neuroanatomist said:
tr573 said:
I edited after - I rechecked the 7D2 manual and it seems to indicate it only does face detection in one shot mode, not AI servo.
It does face detection in AI Servo, as well...
during tracking. The manual defines iTR as: "intelligent Tracking and Recognition: The function that the metering sensor identifies the subject to make the AF points
tracking it."
The point that Rishi fails to grasp is that based on Canon's description of iTR, when in AI Servo AF while allowing the camera to automatically select the initial AF point(s), data from the metering sensor are not used.
From the 1D X II manual, for the second time:
When iTR is ON in 61pt Auto Selection mode: "The AF point is
automatically selected based
not only on AF information, but also the human
face and the
subject's color information."
On the other hand, when iTR is OFF,
only then are: "AF points automatically selected based
only on AF information (The AF does not use facial information or the subject's color information.)"
Do you see the difference between the descriptions for iTR ON vs OFF? And yet you continue to claim that the behavior when it's ON is actually the behavior when it's OFF.
?
neuro continues to post misinformation that misleads the community, simply because he can't now backtrack and admit he was wrong in his all his numerous posts about how our
one motocross piece was supposedly wrong (which it wasn't). Because that would indicate that
DPR was right - which it just
can't be in order to feed his continued character assassination and claims of bias and incompetence.
Furthermore, he continues his red herring argument, ignoring the fact that
even if the AF system were designed to work as he claims (it isn't, according to the
manual), then our observation that the camera tends to focus on the nearest subject - while the other one felt a bit more intelligent in what it automatically selected - is still
valid and useful information.
Hence, not only is neuro's criticism fundamentally flawed because his own position is wrong according to Canon's own 1D X II manual, it's also flawed because the criticism itself - that 'DPR just described the system does what it's supposed to do!' - isn't criticism at all. Even if we were to describe the system does what it's designed to do - and that that behavior is not ideal - how is that grounds for disaccreditation?
The principle behind '
auto' selection is that the camera should intelligently, automatically find the appropriate subject (it's not called '
nearest subject selection', is it?). One camera did it better than the other, we reported that - just as we pointed out where the Canon beat the Nikon.
And yet it's DPR that is biased, incompetent, arrogant... for those asking why I weighed in to begin with, it was to dismiss factually wrong accusations, like 'Rishi now uses AI Servo because of CR's influence' or 'DPR keeps testing iTR in auto mode against Nikon in 3D tracking, therefore all their results are flawed.' (We don't: we test Nikon's Auto against Canon's 61pt Auto w/ auto initial pt selection, and Nikon 3D vs Canon's 61pt Auto with manual initial pt selection). Certain people here keep coming up with false reasons to discredit us so as to forward the thesis that a bunch of camera reviewers - whose daily job is to understand all camera systems inside out - don't actually know what they're doing. Why do they do this? To not have to accept some criticism we leveled at their camera. Rather than check if the experience of a group of expert reviewers who handle these systems every day might have some perspective a more limited user of one system might not have, the gut reaction is to simply discredit us so as to not have to accept the result. Unknowing readers then read those posts and conclude, well they must be right and DPR must be wrong because the counterargument sounds reasonable and comes from an
actual user of the system.
That's when it behooves me to step in and make clear that: (1) no, the counterargument makes an assumption that we didn't consider your point already, which we probably (in most cases) did; and (2) we are users of said system as well. And (3), when we're wrong, and a user of a system does provide us with information we hadn't considered, we're all ears, we update our story, and admit it for full transparency (
http://bit.ly/2bWKSje).