Joe Decker wrote:
I'm not convinced that the variations by serial number indicated in the data so far represent anything but random chance, e.g., I think the data is consistent with the hypothesis that the probability that an owner believes his MkIII has AF issues is (so far) uncorrelated with serial number.
I'd tend to agree. I hope no readers are taking the stats into consideration for a purchasing decision.
Now this is just my opinion...
I know a good number of people are absolutely convinced their camera focuses correctly as intended, in all conditions. And that may very well be true. But I also think a good number of people do not quite understand the issues as reported.
Without going back through messages from various forums and pointing out specific people, I have seen a number of samples posted from owners saying their camera has no AF issue. But the samples were clearly not in the conditions that reveal the trouble so effectively as reported by RG and others. And some of the photos were taken with slower apertures, partially stopped down lenses. AF errors are usually harder to see then.
So I think the statistical numbers in our poll don't really give us a very clear picture of how many cameras are affected. Again, this is just my opinion, I think the trouble is more widespread than many people think.
But I think it's also worth noting if some owners never see the issue because of the conditions they usually shoot in, or the lenses they use, then it doesn't really matter much for them. I think they should continue to be happy with their cameras and don't bother testing in conditions they're never going to shoot in. Or rarely ever.
apdieb wrote:
Mine must have backfocused on this one..
J/King.. But seriously.. At least with this sequence of about 5 shots (this is a crop)..every one was spot on focus wise. I am planning to do a lot more testing though. So far, so good.
I have 1.1.0 and Af set to 1 stop slower than normal. Temp. 101 degrees.
What was your F stop on that shot? Trying with 2.8 I couldn't get it to focus past the water.
Couple of things though you have a strong background and the water I was shooting through was a bit smaller drops but more heavily concentrated. Also at the default AF settings.
I tend to agree also. We're still at a relatively small sample level, especially in some s/n categories. One more "yes" in the newest batch would change that one from 25% to 33%, for example. Given a broad "margin of error," the results are really pretty darn consistent across the board.
Iberian wrote:
What was your F stop on that shot? Trying with 2.8 I couldn't get it to focus past the water.
Couple of things though you have a strong background and the water I was shooting through was a bit smaller drops but more heavily concentrated. Also at the default AF settings.
Garylv - I agree entirely. Many of the people reporting that their MkIII performs faultlessly do not shoot with long, bright lenses wide open or otherwise have not subjected their MkIII to the kind of circumstances under which it has been shown to misbehave.
If one is a landscape photographer who never shoots with anything longer than 35mm or an aperture wider than f/8, the chances of noticing, much less being affected by, the AF problems are slim.
As a side note, many people have stated that the problems manifest themselves most prominently when shooting in bright sunlight and warm conditions. While I can verify that this does indeed cause the MkIIII to malfunction when it comes to AF consistency in my experience, the problem is not limited to those conditions.
I've had extensive problems with the MkIII mis-focusing in one-shot mode indoors toom, both with and without flash, using a variety of lenses from EF 24-70 f/2.8L to EF 85mm f/1.8, despite those lenses performing admirably on my MkIIN. I can also verify a DPR poster's report that his AF microadjustments were thrown completely off by attaching a flash and allowing the AF assist beam to fire...
All in all, the AF on the MkIII seems to have some serious problems judging from my experiences, and it's not just servo mode in bright sunlight that's affected.
I should also add that there seems to be a hard core of people (especially over at DPR) who not only claim that their MkIIIs perform perfectly, but also rubbish the reports of those who say that their own MkIIIs have a problem.
This seems nonsensical to me - if the people who have problems convince Canon to act then it will be to the benefit of all 1D MkIII (and future 1Ds MkIII) owners - Canon are hardly going to make the situation worse by coming up with a solution are they?
Nill Toulme wrote:
I tend to agree also. We're still at a relatively small sample level, especially in some s/n categories. One more "yes" in the newest batch would change that one from 25% to 33%, for example. Given a broad "margin of error," the results are really pretty darn consistent across the board.
Nill
~~
www.toulme.net
Which may indicate a design or software problem rather than sample variation. I will find it interesting and sad if there really is sample variation between cameras. My camera is from the first group, but I'm resisting formal testing at this point.
My prediction is that the AF is too responsive and gets fooled in some situations. Sort of a digital ADD thing.
Be it all that as it may, I think it would be better if we didn't start rehashing the "they're all broken/no they're not" arguments here. That's been flogged to death elsewhere. Jeff set the poll up to establish a factual database, and we should keep the thread on target to that purpose.
In passing I will note (perhaps again) that at this point we can reasonably assume that most of the people actually voting in this poll have a pretty good idea of what the issues are and how to look for them. Remember also that Jeff asked everyone to refrain from voting until they are reasonably sure one way or the other. That's why there's no "I'm not sure yet" category in this poll.
