Canon 1D Mark III Review – AF Introduction

Introduction
As this has been the area that the camera has received the most scrutiny (and also the most misconceptions) and will almost certainly be the lengthiest part of the review. In order to properly review the 1D Mark III, we are going to have to start and focus on this issue significantly.

Rob Galbraith can take credit as being the first person to really publicise, identify and analyze the issues with the 1D’s AF system. He took on an area which is typically measured loosely and qualitatively (with fairly vague opinions) and as much as possible, provided real hard evidence (and full resolution samples) of where the camera had been struggling, and under what conditions. He brought attention and rigour where none had existed previously and deserves great applause for his work in this area. My own review looks to reiterate some of this in the subjects that I’m more familiar with and answer a couple of additional things :

• Whether the issues remain after a camera has been “fixed” by Canon (Rob tested cameras which were fixed coming off the production line rather than ones that had been sent into service centers)
• Whether the AF issues significantly affect the subjects I typically shoot
• What other options are available to photographers to get around issues if they do indeed exist

If you didn’t read the introduction, I suggest you go back and quickly cover it as it outlines what the reported problems were and will set the scene for the following discussion. I will repeat here my setup though, that is the camera which I own and will be reviewing was a body that:

• Fell within the affected serial number range
• Exhibited the AF issues described above
• Has been sent in for the submirror fix
• Is now running on the latest Canon firmware 1.1.3

Another thing I should note is that I have had to microadjust ALL my lenses on my 1D. Typically the adjustment has been to set positive values, varying from +12 on my 24 L, down to +6 on 200mm and 300mm. The microadjustment was absolutely critical in getting good focus accuracy and this was on lenses that did not exhibit front or backfocusing when mounted on a 5D. It was only one the 1D that this issue occurred. Note also that the microadjustment settings I required were the same both before and after the submirror fix. I have not spent a great deal of time trying to understand (or speak to Canon about) what exactly causes this as I’ve been able to completely eliminate any back or front focuses issues through microadjustment. In any case, for this reason and also due to the myriad of custom settings surrounding AF, the 1D has been one of the most complex cameras I’ve had to come to grips with.

Firmware Issues
I’ll get right to the point on the firmware issue – 1.1.3 gets my strongest recommendation. All previous versions of the firmware (including 1.1.0) exhibited major problems on my 1D. Most significantly (and repeatable) was the fact that under all circumstances, versions below 1.1.3 could not effectively hold a focus point on stationary objects when in AI Servo mode. This was a deal breaker for me and led me to immediately upgrade to the latest 1.1.3. All testing below was conducted on this version and I sympathise with anyone who has persisted with 1.1.0 or was stuck with it for any significant period of time (unless they have a VERY different body from mine). In fact, this issue alone has been the single most surprising thing for me in this whole Canon episode – that Canon ever released the 1D on this firmware when the inconsistent focus on stationary objects (when in AI Servo mode) is so easy to replicate.