here we come, 6d AF problems

[Skirball]

My 6D is a donkey...don't get me wrong the image quality, WHEN it focuses properly, is sublime. I too mostly use the centre focus spot. I shoot ceremonies for my local council with the Mayor and participants. The hit rate was less than 20%, so much so that I now resort to manual focus for these.

I also take ID/Passport type images at theses ceremonies, I focus on the right eye, the focus beeps to confirm lock, hold and recompose and when you check the images, a huge percentage are soft, some by quite a margin.

I sent the 6D back to my local Canon service centre and asked them to check it out and they reported no problems but suggested my ageing 28-70 2.8 was probably the culprit. I bought a brand new 24-105 to replace it...guess what, yep still the same issue. And it's like that with my 70-200 2.8 and all my other lenses.

I've had it just 11 months and have totally lost faith in this body and I just cannot trust it turn in the goods when I'm out on a job.

I'll await the rumoured 'new bodies' which we are told will be coming soon to see if there is anything worth getting, if not I'll junk this 6D and look for a decent 1DX

I read this post at least 4 times, as I was certain it was a troll at first. But the more I read it the more I think that you actually believe the poor results are the fault of the camera. Regardless, thank you, it was the most entertaining thing I've read all day.

Good to know that Canon is getting increased 1DX sales due to people not knowing how to use their equipment.

 

[Dylan777]

I want to hear some feedback on AF issues with 6d.

The problem with mine is that it sometimes can't focus (I use 99,9% the central cross point).

I never thought about it like as if it was a problem until last weekend when I mounted my 35mm 1.4L on my friend's 5dm3


I WAS SO SHOCKED. Okay, it can focus perfectly, but hell why it's 10000 times sharper?

I'm considering to sell my 6d in prior to buy 5dm3 this summer and want to know if it's usual and normal to have the AF issue.

It sounds like your friend's 5DIII and your 35L were a much closer match in terms of autofocus micro adjustment (AFMA) - i.e. their tolerances were closer than those of the lens and your camera. With large aperture lenses, AFMA is critical to getting the sharpest focus. In terms of AF performance, the central sensor on the 6D should actually be more accurate, at least in low light. I'd recommend FoCal or another tool to AFMA your lenses and I think you'll be shocked by the improvements.

I agree most, except the high lighted. 5D III has dual-cross type in center, compared to 6D is not. Esp. on large aperture lenses.

True in terms of overall accuracy, I but the 6D is a whole EV [stop] more sensitive in low light, which is what I was referring to. In practical use, I would think the center AF performance is probably very similar between both bodies, but I haven't used the 6D. I'm pretty sure the OP's issue is AFMA, however.

I might have to disagree with you on this mackguyver(doesn't happen very often ;D). The benefit having dual-cross AF is to focus on the subject more accurate.

I do agree with you about OP 6D might suffers with AFMA.

 

[tron]

A few of my lenses looked like rubbish or made the camera seem so until I microadjusted them. I did not even use focal.
A trial and error AFMA for my 85mm 1.2, 135 2 and 35 1.4 worked miracles on all my cameras.

Now something funny:
5DMkIII No1: 85 1.2L II +10
5DMkIII No2: 85 1.2L II +10

5DMkIII No1: 35 1.4L +10
5DMkIII No2: 35 1.4L 0 !!!

 

[candc]

I have a 6d and think the AF is very accurate if you can live with the few and far apart AF points. I don't have a 5diii to compare it to but I think the accuracy is better than on my 70d just clunky and too much space in between the points for action stuff but fine for static shots. I bought a 50l a bit ago and it has a reputation for focus problems wide open, so far I haven't had any.

p.s. what the others are saying is important. You can set up a focus target in the lawn (grass is a good focus indicator) or set a tape measure alongside, check to make sure you are not front or back focused and adjust your afma. You can use live view to focus and check against also.

 

[FEBS]

A few of my lenses looked like rubbish or made the camera seem so until I microadjusted them. I did not even use focal.
A trial and error AFMA for my 85mm 1.2, 135 2 and 35 1.4 worked miracles on all my cameras.

Now something funny:
5DMkII No1: 85 1.2L II +10
5DMkII No2: 85 1.2L II +10

5DMkII No1: 35 1.4L +10
5DMkII No2: 35 1.4L 0 !!!

I think that your trial and error method has something to do with this point. I first tried trial and error, then dot-tune but was never satisfied about the results. The tolerance was there. Focal, which uses statistics behind the screens, is taking care of that error. Focal really looks for the top sharpness. I also think that a 34 1.4 is less sensitive to afma then a 85 1.2. I see that more or less also on my 24 1.4. The longer the lens the more critical. The wide angles have a much wider DOF, so simply by trial and error finding the top level of sharpness doesn't seem simple to me.

