Think of three different vessels, a 16oz glass (sRGB), a Quart jar (cmyk), a 2 quart cake pan about 10 inches in diameter (Adobe RGB 98): I could put a couple more in there but it would make it unnecessarily confusing - as if it isn't already [img]/ubbthreads/images/graemlins/wink.gif[/img] . As you can see the smallest gamut is sRGB, it not only holds the least amount [volume] but it also is narrow in diameter and range. The Adobe RGB is the largest and is spread over a much bigger field of vision, a more capacious range of color, ensuring detail in shadow and highlight areas: An extremely sensitive gamut allowing for maximum image quality, color, saturation, brilliance and contrast. Now the cmyk is between the two and this is the gamut used for the magazines, ink-jet printers, etc., I am illustrating it because most of us take an image from another gamut and then print that image with hopes it will look as good as what is on the screen.
Okay, first scenario, you are shooting in sRGB:
When you shoot in sRGB you are limiting yourself to 16oz of total capture volume, and what is saved to the card is never going to be more than that. When you save that image and convert to Adobe RGB, you pour the 16oz of liquid into a much larger color container, but unfortunately you only have 16oz worth of information - but - when the information spreads in the larger cake pan, the contrast and color saturation usually change because the field is much larger and the gamut provides better range, but only through emaciated data: Unfortunately you cannot create something that wasn't there in the first place. You have more gamut space to enhance the image in new values, some contrast, and performances, but you do not get the precision that was lacking before. sRGB is a limited gamut designed for electronic formatting, a limited gamut of thousands of colors that easily displays on the www and any transportable browser - cross platform as well.
Second scenario, shooting in Adobe RGB:
This time we shoot using the cake pan, a large area, great color volume, contrast detail, dynamic range, 64oz of color information and magic. Once we have this file, we can at any time in the future use this image for anything we want and not have to try and recreate what might have been lost. Here we can change the image to sRGB and the transference keeps all the important information because it was there in the first place, and if we go to any printing device, we can change the image to cmyk of any ICC profile and our original is limitless in information. This is the non-handicapping gamut, so to speak: Millions of colors.
Third scenario, shooting in raw:
This is the best, although the slowest - on some cameras - and card using epidemic. There is no compression of a jpeg nature and the gamut is going to be the largest the image will recreate, on top of that the format is designed to be altered in all aspects of the proprietary [camera] software's capabilities. From here you can save a file out as sRGB, again as Adobe RGB, or offer cmyk a complete gamut to translate from.
Remember at HTL Jim when we were talking about the raw image being flat, low in contrast but high in volume, and when you take the image into your proprietary software you can adjust the contrast curve and the image really pops? Well that is because you have more information to begin with in that flat image, the data is spread and non defined. It is a beautiful thing. [img]/ubbthreads/images/graemlins/grin.gif[/img] After you create a super quality image output from your raw data file, out from the proprietary camera software, you would first save it as an Adobe RGB tiff or .psd file, then if you want to post it to the www, you make a smaller jpeg sRGB version and you have the best of forever worlds.
Note: When you take an Adobe RGB image and convert to sRGB, the saturation and color always shift away from red and orange, your image will become dull and lackluster. To compensate, open (in Photoshop) Image>Adjust>Hue/Saturation, adjust the Hue -4 (move the slider to the left - this will increase the magenta putting warmth back), and push the Saturation to 10 or 12 (slider to the right - this will pump the lost life back into the image).
Having illustrated all of this, the obvious for a camera manufacturer to advise shooting sRGB is because of speed of information, if the gamut is smaller, the bit depth is automatically limited - this btw is where these volumes of information are stored, bit depth; the more information, the more processing is required at the time you shoot. Your burst rate or capture rate is dictated by your ram or cache volume and processing time.
In review, if you are shooting only for the www, you can shoot sRGB, a gamut of thousands of colors. If you intend on using the images as if they were archival fine negatives for your future, to be used in [cmyk] print, to make beautiful Light-Jet prints or print from any die sub or quality rgb output device, then you want to shoot in a richer gamut; sRGB in millions of colors.
Learn more about gamuts at Supershoots San Diego Smackdown, Feb 28-29