My ordeal with Canon's AF

<p>For the sharpness problem is probably the people who are shooting Nikon are uploading via jpeg and the Canon are uploading in Raw. That is my best guess if not start asking your coworkers what they are doing different on their focus and how they are processing their photos. I does not hurt to ask.</p>

<blockquote>

<p>A critical thinker would look in the mirror and ask "Am I <strong>sure</strong> the equipment is faulty?", then test the heck out of it...</p>

<p> </p>

</blockquote>

<p>This is not only the most salient comment in this thread, but one of more perceptive comments I've read on p.net for quite awhile. Permit me to show this by way of example.</p>

<p>Some time ago, I thought that my EF 24-70/2.8 L was quite "soft" compared to my EF 50/1.4. I knew that, generally speaking, primes were sharper than zooms, but I just couldn't believe that an L-zoom would be so relatively soft. So I decided to investigate, and discovered that the soft handheld shots taken with my 24-70 all had shutter speeds below the inverse of the focal lengths I had shot at. Upon "correcting" this problem, the softness miraculously went away, and the lens became almost as sharp as the prime. Further investigation (such as bracing the camera) revealed that the zoom delivered quite astonishing resolution.</p>

<p>Of course, had I not applied a little critical thinking, I might have been pointlessly venting in the same way as the OP...</p>

<p> </p>

<p>This is mostly operator error I have used 70-200mm, 24-70, 300mm f/4 with no problems. It takes practice to get sharp images, you can not expect every image you take to come you sharp. It could be the way you are using your gear. If you want a sharp image of a flying bird use Al Servio and use TV 1/1000 in panning mode, if you want to get an subject that does not move use one shot and use Av in panning mode 1, try shooting with the aperture wide open and narrow. Use countious for subjects that are moving. I hope this helps.</p>

<p>how funny there are folks suggesting me i should try to better test my equipment and, for instance, try not to activate all AF fields when shooting or do not use f22.<br>

this is beyond me. cant you people understand what you are looking at, especially this self-satisfied guy named taylor? as i wrote i have been using canon digital stuff since there was the original 1D out. do you know how long this has been? do you think I had not in the mean time opportunities enough to test my gear to the limit, dont you? tripod, MLU, LiveView, etc. including? and you suggest me not to use f22 when shooting?<br>

now, how brilliant and intelligent is that? <br>

i hate being personal, but dont you think, my dear lee taylor, you could use some classes about logical thinking, making assumptions etc? next time please do not tell me what i should do and what not. how can you know i had not done it already? instead you take it for granted. in your small world what cannot be, has never happened. i am very sorry and hope this will not devastate your ego: yes, there are things possible, even if you cannot imagine them.</p>

<p>ad res: i sent in my stuff a few times and never got the information from the service, like: sorry guy you are a troll (thanks folks for the compliment), the equipment is ok and there is nothing to calibrate. no, there were each subsequent time corrections to be made. </p>

<p>i say:</p>

<p>canon cameras and lens are great, but they do not deliver what they promise. you have to invest a significant amount of care and time to get your canon equipment to work. apparently the chances are very high, that out of the box, there will be troubles. and apparently the chances are, that sending in your new stuff once or even twice will not help. be prepared for more.</p>

<p>i say:</p>

<p>i do not know if nikon is "better" than canon at any reagrd and have no plans whatsoever - how often must I repeat that?- to change brands. however, what I see, is, that at our agency people using Nikon stuff have out of the box significatly higher rate of sharp pictures, than the guys who use Canon.</p>

<p>and the last one, it seems that simple providing a link wont help in this forum, so i will try to point it out explicitly for you:</p>

<p>www.arcurs.com/clash-of-the-titans-canon-1ds-mark-iii-vs-nikon-d3x-for-stock<br>

says:</p>

<h3>Conclusion</h3>

<p>The Canon has lower noise levels and the lenses perform generally better, but what does this matter if your images are more out of focus, if the camera is much harder to work with on a daily basis and if you can get the same results form a Nikon by just choosing the right lenses and get another 3 mega pixel on top?<br>

end of citation</p>

<p>but, if some of the people in this forum think, i am wrong, because the camera they own with the f4,0 (!) zoom lens they own focuses ok, than the better for them and the worse for me.<br>

oh, my last sentence was irony, right?</p>

<p> </p>

<p>Why not switch to Nikon if Canon cameras don't work?</p>

<p>Jan, i think the majority of people considered your statement claiming ownership of the 1d, but then decided to put it in the BS bin when reading the rest of your post.<br>

Trained pros take bad shots, they just do it less often than beginners. As a rental junky, I've never had an issue using a variety of used gear and having it come out sharp. If the shot was messed up, during review, I could see how I did it each time.</p>

