DPReview.com is closing April 10th - Find out more

Can Canon AF do real time tracking of objects (not just faces,eyes,cars,animals etc)?

Started 1 month ago | Discussions thread
ThrillaMozilla Veteran Member • Posts: 7,665
Re: Can Canon AF do real time tracking of objects (not just faces,eyes,cars,animals etc)?
4

Laqup wrote:

ThrillaMozilla wrote:

I would be inclined not to trust any review that shows a camera not being able to do something, unless I knew how the camera was set up -- regardless of how "well-known and fair" the reviewer is. If they don't test your case with known settings, they might be right, or they might be wrong.

Please, no ad-hominem arguments.

You are probably aware that this wasn't the point.

I think it is Alastair Norcross' point, and I believe it's a valid one.

What has been demonstrated by dpreview is still better than someone just posting on a forum saying "it works". Don't you think? That was the point.

I have noticed that DPReview seldom tells what settings they use, and seldom document any effort to set options for the tested case. So no, I don't agree. It's very hard to know without adequate information.

And btw: If a camera is that much rocket science that it works with a relatively "generic set of parameters" only in very limited number of use cases and needs constant (major) fine tuning, then it isn't state of the art either.

They are all optimized for a particular use case, which may or may not match the test. For that reason, because they usually give incomplete information, I usually find a favorable test more persuasive than a negative one.

And yes, the R5 has quite a few options, but some of the issues I demonstrated in my video in this thread (see one of the later posts) are completely decoupled from the AF settings.

I assume you mean the Z6-R5 comparison. I'll tell you what I know about it.

I just tried essentially the same test with my M6 Mark II, and it handled the test about as well as the Z6. I even moved the subject out of the frame as you did, and it passed that test too. I suspect that the R5 works at least that well, or better.

The M6 II has an undocumented(?) feature whereby one can specify the focusing spot on the touch screen and track that spot. It works quite well. I assume the R5 probably does have a similar feature.

The M6 II has one quirk, though. With Servo on, if you continuously activate focusing, it may switch to a wide-area or multiple-spot focusing mode, and under some conditions it could lose the subject. Your R5 did that. But the M6II is plenty good for focus and recompose, and it's not bad for action and BIF.

I don't have enough info to judge the R5, but in any case, the R cameras have an option for sticky tracking that the M cameras lack. By most accounts the R5 works very well for action and BIF. Furthermore, the R3 and R7, with their newer focusing engines, specifically allow tracking with any available spot size.  So I expect that they are quite capable of working just like the Nikon does.

Example: The complete failure to pick up certain elements in the scene as the desired AF reference (e.g. the wooden sticks in my vid). There are no accelerations or any sort of camera movement involved, you just can't track some elements properly or only with a low hit rate. Another example would be the ability to pick up the initial subject after it has very briefly left the field of view. In this regard there seem to be some changes in the later models, would be interestint to see if they help for non-detectable objects too (as far as I know not (yet?)).

 ThrillaMozilla's gear list:ThrillaMozilla's gear list
Canon EOS Rebel SL1 Canon EOS M6 II Canon EF 400mm f/5.6L USM
Post (hide subjects) Posted by
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark MMy threads
Color scheme? Blue / Yellow