The 1D3 is a new high technology product, leading edge stuff, that have gone through serious beta testing and all the rest.
However, having a number of beta testers, say hundreds, (which is how the 1D3 got to firmware version 1.0.8) is not as comprehensive a sampling as the delivery of 20-30,000 units. Some of the complaints are merited, some of the complaints are not; some have not read the operating instructions, some of the remarks are tendered from folks that do not even have the camera.
The factory has to sift through all this input, filter the facts from the fiction, and determine an appropriate hardware and/or software (firmware) solution, then test it, etc.
Canon's next release of the firmware (could be called 1.0.9 or 1.0.10 or whatever depending on the layers of "fixes") is in the hands of their service centers worldwide, so they can supervise the implementation of the new-firmware, and bench-test the results with their engineers and technicians in a controlled environment... after a few weeks of this, and if it is determined that x-percent of the issues are in fact resolved with this "solution," then Canon will decide to move to a broad-based release of the firmware update, by issuing an announcement and placing the files on their Servers for download; If the new firmware does not correct what they expected it to correct, they'll be some more delays; but all for the good of the end result... "simple" as that.. actually not so simple at all... lots of late night hours for the Canon folks, to get this right. Canon does not want too many standing on the sidelines for purchasing the 1D3; Canon wants lots of sales and happy customers. Period.
These guys, the folks at Canon, are of the best developers on the planet. Delivering a very deliberate approach, not reactionary, with no arrogance and certainly not incompetence. Be patient, this is a complex device; they'll get it right.
Hrow wrote:
Just a got a 180mm macro yesterday and was out playing this morning. It was really interesting to see just how small the AF area is - less than the size of AF point rectangles. Using the center point, the AF went shooting off into na na land while 1/3 of the rectangle was still on the subject (the edge of a flower).
Hmmm, in the past, the actual AF area was larger than the squares on the focusing screen.
Some of what is here is venting true, but for those that use the camera a lot, on deadline and with committments to clients, were hoping for more. I would hope that most tech companies would not normally expect to sell thier high-end equipment with the idea in mind to beta test the first users...this is a bad idea with the pro market. This group of people needs equipment they can rely on from the beginning. This is not a consumer or pro-sumer market camera...and yes, Canon is no doubt working on it hard. They spent a lot to upseat Nikon from the high end market and no doubt are worried about this issue. But that is hardly consoling to those that have units with this problem and had work affected by it. Having your keeper rate go down significantly at a sporting event is just one such example.
DavidP wrote:
Hmmm, in the past, the actual AF area was larger than the squares on the focusing screen.
That's what I always thought but I watched it do it over and over again. There was still about a 1/3 of the rectangle on the main subject and it would then try to focus on the background or an object behind it. 100% consistent.
Hrow wrote:
[There was still about a 1/3 of the rectangle on the main subject and it would then try to focus on the background or an object behind it. 100% consistent.
That is an interesting observation! Did you try testing for the same result on a variety of objects?
Hopefully, other folks can either confirm or refute your findings. If nobody else can duplicate what you have found then perhaps you have a faulty body. Just out of curiosity has our camera been problematic from the standpoint of AF (sorry, but I haven't taken the time to check all the posts)?
jkurkjia wrote:
That is an interesting observation! Did you try testing for the same result on a variety of objects?
Hopefully, other folks can either confirm or refute your findings. If nobody else can duplicate what you have found then perhaps you have a faulty body. Just out of curiosity has our camera been problematic from the standpoint of AF (sorry, but I haven't taken the time to check all the posts)?
Regards,
Joe Kurkjian
Erratic is the best word for what I have experienced. Most of the time it has produced very well but when it gets quirky it gets really quirky. I didn't really do a lot of testing because the goal was to play with the 180 and I didn't have a lot of time. It certainly is possible that something is wrong with my body and I have thought of a couple of possiblities. However, if that is the case at least it will be very easy for the techs to replicate.
Thanks for keeping us updated with your findings. In another thread you wrote "No hardware was replaced but it was checked to spec after the reload." Do you know what this means exactly? Thanks.
