The new file contains only one target, a bit larger. It's the same pattern of stripes with differing widths. I revised the target that I described earlier in this thread. I'll describe the details of my test procedure, and the results, in another post. Here is a set of shots showing the setup I used to test the sensors in my D800. This is a 100% crop from a screenshot in ViewNX, showing one of my targets under a focus bracket. ViewNX and CaptureNX can be configured to display the selected focus bracket for a shot. If you position the target on a plane parallel to the image plane of the camera, and at a distance 50x the focal length of your lens, the target will be approximately the right size to fill one of your focus brackets (at least, for the cameras I have tested recently: D700 and D800). The black background of one target should measure about 100 x 60mm. I recommend printing it on a good inkjet. The file contains three of the targets on an 8.5x11" page. I made some targets based on this design, and it seems to work well on the line and cross sensors in the Nikon PDAF system. The width of the bars varied, to prevent errors caused by shifts of one or more pattern cycles.
#Target photo print pdf series#
The focus target in that rig was a simple series of alternating white and black vertical bars.
In a message elsewhere on this forum (I lost track of the link), someone showed a photo of a focus test rig at a Nikon service center. For example, a starburst pattern can be useful for evaluating sharpness, but when used as a target for a phase detection AF system, the results might be somewhat sensitive to the exact aiming position.
Some people are running tests using autofocus targets that might be less than optimal.
0 kommentar(er)
