It's not that the camera can't pick up low light, it's that it can only work within a specific range of light and still capture detail. Stuff outside of that range gets rendered either black or white. A major part of learning to take good pictures is learning to recognize that and exploit it.
Back to my first post:
<a href=showthread.php?s=&postid=15501596#post15501596 target=_blank>Originally posted</a> by beerguy
Usually I'll pick the brightest object in the frame and set my exposure to it. Then I'll look at the darkest portion and see if I'm still within the range of the camera.
Looking at my meter, I know that anything that shows up as +2 will be rendered white with little or no detail. Anything that shows up as -2 will show up as black with little or no detail. In the scene that you described I'd probably also start by metering the sky and set the camera to it. Then I'd meter on the dog. If metering on the dog tells me that the dog is going to be too dark I have a few options.
1. Recompose the shot to eliminate the sky. If you've got something in the frame that's hotter than the rest of the scene, the simplest thing to do is eliminate it.
Example:
http://www.binaryemulsion.com/wordpress/2009/06/22/rock-creek-aspen/
The sky in this shot was bright, boring and at least 3 stops hotter than the scene. I wanted the image of the aspen so I just excluded the problem.
2. Make an artistic decision as to what is important to me in the shot. If the dog is the subject then I should set my camera to expose the dog rather than the sky. In that case you'd need to accept that your sky is going to blow out. Likewise, sometimes it makes sense to allow some of the frame to go black.
Example:
http://www.binaryemulsion.com/wordpress/2009/07/11/simple-is-good/
Here I could clearly see detail in both the tree and the foreground but the image is clearly about the color in the sky, not the foreground.
3. Use split neutral density filters to bring down part of the scene. This doesn't always work as not every scene lends itself to this technique. I use split grads A LOT. If you were to run into me out shooting, I'd have at least 4 different ones in my vest pockets.
Example:
http://www.binaryemulsion.com/wordpress/2009/08/11/missing-yosemite/
The sky in the background was a full 3 stops hotter than then it needed to be to match the foreground exposure.
4. Take two exposures, one for the highlights and one for the shadows and blend them in Photoshop. This isn't the same as HDR. It can also be done by processing your RAW image twice and combining the two. I load both as layers and then use a luminosity mask to blend the images.
Example:
http://www.binaryemulsion.com/wordpress/2009/05/15/image-honesty/
This is what the scene looked like to the naked eye but there's about a 5 stop difference between the moon and the foreground. I recognized that, while shooting, and took two exposures back to back at different shutter speeds knowing that I'd have to combine them.
5. HDR software. The best results here are with at least 3 images, sometimes much more than that, taken at different shutter speeds to expose the whole range of the scene. There are lots of problems with this approach. If anything is moving in the scene it'll be very hard to produce a sharp image. Folks also tend to really overdo the adjustments and end up with a garish image. More often than not, they use HDR to take an otherwise boring and poor photograph and make it look "neat." I'm not saying that it's impossible to create realistic HDR images, it's just that most people don't, mainly because it's a much steeper learning curve than just making it look "neat."
Don't look at it as a limitation. Learn to recognize what your camera can or can't do and they leverage that knowledge into producing a good photograph.
Cheers
