Follow up on focusing issues with Hasselblad 503cm

Posted on April 11, 2012


I have had a good look again at this issue by examining first whether the camera is focusing accurately and secondly  are the focusing problems down to operator error and in what circumstances does this arise. I was testing using my Hasselblad 503cm, 80mm CF lens and Phase One P45+ digital back.

It is not particularly easy to test the focusing accuracy of a lens when one is not sure about if it has been focused correctly. In order to do this I set up an experiment for which I can be pretty sure about my manual focusing accuracy. This involved shooting a Colorchecker target from exactly 36 inches. I focused on the top edge of the target using the split screen which was relatively easy to do as it filled much of the frame. I checked the distance using a Leica Disto meter. After my effort at focusing the  lens ring showed exactly three feet or 36 inches. I shot the image at f2.8 which by my calculations would give me around 1 inch of depth of field. This would show any focusing accuracy problems with the lens/camera as they would not be masked by available depth of field. The image I took is shown below along with a 100% crop.

Test image shot at 36 inches at f/2.8

100% crop of top edge of target

This shows that the top edge appears pretty sharp with the focus falling off rapidly behind. There is some softness but I think that this is more like down to the softness of the lens at f2.8. So it appears that the lens/camera combination is not backfocusing. The issue seems more likely to be operator error.

To test my focusing accuracy using the split screen I made a series of pairs of images. The first of each pair was focused manually by me using the split screen. The second was set at the distance indicated by the Leica Disto meter. I shot many images of different types of subjects at different distances. My focusing was pretty consistent with the meter when the subject either a) had a sharp edge on which I could focus or b) was relatively close i.e. within the range 3-5 feet. For less defined three dimensional objects particularly at distances over 5 feet my accuracy was less consistent. There was no pattern to my lack of accuracy sometimes I was over, and other times I was under. Never by much but sufficient to show as I was shooting at f/2.8.

The process which gave very consistent results was to focus using the split screen, check using the Disto meter,  refocus if there was a discrepancy to the meter reading,check again using the split screen. By using the Colorchecker as my target the focusing is made easier, although at a distance of 7 feet it is very small in the viewfinder. I have been shooting my portraits at distances between 5 and 7 ft using f/8 or f/11. Here is an example of an image shot at 7ft using this process.

Test image shot at 8ft f/8 at iso 200

100% crop of above image

The bottom line is that I need to practice more and to be aware of my limitations in the short term. I will keep posing updates to the blog.