RobGalbraith.com
Go to advertiser website.
     Home
     RSS
     CF/SD/XQD
     About
     Contact
Go
Go to advertiser website.
 
An analysis of EOS-1D Mark III autofocus performance - Continued
October 18, 2007: An EOS-1D Mark III autofocus fix is on the way
 
Beginning October 17 in the U.S. and presumably other regions as well, Canon began informing EOS-1D Mark III owners that a fix for the camera's AI Servo autofocus problems has been developed. Chuck Westfall, Canon USA's Director of Media and Customer Relationship, says that Canon USA staff at their factory service centres, call centre as well as pro market field representatives (including those at the PhotoPlus Expo trade show in New York City starting today) have been briefed on the fix and are able to speak about it to EOS-1D Mark III shooters.
 
We spoke with Westfall this morning about what customers are being told regarding the fix. In FAQ form, here's what we learned.
 
Q. What is the fix that Canon has developed?
 
A hardware fix. Specifically, a component in the camera called the sub-mirror must be adjusted. The sub-mirror - also called the secondary mirror - is a tiny mirror behind the main mirror that passes light down to the autofocus module in the base of the mirror box. The adjustment procedure may include the swapping of one or more parts that make up the sub-mirror mechanism, though the sub-mirror component itself will not be replaced.
 
Canon's EOS-1D Mark III White Paper describes the light path from the lens to the autofocus module this way:
 
As with previous EOS cameras, the TTL-SIR (Through-the-Lens Secondary Image Registration) focusing principle is employed. The light flux passes through the camera lens and passes through the half mirror at the center of the main mirror. The light flux proceeds downward by reflecting off the oblong, concave secondary mirror to a fully reflective mirror in the base of the camera’s mirror chamber. Then it passes through an infrared-absorption glass, a secondary image-forming lens, and a cover glass before reaching the AF sensor.
 
And includes these graphics:
 
eos-1d_mark_iii_light_path.jpg
 
Q. What EOS-1D Mark III autofocus problems does the fix address?
 
It's meant to solve poor AI Servo focus prediction when tracking a moving subject, particularly in warm-to-hot shooting conditions and particularly when shooting bursts of pictures. Westfall says that the information being distributed internally makes no mention of bright sunlight as being a factor (though obviously it would contribute to the heating up of the camera), nor does it provide a lot of detail about how much better the autofocus system will perform once the sub-mirror tweak is made, other than it will enable the camera's autofocus system to perform at the level it was designed to.
 
Q. Are all EOS-1D Mark III cameras affected?
 
No, says Westfall. Canon is planning to provide a simple way for EOS-1D Mark III shooters to identify if their camera was manufactured in production runs that included a sub-mirror assembly that is either adjusted differently or contains one or more different parts than a fixed sub-mirror assembly. Canon is not suspending shipments of the camera; reading between the lines, this is probably because recently-manufactured bodies already contain the sub-mirror fix.
 
Westfall stresses that even if an EOS-1D Mark III falls within the affected production range, it won't necessarily be exhibiting autofocus problems, but Canon doesn't have a way of determining which ones will and which ones won't. As a result, Canon USA will perform the repair procedure on any cameras in the affected production range that EOS-1D Mark III owners choose to send in. This will probably be true of Canon service centres in other parts of the world too.
 
Given that we've shot with nine different EOS-1D Mark III bodies, sporting a wide range of serial numbers, and all nine have exhibited the same sort of poor AI Servo autofocus performance described in this article, we're inclined to think that the pool of affected cameras is going to be darn near all of them.
 
Reading between the lines again, we think that Canon is trying to communicate to owners of the camera that if they're happy with its AI Servo autofocus performance, they're under no obligation to ship it to Canon for the sub-mirror fix. In other words, this is not a product recall. At the same time, they'll perform the procedure for any customer whose EOS-1D Mark III falls within the affected production period and wants the procedure done.
 
Q. When can I send my EOS-1D Mark III to Canon to have the sub-mirror adjustment done?
 
Canon USA is planning to announce the details of this at the end of the month. The information they intend to release at that time will include how to identify an affected EOS-1D Mark III body, their schedule for accepting cameras for sub-mirror adjustment at factory service centres and shipping details. Other Canon sales regions worldwide are probably gearing up to do the same, and their announcements are likely to be towards the end of October also.
 
Q. Have you tested an EOS-1D Mark III that has had the sub-mirror adjustment done?
 
No. As soon as we're able to, we will. An update of this article will follow after that.
 
Q. Has Canon delivered their promised response to your report of the EOS-1D Mark III's autofocus problems?
 
No. And while the announcement of this fix renders a direct response a bit redundant, we're still expecting Canon to live up to the spirit of their promise by explicitly acknowledging that the autofocus problems we described to them were real.
 
Q. Is Canon still intending to release new firmware for the EOS-1D Mark III that includes changes to the camera's AI Servo autofocus performance?
 
At this time, Canon isn't saying what their firmware plans are for the EOS-1D Mark III, including whether the sub-mirror adjustment will coincide with new firmware.
 
It's a safe bet that the v1.1.1 beta firmware, which first emerged at the time of the IAAF World Championships in Athletics in Osaka in late August, won't be released publicly. That is, Canon may issue EOS-1D Mark III firmware called v1.1.1 at some point, but enough time has passed now that it's unlikely that it would be the same as the Osaka beta.
 
Q. Are the sub-mirrors in any other Canon cameras affected?
 
No, says Westfall. Canon has determined that the sub-mirror problem is unique to the EOS-1D Mark III and that no other shipping Canon digital SLR models are similarly affected. Westfall also notes that the EOS-1Ds Mark III will already contain a properly-functioning sub-mirror when it ships and that its release schedule hasn't changed: it's on target for mid-November delivery.
 
eos-1d_mk_iii_volleyball.jpg
Sharper Image: Canon EOS-1D Mark III production model (firmware v1.1.1 beta) + EF 300mm f/2.8L IS, ISO 200, 1/320, f/7.1, Zeus strobes (Photo by Rob Galbraith/Little Guy Media)
 
Revision History
October 18, 2007: Changed various references to the EOS-1D Mark III's sub-mirror to reflect the fact that the hardware fix is being described as an adjustment by Canon, but one that may include the replacement of one or more parts of the sub-mirror mechanism.
October 18, 2007: Updated the firmware answer to indicate that Canon isn't revealing their firmware plans at this time.
October 19, 2007: Clarified the section describing which cameras are affected.
Next Page: October 29, 2007 update
Go to Page: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 
Go to advertiser website.
©2000-2013 Little Guy Media. Not to be reproduced without written permission.