Mk3 Af

[Paul Bestwick]

I believe there is a firmware fix for the MK3 AF issue not too far off. Having said that, my source informed me that there have been no AF problems reported in Australia. Not one !
I can at least confirm that for my own camera.

Cheers,

Paul

 

[Nill Toulme]

Rob Galbraith posted another update to his AF report today; not encouraging at all. I'm really in a quandary over this.

Nill
~~
www.toulme.net

 

[Paul Bestwick]

Hey Nill,

relax man. This is one awesome camera. From what I can see next to my 1DSMKII I am liking it a lot.
Check the 2 crops in my new comparison thread. The highlight priority feature is great. The exposures are the best I have ever seen straight from the camera. The detail is fabulous.

Cheers,

Paul

 

[Nill Toulme]

Well I'm still hanging in, but I'm thinking more seriously about going ahead and picking up a 1DsMkII to hold the hand of the III that I still have on order.

Interesting counterpoint to RG's update here:

http://photography-on-the.net/forum/showthread.php?p=3476589#post3476589

Pekka generally seems to know what he's talking about.

Nill
~~
www.toulme.net

 

[Paul Bestwick]

thanks for that link Nill. Confirms what I have thought all along. No camera problem.
The evidence can be seen in those amazing speedboat shots. Wow.
You had better get out there & buy one Nill.

Cheers,

Paul

 

[Nill Toulme]

Well I'm not going to go out and beat the bushes... I've cast my lot with the order I have pending. It'll come when it comes, and we'll see then .

Nill
~~
www.toulme.net

 

[Nill Toulme]

As of this posting the latest FM Mark III owners' poll is split right down the middle, with 40% of voting owners saying their cameras definitely have AF problems, 40% saying theirs definitely do not, and 20% still unsure.

What I'd really like to see is a few owners in the "definitely problems" camp getting together with a few from the "definitely OK" crowd and exchanging cameras for a shooting session or two. Has anybody seen any indication that anything like that has occurred anywhere? The closest I've seen to anything like it is a post from somebody saying they exchanged their first body for another one and the second body was a lot better.

I'd just like to know what a (preferably knowledgeable and competent) owner of a "bad" unit makes of a "good" one, and vice versa.

Nill
~~
www.toulme.net

 

[Tony Bonanno]

Hi Nil,

Regarding the focus issue. I can tell you that I've compared my 1D3 to one of my old 1D2 bodies going through the same test procedure. The 1D3 performed noticeably worse than the 1D2. That was enough to ruin my day.

Tony

 

[Ben Rubinstein]

Given that the vast majority of people buying the mkIII had a mkII I would assume they are experienced in what they are used to seeing. I also think that many people are missing the point that this is a problem in hot and very sunny/contrasty lighting. Australia is in mid winter and the UK (which also says no problems) is in mid monsoon at present!

 

[Steve Saunders]

Good points about the rubbish weather possibly affecting tests. I wonder if the serial number on the camera makes any difference? Mine is 225XX.

 

[KrisCarnmarker]

Given that the vast majority of people buying the mkIII had a mkII I would assume they are experienced in what they are used to seeing.

But Ben, by the same token, it can be said that because the vast majority had a MkII, they are overconfident in their handling and expectations of the MkIII. There is rarely anything worse than somebody who thinks they know how to do something but in reality does not. I'm not saying this is the case, but just because somebody previously owned a MkII does not make him a better judge. Quite possibly the opposite.

I also think that many people are missing the point that this is a problem in hot and very sunny/contrasty lighting. Australia is in mid winter and the UK (which also says no problems) is in mid monsoon at present!

Yes, but this goes both ways no? There's plenty of people claiming problems when they obviously shot in other conditions. My own tests are hardly proof of anything but it doesn't get much hotter or brighter than here anywhere While the test shots did show some OOF samples, it performed far far better than my 20D, under the same conditions (which where admittedly not ideal)

Only time will tell what is happening, but my guess is that its a combination of both.

 

[Ray West]

A couple of many things that could effect 'tests'. Many folk with focussing issues send the camera and lens back to canon, to get the tolerances sorted out. It may be that they have not yet done that with the mk3. Another situation, concerning objects approaching (or going away from) the camera (assuming you have selected one fixed af point only), although the point you are covering with the af sensor rectangle may be the same, if you can aim that consistently, its size relative to the af sensor will be changing, and since most subjects are usually convex, for a distant shot, I think the af sensor area will include subject area behind where you want it to be. e.g. a car number plate close up, versus plate/bonnet/windscreen line in the distant view.

