Skip to content

INDUSTRY FIELD GUIDE

Dye and pigment data for AI color-reference evidence

Synthetic dyes and pigments form a defined manufacturing category in the Census manual.

By Fiund · Updated 2026-11-05

NAICS 325130 · Synthetic Dye and Pigment Manufacturing

Synthetic dyes and pigments form a defined manufacturing category in the Census manual. An original dataset can test the distinction between a declared material identifier and an appearance comparison made under a specified view. Census dye and pigment scope

Create self-authored color cards, fictional product identifiers and a sample-preparation record. Photograph two similar-looking cards with a shared reference. Add another image taken without that reference and a report tied to a different preparation. Ask the assistant which comparison is supported and which product identity is established by the documents. The reference answer should avoid inferring composition from hue alone.

Label material identifier, preparation, substrate, reference card and view. Hold out a new color family during evaluation. Score wrong-preparation pairing, unsupported comparisons across views and identity inferred from appearance independently. Include a complete case where both the product record and the comparison conditions are available.

Use printed paper cards and invented declarations for the pilot. Actual formulations, customer color standards, licensed reference material and laboratory reports need source-specific permission. A licensing proposal should explain the visual-and-document task, independent key and available context. Confirm buyer fit before preparing a commercial color archive or commissioning further capture.

Discuss this original manufacturing sample with Fiund

Original Fiund exercise

The same pigment code under two viewing setups

An original purple card is photographed under two documented lighting setups. No actual shade approval is simulated.

Fictional records supplied to the evaluator
RecordSourceWhat it records
1Reference cardColor C8 approved for visual comparison only with setup V1.
2Sample image AC8 card photographed in V2 with a warm background.
3Sample image BC8 card and reference photographed together under V1.

Question for the model

Which image can support the specified visual comparison?

Show the reference answer and checks

Use image B because record 3 places the sample and reference in the same V1 setup required by record 1. Image A can document C8's appearance under V2, but matching its product code does not make the lighting or background comparable.

  • Comparing RGB values across uncontrolled setups
  • Treating code equality as a measured shade match

Fields to preserve

lighting_setup
Records comparability conditions
reference_in_frame
Captures simultaneous comparison evidence
color_code
Identifies the sample without replacing observation

Test another case

Show two C8 samples under V1 with the reference clipped off one frame; distinguish setup match from reference availability.

Only owned color cards; no proprietary formulas, color libraries or claims of commercial approval.

Download this fictional example and answer key (JSON)

Discuss a contribution

Describe the sources you control, the task they could support and what has already been checked. Use the existing-data assessment for an archive or the collection brief for a new recording project. The contributor agreement defines Fiund’s licensing authority and compensation. A guide or fictional exercise does not establish an accepted collection or a sale.

Primary sources and further reading