Am I being unreasonable? (Zeiss Sonnar focus issue)

[jaapv]

Advertisement (gone after registration)

What Zeiss needs to do is move the optical cell probably with shims.

That might not be that simple On some Zeiss lenses all elements are shimmed individually.

[Mike Shimwell]

For Zeiss to say tolerences are sloppy with regard to Leica digital register distance is plain absurd. That lens has focus shift and it needs to be calibrated to 1.5, not 2.8 like the original samples , otherwise why have a 1.5 lens. Now all the 5.6 shots are out of focus, but there is more tolerence there. 1.0 Noctilux is the same. 5.6 suffers, but you can bring it back in with an educated guess..

 

 

I think that some care should be taken before assuming the American importer is saying what Zeiss is saying. Zeiss (Germany) has always been completely open about the lens characteristics as far as I know.

 

Also, it doesn't follow that because the maximum aperture is 1.5 that the lens should focus optimsed for f1.5. Zeiss initially chose f2.8 as a practical compromise. However it became apparent that quite a number of users preferred an f1.5 optimisation and Zeiss are very happy to calibrate the lens if owners prefer. It means a short trip back to Oberkochen. Every user needs to decide if it matters and, if so, what they will prefer. The answer won't always be the same.

[Washington]

Just an example of the old 1.5 up in Rocky Mountain Nat. Park:

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

[Scola77]

 

*cough*...

 

Some of us chose to marry a brand new MP4 with a Sonnar last year in preference to a Summilux because they wanted a lens with character and a recognisable "fingerprint" rather than one that delivers sterile clinical precision

 

Regards,

 

Bill

 

Completely agree and I hope you know I was just being a tad sarcastic. I already own 2 other Zeiss lenses and love 'em! Got the 21 f/2.8 and 35 f/2.

[Scola77]

I am actually happy Zeiss America is upfront with what is going on and willing to make it right by offering this service. It is certainly not ideal, but I have the time to get it right. If Zeiss Germany nails the optimization, it will make a fine addition to my arsenal and even make me think twice about getting a lux ASPH (which at cheapest on eBay is 5K+ used).

 

What makes this forum so special is the endless supply of advice, guidance and help with matters such as these. I am the only person I know that uses a Leica M camera and have self taught myself how to use it. There are no Leica dealers in my hometown so pretty much all my knowledge comes from sites such as these,

 

Anyway, I will report back when I test my newly optimized/adjusted 50...hopefully with good news!

 

Thanks again.

[pico]

I remember Leica has a problem with original M digital. The film channel has debth and film uses it, therefore the RF is not calibrated to the pressure plate, but somewhere in the film track.

[...]

What Zeiss needs to do is move the optical cell probably with shims. Digital or film m has nothing to do with the problem. All that was fixed years back.

 

I was thrilled to find that my new M9's focus is right-on with the lenses I use. Even the V2 Summilux 35mm, when locked on infinity shows the image to be not truly at infinity even when the subject is, FAPP, at infinity. I nudge it off a bit until the RF coincides and in 1:1 views, there is a subtle difference - the nudged focus image is sharper. (Images truly at infinity are sharp in the locked position - infinity being very, very far away.)

 

Regarding shimming - do you mean shimming the lens (probably) or the sensor? I'm more for shimming a lens' rear mount rather than the camera.

[pico]

Advertisement (gone after registration)

Just an example of the old 1.5 up in Rocky Mountain Nat. Park:

 

Was that posted to show us how unsharp the image is? I see nothing in focus.

[tobey bilek]

I would assume individual elements are shimmed to make the cell perform to spec. After that it is put in a mount and placed a certain distance from the film either 2.8 or 1.5 correct distance.

[Xmas]

I would assume individual elements are shimmed to make the cell perform to spec. After that it is put in a mount and placed a certain distance from the film either 2.8 or 1.5 correct distance.

 

Hi Tobey

 

I'd suggest fast lenses focus shift cause they are wide aperture, relative shift of lens elements unlikely to help much, they are either milling the mount of the seat of the lens cell in the heliciod or adding a shim to it. It would be the only shim in the lens...

 

Cosina will be 'Deming' all their products using CNC robots such that a girl assembles piece parts without any adjustements, 100% work to spec. Leica should be to, the M4-2 was their first assemble monkey camera.

 

Noel

[Mr_Jones]

Just got a copy of the Sonnar 1.5, new and supposedly optimised for 1.5 at the factory. I'm seeing significant focus shift and back focus. The M8 body is well calibrated with 3 longer lenses so I'm fairly sure the lens is a bit out. Attached are some test card pics. Not sure if I'll adjust to nudging the focus in practice as the shift seems extreme. If you are used to this lens does this behaviour seem normal?

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

[tobey bilek]

Send it in and tell them you want the focus optimized for digital and for 1.5 instead of 2.8.

 

The lens has focus shift so it is impossible to get it to focus at all stops and all distances

[Mr_Jones]

I've just exchanged it for another example, it's supposed to be optimised for 1.5 already as the shop requests that. I've not had a chance to test the second one but I'm guessing from the blurred noses that it is the same or similar.

