SERIES 2. THE SCIENCE OF COLOR
One expression appears frequently when people talk about skin color:
Olive skin.
When a complexion appears slightly greenish, or seems somewhat different from what people commonly think of as reddish or yellowish skin, people often describe it as “olive.”
If you look up olive skin online, it is not difficult to find explanations suggesting that green is mixed into the skin, or that the skin itself contains a greenish quality.
But seeing color and measuring color are not the same thing.
Once skin color is approached as measurable data, a different question appears.
Is there actually green in the skin that people perceive as olive?
And perhaps even more importantly:
What difference in skin color are we actually seeing when we use the word “olive”?
Looking at the idea of “greenish skin” through measurement
COCORY observes skin color through L*, a*, and b* values in the CIELAB color space.
L* represents lightness.
a* represents the red–green axis.
b* represents the yellow–blue axis.
In CIELAB, the positive direction of a* moves toward red, while the negative direction moves toward green.
The positive direction of b* moves toward yellow, while the negative direction moves toward blue.
However, in commonly measured human skin colors, a* and b* generally appear in the positive range.
In other words, skin that people often describe as “olive” is still usually measured within a region where red and yellow are present.
This is important.
A complexion may look slightly greenish to the eye, but that does not mean its measured skin color is actually located in the green direction of CIELAB space.
So there is a difference between saying:
“There is green in olive skin.”
and saying:
“This skin looks olive or greenish.”
The first is a claim about measurable color position.
The second is a perceptual impression created by the observer.
Reading warm and cool through b*−a*
At COCORY, we do not stop at looking at a* and b* separately.
We also examine b*−a* in order to understand the relationship between yellowness and redness in skin.
At COCORY, this value is used as an important reference for interpreting the warm–cool direction of skin color.
When b*−a* is relatively lower, it is interpreted toward the cool side.
When it is relatively higher, it is interpreted toward the warm side.
But one distinction is essential here.
b*−a* does not describe the entire skin color.
It tells us the difference between b* and a*.
That means the difference may remain the same even when the actual values of a* and b* are different.
And this becomes an important clue in rethinking what people call olive skin.
The same b*−a* does not mean the same skin color
Let us compare the following three examples:
All three examples have b*−a* = 6.
| a* | b* | b*−a* |
|---|---|---|
| 20 | 26 | 6 |
| 14 | 20 | 6 |
| 4 | 10 | 6 |
In other words, the difference between b* and a* is the same.
And yet the actual colors do not look identical at all.
Why?
Because although
20 and 26,
14 and 20,
and 4 and 10
all differ by 6, the absolute magnitudes of a* and b* are different.
In the color with a*=20 and b*=26, both the red and yellow components are relatively strong.
By contrast, a*=4 and b*=10 maintains the same b*−a*=6, but both component values are much lower.
So even when the warm–cool relationship is the same, the visible color can still look very different.
The difference also appears in chroma
The difference in the absolute values of a* and b* can also be seen in chroma.
In CIELAB, chroma C* can be calculated as:
C* = √(a*² + b*²)
Applying that to the same three examples gives:
| a* | b* | b*−a* | C* |
|---|---|---|---|
| 20 | 26 | 6 | about 32.8 |
| 14 | 20 | 6 | about 24.4 |
| 4 | 10 | 6 | about 10.8 |
The b*−a* value remains the same in all three cases.
But C* changes considerably, from about 32.8 to 10.8.
This means that skin colors can share the same warm–cool relationship while still carrying very different color strength.
And if L* also changes, the visible difference between those skin colors becomes even greater.
Ultimately, one skin color cannot be fully explained by b*−a* alone.
It is formed by L*, a*, b*, b*−a*, and C* together.
The same difference does not mean the same position in color space
We can go one step further.
a*=20, b*=26
and
a*=4, b*=10
both give us b*−a*=6.
But in the first example, b* is about 1.3 times a*.
In the second example, b* is about 2.5 times a*.
In other words, the difference between a* and b* may be the same, while their ratio is not.
From the perspective of CIELAB color space, the two colors do not occupy the same location.
So saying b−a is the same does not mean every aspect of skin color is the same.
