I really appreciate your testing of this lens! And I also agree, that one should use the lens in real world scenarios. I don't advocate excessive testing except as necessary if issues are found.
But could you do a test for me? One of my typical lengths for use with this lens would be for close portraits and candids of about 30"-36" or 75cm to 90cm. Your test results show 18" and 51" but nothing in between. Would you test at 30"-36" and report your findings and also can you tell us at which point over 18" that the 50L begins to LOSE the backfocus tendancy?
This way, perhaps we would know at which length that AF might be fully dependable.
William, I appreciate the reply and your testing. Nicely done on all of it.
There is no doubt in my mind my lens backfocused. In fact I tried real world testing and truth be told I've never shot a ruler or a chart in my life, because I feel real world is how the lens get used, so use it like that and see.
I then began noticing some OOF shots and really didn't put 2 and 2 together until Sam posted his examples. I then immediately thought back to some similar shots I had taken where it did the same thing and how I also missed the ring shot at a wedding ceremony I shot last weekend (even though it confirmed focusing and I was shooting at 2.8 or 3.2, so depth of field should have been ok).
Any ideas on why the lens would have improved the way it did when setting the focus points to only 11 points and 9 points as opposed to 45 (btw, even with 45 point on, I was using mainly the center point)?
You are all lucky people from the Nikon camp aren't reading this.
We are discussing why its OK that a $1600 prime lens doesn't focus properly at close distances.
Could you see a retail salesperson explaining this to a potential customer? "Well here we have the $3000 1D and the $1600 50L, but don't use this $5000 setup to shoot close portraits! The will be OOF. Tell your people to stand back a bit!"
CMOS, you've repeated this several times and I am really trying to better understand this....
"We are discussing why its OK that a $1600 prime lens doesn't focus properly at close distances"
Yet, I don't see one person here that thinks it's OK... Hence why several of us have already sent their lenses back.... Anyone that seems to be keeping it seems to not have the problems the few of us have with backfocusing.
Heck, I WAS and STILL AM a huge proponent of this lens, even though mine is getting ready to go back on a UPS truck back to the dealer.
i don't find the close-focus issue "ok", either, but feel worlds better about it knowing that the test results are the same on my lens as they are on william's. i now understand this lens' limitations more so than i did less than a week ago when i bought it -- and that's a step in the right direction. this lens isn't perfect and there still could be some issues but if real-world performance proves itself then that's all i care about.
the nikon camp (or anyone else for that matter) can say or think whatever they want. their eyeballs aren't behind my viewfinder when i take a photo or answering my phone when a client calls.
There is absolutely no doubt in mind that my lens is backfocusing terribly. I'm not a fan of test charts and only use them if I see a problem in my practical shooting. The issue is clear in every condition I use it in.
Since the 24-105 f/4 IS "photon torpedo" days I have not seen such varied feedback by the users of a new lens, including W.Castleman's most recent 50 f/1.4 resolution comparison graph (I have also seen the one previously published on his website) as well as another opinion stating that the 50 f/1.2 sharpness is comparable to the 200 f/1.8. All that in itself IMO does not auger well.
Neither does the fact that the Canadian lens deliveries have been effectively pushed into the end of January or even February (?), compared to the initial release date of Nov.2006.
Although based on what I have seen here and elsewhere I have lost much of my genuine interest in the lens itself, I am becoming intrigued and thus inclined now to get a copy as soon as the lens become available in Toronto in order to satisfy my own curiosity first hand.
From all the complaints of QC control and lacking funds, I cancelled my order for the 50L for now. I'll wait a year and a bit for the price to go down and hopefully the quality control will be fixed by then.
braxus wrote:
From all the complaints of QC control and lacking funds, I cancelled my order for the 50L for now. I'll wait a year and a bit for the price to go down and hopefully the quality control will be fixed by then.
Sadly, it may not be as wide-spread as this thread makes it out to be. You should've kept your order and tested it for yourself. I can't imagine the number of orders that have been potentially cancelled because of this thread. Good copies could have landed in the hands of members who would've have been able to confirm or deny back-focusing issues.
I have learned a lot about the EF 50mm 1.2L from reading this thread and testing some of the issues that were of concern with my lens. I am confident that I have a lens that will perform well under the conditions in which I use it.
One of the best ways to determine if a lens is misautofocusing under the conditions of interest is to shoot a target (resolution chart, news print, mannequin, etc) at the working distance you use and then compare image sharpness with autofocus and compare it with image sharpness you get when you manual focus the lens using a 2.5x magnifying angle finder. If your lens is misautofocusing, this should detect it. Run at least 3 replicate tests at each f/stop for manual and autofocus. All autofocus cameras yield autofocus variation. You will find that your "perfect" manual focus efforts vary.
Using the Jackson FocusTestChart will over estimate misautofocusing in my experience.
I have been a little surprised by some of the comments in this thread .... but empathize with those who are passionate about their expensive gear.
