mill4570 wrote:
Today Canon seems to have stumble into the "Kubiashi Maru" scenario. People are mad at them for announcing another fix, but were mad prior to that for not fixing the camera. I believe Canon will fix it, maybe not to everyones liking, but they will sort it out.
I'm definitely in the "fix it again" camp, if that's what it takse to get it right.
DavidP wrote:
I'm definitely in the "fix it again" camp, if that's what it takse to get it right.
In the long run, Canon will get more cheers than jeers even if it takes two more fixes. My bottom line is fix it no matter how long it takes. Make this the camera everybody wanted after reading the white paper.
All of you calling for Canon to "fix it" are ignoring the possibility that there is a fundamental design flaw and that it may remain unfixed. Canon really dosen't need to discontinue this body, just introduce another with a new designation and without the problem. Early adoption has its risks and sometimes rewards. From now on I'll take my chances with new, cutting edge can openers and wait a bit on high priced items.
Red
Red Grainger wrote:
All of you calling for Canon to "fix it" are ignoring the possibility that there is a fundamental design flaw and that it may remain unfixed. Canon really dosen't need to discontinue this body, just introduce another with a new designation and without the problem. Early adoption has its risks and sometimes rewards. From now on I'll take my chances with new, cutting edge can openers and wait a bit on high priced items.
Red
Red,
I'm not ignoring that at all. If there is a major design flaw, why can't that be fixed? You don't have to put the entire body in the trash compactor. It doesn't matter to me if the new fix is to remove the entire AF system and replace it with the AF system of a new MKIV. I just believe Canon can not afford to give up on this camera unless they are willing to take them back as trade on the MKIV.
Red Grainger wrote:
All of you calling for Canon to "fix it" are ignoring the possibility that there is a fundamental design flaw and that it may remain unfixed.
Certainly for electromechnical items like this, there is a distinct possibility that a fix cannot be provided in the field. It is also possible Canon has discovered this and is taking them this long to come up with other work-arounds.
Early adoption has its risks and sometimes rewards. From now on I'll take my chances with new, cutting edge can openers and wait a bit on high priced items.
Red
Sorry but what "early adopters?" This is a third iteration of a camera with incremental changes in each version. We are not talking about the first digital camera they ever made or anything.
My sense is that Canon revises its auto-focus so infrequently that they lack adequate test coverage for the feature. With such a dynamic system, it is hard to have repeatable tests. But hard should not equate impossible. When I first read about RG's focus issues, the engineer in me sure became worried. AF is not something that would go from bad to good from early production to real products.
amirm wrote:
Sorry but what "early adopters?" This is a third iteration of a camera with incremental changes in each version. We are not talking about the first digital camera they ever made or anything.
My sense is that Canon revises its auto-focus so infrequently that they lack adequate test coverage for the feature. With such a dynamic system, it is hard to have repeatable tests. But hard should not equate impossible. When I first read about RG's focus issues, the engineer in me sure became worried. AF is not something that would go from bad to good from early production to real products. ...Show more →
The parts of your quote that I italicized and underlined are a bit at odds with each other. The 1D-3 isn't something I'd consider an "incremental change" from the 1D-2N, precisely BECAUSE of the new AF system.
Still, you'd think that Canon would've figured out the issue by now, seeing that they knew about it (or at least something about the AF was wrong that they knew about) over a year ago.
I have become convinced that Canon has a problem under certain circumstances and/or with certain bodies that they simply can't fix practically. They may be able to minimize it so that it isn't as noticeable and they are hoping that that buys them some time and that they can weather the storm. So far, they are right.
They could, and almost certainly will (IMO), introduce a replacement body for Photokina and claim that they are maintaining their 18 month upgrade cycle and it would sound plausible. Do it at the same time as the 5DII to let that body take most of the press and they come out pretty clean. The people who will get screwed are the current 1DMkIII owners but many have so much invested in glass that it is cheaper to take the hit on replacing the body than it is switching. Ultimately, they don't care if we like them, only that we buy their products.
amirm wrote:
Certainly for electromechnical items like this, there is a distinct possibility that a fix cannot be provided in the field. It is also possible Canon has discovered this and is taking them this long to come up with other work-arounds.
Sorry but what "early adopters?" This is a third iteration of a camera with incremental changes in each version. We are not talking about the first digital camera they ever made or anything.
My sense is that Canon revises its auto-focus so infrequently that they lack adequate test coverage for the feature. With such a dynamic system, it is hard to have repeatable tests. But hard should not equate impossible. When I first read about RG's focus issues, the engineer in me sure became worried. AF is not something that would go from bad to good from early production to real products.
The passage below from the manual indicates, to me at least, that this is not just an incremental change. I'm not a software or engineering guy so I guess one could argue about this. But it seems to me that they've implemented a new imaging engine, two of them in fact, and along with various other changes that qualifies as more than incremental.
