Multiply mix LAB vs RGB

In photoshop, LAB mode has only 2 mixing mode that works the same as RGB (colour and luminance), but everything else is quite different. since I’m painting inside of LAB mode I need to use multiple mode to paint shadows or other kind of darkening effect. And here what I find, Lab multiply is always warmer than RGB multiply, and will maintain saturation, and sometimes more saturation than both colour participating the mix.

the left side is RGB fully saturated red mixed with green in multiply mode, and the right side is the same but in LAB mode. As you see fully saturated red and green will give you black in RGB, but in LAB is no where near as dark, and the result mix is a warmer brown, slightly more yellow than red. The red and green are then compared at various saturation, the LAB version multiply mix always contain more saturation, yet the RGB version losses saturation… Then the same test done with blue vs yellow, to my suprise, the RGB version mix are all grey!! while the lab mix is a purple that get more and more warmer as saturation decrease, so it gets a little grey at saturation 25%.

Now lets look at a more natural colours, imagine skin tone.

Here the base colour is a light skin tone(top), and layered is a very saturated cool purple(bottom) (color shown on top right corner), I layer the purple on top of skin tone with multiplyed linear gradient. The left side is RGB, as you can see the mix at the bottom is purple just like original, but in the mid section of gradient it gets grey!! The right side is LAB version, and the bottom colour no longer resemble the original purple, but it got warmer! infact it got more saturation as the HSB colour info showed me. But the result mid section is very rich and saturated. I personally like the LAB version better, I guess this a matter of opinion and also depends what you are trying to achieve, if you are painting very cold and lifeless stuff the saturation drain is ok, perhaps like shadows casted on objects that absorbs less colour, like…stone, metal etc… Now to the LAB results, this is like Subsurface scattering effect (SSS) as 3D/texture artist can relate to, and it has to do with the skin absorbing more colour and more saturation is apparent during shadow’s transition.

I think you could certainly use it to colour and paint in multiply mode in LAB to get that SSS effect although I must say the fully shadowed area probably will be less saturated. I guess the fact that multiple in LAB is always warmer is a good thing for painting skin.

~ by yinako on March 11, 2007.

4 Responses to “Multiply mix LAB vs RGB”

  1. “Then the same test done with blue vs yellow, to my suprise, the RGB version mix are all grey!!”

    This is normal, Yellow and Blue are opposites on the colour wheel whereas Red and Green are not.

  2. It’s interesting to see what happens when a monitor (which essentially can only work in the RGB colour space) tries to figure out the LAB gamut…

  3. administer says : I absolutely agree with this !

  4. 1. To get a green when you multiply blue and yellow both the blue and the yellow must have a healthy amount of green in them to begin with — same as with real paint.

    http://www.digitalartform.com/archives/images/psColorMixing_1.jpg

    2. ‘Subtractive’ color mixing is really based on multiplication, not subtraction. Take the C, M, Y, (and K, if you have one) channels and make them into layers. What blend mode must you use to combine them all into a correct image? Difference? No. Exclusion? No. Mutiplication? Yes, sir.

    3. An RGB monitor emulating Lab is no more unusual than a Linux machine emulating Windows.

Leave a Reply