 

[tron]

A few of my lenses looked like rubbish or made the camera seem so until I microadjusted them. I did not even use focal.
A trial and error AFMA for my 85mm 1.2, 135 2 and 35 1.4 worked miracles on all my cameras.

Now something funny:
5DMkII No1: 85 1.2L II +10
5DMkII No2: 85 1.2L II +10

5DMkII No1: 35 1.4L +10
5DMkII No2: 35 1.4L 0 !!!

I think that your trial and error method has something to do with this point. I first tried trial and error, then dot-tune but was never satisfied about the results. The tolerance was there. Focal, which uses statistics behind the screens, is taking care of that error. Focal really looks for the top sharpness. I also think that a 34 1.4 is less sensitive to afma then a 85 1.2. I see that more or less also on my 24 1.4. The longer the lens the more critical. The wide angles have a much wider DOF, so simply by trial and error finding the top level of sharpness doesn't seem simple to me.

I would buy FoCal (if you have a lot of gear it is value for money anyway) but I felt it would be too much trouble to get the targets printed at various sizes, hanged 100% vertically and use proper lighting (which I do not have). So I used trial and error and I am satisfied.

Plus a correction: My two cameras are 5DMkIII not II (but the principle is the same).

 

[Loswr]

...I felt it would be too much trouble to get the targets printed at various sizes, hanged 100% vertically and use proper lighting (which I do not have).

Just to point out…if you test at a distance that's a fixed multiple of focal length (I use 25x and 50x), the target is always the same relative size. Personally, I just use an inkjet print of the target on letter-sized heavy matte paper for all my testing. Also, it does not need to be hung perfectly vertical, as long as you have a ballhead it's easy to align the camera to the target at a slight angle, and the software guides you through that.

 

[tron]

...I felt it would be too much trouble to get the targets printed at various sizes, hanged 100% vertically and use proper lighting (which I do not have).

Just to point out…if you test at a distance that's a fixed multiple of focal length (I use 25x and 50x), the target is always the same relative size. Personally, I just use an inkjet print of the target on letter-sized heavy matte paper for all my testing. Also, it does not need to be hung perfectly vertical, as long as you have a ballhead it's easy to align the camera to the target at a slight angle, and the software guides you through that.

Thanks for that. So I guess this leaves only the lighting not being correct. As for distance, I have 14mm to 500mm but up to 200mm I sometimes shoot from not too far away (especially for 85m 1.2) so I guess I can make a compromise and test shoot from a little closer than 25X. But even so when I use the 85 the lighting is even less than my living room's so maybe the lighting is of no issue (it is incandescent type by the way so at least it is the correct type).

 

[mackguyver]

I might have to disagree with you on this mackguyver(doesn't happen very often ;D). The benefit having dual-cross AF is to focus on the subject more accurate.

I understand and I guess the way I'm looking at is that for still subjects using the center point AF in decent light with a subject with decent contrast, I haven't found any real-world difference between my 1D X, 5DIII, 5DII, or even 60D and Rebels. I get the technology of the new bodies is better, but the center point AF on all of the Canon bodies is excellent. If we move beyond the central sensor and/or throw AI Servo into the mix, that's a different story

 

[Loswr]

...I felt it would be too much trouble to get the targets printed at various sizes, hanged 100% vertically and use proper lighting (which I do not have).

Just to point out…if you test at a distance that's a fixed multiple of focal length (I use 25x and 50x), the target is always the same relative size. Personally, I just use an inkjet print of the target on letter-sized heavy matte paper for all my testing. Also, it does not need to be hung perfectly vertical, as long as you have a ballhead it's easy to align the camera to the target at a slight angle, and the software guides you through that.

Thanks for that. So I guess this leaves only the lighting not being correct. As for distance, I have 14mm to 500mm but up to 200mm I sometimes shoot from not too far away (especially for 85m 1.2) so I guess I can make a compromise and test shoot from a little closer than 25X. But even so when I use the 85 the lighting is even less than my living room's so maybe the lighting is of no issue (it is incandescent type by the way so at least it is the correct type).

Not sure where you are or what available there, but home improvement stores here have halogen work lights that are relatively inexpensive ($20) and a couple of them would provide ample light.

 

[mackguyver]

Not sure where you are or what available there, but home improvement stores here have halogen work lights that are relatively inexpensive ($20) and a couple of them would provide ample light.

Those would definitely work well and I always feel silly using my Einsteins' modeling lights for FoCal tests.

 

[aleshaloginov]

I actually noticed that the problem I described appears only when I use wide angle lenses

The wider the angle, the bigger the problem — when I go to 16 on 16-35 it becomes totally obvious that something is wrong.

 

[ifp]

I actually noticed that the problem I described appears only when I use wide angle lenses

The wider the angle, the bigger the problem — when I go to 16 on 16-35 it becomes totally obvious that something is wrong.