As far as early adopters, in marketing speak (yes, I'm in marketing ) these are the types of buyers who jump on any new technology product (perhaps only the first 10% of buyers.) So I'd put most of the of the folks who immediately ordered a MkIII (and have been in return/fix-it hell for almost a year) as early adopters. The problem that I think Canon has created with the MkIII is that technology companies need "early adopters" to get things rolling. Unfortunately, the lesson a number of people have drawn (see some earlier comments in the thread and other boards) is that Canon has not handled this very well and in the future they ought to wait for 6-12 months to make sure the product doesn't have issues. From Canon's perspective that's not too good because it affects the return on investment calculations product managers have to go through (i.e. justifying their product to higher ups) which will affect decisions on product development, investment, etc (i.e. in other words, slow adoption means it takes longer for new products to "break even.") It would have been a different situation if Canon had made its MKIII mistakes but made special efforts to reassure the early adopters that they weren't going to get strung out forever (for instance, it could have told people categorically/publicly that it would ensure they were satisfied, offered to fix OR replace their cameras, or even return their money, etc.) Now, I know some will say this is too expensive, etc., etc. Tough noogies. If you're a technology company you don't screw the early adopters; you make sure people know they're not going to be penalized for buying a new-generation product early.
From MkIII manual:
"The beautiful signal from the sensor is sent via a very fast 8-channel readout to two new DIGIC III imaging engines. In the past, one such device has always sufficed, but with the EOS-1D Mark III, there was to be no compromise of either speed or image quality. The Dual DIGIC III Image Processors perform parallel processing, common nowadays in serious computers but unique among cameras, to speed through information. Their workload is increased by the fact that the new sensor captures 14-bit depth color rather than 12....."
rscheffler wrote:
I think there could be a market for two different full frame cameras. I've seen some side by side tests comparing the D3 against the 1DsIII and you can see that there is a significant difference in image resolution (in favor of the Canon). But the D3 benefits at higher ISOs due to larger pixel sites. There are those who need 21 MP. Give me a 14 MP full frame 10 fps excellent high ISO camera and I would be happy to have only one camera model... but right now I'm contemplating the 1DsIII (also hoping for a D3x to knock it down $1-2K).
To me, you sorta backed my line of think up.. You're contemplating a $7-$8k 1DsMKIII. If there was a $4500 FF Mark III, would you be? Granted, The D3 does put a little pressure on Canon for a lower cost premium series FF camera... But for "most" photographers that I know that require high megapix and Full Frame (typically for studio work) a 5D actually functions fine (as does the 1DsMkII). Sure there is a segment that would prefer a 1D series build (count me as one)...I just don't see it happening anytime soon at that price... I do, however, see a beefed up 5D forthcoming... But not a 1D series in that range.
Without having read all the posts in this thread, I thought I'd chime in with my experience. Since receiving the camera in June, I've had major issues shooting outdoor sports with AI Servo in the sun, just like Rob Galbraith. However, indoors (low to poor light) my results were very good, better than the 1D I had shot with for two years.
Since receiving the submirror fix, though, the AF has fallen apart where it had previously excelled. When shooting basketball, I'm getting many frames where the focus is hitting in front of my subject by a foot or more. This never happened before the repair job. After initially tracking a subject, the first shot I fire off is almost always focused way out in front, and subsequent images in a burst will sometimes be right on the money, sometimes slightly off, and sometimes they're focused way too far in front.
I was ready to call the service center and send it off again when I heard about this latest supposed fix, so I guess I'll wait. Before the initial fix, I was always confident the camera would perform in relative low light. Now it's unreliable. I've not had a chance to shoot anything outdoors since the fix, so I have no idea what the results might be. I no longer enjoy photography because of the whole ordeal. I get paid to shoot sports, and I find myself worrying too much that I'm not going to get the shots I need. I'm still waiting for a happy ending.
Hrow wrote:
I have become convinced that Canon has a problem under certain circumstances and/or with certain bodies that they simply can't fix practically.
So, after reading Rob's latest update, what convinced you that this problem can't be fixed practically? Have you seen or heard of the results from this latest fix?
I'm assuming that you're referring to the latest fix being discussed in this thread, as Rob stated two months ago that there were still unresolved issues after the blue dot/sub-mirror fix.
ben_is_in wrote:
Without having read all the posts in this thread, I thought I'd chime in with my experience. Since receiving the camera in June, I've had major issues shooting outdoor sports with AI Servo in the sun, just like Rob Galbraith. However, indoors (low to poor light) my results were very good, better than the 1D I had shot with for two years.
Since receiving the submirror fix, though, the AF has fallen apart where it had previously excelled. When shooting basketball, I'm getting many frames where the focus is hitting in front of my subject by a foot or more. This never happened before the repair job. After initially tracking a subject, the first shot I fire off is almost always focused way out in front, and subsequent images in a burst will sometimes be right on the money, sometimes slightly off, and sometimes they're focused way too far in front.
I was ready to call the service center and send it off again when I heard about this latest supposed fix, so I guess I'll wait. Before the initial fix, I was always confident the camera would perform in relative low light. Now it's unreliable. I've not had a chance to shoot anything outdoors since the fix, so I have no idea what the results might be. I no longer enjoy photography because of the whole ordeal. I get paid to shoot sports, and I find myself worrying too much that I'm not going to get the shots I need. I'm still waiting for a happy ending.
