New version 50mm L f1.2??

joe mama wrote in post #7406227
I've always found complaints about the 50 / 1.2L's backfocusing when at close focusing using the center AF stopped down to be a bit odd. Do so many people do that with a fast 50mm? I described the issue here:

http://forums.dpreview​.com …rum=1029&messag​e=25117450

Of course, it would be nice if Canon offered a firmware solution for this issue so as not to adversely affect other properties of the lens. Still, how many people use a fast 50mm prime in the prescribed manner so that the issue makes itself so much of a problem that they denigrade the lens on this one minor point?

These are the types of pics I take with the lens:

http://www.pbase.com/j​oemama/50l_gallery

and it just isn't an issue for me.

I suggested something similar and then people explained to me they -cant- because its an optical issue relating to how the lens is constructed and not something a simple software fix can take care of..

I agree it would be nice to have it updated with a floating element, but id rather have a update to the 50 prime that gets close to the .....I cant think of a good nickname for the Sigma 50 f/1.4

(Figma?)

KenjiS wrote in post #7435773
I suggested something similar and then people explained to me they -cant- because its an optical issue relating to how the lens is constructed and not something a simple software fix can take care of..

I agree it would be nice to have it updated with a floating element, but id rather have a update to the 50 prime that gets close to the .....I cant think of a good nickname for the Sigma 50 f/1.4

(Figma?)

In fact, it is a simple software fix. The focus shift is predictable given the aperture, subject distance, and focal point used is known. Since the 50 / 1.2L reports the aperture and subject distance, and the body knows the focal point used, the camera AF algorithms can then simply be programmed to adjust to compensate, wherease adding a floating element adds expense, size, and weight.

By the way, it is my understanding that the Sigma 50 / 1.4 "suffers" from the same focus shift as the 50 / 1.2L.

joe mama wrote in post #7435900
In fact, it is a simple software fix. The focus shift is predictable given the aperture, subject distance, and focal point used is known. Since the 50 / 1.2L reports the aperture and subject distance, and the body knows the focal point used, the camera AF algorithms can then simply be programmed to adjust to compensate, wherease adding a floating element adds expense, size, and weight.

By the way, it is my understanding that the Sigma 50 / 1.4 "suffers" from the same focus shift as the 50 / 1.2L.

It must not be that simple of a software fix since the EF 85 f1.2 II uses a floating element design.
If this issue could have been alleviated by simply programing AF routines in the lens based upon aperture and distance, Canon surely would have saved the additional material and design complexities of the 85 f1.2 II by implementing such software AF corrections in this lens (instead of the more expensive floating element and electronic manual focus.

It's a design limitation...is what I heard directly from Canon when I was battling it out with them with my 50L.

joe mama wrote in post #7435900
In fact, it is a simple software fix. The focus shift is predictable given the aperture, subject distance, and focal point used is known. Since the 50 / 1.2L reports the aperture and subject distance, and the body knows the focal point used, the camera AF algorithms can then simply be programmed to adjust to compensate, wherease adding a floating element adds expense, size, and weight.

By the way, it is my understanding that the Sigma 50 / 1.4 "suffers" from the same focus shift as the 50 / 1.2L.

joe mama wrote in post #7406227
I've always found complaints about the 50 / 1.2L's backfocusing when at close focusing using the center AF stopped down to be a bit odd. Do so many people do that with a fast 50mm? I described the issue here:

http://forums.dpreview​.com …rum=1029&messag​e=25117450

Of course, it would be nice if Canon offered a firmware solution for this issue so as not to adversely affect other properties of the lens. Still, how many people use a fast 50mm prime in the prescribed manner so that the issue makes itself so much of a problem that they denigrade the lens on this one minor point?

These are the types of pics I take with the lens:

http://www.pbase.com/j​oemama/50l_gallery

and it just isn't an issue for me.

I like the photos in your gallery and help me to see this lens in a better light. Thanks for sharing the link.