The link that Nil mentioned has a number of explanations. The important thing is that this stuff is not magic, someone has to work out how/what focussing is, design software and hardware to comply with their algorithms, to get the system to react in micro-seconds Then the marketing guys put it out for the public to do the testing.

You have to treat these digital cameras as consumables/disposable. In a year or two, you will be buying something else, with even better performance. These are not the same as film cameras. Also, one camera does not perfectly fit all.

Best wishes,

Ray

 

[Asher Kelman]

A couple of many things that could effect 'tests'.........

if you can aim that consistently, its size relative to the af sensor will be changing, and since most subjects are usually convex, for a distant shot, I think the af sensor area will include subject area behind where you want it to be. e.g. a car number plate close up, versus plate/bonnet/windscreen line in the distant view.

This is a novel point that no one else has put forward, AFAIK, another reason for not just using the center focus point and the fastest reaction setting.

The link that Nil mentioned has a number of explanations. The important thing is that this stuff is not magic, someone has to work out how/what focussing is, design software and hardware to comply with their algorithms, to get the system to react in micro-seconds Then the marketing guys put it out for the public to do the testing.

All I want is that there is a very large improvement in getting and keeping focus in low light and shade. I do a lot of portrait and street work with no flash. When my 5D or 1DII can't grab focus it's frustrating! If that would be solved, then for me, at least the 1DIII would be worthwhile, except, the 5D is so wonderfully light! But for sure, event work would be easier.

However, according to Rob Galbraith and Mike Sturk, with 20,000 1DIII frames experience, focus on stationary object aren't necessarily grabbed and kept in focus!

Asher

 

[Ben Rubinstein]

But Ben, by the same token, it can be said that because the vast majority had a MkII, they are overconfident in their handling and expectations of the MkIII. There is rarely anything worse than somebody who thinks they know how to do something but in reality does not. I'm not saying this is the case, but just because somebody previously owned a MkII does not make him a better judge. Quite possibly the opposite..

I'm sorry, you're suggesting that the AF of the mkIII is not up to its predecessor and people are going to have to cope with it? The mkIII bloody well better exeed the mkII, Canon says it does, it's marketed as such and cost in line with that. That it doesn't even match the mkII would be a disgrace.

Yes, but this goes both ways no? There's plenty of people claiming problems when they obviously shot in other conditions. My own tests are hardly proof of anything but it doesn't get much hotter or brighter than here anywhere While the test shots did show some OOF samples, it performed far far better than my 20D, under the same conditions (which where admittedly not ideal)

I'm sorry but that means nothing, you can't compare it to a far inferior camera and say that it's better. I should hope it would be far better than the 20D but that is immaterial.

I don't shoot sports but there are a lot of people, well respected people who know what AF performance they should be getting from 100,000 of frames of experience with the inferior previous model, who are seeing and showing examples of problems that they know should not be there. I wonder if they should make a new poll on Fred Miranda (where they've been keeping count) asking how many people are seeing the problem with the mkIII who a) Own the camera, b) shoot professionally c)shoot sports/action photography. I think the numbers might change significantly...

 

[Nill Toulme]

... I wonder if they should make a new poll on Fred Miranda (where they've been keeping count) asking how many people are seeing the problem with the mkIII who a) Own the camera, b) shoot professionally c)shoot sports/action photography. I think the numbers might change significantly...

Ben that might very well be true — but then again it might not. There are indeed people who a) own the camera, b) shoot professionally and c)shoot sports/action photography who say it works great. There's a whole thread on SS devoted to them in fact.

I think both those who say (implicitly or worse yet explicitly) "My camera works fine and therefore they must all be OK," and those who say "My camera sucks and therefore they're all broken" should stand down on both of those propositions until they're proven one way or the other. I remain hopeful that there are both good and bad units out there and that, therefore, sooner or later they'll all be good (and, very selfishly, that the one I have on order will be a "good" one).

Nill
~~
www.toulme.net

 

[Tony Bonanno]

Did some more testing today (July 4).. was a little cooler and I changed some of the AF settings (disabled any "helper" points and used just the center AF point). Have to admit, it performed pretty well, as good as the 1D II. I'm puzzled as to why the change. Have a shoot coming up this weekend that will be a good test for the 1DIII. We'll see what happens. There is so much to really love about this camera. The technology is incredible. I sure hope I don't have to return it.

 

[Scott B. Hughes]

Your experience is good to hear. I may be WAY off base, but have felt that there have been 'setting issues' which aren't 100% clear yet.....

I want to add one to our kit. Who/where is one available? -Scott

Did some more testing today (July 4).. was a little cooler and I changed some of the AF settings (disabled any "helper" points and used just the center AF point). Have to admit, it performed pretty well, as good as the 1D II. I'm puzzled as to why the change. Have a shoot coming up this weekend that will be a good test for the 1DIII. We'll see what happens. There is so much to really love about this camera. The technology is incredible. I sure hope I don't have to return it.

