How we verify Colour accuracy

Arcanium Studios maintains a high quality control tolerance of 97% or greater. Our artists work with AI to maintain this precision and provide customers the transparency of our methods.

We do not rely on instinct or surface aesthetics, we apply measurable validation methods by transferring methodology used in professional design and engineering.

From Claim to Proof: How We Verify Accuracy

Colour calibration workflow — professional studio vs AI × Human (Valehart)

Measurement
Spectrophotometer reading taken; LAB values compared against brand reference by a technician.
Same spectrophotometer reading, but LAB → ΔE computed instantly; AI flags direction of error (too warm/cool, chroma shift).
Adjustment
Pigments adjusted by experience; repeat readings until it “looks right”.
AI proposes pigment ratio tweaks and finish compensation (matte/gloss). Iterate until ΔE ≤ 3, then human confirms under D65 light.
Lighting Control
Checked under house lighting; daylight checks vary by operator.
Standardised viewing: D65 (6500 K) reference or calibrated daylight. WB noted in the record.
Acceptance Criteria
Visual judgement; small variance accepted informally.
Objective tolerances: ΔE < 3 = pass, 3–5 = review, > 5 = reject. Numbers stored with the sample.
Record & Repeatability
Notes kept on sample cards or project files; hard to reproduce exactly.
Digital log: target LAB, measured LAB, ΔE, lighting, material finish, final pigment ratios — re-mixable on demand.

Why it matters: Both workflows use a spectrophotometer.

A standard workflow takes anywhere between 10-30 minutes PER colour correction cycle.

From artist's sample capture to material prep, end-to-end reducing the verification time roughly by 90-95% while maintaining pro-grade precision.

METHOD:

Image capture of finished item next to sample

Formula: ΔE = √((L₁ − L₂)2 + (a₁ − a₂)2 + (b₁ − b₂)2)

SAMPLE OUTPUT:

Yellow → ΔE ≈ 2.3 (visually identical)

Light Blue → ΔE ≈ 4.8 (slightly cooler, but within real-world tolerance)