Same problem here after fix. Take some shots and various micro adjustments and make note of them.. Then go back and see which are in focus. I did this recently, and with adjustments of +5 to +15 I am having much MUCH more consistent results. But like you, I got it back and everything front focused like crazy.
DavidP wrote:
The parts of your quote that I italicized and underlined are a bit at odds with each other. The 1D-3 isn't something I'd consider an "incremental change" from the 1D-2N, precisely BECAUSE of the new AF system.
Still, you'd think that Canon would've figured out the issue by now, seeing that they knew about it (or at least something about the AF was wrong that they knew about) over a year ago.
I sitll don't think that there should have been the issues if Canon had done a proper job testing the AF system. If you go back to the film days, the changes Canon made to the AF system from the 1N or 1RS were to the EOS 3 and 1V HS system were much more drastic then what they have made to the 1DIII.
Morover in the case of the 1DIII, RG had already pointed out there was issues with the AF with his test model. To me it is inexcusable that Canon already knowing there were issues would still release the camera.
On my side, i feel like a fool for making (what I thought was the logical) assumption that there is no way Canon would release a camera without making sure all issues with the AF, since they were already aware of issues with there prototype.
IMO whoever was responsible for approving the camera for release should be fired at Canon. It is inexcusable that a Pro body would have been released without proper QC and thorough testing.
I also think that the only way Canon can resolve this problem to the satisfaction of customers is to suck it up and issue a recall on the camera (the way they did with the 24-105L), at the very least they should offer a refund to those that are not satisfied with their cameras. It is inexcusable that a major issue like AF has not been resolved almost a year after the camera is released (considering a major focus of their marketing campain for the 1DIII was the AF operation touted as being the best of any camera to date).
They have been in spin mode from the start and need to show some integrity if they hope to keep customers like on board. It just seems that after they moved ahead of Nikon, they really let their QC and customer service droop (at least in Canada I saw a drastic change at the level and quality of service in Calgary).
To me the 1D3 AF system is an incremental change in the sense that it is still a 45-point system. It's not as if Canon has no past experience handling the focus info coming in simultaneously from 45 different cross-type and linear sensors - they've been doing it since the EOS 3 was released in the 1990s. It's not as radical a change as it was going from 7- or 9-point AF to 45-point back then.
Also, the change from Digic II to Digic III is largely irrelevant because the Digic computers do not handle the AF in the pro bodies - a separate computer is used for AF and has been at least since the 1V was introduced. That computer used to do exposure calcs as well until the 1D2 was introduced. Digic is used for in-camera image processing and managing histograms and so on.
Worse than any defect in the camera (which they were surely aware of when it was released) has been the blame-the-photographer attitude they took while they looked for the real problem. That is unforgivable. I don't expect many people to jump ship to Nikon as it makes little or no financial sense, but I do hope that more new high-end DSLR users start off with Nikon to make Canon appreciate that they handled things poorly. The trouble is that Nikon is no saint either, and probably even worse.
I'm also convinced that the 40D has a similar AF problem to the 1D3. Something that is not noticed by everyone but is there in too many cameras nonetheless. Mine was a dog and Canon could not fix it. Others have reported similar problems but the big majority of users would not realise whether or not they had a defective camera.
With my history of defective Canon cameras and the 1D3 fiasco I would highly recommend that nobody buy a new model for the first several months after it hits the streets. Let the early adopters and testers prove that it is worth having first. I'm amazed at how many people have ordered a 1Ds3 even though the 1D3 (with the same technology) is still not fixed.
With my history of defective Canon cameras and the 1D3 fiasco I would highly recommend that nobody buy a new model for the first several months after it hits the streets. Let the early adopters and testers prove that it is worth having first. I'm amazed at how many people have ordered a 1Ds3 even though the 1D3 (with the same technology) is still not fixed.
) these are the types of buyers who jump on any new technology product (perhaps only the first 10% of buyers.) So I'd put most of the of the folks who immediately ordered a MkIII (and have been in return/fix-it hell for almost a year) as early adopters. The problem that I think Canon has created with the MkIII is that technology companies need "early adopters" to get things rolling. Unfortunately, the lesson a number of people have drawn (see some earlier comments in the thread and other boards) is that Canon has not handled this very well and in the future they ought to wait for 6-12 months to make sure the product doesn't have issues. From Canon's perspective that's not too good because it affects the return on investment calculations product managers have to go through (i.e. justifying their product to higher ups) which will affect decisions on product development, investment, etc (i.e. in other words, slow adoption means it takes longer for new products to "break even.") It would have been a different situation if Canon had made its MKIII mistakes but made special efforts to reassure the early adopters that they weren't going to get strung out forever (for instance, it could have told people categorically/publicly that it would ensure they were satisfied, offered to fix OR replace their cameras, or even return their money, etc.) Now, I know some will say this is too expensive, etc., etc. Tough noogies. If you're a technology company you don't screw the early adopters; you make sure people know they're not going to be penalized for buying a new-generation product early.