Roy, those two shots are no good as examples. You will get exactly the same result taking those two shots even with the same camera.
This is a moderately difficult exposure situation you have, though common enough. You have a dark(ish) foreground and a bright sky, with (in this case) the horizon almost straight across the middle.
Compare the two shots side-by-side or flick between them. In the first shot, the foreground is exposed correctly, the sky is a bit too bright. In the second shot, the sky is exposed correctly, the foreground is too dark. Notice that in the first shot (foreground correct) the horizon in centred; in the second shot (foreground too dark) the horizon is above centre - i.e., the camera is metering off the sky.
So, in both cases, the camera has done the right thing, it's just metering off different things depending on what you point it at.
I initially found my new 400D was under-exposing around 1/3 or 1/2 a stop (compared to my 20Ds). I started by adding a half-stop of EC, which was often too much. Then I went to a third of a stop, which was better, but still upredictable and annoying. Then I switched from partial metering to evaluative, with no EC, and got better results again. I still don't think the 400D gets exposure as right as often as the 20Ds do, but it no longer annoys me as much as it used to.
|