Thinking about the 50mm Sonnar - but those f/1.5 vs. f/2.8 threads make my head hurt!

[ferider]

I don't think you'll be able to identify if you have a 1.5 or 2.8 version by serial number or anything else but test photos.

John, the C-Sonnar RF cam has 3 visible segments and 120 degrees focus throw. I'm convinced Zeiss just rotates the cam by 120 degrees to get either 1.5 or 2.8 optimization. I would wildly guess the third position is for an f5.6 optimization that nobody ever asked for. Some classic Sonnar variants are optimized for 5.6, like the Canon 50/1.5.

For me personally the shift mattered, but I found it easily correctable on the M3 using the DOF marks in the finder. Here is the difference close up:

Cheers,

Roland.

 

[tbarker13]

Roland,

That amount of shift looks similar to what I found with my copy that was optimized for f/2.8.
I'm guessing the shift is much less noticeable when shooting at things further away - which may be why some people say they don't detect a shift.

I use my 50 primarily for portraits and the like - shot fairly close. Obviously, it's possible to make a mental calculation to adust for the shift. But for me personally, I would prefer to be able to trust the focusing, as seen through the viewfinder.

 

[Sanders McNew]

This one is at best of academic interest
to me as I don't have the lens. But if I
were choosing, I would prefer the lens
optimized for use at f/2.8 for a couple of
reasons. First, most lenses resolve much
better when stopped down a couple of
stops -- that puts the sweet spot of this
50/1.5 (assuming it acts like most other
lenses) around f/2.8. Second, if you are
shooting people, f/1.5 is such a thin DOF
that you will have a hard time hitting
your mark with the lens -- especially if
shooting in close, with a center-spot
rangefinder -- and keeping both eyes
in focus, if you care about that sort of
thing. I do have a Nikkor 50/1.4 Sonnar
and I almost never shoot it wider than
f/2.8 for these reasons.

 

[Steve Bellayr]

I photograph with this lens at the widest apertures from a distances greater than minimum focal distance and have not noticed any shift.

 

[noimmunity]

I use my 50 primarily for portraits and the like - shot fairly close. Obviously, it's possible to make a mental calculation to adust for the shift. But for me personally, I would prefer to be able to trust the focusing, as seen through the viewfinder.

Me too. Wide open/MFD. That's why I had Cosina optimize mine for 1.5 . Works better for me.

The thing is, as Marek has persuasively shown, that the 2.8 optimized version will display, at smaller apertures like 2.8 to 4.0, a kind of creamy OOF rendition that most lenses only do around f/2 or wider.

 

[lawrence]

I have a ZM Sonnar and tested it for focus shift and it's clearly optimised for f2.8. With the lens at f1.5 and the subject 4ft away I need to set the focus at 4ft 3ins.

 

[lawrence]

This one is at best of academic interest
to me as I don't have the lens. But if I were choosing, I would prefer the lens optimized for use at f/2.8 for a couple of reasons. First, most lenses resolve much better when stopped down a couple of stops -- that puts the sweet spot of this 50/1.5 (assuming it acts like most other lenses) around f/2.8. Second, if you are shooting people, f/1.5 is such a thin DOF
that you will have a hard time hitting your mark with the lens -- especially if shooting in close, with a center-spot rangefinder -- and keeping both eyes
in focus, if you care about that sort of thing. I do have a Nikkor 50/1.4 Sonnar and I almost never shoot it wider than f/2.8 for these reasons.

I agree. For most practical purposes I think f2.8 is more useful and no doubt why Zeiss corrected the earlier ones for this aperture.

 

[dof]

How can I distingish the one optimized for 1.5 and the one optimized for 2.8? I would like to buy this lens for my M8 in a couple of months, and would like to get the 1.5 version.

My understanding has been that even the new ones from the factory the are supposed to optimized for f1.5 can still be set up for f2.8. The only way to know for certain is to test the focus on the lens in question, or have your dealer test for you.

I bought a copy from Tony Rose at Popflash. I let him know I was interested in an f1.5-optimized lens and while I had to wait a week or so for it(big deal, right?), he tested the copy I bought and it has been spot on.

As an aside, I love this lens on the M8. The combination of its contrast rendering and the sensor yields color in a manner that really shines for me.

