Page 107 - Friedman Archives

Lenses 107

I did that test because of a blog post I wrote in which I tried to find the A77
II’s focus tracking limits. I found it pretty easily – a zip line test where
subjects were zooming toward me at speeds much faster than 45 MPH, and
the camera only got about 75% of them in focus. (You can read the
original blog post here: http://bit.ly/1rUQdWy . I got a lot of flack for that
post because I was using an older lens, and it’s “simple common
knowledge” that the newer, SSM lenses can follow-focus faster than the
older screwdriver-blade lenses.
Well, it may be common knowledge, but it’s also not true. Once focus has
been achieved, a lens does not have to move all that quickly to track a fast-
moving object, since the lens is already almost at the right place. Only a
small tweak to the focusing position is required, and if driven by the correct
predictive algorithms, even the oldest of Minolta lenses can do it.
Now let's have a look at one of the zip line examples I shot as part of the
research for this book: http://bit.ly/1ovj5tu (99 MB). The settings on the
camera for this sequence were:

• Continuous Shooting Low
• AF-C
• Focus Area = Lock-On AF Wide
• AF Drive Speed = Fast
• (Picture) AF Track Duration = 5
• Priority Setup = AF
(That last one means it won't take a picture unless the camera thinks it's in
focus.)

Now let's look closely at one of the out-of-focus shots; let's say
DSC04446.jpg. The background is in perfect focus! A quick look at the
other out-of-focus shots reveals the same thing. This is clearly NOT a case
of "the lens couldn't keep up"; it is a case of "the camera got confused and
told the lens to focus on something a very different distance away". (And
the lens re-focused very quickly!) A few shots later it recognized its
mistake and found the subject again. THAT's where fast AF matters –
when you're recovering from focus tracking errors. 