Practice creating looks with footage in different codecs from different cameras to understand how each responds to manipulation.
Every camera has a different latitude, different contrast curve, different color response. To compound the possibilities, the codecs that the footage can be digitized as vary just as wildly.
The RED camera responds very differently from a 5DmkII (or an Alexa, or an HVX, or an EX1, or a Phantom, or 35mm, or 16mm…) and you need to know these differences to know how best to manipulate the footage and how aggressive your look can be. Does the footage tend to be warmer? Does it have more contrast? Is there more noise or grain in the shadows?
Codecs have different properties too. HDV tends to be a nightmare to edit, but it also can be a nightmare to color with it’s long-GOP structure. Compression blocks and artifacts can be hiding just out sight in the shadows. ProRes is a personal favorite of mine to color, because I feel it handles high contrast very well. RED R3Ds can be a logistic nightmare when managing colorspace, gamma, log vs. linear, ISO changes, etc., but it’s high dynamic range compared to normal video, plus the extra information it keeps in the highlights and shadows make it great once you make those decisions. The H.264 that the HDSLRs shoot breaks down pretty quickly, so you have to be very careful when stretching contrast.
Knowing the properties of the cameras and footage will help you have an idea of how far you can grade right from the start, and when you have a mix of cameras and codecs, will let you make sure your look will work for them all.