Basically, you need a way to capture the wide range of light intensity in the scene, so you can see colour detail both in (for example) the bright sky and the darker silhouetted foreground. In the above cases this was done with multiple exposures using different shutter speeds.

I didn’t have a tripod, so the shots were hand-held. Later, I aligned them and then turned them into HDR image files (OpenEXR file format).

Finally, I used a process called “tone mapping” to remap them back into a regular (low dynamic range, 8-bit) image file. The tone mapping process I used looks at intensity locally, more the way the human eye does, so that you can (for example) see both the sky and ground with similar “actual” intensity levels and yet get have impression the sky is brighter. It can give a sort of surreal look when viewed all at once! In fact, ironically — after all these years of trying to make computer-generated imagery look “real” — it makes real photos look, well, computer-generated. And that’s…cool? (-;