Lab Colour Space vs AdobeRGB

Can someone please explain the difference between Lab and AdobeRGB colour spaces? Are there benefits to be gained by converting from AdobeRGB to Lab, editing, and then modifying for output?

AdobeRGB like sRGB uses 8, 16, or 32 bits for each color Red, Green and Blue.
Although Adobe RGB is expanded to such that those same millions of colors cover a wider area of the color chart. That simply leaves bigger holes between colors.

Lab on the other hand defines pretty much any color you could imagine and those not possible... How do you envision or print a very bright dark-green in RGB?
Lab uses real numbers (e.g. 1.5, 2.73, etc...) AdobeRGB and sRGB use integers. So although Lab uses values of only 0-100, the possible color values are almost limitless.
A bright green (green value of 100, actually a:100, b:100 I believe) could also have a luminosity of L:1, and therefore still be very dark.

Some things work better in Lab. But some things just make more sense and work better in the RGB modes.
A great book to check out is "The Canyon Conundrum" by by Dan Margulis.

Lab mode offers a useful sharpening technique that is especially good for architecture.

Basically you convert the RGB image to Lab, select the lightness channel only and apply large amounts of sharpening. This sharpens the basic black and white structure of the image without affecting the chrominance in colour channels a and b. So, you get very sharp images without smearing the colour.

After the sharpening, convert back to RGB ready for saving.

You are comparing apples and oranges here...
* "AdobeRGB" is a colourspace (sRGB is another)
* "Lab" is a colour mode. (RGB is another)

Amoringello's comment is, therefore, incorrect as Lab is not a colourspace

The colour mode is how the numbers represent the colour within a given image. If you like, it's the recipe for mixing a particular colour.
In the case of Lab there are three numbers per pixel - "L" for lightness, "a" and "b" for colour.
In the case of RGB there are three numbers per pixel - "R" for red, "G" for green, "B" for blue.

The colourspace (or gamut) defines the range of colours that a given colour mode can represent. In a sense this defines the scale of range (a bit like choosing a ruler that's either a 1m or 30cm ruler and defining the numbers as a percentage of the rule - if you get my drift).

That said, John's comment is accurate... but some of the tricks of converting to Lab and back again are unnecessary in PS. There are two useful tricks for Lab - (1) sharpening, (2) boosting colour contrast.

Both of these can be achieved implicity:
1) Duplicate layer; sharpen (hard); and change layer blend mode to "Luminosity".
This has the same effect (*) as the conversion to and from as Photoshop only adjusts the Luminosity component. (* well, very, very nearly.)

2) Add a "Brightness/Contrast" adjustment layer; Change layer blend mode to "Color"; Increase the contrast in the control.
This does the same trick that can be done in Lab mode where the a+b channels are given extra contrast.

Most experts generally advise against unduly converting from Lab-rgb-Lab because there is a real possibility of introducing errors in the conversion (due to number rounding errors). So the above two methods are helpful

There are more colourspaces too - ProPhoto is a very good one (but needs to be handled in 16-bit only). AdobeRGB is a good general working gamut that suits printing very well.

There are other colour modes too - CMYK being another (useful for creating colour separations for offset-Litho printing).

Hope that clears it up
Matt

And in Lab you have:
CIE Lab
Hunter Lab
etc.

The colour industry is one of the most confusing industries because there are so many different spaces and modes and every mode or space has it's variations.

Thanks Guru.