Monitor Color Accuracy Test
Use reference patches for a visual cast check, or enter measured Lab data to calculate CIEDE2000 ΔE00.
Test settings
One row: reference L,a,b | measured L,a,b. L 0–100; a and b -200–200; up to 100 samples.
Enter valid rows as reference L,a,b | measured L,a,b.
Color-accuracy output
JavaScript is required to render dynamic reference patches.
- Maximum ΔE00
- —
- 95th-percentile ΔE00
- —
- Valid samples
- —
- Current mode
- Visual screen
- Reference space
- sRGB
- White point
- D65
- Ambient light
- Dim light
- Readiness
- 0/4
Lower ΔE00 means the measured color is closer to the reference. The page does not label a display pass or fail.
| Sample | Reference Lab | Measured Lab | ΔE00 |
|---|---|---|---|
| Visual mode has no sample-level quantitative data. | |||
Using and interpreting the result
Using and interpreting the result
Visual mode compares neutral gray, basic-color, and skin/natural-color patches. It can reveal obvious cast or unstable gray balance but cannot provide a reliable ΔE.
CIEDE2000 ΔE00 calculation
Instrument mode compares each reference Lab triplet with its measured triplet using CIEDE2000, then calculates average, maximum, and 95th-percentile values from unrounded results. The page displays two decimals.
Test limits
The page does not read the display's spectrum or chromaticity. Visual results depend on color management, ICC, brightness, white point, ambient light, and visual adaptation; measurements must share conditions.
Tool information
All input and calculations stay in this browser. Lab data is not uploaded and no third-party runtime resource is loaded.
The calculation uses CIEDE2000. The sample follows the public supplementary CIEDE2000 test data by Sharma, Wu, and Dalal.
Frequently asked questions
Can I test color accuracy without a colorimeter?
Only visually. This can find obvious gray or patch cast but cannot produce reliable quantitative color difference.
How should I interpret ΔE00?
Smaller means closer to the reference. Review the average, maximum, 95th percentile, and individual outliers.
Can a screenshot prove monitor color accuracy?
No. It bypasses the physical panel, backlight, actual ICC output, and ambient light.
Can D65 and D50 be compared directly?
Not recommended. Use the same white point and color-management condition, or perform an explicit adaptation conversion.