether shade is a major or minor defect and how a shade group failure is handled. The garment stage follows the same record-keeping logic as fabric inspection.
Worked Example: Controlled Wash Against an Inconsistent Bulk Lot
The example below compares two hypothetical bulks of the same mid-blue five-pocket style, judged against the same golden sample. It is illustrative: there are no real laboratory values, no real factory and no benchmark here, only a way to see how the controls change the outcome. The shade ratings use an indicative 1 to 5 scale defined in the next table, which is not the same as the 1 to 5 grey scale used in colourfastness testing.
Assumptions for the illustrative example
| Item | Assumed value |
|---|---|
| Style and order | One mid-blue five-pocket jean, 6,000 pairs, washed in six lots of 1,000 pairs (hypothetical) |
| Reference | The same golden sample, light and dark limits and viewing conditions for both bulks |
| Controlled bulk | One approved fabric lot, frozen recipe, reference piece logged from every load |
| Inconsistent bulk | Three fabric lots, some mixed in cutting, recipe adjusted by operators without notice |
| Rewash rule | One rewash allowed per pair; a pair that still fails is rejected for shade |
| Rating method | One trained assessor, dry pieces, agreed light, rated against the golden sample (indicative) |
All inputs are hypothetical and chosen for demonstration only. They are not typical industry results.
Indicative shade rating scale used in the example
| Rating | Meaning against the golden sample | Default action |
|---|---|---|
| 5 | Indistinguishable from the golden sample under the agreed light | Accept |
| 4 | Inside the band; a difference is visible only when pieces are held side by side | Accept |
| 3 | At the edge of the band, near a limit garment | Hold for buyer review |
| 2 | Outside the band; visible at normal viewing distance | Rewash candidate |
| 1 | Clearly off shade | Reject, or re-process only after review |
Indicative scale for this illustration only. Real programmes define their own scale, band and actions in writing.
The controlled bulk first, with the six wash lots, their ratings and the quantities sent for rewash or rejected after rewash.
Controlled bulk: lot-by-lot shade table (illustrative)
| Lot | Fabric and recipe status | Indicative rating | Pairs washed | Sent for rewash | Rejected after rewash | Decision |
|---|---|---|---|---|---|---|
| C1 | Fabric lot A, recipe frozen | 5 | 1,000 | 20 | 5 | Release |
| C2 | Fabric lot A, recipe frozen | 5 | 1,000 | 10 | 0 | Release |
| C3 | Fabric lot A, load at top of weight range | 4 | 1,000 | 40 | 10 | Release, log load size |
| C4 | Fabric lot A, recipe frozen | 5 | 1,000 | 10 | 5 | Release |
| C5 | Fabric lot A, last rolls | 4 | 1,000 | 30 | 5 | Release |
| C6 | Fabric lot A, recipe frozen | 5 | 1,000 | 10 | 5 | Release |
| Total | Mean rating 4.67; spread 1 point | 28 points | 6,000 | 120 | 30 | 5,970 pairs shipped |
Illustrative figures. Sum of rewash 20 + 10 + 40 + 10 + 30 + 10 = 120; rejects 5 + 0 + 10 + 5 + 5 + 5 = 30; mean rating 28 divided by 6 = 4.67.
Now the inconsistent bulk, on the same style and the same reference. Here the fabric lots are mixed and the recipe is adjusted on the shop floor.
Inconsistent bulk: lot-by-lot shade table (illustrative)
| Lot | Fabric and recipe status | Indicative rating | Pairs washed | Sent for rewash | Rejected after rewash | Decision |
|---|---|---|---|---|---|---|
| I1 | Fabric lot A, recipe as written | 4 | 1,000 | 150 | 30 | Release after sorting |
| I2 | Fabric lot B (lighter before wash), recipe unchanged | 2 | 1,000 | 400 | 180 | Hold, rewash outside band |
| I3 | Fabric lot A, recipe as written | 5 | 1,000 | 60 | 30 | Release |
| I4 | Fabric lots B and C mixed in cutting | 3 | 1,000 | 250 | 90 | Hold for buyer review |
| I5 | Fabric lot C, cycle time extended and not logged | 2 | 1,000 | 380 | 150 | Hold, rewash outside band |
| I6 | Fabric lot A, new chemical batch not recorded | 4 | 1,000 | 140 | 60 | Release after sorting |
| Total | Mean rating 3.33; spread 3 points | 20 points | 6,000 | 1,380 | 540 | 5,460 pairs shipped |
Illustrative figures. Sum of rewash 150 + 400 + 60 + 250 + 380 + 140 = 1,380; rejects 30 + 180 + 30 + 90 + 150 + 60 = 540; mean rating 20 divided by 6 = 3.33.
The pattern in the second table is the one buyers recognise: no single catastrophic failure, but several unrecorded changes (a different fabric lot, a longer cycle, a new chemical batch) that together widen the spread. The chart and results below put the two bulks side by side.
Reading the Result: What the Numbers Show
The ratings chart shows the spread within each bulk, and the outcome bars show where the 6,000 pairs ended up. Both use the illustrative figures from the lot tables above, not market data.
Indicative shade rating by wash lot, controlled against inconsistent
Higher is closer to the golden sample
Illustrative and indicative ratings on a 1 to 5 scale, taken from the worked example. Not laboratory values.
Controlled bulk: where the 6,000 pairs ended up
Illustrative outcome
5,880 passed first time, 90 rewashed then passed, 30 rejected. Illustrative.
Inconsistent bulk: where the 6,000 pairs ended up
Illustrative outcome
4,620 passed first time, 840 rewashed then passed, 540 rejected. Illustrative.
