As a happy owner of a new D800, I thought it was a good idea to check if some of my lenses needed some adjustments of the auto-focus.
As I didn't want to spend a lot of money on extra gadgets or software, I downloaded the usual test charts available on the web but I was not really satisfied about how they looked and allowed me to gauge the results.
Therefore, I thought about the process and came up with the method I describe on my blog:
#1. "RE: My method for fine tuning the auto-focus" In response to Reply # 0
Very interesting Bernard I like the test. I wonder if the grid would confuse the focussing at all? I assume before the grid is not vertical (perspective) it would not. I have tried it with three of my lenses on my D800 and all focus exactly in the centre. Was a useful exercise so thanks for the idea.
#4. "RE: My method for fine tuning the auto-focus" In response to Reply # 0 Mon 12-Nov-12 04:55 AM by jamesvoortman
I like the method and will try it.
I have a small issue with the advice to shoot at minimum focussing distance. How do we know that the focus point correction applied at this distance will be applicable to normal shooting distances for the respective lens? For zoom lenses, will it be the same at both wide and long settings? I would rhink the best is to carry tests out at normal working distances for the lens involved and for zooms to check the correction required at several points over the focal length range, choosing a correction that it either averaged or appropriate to the most commonly used setting.
Therefore for macro lens : Very close working distance and small image area.....A test page with finer dot pattern might be a good idea.
For 80-400VR : The most critical sharpness requirment is when shooting birds - usually done at 4m to 10m with this lens. Again the target may not be of appropriate size.
Wide angle and standard lenses I use normally for general purpose and landscapes - frequently used with distant focus not practical for this test so obviously I will conduct the test at reasonable working distance and hope the results translate to longer focal distances
#5. "RE: My method for fine tuning the auto-focus" In response to Reply # 4 Mon 12-Nov-12 02:09 PM by Gromit44
>For 80-400VR : The most critical sharpness requirment is when >shooting birds - usually done at 4m to 10m with this lens. >Again the target may not be of appropriate size. > What about printing several copies and taping them together (in a line) to make a larger target?
#6. "RE: My method for fine tuning the auto-focus" In response to Reply # 4
All the points you raise are legitimate and apply to all methods designed to fine tune the auto-focus.
1) Concerning the focusing distance, it is best to choose one for which the depth-of-field is narrow. First, it is only in such a setup that ultra precise focusing is really needed. Second, if the depth-of-field is not small, it is difficult to visually gauge if the focusing was done at exactly the right place.
For longer lenses like your 80-400 with shooting parameters, 400mm, f5.6, distance of 6m to subject, the depth of field is around 8cm (4cm in front, 4cm behind). I therefore suspect that even at 6m, a single test chart will be very efficient in determining proper focusing.
2) For zooms, it is indeed the case that at different focal distances, a different ideal correction might be needed. Hopefully, it is possible to find an average correction that is acceptable for all focal lengths. Otherwise, a Nikon repair center should be contacted.
In any case, whatever correction is applied, the final test (without any sort of chart) comes from trying out the correction on your usual subjects at normal working distances. If the vast majority of pictures turn out good, then the fine tuning is done properly for you.
All the fine focusing methods should be considered as aids to achieve good focusing but they do not replace real field testing.