1DX Mark III

[Jonathan Huyer]

The new family member arrived yesterday, and I was able to get out today on an owl shoot to run it through some paces. This Great Grey was a fair ways away (I should have used an extender on the 500 f/4), but it served as a useful test subject nonetheless. I had all the focus points activated for auto tracking, and when the owl took off to catch a mouse I was quite impressed at how the camera could follow it in spite of the cluttered background. I'm sure my "old" Mark II would have lost tracking against those trees -- it can never seem to lock on in those situations. The image quality is decent for sure, although I can't say how much better it might be than the Mark II.


1/1600 sec @ f/4, ISO 500




1/1600 @ f/5.6, ISO 1000
Flying the wrong way of course, but useful as a test nonetheless




1/1600 @ f/5.6, ISO 1250




The new focus system is quite impressive. The focus points are smaller, so there doesn't appear to be a need to ever have to use the micro-focus mode (where just a tiny dot is activated for precision focusing). The AF-ON trackpad is rather slick. You first press the focus selection button, then slide your thumb over the AF-ON button to move the point around. Pressing the AF-ON button in then locks the selection in. It's certainly better than using the toggle selector, but it will take some getting used to since I have to re-train to this new procedure.

The speed and buffer size are simply ridiculous. The Mark II was already excellent in that regard, but this is just beyond nuts. I could hammer the shutter down for as long as I wanted, and the buffer counter in the viewfinder wouldn't even budge from 99. Luckily I got a 512 GB CF Express card as part of the early bird deal, because I think I'm going to need lots of storage .

[Joel Eade]

That's some good news to hear about AF tracking ..... I have been tempted by the Sony performance for birds in flight but hate to invest in a new system. Would love to hear more when you get more experience. Thanks!

[DavidEccleston]

I'm sure my "old" Mark II would have lost tracking against those trees -- it can never seem to lock on in those situations.

I thought that was what the IR in the RGB+IR metering sensor was about. It will track warm bodies, so if it is temporarily obscured, or against a cluttered background, it will still stand out and be trackable.

Then again, I could just be making things up, as I don't think Canon has ever said how they use the IR data. I seem to recall someone shooting baseball, and being impressed with the tracking still working when the players moved behind some netting, which just reenforced my idea on how the IR is being used.

Assuming my IR idea is right, your owl might be too well insulated and not show up strong enough for older bodies. (I'm definitely into the realm of just making stuff up now!)


edit: http://infrared-birding.blogspot.com...d-birding.html

[Jonathan Huyer]

I thought that was what the IR in the RGB+IR metering sensor was about. It will track warm bodies, so if it is temporarily obscured, or against a cluttered background, it will still stand out and be trackable.

Does the metering sensor work in conjunction with the AF sensor? I mean, in terms of helping it obtain focus? I'm rather vague on how the whole process works so I'm keen to learn more.

By the way, here is an uncropped version of the owl in flight, so you can see how small an object the focus was tracking:

[Jayson]

I was hoping you would get out and give the toy a spin. Excited to see what more you have to do with this new addition!

[DavidEccleston]

Does the metering sensor work in conjunction with the AF sensor? I mean, in terms of helping it obtain focus?

No idea. You'd think they would find a way to leverage any data available, but they don't really discuss how their AF tracking systems work. Even if IR data is not be dense enough to help with focus directly, but it could be used as a guide to help determine which focus points/zones to track with.

Like I said, I'm in "making stuff up" territory. It seems plausible, but I have no idea.

[Joel Eade]

These are brief videos about the AF system .... no mention of IR sensing so not sure about that


https://www.youtube.com/watch?v=eExbVhK26bA


https://www.youtube.com/watch?v=DM6KkQcNI8w

[Jonathan Huyer]

I tested the auto-tracking focusing today, at a local ski race.

Settings:
AI Servo AF Case 'A' - tracking automatically adapts to subject movement
AF priority (people) - enable
Subject switching - enable (slow)
1st image priority - focus
Initial servo AF point - Auto
Focus mode - all points active

Lens used: 100-400, with wide open aperture except where noted

Some examples --- note all photos are uncropped so you can see the relative size of the subject's head in the frame:




Focus continued to track on the back of the head:




Focus worked well when subject was darker than background:




Here's a miss. Failed in strong backlighting.




Another miss - a rare occurrence. Probably operator error!




This was more typical:




Panning shot just for fun --- not hard to get focus since it's at f/18

[Jonathan Huyer]

I've been reading the White Paper and found this tidbit, which explains how the metering sensor helps out with auto focus:

The metering sensor teams up with the new, square pixel AF sensor during viewfinder
shooting, assisting the AF system with its subject detection capabilities. The metering
sensor, with its color recognition capabilities and 400,000 pixel resolution, can recognize an
initial subject, and then update the AF system as to its location during Zone AF, Large Zone
AF, and Auto AF Point Selection shooting through the viewfinder.

[DavidEccleston]

Cool. So they are using the color data as I imagined they would use the IR data, to help decide which focus points to use. Unless the IR data is too low res, noisy, or otherwise unsuitable, I'd imagine they're using it too.