neuroanatomist said:
Ripley said:
My AFMA test results differ between wide open and stopped down to the sharpest aperture, why?
Possibly focus shift (with some lenses, stopping down shifts the focal plane).
Expounding on our discussion, here is what Reikan has to say on page 127 of "FoCal for Windows version 1.8.1"
23.3.3 Test Aperture
...It is generally suggested that the best aperture for an AF microadjustment calibration test is with the lens open widest – this will give the shallowest depth of field and the biggest difference in analysis results for each test point.
However, under some situations, the analysis may not work optimally at widest aperture – particularly with lenses that are very soft or exhibit strong aberrations at widest aperture.
If this setting is changed, it is best to run as close to wide open as possible.
Testing at the sharpest aperture is not best practice, nor does testing wide open always yield optimal results (presumably mostly with fast primes, like the 50L you mentioned).
[/quote
Just to be clear, altering that setting is
NOT changing the aperture at which the autofocus step is performed. It's changing the aperture used to capture the images for analysis. So, in the case of the 50L for example, running the test wide open would give accurate results for shooting wide open. Testing at f/2 would give a different result (at a given distance fairly close to the camera), applicable to shooting at f/2 at that distance. Also, the deeper DoF with a narrower aperture masks misfocused shots, making the analysis more challenging for the software (flat curve, hard to find the peak).
