From Wikipedia, the free encyclopedia
Table visualization of the information from the Motivation and derivation of Oklab section of https://bottosson.github.io/posts/oklab/:
| Color Space |
Opponent Model |
Predicts L/C/H Well |
Good Blending |
D65 Whitepoint |
Good Numerical Behavior |
Assumes Normal Viewing Conditions |
Scale Invariant |
Notes
|
| CIELAB |
Yes |
Poor, especially hue |
Not mentioned |
Yes |
Some issues |
Yes |
Yes |
Poor hue prediction, especially for blues
|
| CIELUV |
Yes |
Poor, especially hue |
Not mentioned |
Yes |
Some issues |
Yes |
Yes |
Similar issues as CIELAB
|
| CAM16-UCS |
Yes |
Good |
No, due to chroma compression |
Yes |
No |
No |
No |
Best perceptual uniformity overall
|
| CIECAM02-UCS |
Yes |
Good |
No, due to chroma compression |
Yes |
No |
No |
No |
Older version of CAM16-UCS
|
| OSA-UCS |
Yes |
Good |
Not mentioned |
Yes |
No analytical inverse |
Yes |
Yes |
Good model, but impractical due to inverse
|
| IPT |
Yes |
Poor lightness/chroma |
Good |
Yes |
Yes |
Yes |
Yes |
Simple structure; strong hue modeling
|
| JzAzBz |
Yes |
Fairly good |
Not mentioned |
Yes |
Yes |
Yes |
Not scale-invariant |
HDR-optimized; scale-dependent
|
| HSV (sRGB) |
No |
No |
No |
Yes |
No |
No |
No |
Included due to popularity only
|
| Oklab |
Yes |
Yes |
Yes |
Yes |
Yes |
Yes |
Yes |
Designed to meet all listed criteria
|
