Canon 1D MKIII Autofocus- Interactive Review

Canon 1D MKIII Autofocus Problems- Interactive Review

Click here to read the entire article

It wasn't the photographers who said the camera could do it, it was Canon! No we are not asking too much; This is a top of the line camera, for a top of the line price, and it should do what they said it would do. Shame on Canon for not testing all situations and program Auto Focus correctly. YMMV

2 more entries added.

Thanks for the compliments.

Dierk, I do think there are some real issues, at least with some bodies. However, I do think some people are now looking for this camera to do a lot more than the the MKII N ever did.

Steve, I have noticed a few strange things in regards to dof at times as well. I haven't done an in-depth analysis. But, I think some of it may have to do with an older 300mm lens that was used in some of the trials. Achieves focus, but other things in the same plane are oof. Curious.

Since the Mk II did/does better in some circumstances, there has to be a real issue. I do not deny that. Considering the limiting factors one has to choose for this to appear, and the fact that your findings after the frmware upgrade are much better than Rob's, combined with your idea why the findings are significantly different led me to my comment.

For one, the camera is set to find the best possible AF lock point, which, by definition, is the one with highest contrast within a certain range of AF points chosen by the user. It is very similar to the Group feature of Nikon's cameras, I guess. Over the years AF systems have become very sophisticated with Canon leading the way to take into account the complete image area; Nikon has always been more conservative, Nikon photographers even more so*. The concept is to give a higher percentage of acceptable images. This does not mean that in a very high speed shooting sequence every single photo will be in focus, much less where we want it.

Essentially I am not saying more than you, Drew, when you compare your tests with what Rob found after firmware update: It may just be different expectations. One of the biggest drawback of virtually every test in the photographic world is that the starting point, the expectation or, as Sir Karl Popper called it, the conjecture isn't laid out clearly.

In this case it should be relatively easy: What exactly does Canon claim for the Mk III on this particular subject? We have to take their claim, probably refine it, re-phrase it in terms making tests possible, then see if the claim can be verified or falsified. Since not all cameras seem to be affected, verifying will not get us far, so we have to phrase the conjecture in a way it is easily falsified by tests.

While your actual test is thorough, I miss the expectation. Are you just interested in the number of in-focus shots between Mk II and Mk III? Add to this confusion on my part my ignorance of the intricacies of Canon's cameras. It's hard enough for me to memorise what they are called [actually futile], much less how they differentiate. For instance, are the AF and "winder" hardware of the Mk II and Mk II the same? That is, are the differences in AF due only to software changes? If both cameras are capable to shoot the same number of photos in a second, how do they achieve that? Could it be that other algorithm and hardware changes [shutter etc.] contribute to the problem?

Just confusions arising in my overactive brain ...



*I use Group with AF-C and find it very reliable and fast, many other photographers stay with Single AF point, often centre.

I have not run into either (gross over/under) yet.

Drew,

I'm sure I'm not alone in saying that your posts are greatly appreciated and very informative. The "scientific" approach has convinced me. My Mark III is ordered and on it's way. (Soon I hope)

It's apparent to me that what some may see as a total "issue" may not affect me. Only until I get behind the lens and start burning up my own memory cards, will I truely understand the limitations of not only my camera but my own ability to take good shots.

Hopefully, I can soon contribute some substance of my own to your reviews.

Again, Thank You!

Hi all,
received my 1D Mk111 two days ago, shot my first wedding yesterday with the camera, just edited aprox 1000 images from the wedding and see no autofocus problems with the camera, if anything the camera locks focus very quickly and from there holds focus superbly, I have not done any extensive test of the focus as done by Drew but have found the focus in the field to be superb.Also the image quality is great and at 800 iso is very very usable and 1600 iso is very impressive, very little noise in the shadow areas compared to any camera i have used before which includes the 2 mega pixel Canon D2000, 1D, 1D Mk 11, as you can see i have had them all so I am well used to digital cameras.
Hope this helps with concerns out there about this camera, i am not disputing that there may be issues with some cameras that have been released but the camera i received two days ago is a superb camera and leaves me with one question, should i buy a second 1D Mk111.
regards and happy shooting
Also before anybody asks the camera is loaded with firmware 1.0.8
Dylan
ps many thanks to Drew for all the testing he did on this camera.

As I understand the problem, it is with AI on moving objects where there is a high contrast area that is not the photographers focus point? That would tend to make me think that a wedding shoot would not be expected to see the problem (not that they have a camera that does not have a problem even with a lower firmware than the one that Canon has indicated as a probable fix for the problem). YMMV