<p>With the time you have taken for all of these rants, you could have easily:</p>

 

<ul>

<li>already exchanged your system for a Nikon system, saving everybody a lot of time</li>

<li>done some more thorough testing</li>

<li>at least posted an image demonstrating the problem you are dealing with</li>

</ul>

 

<p>I certainly don't know about the technical aspects of Canon's autofocus systems as well as most of the above folks, but I want to say that based on my experience with my Canon cameras and lenses, I can believe that a person could have tremendous trouble with autofocusing and it isn't due to user error...and that others can have very little trouble with autofocusing, and it isn't due to user 'smartness' either.</p>

<p>I personally haven't had too much trouble with autofocusing after my 20D/16-35mm f2.8 L combination--my first digital set. I mostly had trouble with backfocusing with the wider focal lengths. This was due to factors on both my end and the camera/lens end. The 20D (and I assume cameras of that time, such as the 1DMkII) definitely had flaky autofocusing. I once watched the focus point change position before my eyes--in One Shot mode, after it had locked on. I could even make it repeat. A colleague purchased the 1DMkII about the same time as I bought the 20D, and he has to watch the autofocusing too.</p>

<p>After I got my 5D, which I paired with a Tamron 28-75mm rather than an L lens, I didn't have much trouble with autofocusing, beyond the extreme wide angle lengths, where it is hard to autofocus or even manually focus, anyway. My teles are Canon primes (not L), and I never had trouble with those. I find, though, that I still have a higher percentage of OOF shots when using the 5D with the 16-35mm L. I have heard of plenty of people who have trouble with the 5D with the 24-70mm f2.8 L. So much trouble, in fact, that they have given up and gone to Nikon. These are professionals, and certainly know how to test and use their gear. Most 'gone to Nikon' users are very happy with their system.</p>

<p>From the above I can only tentatively conclude that, possibly:</p>

<p>1. Perhaps L lenses are so 'well oiled'--their autofocus systems are so finely tuned and the mechanism so smooth--that they tend to jump the gun re commands from the body's autofocus system. By 'jumping the gun', I mean the first attempt to autofocus results in missing the mark, therefore the lens will backfocus. I notice with Tamron lenses, that when the lens misses the mark, they front focus. I figure my Tamron 28-75mm's autofocus is a 'pig' compared to an L lens' autofocus, but it's very coarse nature makes it more reliable with autofocus.</p>

<p>2. Even though one can improve one's autofocus technique by knowing about the camera's system--giving it a good 'target', knowing the actual points are bigger than marked, etc.--there are still small differences in a person's actual use of the gear that might account for autofocus working OK in one person's case, and not in another.</p>

<p>3. In the OP's case, I know the 1DIII had autofocus problems from the get go.</p>

<p>All of the above (save for 3) is not based on any science of any kind, and is all highly speculative on my part. Just thought I'd be one of the voices that didn't immediately say that Jan's problems with autofocus is due to user error.</p>

<p>This is another one of those pontless discussions. Only a few happenings can be construed as facts:<br>

1. The OP has been having recurrent problems with his cameras and/or lenses;<br>

2. No one can tell what is the cause of those problems;<br>

3. I have been using the EOS system for 20 years, and I can tell you for a fact that it has been a problem free experience thus far.<br>

Other than these, it is just speculation, which can be fun, but is pointless.</p>

<p>Here,Here</p>

<p>I don't know, Paulo, I would ask--pointless for whom? I assume the OP is interested in a discussion of his issue with Canon autofocus, to include speculation, because he wants to a) know if anyone else has had issues and b) wants to find solutions, if any and c) wants to know what people think. If this is the case, why do we have to stick with facts?</p>

<p>From the above, I would conclude people think he a) is a whiner and b) doesn't know how to use his gear. This is typical for people asking these type of questions. I belong to a professional wedding forum and there have been similar questions from wedding professionals about Canon autofocus, many times involving the 24-70mm L, a popular wedding lens. The people bringing up the issue are similarly roasted for not knowing what they are doing. But upon examination, they are not doing anything 'wrong'. It isn't user error. As I said above, I think there might be (this is indeed speculation) room for individual interface--how a person uses his or her gear--that might not work well with Canon autofocus.</p>

<p>I also have not had too many problems, as stated above. However, just because I have not, doesn't mean the issue isn't very real for someone else.</p>

<p>Nadine, I believe that if the OP were serious and sincere, he would have posted examples of photos he believed degraded by a dysfunctional AF system, complete with EXIF data. Instead, he quotes blogs and reviews with commentary that is hardly on point.<br /> In any case, this guy seems to want to evangelize a cause, not improve his photography.<br /> Just my $0.02.</p>

