Color management confuses and defeats the VAST majority of its users. There are very good (guru) users out there with good websites. My favorites are Andrew Rodney's
http://digitaldog.net/ and Bruce Fraser's
http://www.creativepro.com/author/home/40.html I am FAR behind these guys in color knowledge! But, I've always felt the subject needed a better "intro course" than is out there. It is so complicated and with so many "exceptions", it is really hard to find the basics for the exceptions. So, here is my attempt at a "BASIC SIMPLIFIED BASIC" intro to color management:
Somewhere in the 1980's Apple came up with a program called "Colorsync". This allowed hardware devices to tell each other how they handle color and to "play well together". The IBM/Microsoft world first ignored this, then liscened it but implemented it half heartedly. This is much of the "Mac is better than IBM for photography" talk. In reality (and I'm a die hard Mac owner), the IBM world has finally caught on, and the two systems are now virtually equal in color management.
There are more colors in the world than we can see. There are more colors that we can see than can be recorder by a typical consumer or pro camera or scanner. There are often more colors our camera or scanner can record than our printers can print. How we handle these "out of gamut" colors is very important. Lets MAKE UP some numbers and situations: You have a camera that clips reds... sound familiar.. this camera when photographing a "red" object that is 94% red, 3% green, and 3% blue sees the object as pure red. It writes the data file as 255,0,0. If you open this file in a computer that is capable of presenting up to a 98% pure red, your file says 255,0,0 so the computer shows you its max, a 98% red. This is a 4% error, your red is now neon and you are unhappy. If the camera file could tell the computer.. "hey, I'm a little weak on reds, my max is 94%", then the computer could compensate and give you the best view (and print) possible. This giving of info is called "tagging" a file.
That is Colorsync's purpose. Apple's Colorsync is still the main component of color management (for Macs and PCs). It has two primary modes "device" and "space".
A device profile should be provided by the device's manufacturer and describes how the device handles color. Each hardware device that handles color (monitor, camera, printer, scanner) should have its own unique profile.
Colorspace is a bit harder to describe. All of these device and profiles have to "play together" in a "space". The space determines the outer limits of color. Just how pure a red can any device be allowed?? The two most common spaces are sRGB (standard for the web) and AdobeRGB (somewhat standard for printing). sRGB is a "smaller" space, it doesn't allow as saturate (bright) colors. Sounds bad, doesn't it? But, if you are trying to get 10 million folks with 10 million different monitors to see your pic, a limited color set is just the "ticket" to a degree of consistency.
There you have the simplest I can describe it. That is dangerous, there are SOOOOO many exceptions and footnotes in color management. They are important! Read Rodney or Fraser for a start on these. I'll mention a few that are specific issues with our Sonys.
First, there is one "colorspace" that is sometimes also a device profile. sRGB was designed as a "universal" profile. It was intended that lots of manufacturers would design there devices exactly to this profile, then folks could just use sRGB for all settings and things would be simple. Sounds good.... but very few manufacturers did it or did it well. Also, sRGB is a small space.
Second and most important: Sony ($%#@ them) doesn't provide a device profile. So, we are rather handicapped from the start. The F707 is "close to" sRGB so many folks use that. I just hate to spend $1000US for "close to right". When you have a device without a profile it creates "untagged" files. There is no info for colorsync to read, so it has to guess. How it guesses differs with different programs and preference settings. You are really in "the wild" with an untagged file.
Many programs will assign (tag) the colorspace in your preferences to any untagged (Sony) file. If this is the case with your program, it is best to assign sRGB as your preference. Other programs will warn you of untagged files and ask what to do. In this case you can assign sRGB or my profile or Pop Photo's. If you use my or Pop Photo's profile you are now in a non standard space. You may want to "convert" to a standard space if your printer requires this or you are mixing images from other sources.
This "Convert to profile" actually rewrites the data in your file. In the above example (weak reds) your 255,0,0, clipped red could get rewriten as a slightly impure red (perhaps 245,5,5) to compensate for the fact that colorsync has been told that is all the original device was capable of. You don't gain or loose anything here. You don't get a bigger set of colors by converting to a bigger space (notice your small space Sony red go rewriten as impure).
Remember that as with layers and judges, it is often the details that get you "off" or convicted. It is often the "details" that are needed to get your system running at 100%. I haven't even begun to talk about details here. I just hope that with some basics, the details will begin to make sense!
PIM tries to simplify and automate many of these choices. It's main limitation is that you have to have both a camera and printer that are compatable. If your work is going to a non PIM printer, you are once again "in the wild". It may be a great choice for the home user.
--
Tom Ferguson
http://www.ferguson-photo-design.com
F707 Studio Info Page:
http://www.pipeline.com/~tomf2468/temp1.html 
You EXIF viewer looks nice, more complete than mine. But, I could not find a Mac version anywhere. 