Three things follow. First, the quantity shortfall (510 more pairs rejected in the inconsistent bulk) is an in-full problem as well as a quality problem, because a short delivery of one wash is hard to replace once the delivery window has passed. Second, the rewash volume is not free: every rewash adds days and exposes the garment to more mechanical and chemical action, which can change shrinkage, strength and hand-feel. Third, even the shipped pairs from the inconsistent bulk contain ratings from 2 to 5, so a store delivery can put visibly different pairs of one style on the same rail. Those pairs may never be counted as rejects but they can come back as customer returns.
Record Acceptable Variation and Corrective Action
The last control is the record. Every lot should carry a short shade record: lot number, fabric lot, recipe version, rating, decision, who approved it and the date. The record does two jobs. It makes acceptable denim shade variation explicit, so a lot at the edge of the band is a documented decision rather than a disagreement. It also builds the history that lets you find a cause when a lot goes wrong.
When a lot falls outside the band, the response should follow a fixed order: contain, find the cause, correct and prevent. Contain means hold the lot and stop packing. Find the cause means checking the fabric lot, recipe version, load size, machine, chemical batch and viewing conditions against the record. Correct means a defined rework, and prevent means a change to the recipe or the control that failed. Agree before bulk how many rewash cycles a pair may have, because repeated washing is a rework with its own risk, and agree in writing who bears the cost of rework and rejects.
Shade fault, likely cause and first response
| Symptom | Likely cause to check first | First response |
|---|---|---|
| Whole lot lighter or darker than the golden sample | Different fabric lot, or recipe time and dosage drift | Hold the lot; compare the unwashed swatch and recipe records |
| Loads differ from each other, same recipe | Load size, chemical batch, machine or water change | Check batch logs; sort into shade groups; restate load rules |
| Panels in one pair differ | Cutting from mixed rolls or broken bundles | Stop cutting; re-establish roll and bundle integrity |
| Piece-to-piece difference in a manual finish | Operator technique or no reference piece at the station | Place a reference piece on the line; train and re-check |
| Pass at the factory, fail at goods-in | Different light, damp pieces, or a different assessor | Re-judge under agreed viewing and conditioning rules |
| Shade matches but rub or wash fastness fails | Wash chemistry or incomplete rinse | Test again; revisit the recipe, not just the colour |
A diagnostic starting point. The actual root cause must be confirmed from the lot records and, where needed, tests.
The KPIs: Shade Rejects, Rewash Percentage and Returns
Three measures describe how well shade control is working. They are only comparable between suppliers or seasons if the numerator and the denominator are defined the same way, so state the definition on the scorecard. If you already use a formal supplier scorecard, add these three as shade measures alongside delivery and quality.
Shade KPIs: formula, denominator and what each signals
| KPI | Formula | Denominator and variations | What it signals |
|---|---|---|---|
| Shade rejects | Pairs rejected for shade after any allowed rewash divided by pairs washed, multiplied by 100 | Some buyers use pairs inspected or pairs cut instead; state which and apply it consistently | Final quality loss and the in-full risk to the order |
| Rewash percentage | Pairs sent for rewash on the first pass divided by pairs washed, multiplied by 100 | Count first-pass rewash separately from second rewash; also track rewash cycles per pair | How stable the process is before rejects appear; an early warning |
| Shade-related returns | Pairs returned for shade or colour mismatch divided by pairs sold in the same period, multiplied by 100 | Returns lag sales; reason codes vary by retailer; lot or batch code is needed to trace them | What reached the customer; the failures inspection did not catch |
Definitions vary by retailer and supplier, so confirm the numerator and the denominator before comparing figures.
Read the three together. A low reject rate with a high rewash percentage means the factory is rescuing lots with rework at a hidden cost to time and garment condition. A low reject and low rewash with rising shade-related returns points at the inspection method, usually the light, the assessor or the band. To make returns traceable, ask the factory to print the wash lot code on an inside label or carton marking, so a complaint can be matched to a lot record.
Treat the numbers as trends within one programme, not as targets copied from another. A target for a dark rinse wash will differ from one for a heavily abraded wash, and the figures in the worked example are illustrative only.
Decision: What to Do Before Approving Bulk
For a new or repeated wash, the practical decision is whether to release bulk now or require another gate. Release bulk when the golden sample and band are signed, the fabric lot is identified and matches the approved swatch, the pilot wash has passed on the bulk machine, and the recipe is frozen with change notification agreed. If any of those four is missing, hold bulk and close the gap first.
A new fabric lot that differs from the approved swatch, a recipe change on the laundry floor, or a pilot done on a different machine type each reopen the approval. A rising rewash percentage in early loads is the signal to stop and investigate before more of the order is washed.
For a new wash route or a new supplier, consider reducing exposure with a staged order. The test-and-scale model is a risk-reduction approach, not a risk-free one, and a test quantity of around 500 pieces per style and colour is a proposed figure where it is commercially and technically feasible, since fabric, dye, trims and factory constraints vary. Monitor the rewash percentage and shade rejects on that first quantity before committing the remainder.
The most useful question to ask at your next wash approval is this: if the sixth lot differs from the first, which written record would tell you why?
How Ruhrose Approaches Denim Shade Variation
Ruhrose develops denim and other casualwear for UK buyers, with development and production in Bangladesh. For denim, shade control is part of the sampling conversation: we agree the wash reference and viewing conditions with you, review the fabric lot and recipe, and sample before bulk. Our jeans sampling process shows where the wash approval sits in the sequence. Whether a particular wash can be repeated at your volume is a question for the pilot, and we will say plainly when a finish looks hard to hold.