Customer Case Study · Colour Match & Print-Run QC
Towels, bed linen and rugs ship to global retailers who reject shipments when shades drift between print runs. HMC built a colour studio for a home-textiles manufacturer's quality and product development teams: it extracts the palette from any design, converts every colour across HEX, RGB, HSB, HSL, CMYK and CIELAB, scores each run against the approved standard with ΔE2000, groups runs into shade bands and creates new colourways.
Figures indicative; validated metrics available on request.
A print run that looks fine under the shop-floor lights can be visibly different next to last month's run on a retailer's shelf.
Shade approval relied on experienced eyes under a light box; two checkers could disagree.
The same design is printed across machines, dye lots and weeks; small drifts add up.
Without shade banding, slightly different runs were packed together and flagged by the buyer.
Designers think in HEX and CMYK, dye labs in LAB, buyers in their own references.
Built with the colour lab, print floor and product development teams.
Select a step to see what happened.
Rejected shipments traced back to their runs.
Deliverable: Shade failure map and tolerance table.
Every colour carried in every format.
Deliverable: Colour data model and banding rules.
One studio for design and quality.
Deliverable: Colour studio.
Instrument readings feed the studio.
Deliverable: Pilot on bed linen.
Towels, rugs and fashion bedding.
Deliverable: Roll-out plan.
Start from a sample design or upload your own image. Everything runs in your browser; the images are drawn illustrations, and in production the colour readings come from a spectrophotometer, not a photo.
Optional. Without a key, colour names, colourways and approval notes come from built-in rules. Add your OpenAI key to name colours like a designer, suggest seasonal colourways and draft the shade approval note for a buyer.
Pick a sample or upload an image. The palette is found by clustering pixels in CIELAB, so it groups colours the way the eye does. Hover the image for a pixel loupe; click to pin a colour. Click any value to copy it.
Screen colours are approximate. CMYK uses a simple device conversion, not an ICC press profile; LAB uses D65 and the 2° observer.
The approved lab dip for the design is the standard. Every extracted colour is matched to its standard colour and scored with ΔE. Move the tolerance to your buyer's limit; the direction words say how the shade has moved.
ΔE under 1 is barely visible to a trained eye; 2 to 3 is visible side by side. Buyers set their own limits.
| Standard → run | Colour | ΔL* | ΔC* | ΔH* | ΔE | Shade moved | Verdict |
|---|
Synthetic standard and colour library. Library codes are illustrative, not a commercial colour system.
Six runs of the same design from different machines and dye lots. Drag the slider across the image to see the difference; each run is scored, given a shade band, and the band decides which runs can be packed together.
Pick a run, then draft the note for the dye house and buyer.
Runs are simulated from the design by shifting lightness, chroma and hue per colour.
Pick a preset or change any colour. The design is recoloured in CIELAB, keeping its shading and texture. Drag to compare with the original, then draft the lab-dip request for the dye house.
Pick a preset, change a colour, or ask for suggestions.
With every run measured against the standard and banded before packing, drift is caught early and buyers see consistent shades.
Swatch matrix: each cell is the run's actual colour; red outline when ΔE is over 1
Interval plot: range and median per programme
Figures indicative; validated metrics available on request.
A ΔE tolerance agreed with the buyer ends most shade arguments before they start.
"Lighter and redder" tells the dye house what to fix; a pass/fail alone does not.
Even passing runs differ slightly; packing by band keeps every carton consistent.