Originally posted by D1N0 Only when you are doing a stupid bike test using a tripod. Not the way AF is meant to be done at all. Handheld shooting following subjects is what you are supposed to be doing. That that gives hard to reproduce results is not our problem. It is real life use, not pseudo-scientific nonsense. Like emission tests on a roller bank
.
They think they need to Order cameras, Score them, Rank them, and declare a 'Best' camera. The Real-World is subjective. A real-world-test conclusion can be argued, so they tried to devise a
controlled test (like IR's bottle label, fabric and embroidery thread scene) that can be scored on % in focus captures.
This is how Ph.D.'s think.
Unfortunately their test, though somewhat repeatable (human rider, changing light conditions, changing clothing), is not representative of real-world use (too much white for color-assisted AF sensors, tripod mounting f/2.8 lenses, different lenses for different bodies, etc.) The information gathered is perhaps correct but not relevant and useful, and the test is not controlled because the testing equipment is not constant.
Therefore, since the test itself is flawed any conclusion drawn is invalid on its face.
A FAIL for the bicycle test. Shame, really.