A source is outside sRGB. Original mode computes from its original coordinates while the preview is mapped; display mode maps before compositing and comparison.
No universal pass/fail or visibility threshold is applied. This is not a physical sample measurement or WCAG contrast test.
Coordinates used in the calculation
Displayed coordinates are rounded; calculations use full precision. Lab and OKLab are different coordinate systems.
The JSON report records source values, comparison basis, backdrop when used, methods and full-precision coordinates. A smaller number is not a certification or a promise that a difference is invisible.
Color calculations run in your browser. The website still loads assets, advertising and analytics.
Compare the numbers in context
A color difference is useful only when you know the method and inputs.
Keep the two swatches together
Compare neighboring colors while retaining both source values. The calculator names each difference formula and exposes the coordinates used, so an apparently similar number is not mistaken for the same scale.
State what you are comparing
Original mode keeps opaque source coordinates, including colors outside sRGB. Display mode compares separately composited sRGB colors over one specified backdrop. Neither predicts how a physical sample will look under another light.
How to compare color differences
Set the basis before interpreting the number.
01
Enter a reference and another color
Use supported CSS color notation or HSV/HSB input. For original-coordinate comparison both must be opaque. For alpha, select display comparison and supply an opaque shared backdrop.
02
Read the named methods and coordinates
Compare CIEDE2000 and ΔE76 in CIELAB D50, plus ΔEOK in OKLab D65. The OKLab result uses its native scale. Look at the swatches while remembering they are rendered as sRGB approximations.
03
Save the context with the result
Copy or download JSON so the inputs, basis, backdrop and methods travel together. Use the contrast checker for actual text/background pairs, and appropriate measurements when your task involves physical material or production tolerances.
Continue your color workflow
Explore a palette or check a real foreground and background.
Color converter
Convert color notation, preserve alpha and inspect sRGB approximations.
Delta E methods, white points and comparison limits.
What does Delta E measure?
It is a family of numerical color-difference methods. A smaller value means closer colors within the same method and calculation context. It is not a similarity percentage or a universal prediction of what every observer notices.
Why do CIEDE2000, ΔE76 and ΔEOK disagree?
They use different spaces and formulas. ΔE76 is Euclidean CIELAB distance; CIEDE2000 adds weighting terms; ΔEOK is Euclidean OKLab distance. Here Lab uses D50 and OKLab uses D65. Their values should not share one universal threshold.
Does a Delta E below 1 guarantee an invisible difference?
No. Visibility depends on the method, viewing conditions, surrounding colors, display, sample size and observer. The tool deliberately provides no universal approval threshold or automatic brand-color certification.
Can I compare colors with transparency?
Yes, in display mode. Each input is mapped to sRGB and composited over the same specified opaque backdrop before comparison. Original mode requires opaque inputs so alpha is never silently ignored.
Are wide-gamut colors mapped before comparison?
Original mode preserves opaque input coordinates for the calculations, while the preview is mapped to sRGB. Display mode maps each source to sRGB first, then composites and compares. The two modes answer different questions and can produce different results.
Which Lab white point and CIEDE2000 factors are used?
CSS lab() and the Lab coordinates here use D50. OKLab uses D65, with color conversion handling the adaptation. CIEDE2000 uses kL = kC = kH = 1. These settings are recorded in the JSON report.
Is color difference the same as text contrast?
No. Delta E describes color separation within a model. WCAG contrast evaluates foreground/background relative luminance. Two noticeably different colors can still be a poor text/background pair; use the contrast checker for that task.
Can this approve paint, print or a production sample?
No. Pasted color values are not measurements of a physical sample. Production decisions need suitable measurement conditions and agreed tolerances. The JSON report records the digital comparison, and the full URL preserves settings for reloads or language changes.