Elie I would have to disagree with you on the affect of changing from a smaller colour space to a larger one and back again, as in from sRGB to AdobeRGB. Both colour spaces have the same number of shades of colour,which depends on the bit depth, either 256 per channel for 8 bit images or just over 65000 for a 16 bit one. The colour space profile is used so that the computer hardware (monitor or printer mostly) can go from the three colour numbers to the same exact shade of colour, independant of the device being used. Doing this and having lots of possible colours means that those colours are further apart. So sRGB being relativly a small colour space the colours are packed closer together. If we have an image with lots of smooth gradients that all fall inside the colour space this is good. Now we convert the image to AdobeRGB, this is a bigger colour space, so each number has to represent colours that are further apart. So now we have to loose those extra inbetween values that the smaller sRGB space allowed us. Now the image only had values that were within the size of the original space, and it can only hold information on the range of colours in the original space anyway. So when we convert back to the smaller space we won't loose any of the colours in the new space (assuming we haven't added any extra colours during processing). But now we only have the "further apart" colours left in the image. So now we actually have less total number of different colours as all of the original inbetween colours got converted to the nearest avilable colour in the larger space. If you are doing this with 16 bit colours then you might not notice the difference, and the conversion to 8 bit will have a much much bigger effect. Doing it with 8 bit's though may get you a noticeable, rather than measurable, effect.