 

[Tony Bonanno]

Still more testing today (July 5).

First series of tests. Static target. Cameras tripod mounted. Target angled with rule taped to it. Target in bright sun. Temp high '70's - low '80's (not especially hot). Tested using a series with both One Shot AF and AI Servo AF. Single Center AF point with all assist AF points disabled. All shots with 70-200 2.8 IS, with IS OFF and all shots wide open at f/2.8, series of images at 70, 135, and 200. Approx 100+ shots on both 1D III and 1D II. On static subject, saw no difference between AI Servo and One-Shot AF. BOTH CAMERAS PERFORMED WELL. A toss up as far as AF goes.. IQ to the MK III.

Next series today was AI Servo AF test on MOVING cars coming down highway with several poles, fences, and other obstacles along the shooting path. Same lens as above, but shot in PROGRAM mode (typically f/7.1, 1/1000 exposure). ISO 400. ALL shots using AI Servo AF. Temp low '80's. Shot a series using Auto focal point selection and a series using just the Single, Center focal point. Again, BOTH CAMERAS PERFORMED WELL, with the edge going to the MK III. Both cameras performed better and locked focus more consistently when I followed the moving vehicles using the SINGLE, CENTER AF point. Although, I didn't have as much success using AUTO AF with either body, the MK III did better than the MK II when using AUTO AF Point selection. Both bodies had the tracking set to STANDARD (or the middle setting on MK III). This suggests to me that when using the AUTO AF (all available AF points) to track and predict focus in AI Servo mode, that the MK III is significantly improved over the MK II. I found that I had 16 out of 112 shots that were either slightly OOF or seriously OOF on the MK III. I had 21 out of 110 shots on the MKII that fell into this catagory. Considering the quick and dirty setup, the obstacles in the shooting path, and some operator error, I consider the results of both bodies to be quite good. (I would expect to do better in a real shoot situation where I'm paying more attention to my subject, the action, and the placement of the focal point).

The main issue for me was trying to determine if my MK III performs at least as well as my MKII. Based on the experience of the past two days, my MK III appears to be a keeper. The MK III's AF appears to be as good, if not better, than my MK II. Overall, the MK III appears to be an exceptional camera technologically in just about every way. I can't wait for the 1DsIII to be released.

I can't explain the inconsistency I obtained the first day of testing. Looking back on it, the only things that are different these past two days are: 1) the temperatures are cooler, 2) I have disabled any "assist" AF points, and 3), I'm not using IS on the 70-200. What bearing these factors may have had on the AF issue is unknown to me.

Admittedly, I'm still curious about what will happen when the temps are back up in the high '90's . Hopefully, nothing will change (I don't want to return it !).

Tony Bonanno

 

[Steve Saunders]

Very interesting Tony, thanks for taking the time to go through all this.

 

[Jane Auburn]

No way am I going to be a beta tester for Canon. Each possession drags life down enough as it is, without problems or worries to deal with regarding the possession.

Well I'm not going to go out and beat the bushes... I've cast my lot with the order I have pending. It'll come when it comes, and we'll see then .

Nill
~~
www.toulme.net

 

[Nill Toulme]

But I'm two possessions down as it is, having already sold both my Mark II's. I wish the III would come so I could go ahead and start worrying with it. ;-)

I don't mind a little beta testing. Both my D60 and my first Mark II were among the very first ones off the boat.

Nill
~~
www.toulme.net

 

[Asher Kelman]

Tony,

Your work is great to hear about. Keep it up, we need more like that!

All I want to know is how to use the 1DIII and have consistently perfectly focused pictures. So for me, at least, learning how to do something right that works is of great value.

Anyone can tell me how something can't be done! It's still frustrating if the natural consequence of the design is t5o go for what doesn't work as well!

Atill, as long as there are simple, reliable, straightforward strategies to learn in using a new camera, that's fine by me.

We're not there yet. Not quite!

If this happenns, then we'll be happy!

Asher

 

[Scott B. Hughes]

Who is having AF issues on OPF?

With several discussions going on various fora, it's hard to keep track.

Are any OPF members having AF problems? After 2700 releases, I can't say that we have.

All our use has been on event coverage, not under any sort of controlled test environment.

 

[Paul Bestwick]

beta testing......... I honestly don't get where that comes from. The camera has been manufactured, the only change is going to be firmware & following that a new model.
I was an early adopter of the D60, 1DS, 1DSMK2, & now the MK3 & have NEVER had one problem over all of those models.
That includes the current MK3.

PB