You haven't even addressed whether you've tried to AFMA your body/lens combinations.

Maybe you could post some full res samples of your issues, or at least answer some of the questions that have been posed. If your problems are increasing as you go wider, I wonder if you're not keeping your shutter speed up enough.

What I can tell you is, a 5d3 isn't likely to solve your issue. There's a very small chance that your 6d is broken, but you'd save a ton of money by getting it repaired instead. There's a slightly larger chance that a 5d3 will just happen to be better suited to your lenses, as far as manufacturing tolerances go, but that also is unlikely. Odds are highest that you either need to AFMA or you're making a technical error while taking shots. But if you don't share more information, nobody here is going to be able to help you.

 

[mnclayshooter]

Just quickly scanned through the OP's Flickr page. The dozen or so photos there are nicely composed and don't show signs of focus issues to me. I agree with ifp It would be interesting to see the results on a photo where there is an issue.

 

[aleshaloginov]

I actually noticed that the problem I described appears only when I use wide angle lenses

The wider the angle, the bigger the problem — when I go to 16 on 16-35 it becomes totally obvious that something is wrong.

You haven't even addressed whether you've tried to AFMA your body/lens combinations.

Maybe you could post some full res samples of your issues, or at least answer some of the questions that have been posed. If your problems are increasing as you go wider, I wonder if you're not keeping your shutter speed up enough.

What I can tell you is, a 5d3 isn't likely to solve your issue. There's a very small chance that your 6d is broken, but you'd save a ton of money by getting it repaired instead. There's a slightly larger chance that a 5d3 will just happen to be better suited to your lenses, as far as manufacturing tolerances go, but that also is unlikely. Odds are highest that you either need to AFMA or you're making a technical error while taking shots. But if you don't share more information, nobody here is going to be able to help you.

Actually all I can share is that quite often my camera can't focus on wide angle, but when it locks everything is just fine. It's not connected with anything like shutter speed, it's only AF system.Besides, I was told by one technician that this is impossible to repair, though I'm not sure now that he's competent.
I'm gonna AFMA it this week, hope it'll help.

 

[Mt Spokane Photography]

I was told by one technician that this is impossible to repair, though I'm not sure now that he's competent.
I'm gonna AFMA it this week, hope it'll help.

Since the other 100,000 owners do not have the issue, its not impossible to repair. However, Canon might need both the lens and the camera body to see which is causing the issue. Since they cannot find a issue in the body with there lens, that does hint that it might be a combination issue.

If AFMA doesn't fix it, send both to Canon.

 

[FEBS]

If AFMA doesn't fix it, send both to Canon.

Fully agree on this.

 

[ifp]

Actually all I can share is that quite often my camera can't focus on wide angle, but when it locks everything is just fine. It's not connected with anything like shutter speed, it's only AF system.Besides, I was told by one technician that this is impossible to repair, though I'm not sure now that he's competent.
I'm gonna AFMA it this week, hope it'll help.

I wish you'd be a little more clear with your problem description. So far I think you've described two very different problems, one with a 35 f1.4 lens, and another with a 16-35.

I might be reading a little too much between the lines here, but it sounds to me like you can't even lock focus with the 16-35? In one shot AF mode, I would expect that to only happen if you're shooting at a very low contrast target (white wall, clear blue sky, no light etc) or closer than the MFD.

If that's the case, I'd send them in, with a clear description of what you're doing and what's going wrong.

 

[ashmadux]

I can tell you my experiences.

After my first 6d, I thought the same thing. The Af was atrocious. It was missing possibly 95% of shots. I dont use focus/recompose. Canon switched the body. The new one - my current one- works a lot better and is 'stable' in the Af department.

When testing the 2nd body, I canvassed the web and review sites for 6d images- straight from camera- as well as 5d3 raw files. What i found was the overwhelming majority of 6d images were in fact not anywhere close to sharp. But they looked good. Similarly, none of the 6d shots i see are anywhere as sharp as even the 5dmk2 shots Ive been seeing on the web for years.

I sent the second one in along with my 24-105 before my vacation to japan, and it did well- the shots on average are well focused. Im now comfortable enough to carry it around, and did a few shoots with it since then.

So even though its better now, I cant say that Im happy. I still dont have the sharpness or 'wow' factor from the images i expected to see with my 6d. The outer points work, but they fail just as much.

So at some point i will sell it and trade up.



Good luck.

 

[Dantana]

Not sure where you are or what available there, but home improvement stores here have halogen work lights that are relatively inexpensive ($20) and a couple of them would provide ample light.

Those would definitely work well and I always feel silly using my Einsteins' modeling lights for FoCal tests.

I bought a pair from Home Depot for this very purpose, and they were just the thing. Two of them clamped onto chair backs pointed at my Focal target. They are also good in a pinch if I need some constant light for some other purpose.