It only means that one specific relationship used by COCORY to read warm and cool is the same.
Each value tells us something different.
L* describes lightness.
a* and b* describe the actual position of the skin color.
b*−a* describes the relationship between those two values.
C* describes chroma.
That is why one value cannot be treated as the whole skin color.
So what are we actually seeing when we say “olive”?
This is where the expression olive skin becomes especially interesting.
If olive were one clearly defined skin color, then skin described as olive should share a relatively consistent measurable structure.
But real skin color is not that simple.
Even skin colors with the same b*−a* may look completely different depending on the absolute values of a* and b*.
Conversely, skin that people describe in a similar way as olive may still have different L*, a*, b*, and C* values.
That is why it is difficult to define olive through one single b*−a* value or one fixed skin-color coordinate.
A more useful possibility is this:
What we call olive may not be one measurement value at all.
It may be a term people use to group skin colors that produce a somewhat similar visual impression through a particular relationship among L*, a*, b*, b*−a*, and chroma.
Why can skin look greenish even without actual green?
Seeing color is not the same as reading numbers.
When people look at skin, they do not think:
“this skin has a certain a* value and a certain b* value.”
They perceive lightness, redness, yellowness, and chroma together as one overall skin color.
Skin is also never seen in isolation.
It is compared with other areas of the face, with other people’s skin, and with surrounding colors such as hair, lips, clothing, and lighting.
In this process, even skin with a positive a* value may look relatively less red, or may seem visually different from surrounding skin colors.
People do not describe that difference through numerical measurement.
Instead, they use words such as:
“It looks slightly greenish.”
“It looks olive.”
So the word olive may be better understood not as the presence of actual green, but as a visual interpretation of a certain skin-color structure.
It is difficult to assume that one single “olive color” exists
The fact that people use the same word olive does not mean they are all describing the same measurable skin color.
One complexion may have a relatively high L* while keeping both a* and b* lower.
Another complexion may have a lower L* while both a* and b* are much higher.
Their b*−a* values may be the same, or different.
And yet both may still be perceived as olive.
That suggests that within the single word olive, there may be multiple different skin-color structures.
So rather than thinking:
Olive = one skin color
it is more natural, from a measurement perspective, to think:
Olive = a visual impression that may be similarly perceived across different skin colors
Skin color is not fixed by one name
This is also why COCORY does not begin by placing skin into one fixed name.
The skin color we actually see is not always identical.
Sun exposure, blood circulation, physical condition, and environment can all affect the color that is actually visible and measurable in the skin.
So what COCORY looks at is not one permanent color label attached to a person.
It looks at the color that is actually present in the skin now.
What is the current L*?
What is the current a*?
What is the current b*?
What kind of relationship appears in b*−a*?
How does the chroma appear?
All of these must be read together.
No single value can replace the whole skin color.
The important question is not “Is it olive?” but “What color is it?”
There is nothing inherently wrong with the phrase olive skin.
It is a form of language people use to describe the features they perceive in skin color.
But that is different from concluding that:
“There is actual green in the skin.”
or that:
“Olive is one fixed skin color.”
Measured skin color is more complex than that.
Even with the same b*−a*, changing the absolute values of a* and b* can produce completely different colors.
Chroma can vary within the same relationship, and when L* changes, the overall impression of the skin changes as well.
Conversely, skin colors with different measured values may still create a similar olive impression to the eye.
So the main question for COCORY is not:
“Is this skin olive?”
What matters more is:
“What color is actually present in this skin right now?”
Once we stop defining skin through a name first and begin looking at its actual measured values and relationships, the familiar word olive can also be understood in a more precise way.
And in that process, the goal is not to invent one more color label.
It is to establish a more accurate standard for seeing human skin color.
NEXT JOURNAL
02-4. Why Was CUBE Developed?
If skin color is observed only by eye, its interpretation can easily shift depending on the observer and the environment.
So how can small differences in skin color be measured through a more consistent standard?
In the next journal, we explore why COCORY developed CUBE as a tool for measuring human skin color, and why recording skin color as data matters.
COCORY JOURNAL
Color, People, and the Way We See Them.
Setting the standard of color for people.
