mbailey wrote:
Here is a manually focus shot from 50L at f/2.8. This was not posted earlier
That looks great! Is it a 100% crop?
Edit: Saw your next post, now. Yes, I agree, the shot from the 50L is a little sharper. I think this is conclusive... when they get the focus problem fixed, the 50L will be a superior lens all around. I'm sure they will fix this.
Edited by RoyPertchi on Feb 18, 2007 at 11:43 PM GMT
Well well welll... Now we see that the 50L is sharper at 2.8! And probably up through the rest of the f/stops as well. Good job, Mike. Looks pin sharp to me.
But now it is a common belief that the 50 1.4 and 1.8 is sharper at the f/stops where the 50L suffers from backfocusing. We need to relook at everything.
ghuff wrote:
That's what my initial testing showed in early December, and why I have been waiting around for Canon to fix the lens.
I guess everyone has to see it for themselves...
You did? Man, that had to be like 150 pages back or something. Most of us probably forgot you posted that.
And most of us do have to see for ourselves. I was convinced that I had a good copy and you guys were just paranoid. Well, add me to list of paranoids.
lord_malone,
I think you were right to test for yourself. What I meant was everyone had to see that the lens was sharp when focused manually. And yes it was probably on about page 40 something...
It was sharp enough that I sent my first one back 3 times to Canon trying to get them to fix it where it would AF as well as it would manually focus...
I was impressed and really wanted it to work right.
One of these shots was with the old 50L f/1.0...so don't confuse it....it is clearly labeled....I included it in the series for comparison.
But this sequence was essentially manually focused...and shows that the lens was fairly sharp through all apertures....I focused once and left AF off for all consecutive shots..
This was my first lens....distance from the wall was 10 feet.
ghuff wrote:
lord_malone,
I think you were right to test for yourself. What I meant was everyone had to see that the lens was sharp when focused manually. And yes it was probably on about page 40 something...
It was sharp enough that I sent my first one back 3 times to Canon trying to get them to fix it where it would AF as well as it would manually focus...
I was impressed and really wanted it to work right.
One of these shots was with the old 50L...so don't confuse it....it is clearly labeled....
But this sequence was manually focused...and shows that the lens is sharp through all apertures....
I'm just glad that Mike was able to confirm what I began to suspect to be true. After reading this thread starting from day one to now, I just had this nagging feeling that I couldn't shake and didn't have confidence in my own lens testing. When I get my lens back from Mike I'm going to be getting in touch with Canon. Spread the word!
Edited by lord_malone on Feb 19, 2007 at 09:06 AM GMT
I've been following this thread for some time now and have confirmed my lens (UU1100, ser# 694xxx) behaves as has been described by many here (on my XTi, someone ask about a 30D and the XTi has the same AF sensor). I too have found that the problem is often missed in real world shots. The backfocus of my lens isn't as dramatic as some I've seen posted and disappears pretty much with manual focus (focus at f/1.2 and then take a series in MF without changing the focus). Here is the f/2.8 shot in a manual focus series (taken at about 2 feet):
Well mine was one of the UU1100 and I returned it. It displayed very irregular focusing issues. Sometimes it looks ok and sometime it back focus very badly. I returned to my dealer and he did their own test and confirmed my issues. I decided not to get a replacement.
Here is a sample. I focus on my son's left eye and see how sharp his right eye is. Would you believe it. I have confirmed the focus point as well so I did not make any mistake about which eye I focussed on. I was totally shocked when I saw this result. I really didn't want to return the lens as I keep convincing myself it was my focusing error until I did more test to display this inconsistancy
Well, I think we've all already have come to the conclusion that there is an issue and Canon needs to fix it. A can see a firmware update happening in the near future. However, there no denying that when this lens is on, optically it's a solid performer.
tmr4, the XTi, as well as the 30D, are both the most recent incarnations of their respective bodies and may have the same software anomally as the 5D and 1 series bodies. The more variables we throw into the mix, the less/more we seem to isolate the problem, or create more unanswered questions.
