The Canon EF 50mm f1.2 L lens is another good example of the complete lack of quality control at Canon for the last few years. If you thought the 1D Mark III autofocus fiasco was a big problem, this seems to be much worse and up till now, after about 2 years, Canon simply refuses to comment on the issues hundreds, if not thousands, of users have mentioned with this lens. Apparently this case lacked a Rob Galbraith type of person who could really persist about the issues. In the case of the 1D Mark III autofocus fiasco, at first Canon didn’t seem too interested to look at the issues Galbraith was mentioning. After much feet dragging, they finally admitted there was a problem with the autofocus system of the camera, and then struggled for over a year to try and fix it (but failing each time), and up till now the camera still isn’t officially fixed. God knows what Canon Japan is up to. The amount of arrogance they’re displaying is simply astounding. As one user said:
However, for the last 18 months, Canon has displayed an arrogant, careless disregard for honesty and straight-forward practices. They have refused to replace cameras that are clearly defective (as demonstrated for them beyond doubt in my case, and in MANY other cases), and are churning out “L” lenses that ROUTINELY are WAY out of whack due to an apparent COMPLETE lack of testing.
Now on to the topic. From the date of release (late 2006) of the Canon EF 50mm f1.2 L lens, it has been known to have a rather serious backfocusing issue. Many users reported this around the Internet very early on. If you just search on Google, you’ll find loads of information and confirmation on this problem. From here:
back focus with center AF point at f/2.8-f/4
first the closing: I did return this lens.
At the store, I tried 2 other copies but those were form the same batch so I did not expect anything better.
I did try this lens on 2 different 1D series bodies FF and 1.3X (1Ds MKII and 1D MKII) and the result was the same: back focusing on distances 1-3 m. Now, I do have 3 other high speed EOS lenses (24/1.4, 35/1.4, 85/1.2II) and none of those have this problem. So the statement from the previous reviewer on “accuracy” of my cameras sensors does not stand. The only lens with problem was the 50/1.2!
Sad, but I can not keep this lens at this time. Even with excellent results when manual focusing, I can not keep 50/1.2 just because those rare times when I really need critical AF (full open) ability…
i will keep my canon service center story short but will say that i was very, very unhappy with how my problem was handled. lots of driving back and forth and, finally, the lens returned to me in the same condition as when i dropped it off.
finally, someone head my complaint and i brought the lens back to canon 1 final time.
HERE IS WHERE IT GETS INTERESTING:
1. the technician DID find that it backfocused.
2. he said that 4 CM was “within the spec” for the product.
i found this preposterous and asked him to have a crack at calibrating it.
in 20 minutes he was able to get the lens focusing properly on my 5D.
this 20 minutes was the culmination of 2 weeks of persistance.
– my copy of the lens did backfocus (i actually tried TWO copies: both with this problem)
– it was correctable.
– it was a terrible ordeal
– i have yet to determine if it was worth the time and aggravation.
this lens may yet go back to the store.
i am not a pixel-peeper and i rarely put my lenses through precise testing. this lens had GLARING issues, to my eye. perhaps my “review” will spare someone else a bit of grief.
I tend not to harshly evaluate lenses, as I know that no lens will ever be perfect. However, this lens has been a very large disappointment for me (in addition to a waste of time and money). I heard plenty of talk about focus issues with this lens and how it’s not much sharper than the EF 50mm f/1.4 USM lens. Still, I purchased it because I expected very good contrast and wide aperture effects (thin dof, pleasant bokeh).
I did experience the above characteristics, but the EF 50mm f/1.2L USM lens suffers from a design flaw: focus drift and backfocusing. The lack of a floating aspherical element means there will be depth of field aberrations when stopping the lens. That is, the depth of field will appear to shift back when narrowing the aperture.
I have evaluated four copies of the lens on two camera bodies (one old calibrated body and a relatively new body), and each copy exhibits backfocusing. Even after compensating for the backfocusing, the lens sharpness at the center of the image circle within the depth of field isn’t too good. I compared it against the 50mm f/1.4 USM lens at f/1.4 and f/8. While the 50mm f/1.2L is the clear winner at f/8 in terms of contrast and sharpness, at f/1.4, the 50mm f/1.4 lens actually wins out in sharpness. I couldn’t believe it so I exchanged the lens once in May. Same deal. When I decided to give the lens another chance in October, I was disappointed yet again.
