If you're not an imaging nerd, move along.
Some guy did a video on setting some custom processing profiles on the Canon 7D in order to get more dynamic range out of it. I uploaded the profiles to my camera, and wow do they make a big difference. The win here isn't better images, per se, but the ability to fix under and over exposed stuff in post. If I can learn how to use the Color app in Final Cut, this would make an enormous difference. Yes, much of the work comes from what you do after you record.
The discussion around this is still pretty interesting, in that if (or the optimist in me says "when") Canon comes up with a way to record in a "raw" format, the way it does still images, you've essentially got Red by the balls, and empower every jerk with a (relatively) cheap Canon able to create images just as solid.
But until those days come, I have to say that this helps the situation. These flat profiles retain a lot more detail in light and dark areas, which makes it a lot easier to do your own processing. The camera records in H.264 at a bit rate of like 45 mbits, which is pretty damn good for that codec (think JPEG quality 90 in Photoshop... it certainly approaches that). The pro codecs like DVCPRO HD and XDCAM I think are 100 mbits with higher color sampling and far less data loss, but they also assume you're going to be compositing or something.
You can still correct for some of the over processing by the camera by simply lighting correctly, but when you're shooting out in the wild, that's rarely possible. I'm going to shoot with this flat profile this weekend, and see what I can do about learning to use Color after that.