The only real possible disconnect I see at this point is for those who voted No with confidence, and then later have determined either that they maybe had problems all along or else that their cameras have seemingly started to develop problems that they didn't have at first. There are some of those people, and some of them have commented to that effect in the thread.
No poll is perfect, but I think this one is serving a useful purpose so far.
I agree that we should keep this thread on target as to get a good database regarding the Mk3 AF. As of today, the sample size of the database is still too small to be useful. In addition, I still supspect that some of the so called 'AF issues' or 'No AF issues" votes are more a reflection of the user experience/technique on a new and rather complicated camera. I will be very surprised that some users are not going to change their votes after they get more familiar of their new cameras and in some cases, have better technique/confidence in high speed Al Servo shooting with their new cameras.
Even after I have been using my 1DMk2 for 3 years and thousands of BIF shootings, the 1DMk3 makes me feel like a newbie all over again . I still have so much to learn and so far I would consider majority of the mistakes are due to user errors. I believe that the 'optimium' CFn settings would be different for different users, based on their shooting style, subject size/background, speed, lighting, lenses, etc. There are too many possible different settings with so little experience on this camera at the current stage. I am not saying anyone's recommenede settings are wrong, but it will be more than one set of 'Perfect' setting for everyone. That is also why it is so tough for the Canon engineers to know what is the problem, not even talking about a 'fix'.
That is why I refrain from voting as I really don't know the new camera at all. Once I figure out how to use this camera, then I will shoot my 1DMk2 with it side-by-side to do a meaningful comparison. After that, may be I am qualified to vote.
Has anyone sent their mkIII in and had it come back fixed and working to their satisfaction?
I know there are a few who got a new, later serial#'d body on exchange that worked well but I'm not sure I've seen anyone who has actually had theirs repaired and fixed.
Kier wrote:
Garylv - I agree entirely. Many of the people reporting that their MkIII performs faultlessly do not shoot with long, bright lenses wide open or otherwise have not subjected their MkIII to the kind of circumstances under which it has been shown to misbehave.
If one is a landscape photographer who never shoots with anything longer than 35mm or an aperture wider than f/8, the chances of noticing, much less being affected by, the AF problems are slim.
As a side note, many people have stated that the problems manifest themselves most prominently when shooting in bright sunlight and warm conditions. While I can verify that this does indeed cause the MkIIII to malfunction when it comes to AF consistency in my experience, the problem is not limited to those conditions.
I've had extensive problems with the MkIII mis-focusing in one-shot mode indoors toom, both with and without flash, using a variety of lenses from EF 24-70 f/2.8L to EF 85mm f/1.8, despite those lenses performing admirably on my MkIIN. I can also verify a DPR poster's report that his AF microadjustments were thrown completely off by attaching a flash and allowing the AF assist beam to fire...
All in all, the AF on the MkIII seems to have some serious problems judging from my experiences, and it's not just servo mode in bright sunlight that's affected....Show more →
Under the bright sunlight especially backlit situation I will have the most obvious AF issue couple with wide aperture and long lens. I got more oof when I am going for high speed (8-10 FPS) AI Servo tracking IMHO.
The FW 1.1.0 with expansion off and one notch down AF speed did tame down the unpredictable AF though.
Although I really like the pole... I am an advocate of posting examples. I am amazed to see so many comments, but really haven't seen all that many examples. I posted mine which led to discussion on how the scene was composed, contrast, etc.. and it's effect. I am just wondering if there could be some explanation to some of the OOF reports if we actually saw some examples from a wider base of users. RG and a handful of others have graciously posted their experiences, but I just haven't seen enough (GOOD and BAD) real world examples...just lots of comments.
I hope I am not off-base asking for this...And perhaps I may just be overlooking a thread somewhere that is full of examples from a multitude of users.
Kier wrote:
I should also add that there seems to be a hard core of people (especially over at DPR) who not only claim that their MkIIIs perform perfectly, but also rubbish the reports of those who say that their own MkIIIs have a problem.
There's many over they who say that their camera's don't suffer the same problems as RG's. It's in response to people who've never used it blabbing away bout the Mk III's problems. It seems like a number have taken RG's word and passed it down as etched in stone for all Mk III's which certainly isn't the case. Further out, one guy refused to check out his camera because the serial number was 514XXX. Apparently he thought it was a refurb in new packaging and feared it would behave like Rob's.
Personally, and I fear this may be taken the wrong way, but I think a number of the bad bodies out there might not have been deemed so if it weren't for Rob's review. I think the info is good, but it may have started a few hypochondriac photographers.
