I have been working on a concept that combines two different computing methods that tries to solve two limitations of photography. The first issue is the limit in dynamic range of light that can be captured with a single picture. The human eye has solved this problem with adjusting the iris dynamically as more or less light comes into the eye. Still cameras do not have this ability. The second issue lens optics of SLR cameras limit the focus depth. The best focus depth is obtained at the hyperfocal point, but in many cases this is insufficient to have the whole picture in focus.
So two techniques have been developed in an effort to address these limitations. The first is high dynamic range imaging. This approach allows the photographer to recombine different exposures of the same picture in software, thus extending the light range in the final image. The second is focus stacking. In this technique, a series of photographs are taken at different focus points in the image. Subsequently, these are recombined to form a single image with a much deeper depth of field.
All of the pictures were taken with a D800 and either a 24-70 or 70-200mm f2.8 lens at f11. At each focus point, three exposures (- 1 ev, 0 ev, 1 ev) were taken. First, I processed the three exposures to generate a HDR image for each focus point. I manual selected the focus point and adjusted focus using live view. I used Photomatix software to combine the different exposures. Then I recombined the HDR images into a focus stacked image using a test version of Helicon Focus (v 4.2.9).
1. The best image will have significant depth of field interest and significant dynamic range needs and,
2. The elements being photographed really need not to move. A tripod and release cord are essential to obtaining the best results.
The pictures below are my second attempt with this approach. Obviously, there is a tremendous amount of post processing and whether it is worth the effort is subjective. I believe with the best subject images, it will likely produce a pleasing result.