I am going to bow out of the discussion until next Monday or Sunday when I will have photos from a gymnastics meet at the University of Georgia where I will shoot the 50 1.2L under severe stadium backlighting condtions.
wcastleman wrote:
One of the best ways to determine if a lens is misautofocusing under the conditions of interest is to shoot a target (resolution chart, news print, mannequin, etc) at the working distance you use and then compare image sharpness with autofocus and compare it with image sharpness you get when you manual focus the lens using a 2.5x magnifying angle finder. If your lens is misautofocusing, this should detect it. Run at least 3 replicate tests at each f/stop for manual and autofocus. All autofocus cameras yield autofocus variation. You will find that your "perfect" manual focus efforts vary.
Are you using DoF preview to stop the lens down when manually focusing? If not, comparing "sharpness" isn't really doing anything due to the problems of focus shift that were discussed earlier in the thread. AF and manual focus would yield the same level of misfocus without stopping down for manual focus, yielding (erroneously) identical sharpness. And of course, identical sharpness does not mean that the lens is focusing "correctly".
Edited by Sam Bennett on Jan 23, 2007 at 02:08 PM GMT
Possible explaination- I looked at wcastleman's auto focus vs manual focus tests- is it possible the AF is still driving the lens during exposure? Perhaps things get locked in properly when composing, but when the mirror flips up, and the AF sensors are suddenly "in the dark", they get confused, and send spurrious signals to the lens motor, causing it to hiccough enough to throw it out of focus during actual the exposing? Another explanation, maybe some lenses do not have enough "electronic brakeing" of the focus motors, so they are always wiggling about a bit, causing lens to be out of focus in critical applications?
jjlphoto wrote:
Possible explaination- I looked at wcastleman's auto focus vs manual focus tests- is it possible the AF is still driving the lens during exposure? Perhaps things get locked in properly when composing, but when the mirror flips up, and the AF sensors are suddenly "in the dark", they get confused, and send spurrious signals to the lens motor, causing it to hiccough enough to throw it out of focus during actual the exposing? Another explanation, maybe some lenses do not have enough "electronic brakeing" of the focus motors, so they are always wiggling about a bit, causing lens to be out of focus in critical applications?...Show more →
John, you make a point that I didn't mention before but that I might have experienced last night while testing with the little tag on the ball. The lens had a hard time focusing sometimes and it seems like the lens was shifting ever so slightly as the shot was being taken (I was shooting in One Shot mode). Either way, water under the bridge now. I wish I could have held onto it longer and knew about the backfocusing sooner and between all of us here we seem to be coming to some conclusions with an explanation as to what is really happening. Perhaps Canon can take this data and figure out whats what.
For what it's worth, I've gone ahead and ordered a Canon Ec-B Focusing Screen for my Mark II to take AF and camera-assisted manual focus out of the equation. I'm doing this for the following reasons:
To try manual focusing without the camera's aid at all.
To manually focus at any aperture I like (using the DoF preview button)
To make a comparison between AF, camera-assisted (wide-open) manual focus, and stopped-down manual focusing results.
This should establish conclusively whether the topic of "focus shift" is relevant to this discussion.
i've got a third 50L on the way and should see it tomorrow. it'll be interesting to see what happens when comparing it side-by-side with my current copy.
btw, i haven't seen any focus shift with this copy. 11-point AF has fixed any focusing issues i had before. nothing crazy has been happening with the focus point stopped down using MF or AF.
good luck, sam. if you can't get it worked out then i'd be willing to sell you one of my two copies at my cost if requested.
Maybe this was mentioned several pages ago, but I can't recall (would be nice if these forums allowed a single page view!) ...
I'm assuming (quite sure) when people here refer to AF, you're referring to One-Shot mode. I was wondering whether it was worth trying AI Focus to see if that has an impact (had the thought a few minutes jjlphoto's post four posts above)? I know in theory it shouldn't make a difference when shooting test charts and adds extra variables to scientific testing (and not all of them can be controlled by the user!) - but simply wondering if results seem to be the same or different. If the focus is being driven during the black-out period, then AiF algorithms might just have an impact (for better or worse)...
Sam Bennett wrote:
Are you using DoF preview to stop the lens down when manually focusing? If not, comparing "sharpness" isn't really doing anything due to the problems of focus shift that were discussed earlier in the thread. AF and manual focus would yield the same level of misfocus without stopping down for manual focus, yielding (erroneously) identical sharpness. And of course, identical sharpness does not mean that the lens is focusing "correctly".
Edited by Sam Bennett on Jan 23, 2007 at 02:08 PM GMT
DoF preview wasn't used. The photos and data on the page I cite show that "focus shift" isn't an issue.
Sorry, I didn't see this earlier. I really am out of here until the weekend. Cheers
Thanks for the clarification, William. I've been skimming lately since I'm slammed at my real job, and so I missed out on your more lengthy explanation - sorry about that.