<p> </p>

<p>Point taken, Robert. I won't argue there. I am not convinced, however, that there isn't some sincerity in the question.</p>

<p>I thought I had logged on to photo.net. Apparently I logged on to DP review by mistake.</p>

<p>I really don't even understand this thread.</p>

<p>Jan, I'm sorry you've had a run of bad luck. However, to simply assume that this is some how representative of the average Canon user's experience is downright silly. Surely if Canon equipment was as bad as you claim the internet would be downright full of complaints and Canon would have long since shut down its photography unit. You also seem to be implying that the average Canon user is downright ignorant and is completely unaware of how poor their equipment is. Frankly, I'm surprised your trollish post has been treated so well. </p>

<p> </p>

<p>Jan is correct. There is a systemic problem with the canon EOS and I am surprised that more photographers have not realized this. The reasons have to do with manufacturing tolerances and the way these loose tolerances are exacerbated when bodies and lenses are combined. If you photograph a test target with a couple fast Canon zooms, like 2.8s, at a 45 degree angle to the wall while on a tripod you will discover that although you can manually micro adjust the lens to the body at a specific distance there will be significant focus shift as you change distances. The lens will be front or back focused at different distances and is not connected with ones photographic technique. This is sometimes only evident when you make large 20x24 inch or bigger prints. You can micro adjust each of your bodies for each lens at different distances but this is not what you want to be doing when you are photographing and defeats spending $1500 on a lens and $25000 for the EOS system. In addition, once you place the subject outside the very center of the frame the perceived softness gets even worse. Canon actually knows about all of this but they would have to tighten their quality control and that costs money. More photographers need to complain so that they tighten the manufacturing tolerances and we as photographers get what we are paying for. I come from the large format and medium format world so I am very sensitive to issues of soft focus. You can look at my web site if you question any of the above statements or my qualifications for making them. Jimmy Katz <a href="http://www.jimmykatz.com">www.jimmykatz.com</a></p>

<p>Jimmy, if there was a systematic problem then, by definition, most photographers would realize this. Or, most Canon users are ignorant fools. And Jimmy, unless you have handled hundreds of Canon lenses and bodies, you really aren't qualified to make any sort of sweeping statement as extrapolating from your limited experience (in terms of the <strong>number of lenses</strong> you have used, not your talent/experience as a photographer) isn't really relevant. </p>

<p>People get bum equipment all of the time. Sometimes the gear is simply bad, sometimes it is simply that both the lens and the camera are individually within tolerance, but together out of tolerance. This is a problem that afflicts every camera manufacturer as a brief reading of any number of camera-orientated websites will quickly show. </p>

<p>Perhaps Canon is worse than everyone else when it comes to autofocus performance or quality control. However,<strong> none</strong> of us can really say as <strong>none</strong> of us has tested a <strong>large</strong> set of gear. My gear all works (Canon and Nikon), Jan's obviously doesn't and I guess Jimmy has had problems too. What does this show... well absolutely nothing as it is three user experiences out of a million. Even the dumbest among us can see that this sort of anecdotal "evidence" is useless. </p>

<p>@ Craig: the problem with your and similar reasoning is, that it pretends to be scientific, but it is not quite so. rob galbraith did not test statistically significant amount from the complete production of canon's 1d mk III and yet he alone was able to show some problems.<br>

craig, how many cameras and lenses should test OK to acknowledge the whole system as "good"? 70% or 80% or 100%? of course, this question is purely hypothetical, and shows the limits, or if you will the "uselessness", of your argument.<br>

from my experience, from watching at the daily input of images at our agency and from the questioning the pros I know personally, I see that there apparently is something wrong with the system as a whole. <br>

nobody here has yet mentioned big inter- or national agencies switching from canon to nikon in the last three years wheras before it was mostly all-canon game. this could be another hint.<br>

please google for "inconsistent outer AF points" and you will find lots of stories of people complaining about having to send in their canon stuff many times, with no help. i dont think they all have no idea how to use the gear. we are talking about 85mm F1,2 lenses and similar expensive high end gear used wide open and not of consumer f4,0 zooms. <br>

please have a look at the link I have posted here above, this guy seems to know what he is talking about. he says, he has used numerous canon bodies and states, the af results were inconsistent.<br>

well, you would probably say - he cant be right, he's simple used too few bodies to make such an statement. but how many cameras should he test then, in your opinion? five, seven or may be twenty seven? and where the bad luck ends and the systemic problem begins?<br>

also as i have written before, nobody builds his system from a scratch overnight. most, if not all, canon users accumulated their lenses over the years. and now it turns out, for instance, the level of accuracy that was OK with a 4MP camera is not enough for a 15MP or more camera. but the money on lenses has been spent already, so you will look for solutions how to live with the problems, rather than change the brand. <br>

in other words the amount of canon users has not much to do with proven reliability (or unreliability, for that matter) of the system.<br>

and as Jimmy Katz said, it is surprising that not even more photographers are complaining. lets hope then, that with the MK IV generation of canon cameras the problems will be less.</p>

