Galibier, one way of finding out if the focus is good, is taking pictures in RAW and then opening them in ZoomBrowserEx. Once you have opened the CR2 file, you can select "View - Show Auto Focus POints" from the menu. You can then see where the camera auto focused, and then look if the focus in the picture is really there. For this test you should also take pictures in manual mode and manually select a specific focus point to auto focus on.
Try it on near and distant subjects, all at f1.2 for max dept of field so you will immediately notice if the focus is off. Use a tripod and remote control. You could put a ruler on a wall, and take a picture of it from a 45 degree angle, auto focusing on a number in the middle and see what you get. Before taking each shot, make sure you turn the focus ring all the way so that nothing is in focus and when you press the shutter button, the camera is then forced to focus the lens. Immediately after the first shot, take a second and third without changing anything, and compare those 3 shots.
It is amazing that this lens brings so many contradictory opinions. Interestingly its rating on FM is 8.3 (less than many cheap zooms). I had great hopes for this lens last year and I am disappointed that it did not match 85 f/1.2 L or 35 f/1.4 L as far as quality of design is concerned
There definitely seem to be umm, "variances" out there... But to be honest, mine has worked well for me. It focuses on what I want and delivers images that are very sharp, contrasty and have saturated colors. The bokeh is excellent.
I don't know about all this "danger zone" stuff. I've pushed the lens into this "zone" and didn't see any issue. But then with f/1.2 lenses I either tend to shoot at/near wide open - or stopped down a good ways to the sweet spot. But f/2-2.8 is quite common.
In the interest of full disclosure, I got the lens used and the previous owner had sent it to be calilbrated by Canon in January of 2008 - it's a UV07xx date code. Used on a 1D2N.
There's talk about there being a "silent fix" via software on calibrated lenses or those with a UV07xx or later date code. There also seems to be a bunch of lenses exhibiting backfocus out-of-the-box, requiring calibration (common Canon QA issue). Two different issues technically, but perhaps related in the user experience and reports or just confused.
So what can I say? *I* love *my* copy. Maybe it's my lens. Maybe it's my technique. Who knows. What I can say is that I don't spend my days shooting brick walls, rulers, batteries and test charts and viewing the resulting images at 100% for hours. That makes for very boring photos. But I do know when a lens nails focus and delivers a sharp image.
Edited on Jul 17, 2008 at 09:59 AM
Jul 17, 2008 at 09:57 AM
Lars Johnsson Offline • • • • • • • Upload & Sell: Off
I got mine a couple days ago, new from Adorama its a UW06 date code. So far it seems great, although I haven't done any extensive testing.
I was curious and took a picture of a static object, with the camera in a static position. I started at f/1.2 and ran through to f/5. Sharpness is good at f/1.2 and doesn't change appreciably though the aperture range. So that's good enough for me.
Now I'm not saying everyone that claims they are satisfied with their lens has a defective one. Who knows they are maybe exceptions or in the case of newer lenses even silently fixed. But it wouldn't surprise me if some of the people claiming to have a good copy actually have a bad one that just doesn't give them problems for what they are using it for. For example, Castleman was shooting sports with his and was satisfied, but he confirmed later that his lens was backfocusing.
^ A valid point, kkdd. It could very well be the way the lens is used by people both in technique and/or subject matter.
I only have 41 pics (out of quite a lot more) online, shot with the 50L - but they cover a variety of situations from macro to far and have to say, so far no complaints:
FWIW, I might add that I rarely use the center AF point. As with any bright prime, and in general - I select the AF point according to subject placement in the frame. I use Single Shot AF.
Jul 17, 2008 at 01:31 PM
Lars Johnsson Offline • • • • • • • Upload & Sell: Off
Now I'm not saying everyone that claims they are satisfied with their lens has a defective one. Who knows they are maybe exceptions or in the case of newer lenses even silently fixed. But it wouldn't surprise me if some of the people claiming to have a good copy actually have a bad one that just doesn't give them problems for what they are using it for. For example, Castleman was shooting sports with his and was satisfied, but he confirmed later that his lens was backfocusing....Show more →
He was able to duplicate backfocus at 18 inches focus distance after trying hard
Carnathan's and Castleman's observations appear to be the same as mine: they can detect focus shift at very close distances if they test for it, but the focus shift doesn't affect their real-world shooting. Note also that Castleman detected focus errors in the two other lenses he subjected to close-focus testing, his 50/1.4 and 85/1.8 -- lenses he used (and continues to use) without issue. Do these test results make the lenses "defective" or "bad" even when the lenses work well in real-world shooting? I do not doubt those of you who have had bad experiences with the 50/1.2, but if my lens works well for me as I use it, I care very little whether testing for flaws reveals a flaw.
Is all this realy a surprise to Canon? No I think not...every sub 85mm L prime since the late 70's has had a floating element. In fact Canon like to brag about their superior "floating" designs in Lens works III. The problem seems to have originate from from the 50 f1.2L ancient optical design. It's a re-cobbled 55mm f1.2 from the early 70's. Which was pre-floating element design. Canon have wrapped a pro build, modern AF, weather sealed structure around some really old optics.
In fact a lot of Canon's 50mm are longer, nearer 55 and 60mm.
Jul 18, 2008 at 02:47 PM
Lars Johnsson Offline • • • • • • • Upload & Sell: Off
gazzajagman wrote:
Is all this realy a surprise to Canon? No I think not...every sub 85mm L prime since the late 70's has had a floating element. In fact Canon like to brag about their superior "floating" designs in Lens works III. The problem seems to have originate from from the 50 f1.2L ancient optical design. It's a re-cobbled 55mm f1.2 from the early 70's. Which was pre-floating element design. Canon have wrapped a pro build, modern AF, weather sealed structure around some really old optics.
In fact a lot of Canon's 50mm are longer, nearer 55 and 60mm.
So tell us a few of those canon 50mm lenses that are 55 and 60mm......wich ones are they
gazzajagman wrote:
The 50mm f1.4 and 50mm f1.2. Compare them to the Sigma 50mm f1.4, there's a noticable jump in focal length.
The Sigma may be "short". I'm sure there's fractional variation in focal length from one sample to another, but I recall a Canon 50/1.4 testing out at just about 51mm, nowhere near 55mm.
Have any manufacturers produced a 50 - other than a macro - with floating elements? Canon introduced it in the WA lenses because it allowed wide aperture designs without excessive focus shift. I'm sure a floating element could improve the design of the 50/1.2, but it certainly isn't needed as much as it is in the 24/1.4.
I'd still welcome a firmware solution in the form of slight focus adjustment based on aperture and subject distance.
Edited on Jul 18, 2008 at 10:44 PM
Jul 18, 2008 at 10:37 PM
Lars Johnsson Offline • • • • • • • Upload & Sell: Off
gazzajagman wrote:
The 50mm f1.4 and 50mm f1.2. Compare them to the Sigma 50mm f1.4, there's a noticable jump in focal length.
No they are not 55-60mm as you wrote. I know the exact (effective focal lenght) on all Canon 50mm lenses except the 50/1,2. But I own the 50/1,2 and haven't noticed any big difference from the others. Here are the numbers. And most lenses from all brands are not exactly 50, 100,300 or 500mm like the numbers on the lens say. Zoom lenses are a lot worse also.
