A basic step in HDR (I guess) is aligning the bracketed images together, but what I don't understand is why we need this step. The images are all the same, since they were taken from the same location (maybe on a tripod) with time lag of only a fraction of a second. So, it looks like all the images have the same settings except the exposure. So, what is alignment all about?
With high resolution images, even the tiniest vibrations will be visible in the image when viewed at the pixel level. The same holds for slight movement of the subjects in the image (from wind, motion, etc.). Misaligned images degrade the quality of the result of the HDR processor.
Note that if you shoot RAW, you can use a single image to generate a few bracketed jpegs of different exposure levels. These jpegs are perfectly aligned, so if the dynamic range you want to encompass in your HDR image is small enough, you can use these aligned images as input to the processor.
It may happen that your pictures are perfectly aligned to begin with. It may happen. But pixels are really small, and it doesn't take very much in the way of camera movement to change the pixel location of details in the subject. With typical field gear, your images are very likely to be subtly different -- just off by a pixel or two -- and that will affect the sharpness of the final image.
You can shoot multiple pictures in perfect registration with the right gear and technique. Medium and large format 4-shot cameras rely on that -- but that usually means relying on a camera stand (a multi-hundred pound behemoth that is related to a tripod in sort of the same way that a Peterbilt is related to a Mini) and a geared head in a studio environment, with the mirror (in a MF SLR) staying locked up throughout the sequence.
With an ordinary field tripod and a typical head, pixel-perfect registration will be a matter of luck. With a hand-held camera, it's more a matter of Pollyanna optimism gone mad. You may be able to create decent web-sized pictures without alignment, but you won't be able to create a picture that will stand up to high-resolution display without a lot of unnecessary and artificial sharpening (which is, really, trying to perform the alignment after the fact).
You're right, with a use of a robust tripod the pictures should be all aligned by default, but not all tripods are stable, some cameras don't shoot automatically so it requests you are pressing the shutter button constantly causing shaking and there is the chance of the photographer not having a tripod, for this the alignment is required. In my case I shot HDR before I had a tripod and this was a necessary step for me.
I just use Photomatix to generate HDR and there you can skip the alignment phase, if that is your question here.
In the end this step is just to make sure that all of the photos are set in the correct position before go to in the tonemapping.
In another scenario, although the camera might not move, the objects in the scene might and this allows the possibility to keep what you want to be aligned to keep it that way.