<p> </p>

<p>Okay, since I have never had any AF issues with any of my Canon EOS gear (EOS-1V, 3, and 5DII bodies; seven of the best EF primes and four of the best EF zooms), I can safely conclude from your very <em>scientific</em> brand of <em>reasoning</em> that the EOS sytem has no AF problems whatsoever. But where does that leave us? You think the gear's junk, and I think it's near perfect. Sounds like a contradiction to me.</p>

<p>So much for your "scientific reasoning."</p>

<p>I should have clarified a few facts. First, the sample of bodies and lenses that I have observed front or back focusing in has been limited to a total of about fifteen new Canon USA pieces purchased at separate times from different authorized USA dealers. I confirmed with a tripod and wall charts, that not one piece focused precisely out of the box and again I have not tested hundreds of pieces of equipment. To be clear, my standards are that an F/ 1.2 lens or a F/ 2.8 zoom lens should focus “Dead On” wide open, at all distances since that is what we are all paying for.<br>

A few years ago I took some of my gear to a Canon service center and had it calibrated by them while it was under warranty. I made photographs with one of the zooms in front of a Canon employee and was still able to confirm that the zoom lens was front focusing at some distances and back focusing at others. The employee told me that I was “looking too closely at the image” and that for most photographers their lenses and system are fine. I do not claim that Canon’s system is “worse than others”, but I am not sure that aspiring to be “no worse than others” would be a lofty slogan for a company that has access to such superb technology.<br>

Why do you think Canon would enable the 5D Mark 2 to micro adjust the focus forward or backwards, on each lens, to each body, if they were NOT having focus problems with the original 5D which did not have this feature. I think this feature was added because their cameras were having focusing issues. This function does help, but does not solve the front and back focus problem.<br>

I should mention that I shoot all my commercial jobs like magazine covers and CD packages with the Canon system and I actually think the system is a lot of fun to shoot with. My assistants know to remind me to stop certain lenses down so that we get acceptable sharpness for clients and we also sharpen selectively in post. If you are happy with the images that you get from your Canon system and are convinced that your fast lenses, focus accurately, wide open, at all distances, then you should be quite pleased. Focus is subjective and depends a lot on what you have experienced and I would never contradict a photographer who claims to have sharp focused images. I also admit that using medium and large format gear has shaped my point of view.<br>

I am only reporting what I have empirically experienced. I would encourage all photographers to test their fast zooms (F/ 2.8) and prime lenses (F /1.2-1.4) wide open, on a tripod, at 45 degrees to a test chart on a wall for focus accuracy. This way if an image is slightly front or back focus when you photograph in the "real world" you will know if it was your own mistake or an alignment issue with the camera. Be prepared to be surprised. jimmykatz.com</p>

 

<p>Why can't you end this dribble and go take some out of focus Canon photographs.</p>

<blockquote>

<p><em>"I am only reporting what I have empirically experienced. I would encourage all photographers to test their fast zooms (F/ 2.8) and prime lenses (F /1.2-1.4) wide open, on a tripod, at 45 degrees to a test chart on a wall for focus accuracy. This way if an image is slightly front or back focus when you photograph in the "real world" you will know if it was your own mistake or an alignment issue with the camera. Be prepared to be surprised. "</em></p>

 

</blockquote>

<p>Whatever. Whatever whatever...<br>

I shoot real things not stupid test charts.<br>

Test charts are for engineers and non photographers. They are for QA people.<br>

Canon definitively does not have any 'systematic" focusing issues with their EOS product line (now 22 years old). Maybe you should hype your lame theory for Oliver Stone's next film?</p>

<p> </p>

<p>Alan, don't you mean "drivel"?</p>

<p>I absolutely can't believe this is the longest most popular thread on this site. Very sad as it has absolutely 0 viable information.</p>

<p>I've most of the Canon fast L lenses such as the 24 II, 35, 85, 135 and 200/2. Getting them (especially for the 85mm and longer lenses) to focus accurately, at large aperture, is often a hit and miss. Typically, u can try shooting portrait with these lenses with the eye as the point of focus. My hit rate whereby the focus is spot on, is far less than 50%. Such low hit rate is not limited to my current models of the 5D2 and 7D bodies, but also many of the previous models including the 1Ds.<br>

I don't know if this is a problem with the Canon system of is it due to the extremely narrow depth of field. But I really hope Canon could improve upon the focus accuracy.</p>