By redesigning the 85 / 1.2L II without a rear element, they would have introduced the backfocus issue into older bodies that did not have the new AF algorithms. Since the 50 / 1.2L never had a floting rear element, there is no reason not to correct this design limitation by a new AF algorithm.

Yes. But this limitation can be overcome via software. Of course, perhaps that would be asking too much from the company that designed the 1DIII AF algorithms? : )

Kind of you to say!

It can't when you've got so many copies with various degrees of FF/BB and Focus shifting.
The 5 copies that I have either owned or tried shifted differently at any given aperture...so how do you go about fixing for something like that?

They've got certain things that they do (electronic adjustments, adjusting the PCB, etc.) but none of them are solid fixes.

They know about the problem. They chalk it up to product limitation. They won't say that publicly, but at least I got them to admit it to my face.

It is what it is...

joe mama wrote in post #7438792
Yes. But this limitation can be overcome via software. Of course, perhaps that would be asking too much from the company that designed the 1DIII AF algorithms? : )
Kind of you to say!

It is definitely a limitation of the design, that's why it's not considered a "flaw". In fairness, fixing it would almost certainly involve throwing in a floating element that would just increase the minimum focusing distance, which would eliminate focus issues close to MFD on the current 50 1.2L....but what would be the point?

nicksan wrote in post #7441668
It can't when you've got so many copies with various degrees of FF/BB and Focus shifting.
The 5 copies that I have either owned or tried shifted differently at any given aperture...so how do you go about fixing for something like that?

That's a different issue -- that's Canon's infamous QC at work, not the lens design.

The design limitation is the systematic backfocus, the other problems are poor QC, inaccurate AF at such shallow DOFs due to the body's AF algorithms, or user error (e.g. handheld shots wide open at close focusing distances where the DOF is 1/4 inch, and even the slightest motion throws the focal point outside the DOF).

But it is fixable through software.

Right...but the problem once again is that it's a Product Limitation, software fixable or not, according to Canon, and that's all that matters isn't it? This is directly from the horse's mouth. I got them to at least admit to me there is nothing they can or are willing to do about it.

So might as well not talk about fantasies...they ain't gonna do it.

joe mama wrote in post #7442211
That's a different issue -- that's Canon's infamous QC at work, not the lens design.



The design limitation is the systematic backfocus, the other problems are poor QC, inaccurate AF at such shallow DOFs due to the body's AF algorithms, or user error (e.g. handheld shots wide open at close focusing distances where the DOF is 1/4 inch, and even the slightest motion throws the focal point outside the DOF).



But it is fixable through software.

nicksan wrote in post #7442977
Right...but the problem once again is that it's a Product Limitation, software fixable or not, according to Canon, and that's all that matters isn't it? This is directly from the horse's mouth. I got them to at least admit to me there is nothing they can or are willing to do about it.

Well, my interpretation of Canon's acknowledgement that it is a "product limitation" simply means they are not going to recall the lens to add floating rear elements. Surprise, surprise. : )

The viability of a software solution to the limitation is not a fantasy. In fact, it is less a fantasy than vignetting correction and PF removal. But the implementation of a software solution may well be a fantasy, given the 1DIII AF fiasco.

In any case, as I discussed and demonstrated earlier in this thread, it is a limitation that so rarely comes up for me, that I am not losing any sleep over it. Do that many people really use a 50 / 1.2L for stopped down close focusing so much that using an off-center AF point on those occasions is an issue? That would be very surprising to me. Then again, it's surprising to me how many talk about soft corners with lenses wide open, given that the corners will rarely be in the DOF at such wide apertures.

I don't really have anything to add to the discussion you guys are having, but I wanna say that the 50L is just incredible.

perryge wrote in post #7443185
I don't really have anything to add to the discussion you guys are having, but I wanna say that the 50L is just incredible.

I agree. Why fix it if it ain't broke

*ducks and runs for cover*

they need to bring back the 1L

paddycook wrote in post #7443242
I agree. Why fix it if it ain't broke

I agree! And I'm the OP! (But now that I have the lens and have had a few weeks of experience with it...it turns out it's great!)