 

[john_s]

I have a ZM Sonnar and tested it for focus shift and it's clearly optimised for f2.8. With the lens at f1.5 and the subject 4ft away I need to set the focus at 4ft 3ins.

And since the DoF is about +/-1 one inch even at the old slack criterion (30um) this leads to noticeable unsharpness.

I think Zeiss were naughty publishing impressive MTF curves at f1.5 without saying that such figures were not available using rangefinder focusing.

 

[ferider]

MTF is measured at infinity.

 

[rxmd]

In addition, all other camera makers do the same AFAIK.

 

[Sonnar Brian]

3" shift at 4ft. In the 21st century. What did they do to it?

Maybe it's just me, but my old Sonnars just do not shift that bad...
J-3, optimized for wide-open and close-up:
At F1.5,

At F4:

And cost about 10% of the new Sonnar.

 

[tbarker13]

No doubt Brian. That J-3 of mine that you worked on has taken away any regrets I had over selling my copy of the ZM sonnar.

 

[jmkelly]

3" shift at 4ft. In the 21st century. What did they do to it?

Brian I believe the older Sonnars were softer everywhere than the new ZM, and that the "shift" in the old lenses was lost in the softness.

Someday, if I ever have some spare time, in a fit of total boredom, I might do a comparison roll with all the various 5cm/50mm f1.5 Sonnars in my cabinet. I believe that compared to the older lenses, the ZM is sharper across the whole frame at all apertures, and shows greater contrast and color saturation.

If anyone is interested, back in January I did a mini-comparison with the R-D1 on a tripod, shots here.

 

[MCTuomey]

I have a ZM Sonnar and tested it for focus shift and it's clearly optimised for f2.8. With the lens at f1.5 and the subject 4ft away I need to set the focus at 4ft 3ins.

Or just bend/nod forward to the subject very slightly once the finder does its usual "pop" - only necessary when within, say, 3-6 feet and shooting f1.5-2.0 assuming the lens is f2.8 optimized.

I'm a newb to the lens but so far this "adjustment" works well for me.

 

[john_s]

MTF is measured at infinity.

Is there a focus shift at infinity with this lens? If the adjustment that Zeiss do to the original lens to make it focus at f1.5 involves moving the optical cell relative to the mount, I'd say that there would be a difference in its focus at infinity too. I know that at about 4m there is a very big discrepancy.

 

[sepiareverb]

Get an Elmar-M. Cheaper, and beautiful 3-D rendition.

 

[ferider]

Is there a focus shift at infinity with this lens? If the adjustment that Zeiss do to the original lens to make it focus at f1.5 involves moving the optical cell relative to the mount, I'd say that there would be a difference in its focus at infinity too. I know that at about 4m there is a very big discrepancy.

I don't think there is, and believe they just change the RF coupling (rotate the cam) when adjusting the lens; with identical infinity position. If you own the lens, have a look at it's rear and its segmented RF cam.

Best,

Roland.

 

[jmkelly]

Roland - thanks for pointing out that cam segmenting! I had not noticed it before.

I have a ZM Sonnar and tested it for focus shift and it's clearly optimised for f2.8. With the lens at f1.5 and the subject 4ft away I need to set the focus at 4ft 3ins.

Looking at Roland's self-portraits the shift on his lens looks to be about the thickness of the M3 body.

So we have violated the corpse of this horse once again . No matter that the experience of so many is that the ZM focus shift is 1) easily compensated for, 2) not a consideration in practice, or 3) a figment of a fervid imagination, there are a few that continue to insist that this issue is a BIG DEAL. I guess it is time to start a poll.

 

[Doug]

My C-Sonnar is one of the originals set for f2.8. It's at DAG's shop now and has been 6-bit coded. As a final step he's checking/adjusting the focus according to our correspondence. I've been content with the f2.8 optimization, as wider is a special case for me and I'm willing to make the small manual adjustment under those circumstances as long as it's "on" at f2.8 and in the mid-range.

DAG found another compromise, apparently... He managed to improve focus accuracy at f1.5 on my lens while retaining sufficient accuracy in the mid-range. I gather it's optimized at about f2 now. I'll see when it gets back.