: Interpreting a* and b* in L*a*b* color space In L* a* b* color space, I know L* corresponds to "lightness" so that as L* increases I can say colors are lighter. Is there a similar term that

@Si6392903

First, we need to understand what is the concept of the different Lab modes.

The premise is that one perceived color, as humans do, can not be at the same time its complementary color.

Let us use yellow as an example.


Yes, you can have a "redish" yellow
Yes you can have a "greenish" yellow
But you can not have a bluish yellow, because before it turns into blue it needs to pass thru green.


Yellow is the complementary color of blue.



On the other color models you add a component, for example RGB you can have no red (0) or a lot of red (255)

But in Lab you have two "complementary" colors on the same scale. This means that the a and the b are not representing one color but an axis of complementary colors.

I am going to steal @Billy Kerr's image



Two colors that, if mixed in this model, the cancel each other and make a neutral gray.

So the scale is not 0 to 255 but -128 to +128 where the 0 is at the middle indicating that you have a gray.

If you go off this axis you simply move away from gray to yellow or blue.


increases can I say that a color's red shades are more intense or saturated?


Close but not exactly.


Yes if you are further away from 0 you have more "saturation" on that direction.
No, because "Increase" can render a color less saturated, if you consider that you have negative values. -50 is a smaller number than -20. It increased but the color is less saturated. A "correct" term would be "further away from 0" in either direction.


Some Lab models do have Red as a component. (But in other models, the Magenta is used). I will use the Red as an example.

It is not shade, but a component. The difference is that if you say "Red shade", you expect to see Red or Redish.

But if you also have a Blue component you will see Purples and wine shades, not exactly Red shades but Red compoent.

That is why they used a letter. It is an axis.

It could be called L-GR-BY or L-GM-BY. But it is a bit harder to say.

Vote Up Vote Down

@Kevin459

The A channel controls Green/Magenta

The B channel controls Blue/Yellow

You can see this quite clearly if go into LAB mode in Photoshop, and then add a Colour Balance adjustment layer, and open up the properties

Vote Up Vote Down

@Alves566

Describing them as more intense would be potentially misleading.

As the line gets steeper the colors get more intense. If you only raise one side it would get intense but also entirely shifted.

As for terminology I don't know of anything to use. Intensity could describe the slope of the a* and b* while contrast could describe the slope of L. You can also think of them as a* being closely related to Tint and b* being closely related to Temperature.



Some visuals using an old photo of mine. Here we have the original a* channel.



Now if I increase the slope and equal amount we get more intense colors without shifting it:



But where you say, "as a* increases" does it get more intense. Well yes, but it also will shift everything else, because while you have a more vibrant a* you also have flatter colors:



See I can raise the left; which you'd think as we just said would be more intense. But if I lower the right side to keep the image colors normalized we actually make it less intense because the colors are all flatter (making a flat line across 0 on a* removes all color from that channel):



Another catch to saying, "If I increase a* or b* does it become more intense?" is that its based around a zero point. If I raise the left it becomes a very intense red. But if I lower the right it becomes a very intense green. This image isn't the best to show that on so instead if lowered the right on the b* channel making it a very intense blue.



So, do I raise it to make it more intense, or do I lower it to make it more intense? It's more about the overall slope and how far from zero we get.

Vote Up Vote Down