People with a need for speed will buy the 50mm f/1.2L USM because of its strengths and in part due to a need for speed. But with severe issues in terms of the focusing and focus drift, I don’t think any price, no matter how reasonable, can justify this lens’ flaws. If you’re looking for a great f/1.2 lens, the EF 85mm f/1.2L II USM will do everything you ask of it. But as for the EF 50mm f/1.2L USM, buyer beware.
When it focuses on the subject, it’s sharp, provides excellent contrast and beautiful bokeh. It’s really well built and feels great on a 5D. When it misses the focus on the subject … which seems to be most of the time … it’s not sharp, and is intensely frustrating to use. All 4 copies I’ve used demonstrate consistently inaccurate auto-focus with varying amounts of backfocus at all apertures – plus focus shift at f/2 – f/4 for close-in subject distances. My calibrated 5D focuses just fine with 85/1.2L II, 100/2.8 USM Macro, 70-200/2.8L IS, etc…
I returned the lens because of repeatable mis-focusing/softness at about four feet, using an aperture of f2, that made purchase of the lens pointless. I have rated the lens at 7 but in truth I would not rate it at all-except as a paper weight(is that too mean?) Perhaps I was doing something wrong-but I don’t think I was – I am a very experienced photographer. I may test a different 50mm 1.2 in the future. But should one have to do quality control on such expensive optics?
And from here:
After my first shot with the 50 1.2L, I went WFT?!?!? It was seriously backfocusing… and I never experience that with my famed 85 1.2L lens.
On the 50/1.2 Canon decided to save a few bucks and over-simplified the mechanics of the focusing mechanism.)
And from here:
Ok, so I need some input. I just received my 50 1.2L and took some test shots. It has focus issues as described by many, many people.
I got a 50L. This one had something really wrong with it and all shots were fuzzy. I’ve never seen anything like it. It wasn’t OOF….it was just all fuzzy. I sent it back.
The second 50L I got consistently backfocussed about 4″ at all apertures. This was not the famous focus shift issue….just a straight up backfocus. I sent it back.
I went through two different 50Ls before returning to the 85L. Like Jeffrey said, they all have the focus shift, it’s just part of the lens.
I got rid of mine as well. Even using the alternate focus points didn’t seem to make much difference. I’m much happier with my 35L, 85L, and 50 1.4 than I was with the 50 1.2. I tried at least five and they all had the same issues. Really wanted the lens to be decent, but no such luck.
Surely, the chances of two copies being THAT bad are very slim, right?
I thought the same, but gave up after 5 or 6.
From the review here:
It was been mentioned by others that the 50 f/1.2 exhibits slight back-focusing at these maximum magnification, near-minimum focus distances. Since I don’t shoot with a lens like this at these distances very often, I had to go back and check mine. And my aditional tests confirmed a slight backfocusing at 10-20″ or so. The mis-focusing isn’t dramatic, but will have a slight negative effect on image sharpness at these distances.
William Castleman did some tests with this lens, and you can clearly see the issues here:
I concluded that the Jackson 2004 FocusTestChart detected real focus problems at the very short working distance (46cm). These misfocus problems are of questionable significance because I almost never try to produce wide aperture images at the shortest possible working distance to produce tack-sharp images. With the EF 50mm f/1.2L focus function was normal at longer working distances (129 and 258 cm).
Let me make this very clear: THIS LENS BACKFOCUSES / MISFOCUSES AT ALL DISTANCES AND ALL APERTURES AND THE PROBLEM IS SIGNIFICANT. At longer distances it is just harder to detect (but if you look for it, you find it), as well as smaller apertures because the focus plane is deeper. But the back focus problem remains. I have done my own tests which confirm this. The backfocusing happens only when using autofocus. Manual focus works OK.
If you’re wondering what this means in real world shooting, have a look at this picture:
At this size, it looks ok, but when you look at the actual pixels, you see that the focus is WAY off. I focused on the left eye, and the camera decided to focus around the ear instead, while still giving the focus confirmation on the left eye! Check the 100% crops below:
This is absolutely unacceptable behaviour. Imagine having to work with this lens, and find out later that the focus is off on almost all your pictures. You focus on the eyes, and you get the ears in focus instead. There’s a lot more on this here with some sample pictures of backfocusing. With this kind of performance, you can’t consider this lens for any kind of serious work. Just think of what will happen when you print such a picture large enough. Commercial work is out of the question. And with a price of about $1500 you have to wonder what the fuck they’re smoking at Canon HQ to expect to sell this to people. The only reason why they’re still selling it seems to be because they keep quiet about the issues and people buy it not knowing better, and finding out about these issues later.