 

It is going to be a learning curve and I'm sure I will get a lot of blurred noses if I've been drinking!

[gyoung]

Leaving aside for a moment the question of focus shift, something that still puzzles me is that the focal plane of the digital camera is more precise than with film because film can bow in the channel.

In order to cope with this lack of precision in the focal plane, did leica (and others) pick a mid point of this range as the register? If so why did they not stick to the same figure with the placement of the flatter sensor. If they did this there would be no difference in performance between film and digital except that many more film images would show inaccurate focus at wide apertures and close distances.

Indeed its a wonder that we ever got sharp eyeballs at f/1.5 and 3 feet on film, mere chance? And how on earth do you make use of a Noctilux on film.

 

Gerry

[shootinglulu]

I find the 50 Zeiss C Sonnar as pleasing as the 50 Lux asph in it's rendering and I enjoy carrying a less expensive lens around in my coat pocket..

The Zeiss does not focus as closely which can be fustrating if you're used to a closer focusing lens but I find the Zeiss easier to focus than the Leica and it's smaller.

I don't feel the Zeiss has any more character than the Leica..

On balance I kept the Zeiss and love it on the film M..

[ho_co]

Gerry--

Leica did use the same focus point for digital as they had for film.

 

Since color film has three color-sensitive layers, some parts of the image might be pretty much in focus on two of the three, while other parts might be pretty much in focus at another depth. Overall, the image will still look sharp.

 

In other words, the fact that film doesn't lie flat is largely compensated by the very thickness of the film.

 

Leica always took film characteristics into consideration. That's a big part of why the aspherical 35/1.4 was redesigned to have a floating element after the digital cameras were added to the stable.

 

The accuracy of focus needed for the digital sensor is greater than that needed for film. That's why if a lens focuses properly on a sensor, it will also focus accurately on film.

[jaapv]

Leaving aside for a moment the question of focus shift, something that still puzzles me is that the focal plane of the digital camera is more precise than with film because film can bow in the channel.

In order to cope with this lack of precision in the focal plane, did leica (and others) pick a mid point of this range as the register? If so why did they not stick to the same figure with the placement of the flatter sensor. If they did this there would be no difference in performance between film and digital except that many more film images would show inaccurate focus at wide apertures and close distances.

Indeed its a wonder that we ever got sharp eyeballs at f/1.5 and 3 feet on film, mere chance? And how on earth do you make use of a Noctilux on film.

 

Gerry

It is not a question of placement, that is unchanged. It is a matter of adjustment tolerances. As film is less precise than a sensor, due to its thickness, the lens can be more "off" on film. So whilst many old lenses will be spot-on, even on a digital camera, some , that are at the outer limits of the tolerance span, will need adjustment to the tighter present-day values.

[gyoung]

Surely film emulsion thickness varies, ok colour is thick, but Ektachrome is thicker than Kodachrome because of the extra coupler layers, and what about b&w, much thinner. If you add that variation to lower tolerances on registration of of bodies and lenses I would have expected a lot of people to have problems with film bodies at large apertures and close distances, yet it seems to be a problem of digital bodies, even with new lenses.

Emulsion thickness is surely small, microns for each layer? I would have thought there could be more variation in the film position in the focal plane.

An uncharitable thought might be that the problems with the digital bodies stem from poor quality control rather than inherent differences, do other manufactures have similar problems?

 

Gerry

[wlaidlaw]

Gerry,

 

I suspect it is just as simple as you first thought - sloppy adjustment/assembly at the factory, probably for both examples. In my experience, this is rare on Zeiss lenses. All four of the ZM's I have had (21/2.8, 25/2.8, 35/2 and 50/2) have been spot on when checked with the same chart as you use. In fact I have had a better track record with Zeiss ZM's than I have had with modern Leica lenses. Sure the Sonnar is designed to be a bit soft wide open, but it should still focus accurately. I also have not heard about a reputation for aperture shift for it, like many examples of the Leica 35/1.4 ASPH Summilux suffer from. The very first thing I always do with a new lens is run a focus check on it. If you don't know that your tools work properly, not just for photography, then you cannot depend on results of any work.

 

Wilson

[gyoung]

Thanks Wilson, I tend to test new items more subjectively, try them out for the sort of thing I normally do but stretching them to large apertures etc. if that shows up problems I will do more 'scientific' tests.

With my own stuff, 2 M3s, an M6, a III and an R2 with a dozen leica and half that number of Voigtlander lenses have not thrown up any such issues, likewise over the years 5 different Nikon film bodies with quite a few lenses of various makes have been ok with film, but the digatal body bought recently had problems, quickly fixed under warranty.

I don't have any really exotic lenses though!

 

Gerry

[Advertisement]

Hallo,

Look here: Leica M

[TomB_tx]

I agree with Wilson - possibly the lens, although the focus shift tendency is more a matter of design. I did get a new Zeiss Biogon 35 f2.0 that front focused from the factory, while all the other Zeiss and Voigtlander lenses I've bought have been right-on for focus. So an "off" lens can happen even from a good production facility.