Thanks for the example. The second image looks like motion blur to me...not simply oof.. Interested to hear what others think. Definitely not a good result though.. I agree.
Yes, that particular example of the couple looks like motion blur to me. 1/125 should have been ok, but certainly not for everyone. The full frame image really didn't look too bad to me.
In my photos, you don't need to enlarge it at all to see how far off the focus gets. It's very obvious.
These are two samples I posted at DPR in a response to Ben. Ben has also posted samples that look just like RG's results. Maybe worse.
Here are the two samples with the copied text:
These two photos are back to back in a sequence, at a slower fps. It shows how the camera drifts in & out of focus while tracking. The first one looks perfect. The photos are cropped, resized AND sharpened, and yet it's still so obvious. Ignore the framing of my crops, the center of the photos is actually the shoulder/chest area of the players. I'd say about 30-40% look very bad in these conditions. Not what I expect from a new "World Class Focusing System" in a $4500 camera. http://www.pbase.com/gary/image/84202738/original.jpg http://www.pbase.com/gary/image/84202742/original.jpg
Garylv wrote:
Yes, that particular example of the couple looks like motion blur to me. 1/125 should have been ok, but certainly not for everyone. The full frame image really didn't look too bad to me.
In my photos, you don't need to enlarge it at all to see how far off the focus gets. It's very obvious.
Yes, it is so obvious that the AF seemed to be on a coffee break and forgot to follow the soccer player . Looking at the grass actully tells a lot better where is the focus.
Well, this is the problem I see: We have two set of samples from two Mk3 users supposedly to demonstrate the AF problem of the 1DMk3. One is legitimate AF problem and one is very obviously camera shake instead of AF problem. Both users are no strangers to the 1 series cameras. May be we are trying too hard to LOOK for the AF problem and it actually affects our judgement? If this is any indication of the Mk3 AF issue, you have to draw your own conclusion from this poll. I am not saying there is AF issue or there is no AF issue, and I still have to find out for myself.
That's quite a bit back focus in the 2nd frame. Do you suppose it's locked on the previous frame?
I'm trying to get my 3rd body. My first two from B&H were off; one was just terrible in One Shot and the second seemed great except for low light conditions where it paled next to my 30D. I shoot music photography, so that's why I'm buying this camera. Anyway, I didn't want to wear out my welcome with B&H in case no.3 was bad so I'm hoping a thread pops up here or on DPR when some are available someplace.
Garylv wrote:
Yes, that particular example of the couple looks like motion blur to me. 1/125 should have been ok, but certainly not for everyone. The full frame image really didn't look too bad to me.
In my photos, you don't need to enlarge it at all to see how far off the focus gets. It's very obvious.
These are two samples I posted at DPR in a response to Ben. Ben has also posted samples that look just like RG's results. Maybe worse.
Here are the two samples with the copied text:
These two photos are back to back in a sequence, at a slower fps. It shows how the camera drifts in & out of focus while tracking. The first one looks perfect. The photos are cropped, resized AND sharpened, and yet it's still so obvious. Ignore the framing of my crops, the center of the photos is actually the shoulder/chest area of the players. I'd say about 30-40% look very bad in these conditions. Not what I expect from a new "World Class Focusing System" in a $4500 camera. http://www.pbase.com/gary/image/84202738/original.jpg http://www.pbase.com/gary/image/84202742/original.jpg
Yep, it's pretty bad. Rather unpredictable too. Some short sequences look ok, others look terrible. I have some sequences in soft sunlight that look good, where only a couple photos were off just slightly. Bright sun gives me the worst mixed results.
Over at the Naturescapes forum I saw this reply by Charles "Chas" Glatzer, who has been working with the new firmware also. And although he commented on how the MarkIII's AF is different than the MarkII, he also mentioned this:
"However, overall the AF remains well below the Mark IIN in back-lighting, and locking focus with moving subjects is most difficult and erratic at best, with some images sharp in a burst and others not. It is this that causes the most concern."
That's how mine performs also, and appears to be right in line with RG's article. RG mentioned the Mark IIn easily trumps the Mark III in those shooting conditions.
I sure hope Canon will get the Mark III's AF working at least as good as the Mark II. Otherwise it's turning out to be a pretty big disappointment. I'm a little gun-shy about buying a new camera as soon as it's released now. Rather expensive lesson.
. I still have so much to learn and so far I would consider majority of the mistakes are due to user errors. I believe that the 'optimium' CFn settings would be different for different users, based on their shooting style, subject size/background, speed, lighting, lenses, etc. There are too many possible different settings with so little experience on this camera at the current stage. I am not saying anyone's recommenede settings are wrong, but it will be more than one set of 'Perfect' setting for everyone. That is also why it is so tough for the Canon engineers to know what is the problem, not even talking about a 'fix'.