I wanted to determine hyperfocal distances for various lenses on the Alpa since focusing lenses on it are awkward at best. Hyperfocal distances would be useful. The problem is there are no scales and the DOF varies with sensor dimensions and pixel size (discussed below).
Equipment: Schneider 24XL Digitar lens on Alpa with a Leaf 65 back at ISO 100, processed in Raw Developer (all the same settings for all images). Alpa on tripod, subject was an outdoor scene with natural light, f-stops were sequentially tested at f5.6, f8 and f11. The focus on the 24XL was set at infinity, 10m, 5m, 3m, 2m, and 1.5m. (These markings, except for the 2m distance, are present on the helicoid mount, so were easy to set.) Images were taken at all focal points with all 3 f-stop settings. Also note that I have not shimmed the Alpa since the lens, as delivered, seems equally sharp in all corners.
Images were then enlarged to 100% and the central areas of each image were compared. The basic question I wanted to answer was "at each f-stop, which was the minimum distance that produced images with infinity sharpness equal to the image shot at the same f-stop, but focused at infinity?" Such a distance is the hyperfocal distance for that f-stop and lens combination. [One very important point, is that the images are being viewed at 100%; this is not the way things used to be viewed for prints when discussing DOF, circle of confusion (COC), etc.]
The results in reverse f-stop order:
At f11, images focused at 3m were hyperfocal. When focused more closely, the objects at infinity became softer. No CA was noted at any distances.
At f8, hyperfocal distance was 5m. No CA was noted at any distances.
At f5.6, 10m gave the closest image quality to the infinity setting. This 10m distance was not as definitive as the hyperfocal distances at the other f-stops. Closer distances were definitely softer at infinity. However, all images taken at f5.6, which were not focused at infinity, produced CA at infinity. Further, when focused at infinity, images taken at f5.6 had objects at 8-10m that were not critically sharp.
If you try to run this data through various on-line DOF calculators that take into account the COC, which was initially defined for a specific size of print enlargement when viewed at a specific distance, the results would suggest a COC behaving somewhere around 10 to 15 microns. This is in agreement with a COC for digital sensors being about equal to their pixel size (around 7 microns for the Leaf 65/75 backs). These findings are therefore only useful for sensors with similar dimensions and pixels in the range of 7-9 microns.
The Digitar 24XL seems to be hyperfocal at 5m at f8 or 3m at f11 on the Leaf/P45 backs. The other finding is that f5.6 has too much CA to be useful. Another conclusion would be that one cannot use DOF indicators on lenses designed for film when using digital backs on MF cameras (they'll be around 3x narrower, taking into account the sensors being smaller than film, but the COC of film being larger than the pixel size of digital sensors). A useful on-line calculator can be found at Bob Atkins' site at
http://bobatkins.com/photography/technical/dofcalc.html For MF digital backs, use the "2x DSLR - 15 microns" for the "Film Format" setting. As an example, by putting in 35 for the lens (such as the 35XL Digitar), the hyperfocal distance at f8 is about 10m and at f11 it is 7.4m. For the 47XL, these values would be 18 and 13m, respectively.