In my previous life I was a EE. Yes, I know exactly what that means.
In ME (Mechanical Engineering terms) this means the EE screwed up and they had to replace parts but will never admit it. I suggest you start reading Dilbert for this type of humour!
ben_is_in wrote:
Thanks for keeping us updated with your findings. In another thread you wrote "No hardware was replaced but it was checked to spec after the reload." Do you know what this means exactly? Thanks.
Lastly I have done a lot of software updates in my work (Medical equipement) and this software updates also involved uProgram update via JTAG for specific uControllers..
One idea about it could be (my speculation) that if the AF system of the 1 series bodies are achieved with a separate RISC uComputer (as far as I know, if I am not wrong), it could be necessary also to update their/s EEPROM uCode memories via JTAG device. Such device is in the hands of the Tech Service (like myself). For that reason, the customer can not make the update by itself.
This could be explaining better also the possible solution to the AF problem for me ...
The update comprise two parts: uCode at AF system + new firmware v1.0.9 to fix some bugs + control.
This theory will be only true if the update has to be done at Canon Service ONLY, of course ...
Jeff wrote:
That's interesting, as mine has rarely focused on the background when I get the active AF point off the main subject (in AI Servo). It usually ends up front-focusing in this situation, though it makes no sense to me why this would be.
RG's last additions to his continuing article re: 1.09 describes exactly what I saw. I think the slow focusing of the 180mm and the higher magnification just made the problem more obvious.
Juan55 wrote:
Lastly I have done a lot of software updates in my work (Medical equipement) and this software updates also involved uProgram update via JTAG for specific uControllers..
One idea about it could be (my speculation) that if the AF system of the 1 series bodies are achieved with a separate RISC uComputer (as far as I know, if I am not wrong), it could be necessary also to update their/s EEPROM uCode memories via JTAG device. Such device is in the hands of the Tech Service (like myself). For that reason, the customer can not make the update by itself.
This could be explaining better also the possible solution to the AF problem for me ...
The update comprise two parts: uCode at AF system + new firmware v1.0.9 to fix some bugs + control.
This theory will be only true if the update has to be done at Canon Service ONLY, of course ... ...Show more →
If Juan is right (and he probably is, at one point in my life I did write SW both soft and firm ) all the cameras will need to go back for fixes which will not be easy or quick to implement. On a even more depressing note this will delay the 1ds MK III (since they probably share the same AF system) and more importantly the 5D mk II ( I am interested in the 5D mk II since that's what the budget has allocated). Does anybody have any idea how many mkIII's have been made to date world wide?
What a mess, what is interesting is only half the people are having trouble (or so the poll says) which if it's a ucode problem doesn't make any sense at all. This is worse than the Leica M8 magenta/black issue.
The Altavista translation:
Out well dealer circles in the south Deutschalnd informed now the message:
Canon found the error in the autofocus system of the MK III and calls all cameras back in the Service.In approx., to 4 weeks then the new revised supply at the German market will be probably!
If i had seen this rumor this morning (my company firewall blocks news forums unfortunately), i would have declined buying my MIII (2 dealers called me today & Adorama said yesterday that they had more bodies coming in than people on their waiting list where i was #30ish).
My MIII will be delivered by UPS by noon tomorrow & will have a 5209xx serial number. I had assumed that this latest shipment would have the 1.09 firmware (but that remains to be verified) since the supply seemed to have shrunken in the past few weeks only to be replenished this week.
Hopefully, it will have the latest hardware, tho other conflicting rumors say the service action is only firmware while other rumors say some part was replaced. This German rumor & follow-on posts suggest that a memory chip may need to be replaced - possibly via a daughter-board swap.
Edited by biodan on Jul 21, 2007 at 03:07 PM GMT (Reason: Updated serial number)
) all the cameras will need to go back for fixes which will not be easy or quick to implement. On a even more depressing note this will delay the 1ds MK III (since they probably share the same AF system) and more importantly the 5D mk II ( I am interested in the 5D mk II since that's what the budget has allocated). Does anybody have any idea how many mkIII's have been made to date world wide? 
This is worse than the Leica M8 magenta/black issue.