Representatives from Canon were also sent information about this problem early 2007 if not much earlier. But up till today, Canon has still not officially commented on this problem, provided no fix, and simply pretends there’s nothing wrong. Meanwhile they are perfectly happy to keep selling a defective product to everyone without any kind of warning about the issues which they KNOW exist. The only comments about this lens were made unofficially by Chuck Westfall from Canon USA in emails to some users:
I asked chuck westfall for an update last week and this was his esponse:
“Technical discussions about the EF50mm f/1.2L USM are continuing to occur between Canon USA and Canon Inc.. The matter is not resolved yet, but it is most certainly not being ignored. Unfortunately, I can’t say much more about it until there is an official conclusion.”
If you look through all of the threads on the 50L you’ll find that a bunch of us wrote Chuck last May and got the response back that an announcement was expected from Canon in the near future. Follow-ups with Chuch haven’t gone anywhere. Here was his response:
“Canon Inc. is definitely aware of the claims concerning the AF accuracy of the EF50/1.2L USM. They’ve been investigating for the past couple of months, and I am told that there will be some kind of public announcement forthcoming in the near future. Stay tuned, and thanks for using Canon equipment!” – email from Chuck Westfall 5/24/2007
Since we’re now in July 2008, and that email from Westfall was in May 2007, I have to wonder what Canon’s definition of the “near future” is. And again, meanwhile they are perfectly happy to keep selling a defective product to everyone without any kind of warning about the issues which they KNOW exist. They even thank you for using their defective equipment.
At this point I have to ask myself, who the fuck tests these products at Canon? What kind of engineers do they have working there? When I got my copy of the 50mm f1.2 lens, I noticed within a day of using it that it was back focusing. And this was without doing real focus tests, just from normal use. How in god’s name, does Canon manage not to notice this big problem, and proceed to manufacture the lens? Seriously, how?? Is there even some kind of quality control at Canon? It’s the same with the 1D Mark III autofocus issues. Who the fuck tested that camera? How did they manage to release it with an obvious defect in the autofocus system, one which they still can’t seem to fix after a year, even after a few attempts?
And what is even more astounding, is that these issues aren’t small issues. They are CRITICAL issues. If you can’t take a picture that is in focus, that should be a MAJOR showstopper. Focusing well is one of the most important things that this equipment should be able to do. What is the use of having a lens that can’t focus well?? It’s practically unusable! If this is Canon’s idea of delivering good image quality, they might as well rub some vaseline on the sensor of every body they release from now on.
This is what Canon says about this lens:
The EF 50mm f/1.2L USM is suitable for any shooting situation; its lens coating and construction are optimized to minimize the ghosting and flare that frequently occurs when lenses are used with digital cameras. This high-performance, weather-resistant lens delivers all the superb image resolution and contrast you expect in a Canon L Series Lens.
What is even more incredible, is that Canon seems to get away with this kind of behaviour as well. I look at this and keep asking myself how it is that companies seem to get away with such behaviour these days:
Another thing that has me amazed, is that I cannot understand how consumers seem to be so tolerant these days. It seems companies can do to consumers what they want and even rip them off, like Microsoft is now doing, without them even saying anything about it. Such a move from Microsoft should at least have sparked some major PR issues for them if not lawsuits. But consumers these days appear to have a very high level of tolerance. I recently also wrote about the bad quality control at Canon, with regard to their DSLR products. And there you see a similar problem, consumers are having issues with the products not working well, even out of the box, and many are complaining, but it looks like they just accept it for some reason as being normal. When has it become normal for a company to massively screw so many customers on such a large scale, as Microsoft is doing? When has it become normal to buy a DSLR and finding out that it does not work (well) as soon as you power it on? When has it become normal to buy a lens for your DSLR and finding out it only takes soft pictures and doesn’t focus well?
Indeed, when? And why wasn’t I informed? Did I miss a memo or something?
It looks like Canon might have taken their position as market leader for granted, and got sloppy for a few years now. With Nikon coming back into the game, it’ll be interesting to see how Canon will behave from now on. Many pros are switching to Nikon these days, and black Nikon lenses are replacing the white Canon lenses (image from Tour de France 2008 from here) everywhere. With their DSLR division operating at DEFCON 1 right now, maybe they’ll show some improvement. Meanwhile, their customers are still being screwed around the globe, and trust me, they’re not using any kind of lube.