Color Analysis: From ColorChecker to ΔE00
Browser-based ColorChecker 24-patch analysis with ROI adjustment, Delta-E00/Delta-C00 calculation, white balance assessment, and visual noise analysis.
How accurate are your phone's colors? Not a subjective question — it needs a standard chart and a color difference formula to quantify.
What is ColorChecker
The ColorChecker (X-Rite 24-patch) is the most widely used color reference in the imaging industry. Its 24 patches cover common natural colors: skin tones, blue sky, green foliage, flowers, and neutral grays.
Take a photo that includes the chart, and you can use a color difference formula to answer: how much does the captured color differ from the reference?
ΔE00: The Industry Standard
| Version | Year | Characteristic |
|---|---|---|
| ΔE76 | 1976 | Simple Euclidean distance, not perceptually uniform |
| ΔE94 | 1994 | Added weighting factors, significant improvement |
| ΔE00 | 2000 | Current industry standard, accounts for lightness, chroma, and hue weighting |
ΔE00 interpretation:
- < 1: Imperceptible to human eye
- 1–3: Noticeable by trained professionals
- 3–6: Noticeable by general users
- > 6: Obvious color shift
Reference Standard
This tool uses the X-Rite default: post-Nov2014 D50 reference, aligned to Imatest Colorcheck output.
How It Works
Upload photo → Adjust ROI frame → Sample 24 patch regions
→ Average RGB per patch → sRGB → CIELAB D50
→ ΔE00 / ΔC00 per patch
→ Summary: mean, max, WB deviation
All computation is done in-browser — nothing is uploaded.
ΔC00: Chroma-Only
ΔC00 is the chroma/hue component of ΔE00 without lightness. It tells you whether the colors are too saturated or too muted, independent of brightness.
Practical Cases
Case 1: Compare Two Phones
- Shoot a ColorChecker photo with Phone A and Phone B under the same light
- Upload each to the tool and adjust the ROI frame
- Compare Mean ΔE00 and Mean ΔC00 values
Case 2: Check White Balance
- Look at the white balance panel (patches 19–24, the 6 neutral grays)
- a*/b* values far from 0 indicate a color cast
Case 3: ISP Color Style
- Check ΔC00 per patch
- Consistently positive ΔC00 → ISP tends to boost saturation
- Large ΔE00 on skin tone patches → aggressive skin tone processing
Noise Metrics
- Visual Noise (VN): CIELAB-based weighted standard deviation
- Spatial SNR: Per-gray-patch RGB channel standard deviation