I think we all, as 3D artists, fall into the trap of set it up and hit click. Perfect. In reality, we need to think like photographers, and I have worked with some, I will guarantee you that shots like you describe, no image ;-), are either heavily edited post-shot or are meticulously set up to cheat reality with out-of-site lighting to fill dark areas as needed. Even darken areas with large scrims (translucent fabric) to soften harsh sun.
We're assuming you are creating stills (I guess this could work with animation too). One way is to render the main pass for optimal lighting for the majority of the shot, then you can render a different exposure for the areas in question. I believe there's a way to just render a defined area to composite later so you don't have to re-render the entire image twice. Saves time.