Edited by lord_malone on Feb 19, 2007 at 11:00 AM GMT
lord_malone wrote:
...
tmr4, the XTi, as well as the 30D, are both the most recent incarnations of their respective bodies and may have the same software anomally as the 5D and 1 series bodies. The more variables we throw into the mix, the we seem to isolate the problem, or create more unanswered questions.
Hey, I just remembered that I still have my D60 sitting around unused!
Should I test the 50L on it?
Why not, all I need is time!
Canon is certainly well aware by now. I'd almost suggest waiting for an official statement from Canon before you bang down their doors, but then again, the more emails they get the better it is for us. I wrote Chuck again on Thursday afternoon and he informed me testing is still going on at Canon Japan and had no official word.
Just for kicks and perhaps this is a novel idea, but has anyone tested this lens on a film camera? I might just pull out the old Rebel Xs today to try it. Worse comes to worse, I've wasted a roll of film.
I did a few quick shots with the D60, but was only reminded:
1. ISO 1000 is the limit
2. ISO 800 is noisy, to the extent that resolution is reduced
3. AF in dim light can't compare to the 1D, unless the assist light is enabled
With the AF assist, it seemed accurate at 1.4, but that's not where we have problems anyway.
If I get a chance today, I'll try the chart and include a range of apertures.
Ed, good idea - maybe I'll run a roll of film through the EOS 3. Who knows when I'll get to scan it, though...
If Canon is running tests, I don't see how they could fail to identify the problem and quantify it better than any of us could. If a firmware fix is a possible solution, it ought to be on the way!
mbailey wrote:
6. In real world shooting I get a surprising number of keepers given the obvious systematic problem with the lens. Are there so many variables in a real world shooting that we can't control that this focusing problem gets lost part of the time?
This is pretty much my assessment of the 50 1.2. I spent over an hour in the store testing it with one of my 1DIIN bodies and saw the same back focus problems discussed throughout this thread. At very close distances, less than one meter, working under very controlled parameters, AF results indicated back focusing. MF on one point and varying the f/stop from 1.2 to 5.6 showed a shift in the plane of focus progressively farther behind the initial point of MF. (BTW, the camera was tethered to a laptop and images were examined on the computer, not the camera's LCD). This was discouraging to see.
However, after analyzing numerous spontaneous images made around the store, including relatively tight head shots, I found AF was nearly always spot on. For this reason I decided to take the lens as is and try to learn its quirks.
I've had it now for almost a month and find AF results at 1.5-2m and farther are nearly always more accurate than my attempts to MF. At closer distances I haven't really noticed AF inaccuracy, but at those distances it's also a lot easier to accurately MF, which I tend to do out of habit carried over from the 85 1.2 (I find it's usually much faster & easier to MF off centre subjects than focus & recompose... I absolutely do not trust the off centre AF points with fast lenses). At the minimum focusing distance though, it is clearly evident in the viewfinder that AF back focuses - and also evident in final images. Here I'll use MF almost always.
My lens is UU1200 with ser# 714xxx.
BTW, I've been using the Ec-S screen now for a year and have found it extremely invaluable with the 85 1.2. It also seems to work well with the 50 1.2. I would highly recommend it for these fast lenses, though I will add that I personally don't find much difference in focusing acuity between it and the standard screen with 2.8 lenses and slower. As well, those unfamiliar with the "S" screen should be warned that it is somewhat darker than the standard screen, particularly with lenses slower than 2.8. Some may find it to be intolerably dark, though I still use it with f/4 lenses.
I use the Ec-S on my 1D2N. That may account for why I seem to have more hits than misses, and the illusion that my lens didn't suffer from this problem. My 1D has the standard OEM screen, and I noticed more misfocusing when used on that body.
tkjaer wrote: Just received my copy today (after about 2˝ months of wait). I too see AF problems with my lens.
Hi Thomas, Sorry your lens has issues also, and thanks for including some test shots so that we can compare and see the issue.
I don't know how long Canon will continue to test and / or ignore the issue with it seeming to be so widespread. Hopefully they will announce a direction soon.
Just curious, what is your date code and first 3 digits of your serial #?
Most of us probably forgot you posted that. 
